diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18132 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9997583723964626,
+  "eval_steps": 500,
+  "global_step": 2586,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 481.5364157889796,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 12.6553,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 373.615837696346,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 10.8926,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 541.3521511663153,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 13.2363,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 523.5132315653206,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 12.9629,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 515.0572137863752,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 12.7852,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 526.3811596094474,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 12.8916,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 407.57501180989584,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 11.002,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 482.648026875231,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 12.6035,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 504.92910208576893,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 12.6074,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 393.6360870839124,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 12.0127,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 246.30275749513527,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 10.6406,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 316.79595857347636,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 11.5801,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 179.76546501029475,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 11.0068,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 105.34604762825141,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 10.9961,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 95.20928280649406,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 10.9141,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 97.71161584750286,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 10.8096,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 84.80974418089382,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 10.6816,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 66.99584392713597,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 10.4033,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 77.59980258457645,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 10.4561,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 71.97784133450288,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 10.3604,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 60.71096373084412,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 10.2812,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 52.01823702478759,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 10.1553,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 501.30437305720193,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 10.7451,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 50.77186630197056,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 9.9365,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 123.06902752739383,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 10.3555,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 64.4392635165463,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 9.7417,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 46.916655698121126,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 9.5879,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 90.32706977577234,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 10.2334,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 88.8653324960376,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 9.6416,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 74.11904460680567,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 10.3936,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 48.289134449316116,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 10.3271,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 82.4031839271886,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 10.3516,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 68.1926298316496,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 9.5762,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 44.042915685465566,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 9.5054,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 53.10829325161672,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 9.4141,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 45.390533256215136,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 9.2974,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 53.02285450932354,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 9.2637,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 57.3374438362884,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 9.0791,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 56.815722763458126,
+      "learning_rate": 5e-06,
+      "loss": 9.1606,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 80.19958779912238,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 8.9976,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 85.48983399371168,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 8.8926,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 56.04763061491995,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 8.73,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 74.50586478188413,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 8.4473,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 73.04874972281584,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 8.0718,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 264.9715361301418,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 8.4507,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 152.69010501557233,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 6.9888,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 110.88579858458279,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 5.6064,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 120.12136882420224,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 4.519,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 79.99404201169071,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 3.5048,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 85.44304149483389,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 3.7528,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 66.550368967976,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 2.9167,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 125.42474038721963,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 2.1831,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 50.615899773503365,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 2.5865,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 47.90170223928453,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 2.9435,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 41.24265758793064,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 2.2693,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 38.379748635834446,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 1.8779,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 52.765102741075474,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 1.4181,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 45.376548508681665,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 2.2791,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.886474161668495,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 1.9019,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 33.882692272856325,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 2.1064,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 31.80399881176114,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 1.6717,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 30.431479126124767,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 1.46,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.290390440704332,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 1.1166,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 37.69888192511318,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 1.0918,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.44543810080971,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.3311,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 27.742406389524945,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 1.2252,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 28.651983435707272,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 1.2317,
+      "step": 67
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.127454566695906,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 1.2144,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 26.80788136857936,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 1.1363,
+      "step": 69
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.84510398734904,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 0.9485,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 27.249410271667543,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 1.2925,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 26.35779120732198,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.3286,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 27.48592890019767,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 1.1482,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.382578820647375,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 1.398,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.256158986483786,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 1.09,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.26324817474908,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 0.9597,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.552393719923757,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 1.2243,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 28.839364422122316,
+      "learning_rate": 1e-05,
+      "loss": 1.5553,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.65991391720404,
+      "learning_rate": 9.999996077304179e-06,
+      "loss": 1.1489,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.516637811440408,
+      "learning_rate": 9.999984309222866e-06,
+      "loss": 0.7853,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.110210236404964,
+      "learning_rate": 9.999964695774527e-06,
+      "loss": 1.0838,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.769777062435416,
+      "learning_rate": 9.99993723698994e-06,
+      "loss": 1.1767,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.92559374442647,
+      "learning_rate": 9.999901932912188e-06,
+      "loss": 0.8008,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.632354954328026,
+      "learning_rate": 9.999858783596665e-06,
+      "loss": 1.1066,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.110456168304566,
+      "learning_rate": 9.999807789111075e-06,
+      "loss": 0.694,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.070868787310946,
+      "learning_rate": 9.999748949535436e-06,
+      "loss": 1.2248,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.460128994376806,
+      "learning_rate": 9.99968226496207e-06,
+      "loss": 1.267,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.038548464267855,
+      "learning_rate": 9.999607735495609e-06,
+      "loss": 0.6299,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.687490030021358,
+      "learning_rate": 9.999525361252996e-06,
+      "loss": 0.7391,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.75441205219283,
+      "learning_rate": 9.999435142363484e-06,
+      "loss": 0.8803,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.87630811704553,
+      "learning_rate": 9.999337078968633e-06,
+      "loss": 0.688,
+      "step": 91
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.159533517580149,
+      "learning_rate": 9.999231171222312e-06,
+      "loss": 0.6074,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.99879481987676,
+      "learning_rate": 9.999117419290698e-06,
+      "loss": 1.096,
+      "step": 93
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 22.861060099644742,
+      "learning_rate": 9.998995823352276e-06,
+      "loss": 1.0218,
+      "step": 94
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.95571036176603,
+      "learning_rate": 9.998866383597842e-06,
+      "loss": 0.7592,
+      "step": 95
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 21.033741483410342,
+      "learning_rate": 9.998729100230497e-06,
+      "loss": 1.2447,
+      "step": 96
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.823778921682337,
+      "learning_rate": 9.998583973465647e-06,
+      "loss": 0.5127,
+      "step": 97
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.08061210135059,
+      "learning_rate": 9.998431003531008e-06,
+      "loss": 0.7337,
+      "step": 98
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.995139849703078,
+      "learning_rate": 9.998270190666602e-06,
+      "loss": 0.5557,
+      "step": 99
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.124092791792007,
+      "learning_rate": 9.998101535124758e-06,
+      "loss": 0.5056,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.368533101873233,
+      "learning_rate": 9.99792503717011e-06,
+      "loss": 0.6619,
+      "step": 101
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.14285505262652,
+      "learning_rate": 9.997740697079595e-06,
+      "loss": 0.9017,
+      "step": 102
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.000733016447228,
+      "learning_rate": 9.997548515142457e-06,
+      "loss": 0.6348,
+      "step": 103
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.11313981069571,
+      "learning_rate": 9.997348491660247e-06,
+      "loss": 0.6906,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.23633304575936,
+      "learning_rate": 9.997140626946815e-06,
+      "loss": 0.4467,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 21.567298431907933,
+      "learning_rate": 9.99692492132832e-06,
+      "loss": 0.8592,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.329760629248835,
+      "learning_rate": 9.996701375143216e-06,
+      "loss": 0.4972,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.16287909184406,
+      "learning_rate": 9.99646998874227e-06,
+      "loss": 0.8043,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 21.62899621363284,
+      "learning_rate": 9.996230762488544e-06,
+      "loss": 1.1046,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.257522808534574,
+      "learning_rate": 9.9959836967574e-06,
+      "loss": 0.8829,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.309636548333255,
+      "learning_rate": 9.995728791936505e-06,
+      "loss": 0.5758,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 21.881094638157435,
+      "learning_rate": 9.995466048425825e-06,
+      "loss": 0.8626,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.005222348952602,
+      "learning_rate": 9.995195466637629e-06,
+      "loss": 0.5917,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.71551135268679,
+      "learning_rate": 9.994917046996472e-06,
+      "loss": 1.1734,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.144869808442866,
+      "learning_rate": 9.994630789939223e-06,
+      "loss": 1.0547,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.962788937821886,
+      "learning_rate": 9.994336695915041e-06,
+      "loss": 0.9223,
+      "step": 116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.0577013636892,
+      "learning_rate": 9.99403476538538e-06,
+      "loss": 1.1817,
+      "step": 117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.86960241364737,
+      "learning_rate": 9.993724998823995e-06,
+      "loss": 0.924,
+      "step": 118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.515776024934624,
+      "learning_rate": 9.993407396716935e-06,
+      "loss": 1.0204,
+      "step": 119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.929496531011143,
+      "learning_rate": 9.993081959562539e-06,
+      "loss": 0.8133,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.849782696338378,
+      "learning_rate": 9.992748687871445e-06,
+      "loss": 0.8117,
+      "step": 121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.908001749557238,
+      "learning_rate": 9.992407582166582e-06,
+      "loss": 0.3049,
+      "step": 122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.547013586967662,
+      "learning_rate": 9.992058642983173e-06,
+      "loss": 0.7054,
+      "step": 123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 21.972679645643126,
+      "learning_rate": 9.991701870868731e-06,
+      "loss": 1.1915,
+      "step": 124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.365040700379357,
+      "learning_rate": 9.991337266383056e-06,
+      "loss": 1.0072,
+      "step": 125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.357110092218164,
+      "learning_rate": 9.990964830098246e-06,
+      "loss": 0.8587,
+      "step": 126
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.56485575350805,
+      "learning_rate": 9.990584562598679e-06,
+      "loss": 0.9168,
+      "step": 127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.649284451546533,
+      "learning_rate": 9.990196464481025e-06,
+      "loss": 0.7338,
+      "step": 128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.044217159005804,
+      "learning_rate": 9.989800536354243e-06,
+      "loss": 0.8382,
+      "step": 129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.356186603100785,
+      "learning_rate": 9.989396778839572e-06,
+      "loss": 0.5108,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.871602416017183,
+      "learning_rate": 9.988985192570541e-06,
+      "loss": 0.9732,
+      "step": 131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.269214745384646,
+      "learning_rate": 9.98856577819296e-06,
+      "loss": 0.6414,
+      "step": 132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.605076125957375,
+      "learning_rate": 9.988138536364922e-06,
+      "loss": 0.402,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.32994513897714,
+      "learning_rate": 9.987703467756807e-06,
+      "loss": 0.9181,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.052228695935232,
+      "learning_rate": 9.987260573051268e-06,
+      "loss": 0.7179,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.88336521964054,
+      "learning_rate": 9.986809852943244e-06,
+      "loss": 0.4574,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.720367253178598,
+      "learning_rate": 9.986351308139948e-06,
+      "loss": 0.6479,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.809162454287666,
+      "learning_rate": 9.985884939360873e-06,
+      "loss": 0.8713,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.029722474256518,
+      "learning_rate": 9.985410747337788e-06,
+      "loss": 0.6597,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.629873415137338,
+      "learning_rate": 9.984928732814738e-06,
+      "loss": 0.8715,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.495969338119345,
+      "learning_rate": 9.984438896548043e-06,
+      "loss": 1.0109,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.557303172161872,
+      "learning_rate": 9.983941239306291e-06,
+      "loss": 0.7137,
+      "step": 142
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.100360057273956,
+      "learning_rate": 9.98343576187035e-06,
+      "loss": 0.603,
+      "step": 143
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 23.31303623872249,
+      "learning_rate": 9.98292246503335e-06,
+      "loss": 0.5843,
+      "step": 144
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.748929410677942,
+      "learning_rate": 9.982401349600695e-06,
+      "loss": 1.1261,
+      "step": 145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.027377085609972,
+      "learning_rate": 9.981872416390055e-06,
+      "loss": 0.6,
+      "step": 146
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.149135468259303,
+      "learning_rate": 9.981335666231369e-06,
+      "loss": 0.9303,
+      "step": 147
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.743095915764165,
+      "learning_rate": 9.98079109996684e-06,
+      "loss": 0.8395,
+      "step": 148
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.284198483372663,
+      "learning_rate": 9.980238718450934e-06,
+      "loss": 0.7735,
+      "step": 149
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.89526777115439,
+      "learning_rate": 9.979678522550382e-06,
+      "loss": 0.5915,
+      "step": 150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.080587318662797,
+      "learning_rate": 9.979110513144175e-06,
+      "loss": 0.8949,
+      "step": 151
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.521683815398124,
+      "learning_rate": 9.978534691123563e-06,
+      "loss": 0.8933,
+      "step": 152
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.32984954555658,
+      "learning_rate": 9.977951057392057e-06,
+      "loss": 0.715,
+      "step": 153
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.846840360048837,
+      "learning_rate": 9.977359612865424e-06,
+      "loss": 0.9738,
+      "step": 154
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.66118183098209,
+      "learning_rate": 9.976760358471687e-06,
+      "loss": 0.6333,
+      "step": 155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.70839351701734,
+      "learning_rate": 9.976153295151123e-06,
+      "loss": 0.6789,
+      "step": 156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.0894686477276,
+      "learning_rate": 9.975538423856261e-06,
+      "loss": 0.7859,
+      "step": 157
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.954417139226864,
+      "learning_rate": 9.974915745551882e-06,
+      "loss": 0.61,
+      "step": 158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.537000612431887,
+      "learning_rate": 9.97428526121502e-06,
+      "loss": 0.9158,
+      "step": 159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.613435302994104,
+      "learning_rate": 9.97364697183495e-06,
+      "loss": 1.0159,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.030764596896761,
+      "learning_rate": 9.973000878413203e-06,
+      "loss": 0.6882,
+      "step": 161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.686888525158768,
+      "learning_rate": 9.972346981963546e-06,
+      "loss": 0.9044,
+      "step": 162
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.28484214901929,
+      "learning_rate": 9.971685283511996e-06,
+      "loss": 0.3627,
+      "step": 163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.036524213860528,
+      "learning_rate": 9.971015784096808e-06,
+      "loss": 0.4503,
+      "step": 164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.944141583254748,
+      "learning_rate": 9.970338484768481e-06,
+      "loss": 0.5213,
+      "step": 165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.334431234076161,
+      "learning_rate": 9.969653386589749e-06,
+      "loss": 0.8745,
+      "step": 166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.756997212862416,
+      "learning_rate": 9.968960490635584e-06,
+      "loss": 0.8582,
+      "step": 167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.500935474530356,
+      "learning_rate": 9.968259797993197e-06,
+      "loss": 0.4101,
+      "step": 168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.131658386925906,
+      "learning_rate": 9.967551309762028e-06,
+      "loss": 0.7268,
+      "step": 169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.962827814051503,
+      "learning_rate": 9.96683502705375e-06,
+      "loss": 0.8146,
+      "step": 170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.496181084624276,
+      "learning_rate": 9.966110950992267e-06,
+      "loss": 0.6242,
+      "step": 171
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.148046193569224,
+      "learning_rate": 9.965379082713711e-06,
+      "loss": 0.5653,
+      "step": 172
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.171128302657074,
+      "learning_rate": 9.964639423366442e-06,
+      "loss": 0.8087,
+      "step": 173
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.193272560803273,
+      "learning_rate": 9.963891974111042e-06,
+      "loss": 0.6207,
+      "step": 174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.543872733390224,
+      "learning_rate": 9.963136736120318e-06,
+      "loss": 0.8392,
+      "step": 175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.248084489796383,
+      "learning_rate": 9.962373710579296e-06,
+      "loss": 0.8543,
+      "step": 176
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.484173234833333,
+      "learning_rate": 9.961602898685225e-06,
+      "loss": 0.3479,
+      "step": 177
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.28297351527292,
+      "learning_rate": 9.960824301647569e-06,
+      "loss": 0.6513,
+      "step": 178
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.253644262368155,
+      "learning_rate": 9.960037920688007e-06,
+      "loss": 0.7209,
+      "step": 179
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.214769308290915,
+      "learning_rate": 9.959243757040434e-06,
+      "loss": 1.1318,
+      "step": 180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.700363937070268,
+      "learning_rate": 9.958441811950952e-06,
+      "loss": 0.6636,
+      "step": 181
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.598001384260753,
+      "learning_rate": 9.957632086677876e-06,
+      "loss": 0.5206,
+      "step": 182
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.919790411513354,
+      "learning_rate": 9.956814582491731e-06,
+      "loss": 0.6174,
+      "step": 183
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.624843976839964,
+      "learning_rate": 9.955989300675243e-06,
+      "loss": 0.6566,
+      "step": 184
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.814502003312239,
+      "learning_rate": 9.955156242523345e-06,
+      "loss": 0.4054,
+      "step": 185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.4334075653115,
+      "learning_rate": 9.95431540934317e-06,
+      "loss": 0.995,
+      "step": 186
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.99469402177115,
+      "learning_rate": 9.95346680245405e-06,
+      "loss": 0.4368,
+      "step": 187
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.497301252471559,
+      "learning_rate": 9.952610423187516e-06,
+      "loss": 0.5486,
+      "step": 188
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 18.203010945056977,
+      "learning_rate": 9.951746272887298e-06,
+      "loss": 0.6987,
+      "step": 189
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.266313412060434,
+      "learning_rate": 9.95087435290931e-06,
+      "loss": 0.7963,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.453844442747691,
+      "learning_rate": 9.949994664621664e-06,
+      "loss": 0.8017,
+      "step": 191
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.018840099119014,
+      "learning_rate": 9.949107209404664e-06,
+      "loss": 0.8221,
+      "step": 192
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.435658058734695,
+      "learning_rate": 9.948211988650794e-06,
+      "loss": 0.6759,
+      "step": 193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.589226269743104,
+      "learning_rate": 9.947309003764723e-06,
+      "loss": 0.8476,
+      "step": 194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.37013746184168,
+      "learning_rate": 9.946398256163307e-06,
+      "loss": 0.9391,
+      "step": 195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.383288098836749,
+      "learning_rate": 9.94547974727558e-06,
+      "loss": 0.4744,
+      "step": 196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.917621598757963,
+      "learning_rate": 9.944553478542757e-06,
+      "loss": 0.8883,
+      "step": 197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.83605170223955,
+      "learning_rate": 9.943619451418225e-06,
+      "loss": 0.7551,
+      "step": 198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.769255626775342,
+      "learning_rate": 9.942677667367541e-06,
+      "loss": 0.6543,
+      "step": 199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.851763021841114,
+      "learning_rate": 9.941728127868446e-06,
+      "loss": 0.4974,
+      "step": 200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.36157853324669,
+      "learning_rate": 9.940770834410836e-06,
+      "loss": 0.8592,
+      "step": 201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.598908676624283,
+      "learning_rate": 9.939805788496778e-06,
+      "loss": 0.8721,
+      "step": 202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.86750627304113,
+      "learning_rate": 9.938832991640512e-06,
+      "loss": 0.8129,
+      "step": 203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.869386375404257,
+      "learning_rate": 9.937852445368427e-06,
+      "loss": 0.5172,
+      "step": 204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.111611284149054,
+      "learning_rate": 9.936864151219077e-06,
+      "loss": 0.6053,
+      "step": 205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.453502771706678,
+      "learning_rate": 9.935868110743175e-06,
+      "loss": 1.2113,
+      "step": 206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.846093866735691,
+      "learning_rate": 9.934864325503584e-06,
+      "loss": 0.711,
+      "step": 207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.404985998639587,
+      "learning_rate": 9.933852797075325e-06,
+      "loss": 0.749,
+      "step": 208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.806844820304013,
+      "learning_rate": 9.932833527045563e-06,
+      "loss": 0.7894,
+      "step": 209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.057134555889663,
+      "learning_rate": 9.931806517013612e-06,
+      "loss": 0.3543,
+      "step": 210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.26916959315829,
+      "learning_rate": 9.930771768590934e-06,
+      "loss": 0.8822,
+      "step": 211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.329969646979569,
+      "learning_rate": 9.929729283401127e-06,
+      "loss": 1.0573,
+      "step": 212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.951752175310512,
+      "learning_rate": 9.928679063079934e-06,
+      "loss": 1.1208,
+      "step": 213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 18.867489025286066,
+      "learning_rate": 9.927621109275233e-06,
+      "loss": 0.8248,
+      "step": 214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.825573624560661,
+      "learning_rate": 9.926555423647035e-06,
+      "loss": 0.7668,
+      "step": 215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.151796481318033,
+      "learning_rate": 9.925482007867485e-06,
+      "loss": 0.6053,
+      "step": 216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.574192384305192,
+      "learning_rate": 9.924400863620857e-06,
+      "loss": 0.4994,
+      "step": 217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 25.487482582659652,
+      "learning_rate": 9.92331199260355e-06,
+      "loss": 0.4207,
+      "step": 218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.706432457880826,
+      "learning_rate": 9.922215396524089e-06,
+      "loss": 0.8241,
+      "step": 219
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.046136375492727,
+      "learning_rate": 9.921111077103118e-06,
+      "loss": 0.8794,
+      "step": 220
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.216130467061092,
+      "learning_rate": 9.9199990360734e-06,
+      "loss": 0.4697,
+      "step": 221
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.790430832864136,
+      "learning_rate": 9.918879275179819e-06,
+      "loss": 0.9238,
+      "step": 222
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.60646458206681,
+      "learning_rate": 9.91775179617936e-06,
+      "loss": 0.4365,
+      "step": 223
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.50945580415778,
+      "learning_rate": 9.916616600841133e-06,
+      "loss": 1.0106,
+      "step": 224
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.124165314781543,
+      "learning_rate": 9.915473690946345e-06,
+      "loss": 0.9838,
+      "step": 225
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.33921135486922,
+      "learning_rate": 9.914323068288312e-06,
+      "loss": 0.6934,
+      "step": 226
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.773135279302,
+      "learning_rate": 9.91316473467245e-06,
+      "loss": 0.3533,
+      "step": 227
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.712076656888206,
+      "learning_rate": 9.911998691916275e-06,
+      "loss": 0.8798,
+      "step": 228
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.321034050673962,
+      "learning_rate": 9.910824941849401e-06,
+      "loss": 0.6144,
+      "step": 229
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.807512556600663,
+      "learning_rate": 9.909643486313533e-06,
+      "loss": 1.0085,
+      "step": 230
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.0716331281851,
+      "learning_rate": 9.908454327162469e-06,
+      "loss": 0.8295,
+      "step": 231
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.832281460936676,
+      "learning_rate": 9.90725746626209e-06,
+      "loss": 0.5636,
+      "step": 232
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.313004994286489,
+      "learning_rate": 9.906052905490366e-06,
+      "loss": 0.8005,
+      "step": 233
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.278226823757532,
+      "learning_rate": 9.904840646737346e-06,
+      "loss": 0.6354,
+      "step": 234
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.53374789262419,
+      "learning_rate": 9.90362069190516e-06,
+      "loss": 0.717,
+      "step": 235
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.998858736575075,
+      "learning_rate": 9.902393042908015e-06,
+      "loss": 0.6212,
+      "step": 236
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.793993487941199,
+      "learning_rate": 9.901157701672183e-06,
+      "loss": 0.5298,
+      "step": 237
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.63190031707497,
+      "learning_rate": 9.899914670136016e-06,
+      "loss": 0.9484,
+      "step": 238
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.535962960887657,
+      "learning_rate": 9.898663950249925e-06,
+      "loss": 0.5316,
+      "step": 239
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.541903781575932,
+      "learning_rate": 9.89740554397639e-06,
+      "loss": 0.7485,
+      "step": 240
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.733705076482586,
+      "learning_rate": 9.896139453289946e-06,
+      "loss": 0.8448,
+      "step": 241
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.637158660226499,
+      "learning_rate": 9.89486568017719e-06,
+      "loss": 0.4599,
+      "step": 242
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.770280668620607,
+      "learning_rate": 9.893584226636773e-06,
+      "loss": 0.9209,
+      "step": 243
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.191150363467036,
+      "learning_rate": 9.892295094679394e-06,
+      "loss": 0.9502,
+      "step": 244
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.440332805520622,
+      "learning_rate": 9.890998286327802e-06,
+      "loss": 0.7562,
+      "step": 245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.183204852337655,
+      "learning_rate": 9.889693803616793e-06,
+      "loss": 0.805,
+      "step": 246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.372686801084296,
+      "learning_rate": 9.8883816485932e-06,
+      "loss": 0.8335,
+      "step": 247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.312330309808992,
+      "learning_rate": 9.8870618233159e-06,
+      "loss": 0.6341,
+      "step": 248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.965185075220083,
+      "learning_rate": 9.885734329855798e-06,
+      "loss": 0.9597,
+      "step": 249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.330050676547078,
+      "learning_rate": 9.884399170295839e-06,
+      "loss": 0.307,
+      "step": 250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.853807179695082,
+      "learning_rate": 9.883056346730992e-06,
+      "loss": 0.8651,
+      "step": 251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.37943395796441,
+      "learning_rate": 9.881705861268252e-06,
+      "loss": 0.9053,
+      "step": 252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.002233079622927,
+      "learning_rate": 9.880347716026635e-06,
+      "loss": 0.5673,
+      "step": 253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.13571657427554,
+      "learning_rate": 9.878981913137178e-06,
+      "loss": 0.9166,
+      "step": 254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.28773017728128,
+      "learning_rate": 9.877608454742935e-06,
+      "loss": 0.7975,
+      "step": 255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.91888276894326,
+      "learning_rate": 9.87622734299897e-06,
+      "loss": 0.6809,
+      "step": 256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.112063987479349,
+      "learning_rate": 9.87483858007235e-06,
+      "loss": 0.6315,
+      "step": 257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.717478338478736,
+      "learning_rate": 9.873442168142158e-06,
+      "loss": 0.7987,
+      "step": 258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.976682853641545,
+      "learning_rate": 9.872038109399474e-06,
+      "loss": 0.7399,
+      "step": 259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.019130092345165,
+      "learning_rate": 9.870626406047372e-06,
+      "loss": 0.6962,
+      "step": 260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.544973209726715,
+      "learning_rate": 9.869207060300929e-06,
+      "loss": 0.6297,
+      "step": 261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.56822630578895,
+      "learning_rate": 9.867780074387207e-06,
+      "loss": 0.3392,
+      "step": 262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.414663456945581,
+      "learning_rate": 9.86634545054526e-06,
+      "loss": 0.4267,
+      "step": 263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.80463976701528,
+      "learning_rate": 9.864903191026125e-06,
+      "loss": 0.4679,
+      "step": 264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.22991742072056,
+      "learning_rate": 9.86345329809282e-06,
+      "loss": 0.7036,
+      "step": 265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.554718755864346,
+      "learning_rate": 9.861995774020341e-06,
+      "loss": 0.753,
+      "step": 266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.14537913565357,
+      "learning_rate": 9.860530621095659e-06,
+      "loss": 0.8066,
+      "step": 267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.434238244590858,
+      "learning_rate": 9.859057841617709e-06,
+      "loss": 0.68,
+      "step": 268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.236945029363593,
+      "learning_rate": 9.857577437897401e-06,
+      "loss": 0.9234,
+      "step": 269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.971268013963655,
+      "learning_rate": 9.856089412257605e-06,
+      "loss": 0.8905,
+      "step": 270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.180266548897267,
+      "learning_rate": 9.854593767033147e-06,
+      "loss": 0.9239,
+      "step": 271
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.900165311429614,
+      "learning_rate": 9.853090504570813e-06,
+      "loss": 0.7079,
+      "step": 272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.630107114784986,
+      "learning_rate": 9.85157962722934e-06,
+      "loss": 0.6955,
+      "step": 273
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.809851250786709,
+      "learning_rate": 9.850061137379414e-06,
+      "loss": 0.7326,
+      "step": 274
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.763338779408269,
+      "learning_rate": 9.84853503740366e-06,
+      "loss": 0.5795,
+      "step": 275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.867602488164216,
+      "learning_rate": 9.847001329696653e-06,
+      "loss": 0.6658,
+      "step": 276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.926655157463944,
+      "learning_rate": 9.845460016664899e-06,
+      "loss": 0.5341,
+      "step": 277
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 21.861487938034404,
+      "learning_rate": 9.843911100726838e-06,
+      "loss": 0.9161,
+      "step": 278
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.578651892796103,
+      "learning_rate": 9.842354584312841e-06,
+      "loss": 1.1774,
+      "step": 279
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.95235599485366,
+      "learning_rate": 9.840790469865205e-06,
+      "loss": 0.6807,
+      "step": 280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.273416917093268,
+      "learning_rate": 9.839218759838146e-06,
+      "loss": 0.7789,
+      "step": 281
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.660305939473908,
+      "learning_rate": 9.837639456697802e-06,
+      "loss": 0.6102,
+      "step": 282
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.674321658886594,
+      "learning_rate": 9.836052562922224e-06,
+      "loss": 0.4836,
+      "step": 283
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.446557932269133,
+      "learning_rate": 9.83445808100137e-06,
+      "loss": 0.2829,
+      "step": 284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.95429896915303,
+      "learning_rate": 9.83285601343711e-06,
+      "loss": 0.4486,
+      "step": 285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.182400263060108,
+      "learning_rate": 9.83124636274321e-06,
+      "loss": 0.9535,
+      "step": 286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.325480362529039,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.4282,
+      "step": 287
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.955293072524547,
+      "learning_rate": 9.828004322081067e-06,
+      "loss": 0.6136,
+      "step": 288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.68063847364157,
+      "learning_rate": 9.826371937199837e-06,
+      "loss": 0.7924,
+      "step": 289
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.752375694482755,
+      "learning_rate": 9.824731979362991e-06,
+      "loss": 0.6666,
+      "step": 290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.558418540345379,
+      "learning_rate": 9.823084451143755e-06,
+      "loss": 0.6402,
+      "step": 291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.289831377102262,
+      "learning_rate": 9.821429355127225e-06,
+      "loss": 0.475,
+      "step": 292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.666137689193855,
+      "learning_rate": 9.819766693910382e-06,
+      "loss": 0.6295,
+      "step": 293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.101404916196465,
+      "learning_rate": 9.818096470102067e-06,
+      "loss": 0.5609,
+      "step": 294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.31598032284623,
+      "learning_rate": 9.816418686322995e-06,
+      "loss": 0.6307,
+      "step": 295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.379107161465395,
+      "learning_rate": 9.814733345205736e-06,
+      "loss": 1.0251,
+      "step": 296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.486894401842418,
+      "learning_rate": 9.813040449394728e-06,
+      "loss": 0.4821,
+      "step": 297
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.485431455332018,
+      "learning_rate": 9.811340001546252e-06,
+      "loss": 0.6005,
+      "step": 298
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.324218285023527,
+      "learning_rate": 9.809632004328448e-06,
+      "loss": 0.6256,
+      "step": 299
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.630664964848293,
+      "learning_rate": 9.807916460421294e-06,
+      "loss": 0.7766,
+      "step": 300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.180468979520766,
+      "learning_rate": 9.806193372516615e-06,
+      "loss": 0.4442,
+      "step": 301
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.992493754759456,
+      "learning_rate": 9.80446274331807e-06,
+      "loss": 0.5413,
+      "step": 302
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.024474002254971,
+      "learning_rate": 9.802724575541152e-06,
+      "loss": 0.8754,
+      "step": 303
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.262595712517527,
+      "learning_rate": 9.800978871913181e-06,
+      "loss": 0.5826,
+      "step": 304
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.49732050408989,
+      "learning_rate": 9.799225635173305e-06,
+      "loss": 0.6971,
+      "step": 305
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.102131145811159,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 0.7166,
+      "step": 306
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.352157818643372,
+      "learning_rate": 9.795696573373511e-06,
+      "loss": 0.4842,
+      "step": 307
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.872864281797748,
+      "learning_rate": 9.793920753850972e-06,
+      "loss": 0.5209,
+      "step": 308
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.449784634753554,
+      "learning_rate": 9.792137412291265e-06,
+      "loss": 0.4941,
+      "step": 309
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.50302358883469,
+      "learning_rate": 9.790346551492594e-06,
+      "loss": 0.6571,
+      "step": 310
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.583378528777525,
+      "learning_rate": 9.788548174264961e-06,
+      "loss": 0.4422,
+      "step": 311
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.777028371125699,
+      "learning_rate": 9.78674228343016e-06,
+      "loss": 0.6663,
+      "step": 312
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.767121980627469,
+      "learning_rate": 9.784928881821776e-06,
+      "loss": 0.7057,
+      "step": 313
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.643690689173766,
+      "learning_rate": 9.783107972285177e-06,
+      "loss": 0.5859,
+      "step": 314
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.58718992797939,
+      "learning_rate": 9.781279557677514e-06,
+      "loss": 0.6565,
+      "step": 315
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.61723438008309,
+      "learning_rate": 9.779443640867712e-06,
+      "loss": 0.4149,
+      "step": 316
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.903051301626181,
+      "learning_rate": 9.777600224736468e-06,
+      "loss": 0.8454,
+      "step": 317
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.27820238705729,
+      "learning_rate": 9.775749312176249e-06,
+      "loss": 1.0296,
+      "step": 318
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.561919459941654,
+      "learning_rate": 9.773890906091275e-06,
+      "loss": 0.54,
+      "step": 319
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.90625088725865,
+      "learning_rate": 9.772025009397538e-06,
+      "loss": 0.6082,
+      "step": 320
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.795931318192698,
+      "learning_rate": 9.770151625022772e-06,
+      "loss": 0.4639,
+      "step": 321
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.362255993627048,
+      "learning_rate": 9.768270755906467e-06,
+      "loss": 0.737,
+      "step": 322
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.404405821505582,
+      "learning_rate": 9.76638240499985e-06,
+      "loss": 0.3936,
+      "step": 323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.226985176463659,
+      "learning_rate": 9.764486575265893e-06,
+      "loss": 0.7815,
+      "step": 324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.213290735562698,
+      "learning_rate": 9.762583269679304e-06,
+      "loss": 0.7493,
+      "step": 325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.855612829557892,
+      "learning_rate": 9.760672491226515e-06,
+      "loss": 0.3244,
+      "step": 326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.859494898111436,
+      "learning_rate": 9.758754242905688e-06,
+      "loss": 0.7364,
+      "step": 327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.049143075027642,
+      "learning_rate": 9.756828527726706e-06,
+      "loss": 0.6593,
+      "step": 328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.501315360876768,
+      "learning_rate": 9.754895348711167e-06,
+      "loss": 0.6186,
+      "step": 329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.386001069029195,
+      "learning_rate": 9.752954708892379e-06,
+      "loss": 0.7985,
+      "step": 330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.943682142315781,
+      "learning_rate": 9.751006611315357e-06,
+      "loss": 0.8423,
+      "step": 331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 22.073978735800452,
+      "learning_rate": 9.749051059036821e-06,
+      "loss": 0.4766,
+      "step": 332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.383001161296834,
+      "learning_rate": 9.747088055125186e-06,
+      "loss": 0.6212,
+      "step": 333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.559979159684808,
+      "learning_rate": 9.745117602660556e-06,
+      "loss": 0.4283,
+      "step": 334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.043539178696513,
+      "learning_rate": 9.743139704734729e-06,
+      "loss": 0.6991,
+      "step": 335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.853982916157694,
+      "learning_rate": 9.741154364451179e-06,
+      "loss": 0.7775,
+      "step": 336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.36841904507794,
+      "learning_rate": 9.739161584925061e-06,
+      "loss": 0.3944,
+      "step": 337
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.797556965423516,
+      "learning_rate": 9.737161369283201e-06,
+      "loss": 0.6094,
+      "step": 338
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.590158913408084,
+      "learning_rate": 9.735153720664096e-06,
+      "loss": 0.6999,
+      "step": 339
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.608115858367475,
+      "learning_rate": 9.733138642217905e-06,
+      "loss": 0.9479,
+      "step": 340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.638662697374196,
+      "learning_rate": 9.731116137106441e-06,
+      "loss": 0.8627,
+      "step": 341
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.81168040552755,
+      "learning_rate": 9.729086208503174e-06,
+      "loss": 0.4339,
+      "step": 342
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.71572159293172,
+      "learning_rate": 9.727048859593223e-06,
+      "loss": 0.622,
+      "step": 343
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.487851563876756,
+      "learning_rate": 9.725004093573343e-06,
+      "loss": 1.0463,
+      "step": 344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.034321878936248,
+      "learning_rate": 9.722951913651937e-06,
+      "loss": 0.5409,
+      "step": 345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.398187882091447,
+      "learning_rate": 9.720892323049034e-06,
+      "loss": 0.5344,
+      "step": 346
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.410384932838532,
+      "learning_rate": 9.718825324996294e-06,
+      "loss": 0.7071,
+      "step": 347
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.801512622825909,
+      "learning_rate": 9.716750922736998e-06,
+      "loss": 0.6721,
+      "step": 348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.101521323511989,
+      "learning_rate": 9.714669119526043e-06,
+      "loss": 0.4287,
+      "step": 349
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.898914853678722,
+      "learning_rate": 9.712579918629947e-06,
+      "loss": 0.6616,
+      "step": 350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 16.094505991405192,
+      "learning_rate": 9.710483323326826e-06,
+      "loss": 0.6844,
+      "step": 351
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.570602742162924,
+      "learning_rate": 9.708379336906404e-06,
+      "loss": 0.4917,
+      "step": 352
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.49990923034106,
+      "learning_rate": 9.706267962669999e-06,
+      "loss": 0.3858,
+      "step": 353
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.17415916295187,
+      "learning_rate": 9.704149203930522e-06,
+      "loss": 0.5101,
+      "step": 354
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.070323616243257,
+      "learning_rate": 9.702023064012473e-06,
+      "loss": 0.8231,
+      "step": 355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.442825796338978,
+      "learning_rate": 9.699889546251934e-06,
+      "loss": 0.7589,
+      "step": 356
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.434447797677613,
+      "learning_rate": 9.697748653996558e-06,
+      "loss": 0.5636,
+      "step": 357
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.4998605008175,
+      "learning_rate": 9.695600390605573e-06,
+      "loss": 0.5129,
+      "step": 358
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.484249991544537,
+      "learning_rate": 9.693444759449772e-06,
+      "loss": 0.8224,
+      "step": 359
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.236634532454186,
+      "learning_rate": 9.691281763911513e-06,
+      "loss": 0.3444,
+      "step": 360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.203837127604363,
+      "learning_rate": 9.689111407384703e-06,
+      "loss": 0.8957,
+      "step": 361
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.079565652033697,
+      "learning_rate": 9.686933693274801e-06,
+      "loss": 0.5862,
+      "step": 362
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.943981832702935,
+      "learning_rate": 9.68474862499881e-06,
+      "loss": 0.9032,
+      "step": 363
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.657116793762643,
+      "learning_rate": 9.682556205985274e-06,
+      "loss": 0.788,
+      "step": 364
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.115887955496087,
+      "learning_rate": 9.680356439674272e-06,
+      "loss": 0.7415,
+      "step": 365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.776991279641251,
+      "learning_rate": 9.67814932951741e-06,
+      "loss": 0.3209,
+      "step": 366
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.631349905061063,
+      "learning_rate": 9.675934878977814e-06,
+      "loss": 0.5519,
+      "step": 367
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.77166694676649,
+      "learning_rate": 9.67371309153013e-06,
+      "loss": 1.0513,
+      "step": 368
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.607362668588861,
+      "learning_rate": 9.671483970660519e-06,
+      "loss": 0.6216,
+      "step": 369
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.986919671184642,
+      "learning_rate": 9.669247519866645e-06,
+      "loss": 0.7881,
+      "step": 370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.441810654478047,
+      "learning_rate": 9.667003742657674e-06,
+      "loss": 0.7932,
+      "step": 371
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.832073228219938,
+      "learning_rate": 9.664752642554272e-06,
+      "loss": 0.4483,
+      "step": 372
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.877378843428623,
+      "learning_rate": 9.662494223088586e-06,
+      "loss": 0.5046,
+      "step": 373
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.574846710340779,
+      "learning_rate": 9.660228487804254e-06,
+      "loss": 0.2629,
+      "step": 374
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.031790372297301,
+      "learning_rate": 9.657955440256396e-06,
+      "loss": 0.5388,
+      "step": 375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.509246362025129,
+      "learning_rate": 9.655675084011597e-06,
+      "loss": 0.8608,
+      "step": 376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.514147700901432,
+      "learning_rate": 9.653387422647918e-06,
+      "loss": 0.4201,
+      "step": 377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.928736296468973,
+      "learning_rate": 9.651092459754879e-06,
+      "loss": 0.7958,
+      "step": 378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.406831543298647,
+      "learning_rate": 9.648790198933452e-06,
+      "loss": 0.3303,
+      "step": 379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.176847860722598,
+      "learning_rate": 9.64648064379607e-06,
+      "loss": 0.5674,
+      "step": 380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.594114749167495,
+      "learning_rate": 9.644163797966604e-06,
+      "loss": 0.7672,
+      "step": 381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.508920187739042,
+      "learning_rate": 9.641839665080363e-06,
+      "loss": 0.4247,
+      "step": 382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.312865236525589,
+      "learning_rate": 9.6395082487841e-06,
+      "loss": 0.3282,
+      "step": 383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.420897576003357,
+      "learning_rate": 9.637169552735985e-06,
+      "loss": 0.7396,
+      "step": 384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.918836408288577,
+      "learning_rate": 9.634823580605616e-06,
+      "loss": 0.3609,
+      "step": 385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.501390640773282,
+      "learning_rate": 9.632470336074009e-06,
+      "loss": 0.6796,
+      "step": 386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.778029652859374,
+      "learning_rate": 9.63010982283359e-06,
+      "loss": 0.5471,
+      "step": 387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.904114974999754,
+      "learning_rate": 9.627742044588183e-06,
+      "loss": 0.4143,
+      "step": 388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.193083108023753,
+      "learning_rate": 9.625367005053023e-06,
+      "loss": 0.372,
+      "step": 389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.20304839895969,
+      "learning_rate": 9.622984707954732e-06,
+      "loss": 0.2684,
+      "step": 390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.153730677614503,
+      "learning_rate": 9.62059515703132e-06,
+      "loss": 0.4927,
+      "step": 391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.756405811604829,
+      "learning_rate": 9.61819835603218e-06,
+      "loss": 0.53,
+      "step": 392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.644484950473664,
+      "learning_rate": 9.61579430871808e-06,
+      "loss": 0.5715,
+      "step": 393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.845144890159395,
+      "learning_rate": 9.613383018861159e-06,
+      "loss": 0.5069,
+      "step": 394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.453668545128725,
+      "learning_rate": 9.610964490244921e-06,
+      "loss": 0.4631,
+      "step": 395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.855865332958272,
+      "learning_rate": 9.608538726664224e-06,
+      "loss": 0.6215,
+      "step": 396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.034475725087283,
+      "learning_rate": 9.606105731925284e-06,
+      "loss": 0.7627,
+      "step": 397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.395815515587842,
+      "learning_rate": 9.603665509845657e-06,
+      "loss": 0.3597,
+      "step": 398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.960200424167688,
+      "learning_rate": 9.601218064254245e-06,
+      "loss": 0.4959,
+      "step": 399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.59654187388575,
+      "learning_rate": 9.598763398991283e-06,
+      "loss": 0.9339,
+      "step": 400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.625930458057967,
+      "learning_rate": 9.596301517908329e-06,
+      "loss": 0.7191,
+      "step": 401
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.512562111236216,
+      "learning_rate": 9.593832424868271e-06,
+      "loss": 0.7831,
+      "step": 402
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.37473568972198,
+      "learning_rate": 9.591356123745307e-06,
+      "loss": 0.7774,
+      "step": 403
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.020653210117667,
+      "learning_rate": 9.588872618424949e-06,
+      "loss": 0.6464,
+      "step": 404
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.060061292103656,
+      "learning_rate": 9.58638191280401e-06,
+      "loss": 0.3587,
+      "step": 405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.150232973711836,
+      "learning_rate": 9.583884010790605e-06,
+      "loss": 0.7632,
+      "step": 406
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.2848040813398,
+      "learning_rate": 9.581378916304134e-06,
+      "loss": 0.6675,
+      "step": 407
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.190289843811149,
+      "learning_rate": 9.578866633275289e-06,
+      "loss": 0.3964,
+      "step": 408
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.778028011968894,
+      "learning_rate": 9.576347165646038e-06,
+      "loss": 0.7711,
+      "step": 409
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.566273949226623,
+      "learning_rate": 9.573820517369623e-06,
+      "loss": 0.3951,
+      "step": 410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 15.956435241649336,
+      "learning_rate": 9.571286692410553e-06,
+      "loss": 0.3285,
+      "step": 411
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.955462517971391,
+      "learning_rate": 9.568745694744599e-06,
+      "loss": 0.6285,
+      "step": 412
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.310505634010198,
+      "learning_rate": 9.566197528358785e-06,
+      "loss": 0.8301,
+      "step": 413
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.87669860942439,
+      "learning_rate": 9.563642197251382e-06,
+      "loss": 0.871,
+      "step": 414
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.345113391588745,
+      "learning_rate": 9.561079705431907e-06,
+      "loss": 0.503,
+      "step": 415
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.335777437639864,
+      "learning_rate": 9.55851005692111e-06,
+      "loss": 0.5667,
+      "step": 416
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.131097885426687,
+      "learning_rate": 9.555933255750968e-06,
+      "loss": 0.4063,
+      "step": 417
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.004740290120754,
+      "learning_rate": 9.553349305964687e-06,
+      "loss": 0.8546,
+      "step": 418
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.012277781600202,
+      "learning_rate": 9.550758211616684e-06,
+      "loss": 0.7395,
+      "step": 419
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.557615712028966,
+      "learning_rate": 9.548159976772593e-06,
+      "loss": 0.3626,
+      "step": 420
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.777269558781775,
+      "learning_rate": 9.545554605509244e-06,
+      "loss": 0.7731,
+      "step": 421
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.043083310424668,
+      "learning_rate": 9.54294210191467e-06,
+      "loss": 0.3907,
+      "step": 422
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.5211037237441,
+      "learning_rate": 9.540322470088094e-06,
+      "loss": 0.4002,
+      "step": 423
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.203562331802821,
+      "learning_rate": 9.537695714139925e-06,
+      "loss": 0.5848,
+      "step": 424
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.589322406294142,
+      "learning_rate": 9.535061838191746e-06,
+      "loss": 0.5265,
+      "step": 425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.935530972502687,
+      "learning_rate": 9.532420846376316e-06,
+      "loss": 0.2139,
+      "step": 426
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.581798965328279,
+      "learning_rate": 9.529772742837557e-06,
+      "loss": 0.9191,
+      "step": 427
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.69027010639258,
+      "learning_rate": 9.527117531730552e-06,
+      "loss": 0.5804,
+      "step": 428
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.539886877099754,
+      "learning_rate": 9.524455217221537e-06,
+      "loss": 0.9406,
+      "step": 429
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.317674263528616,
+      "learning_rate": 9.521785803487888e-06,
+      "loss": 0.4932,
+      "step": 430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.719454079724583,
+      "learning_rate": 9.519109294718127e-06,
+      "loss": 0.6043,
+      "step": 431
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.474475327206846,
+      "learning_rate": 9.516425695111906e-06,
+      "loss": 0.4548,
+      "step": 432
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.833652739551171,
+      "learning_rate": 9.513735008880001e-06,
+      "loss": 0.5104,
+      "step": 433
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.54003140540115,
+      "learning_rate": 9.51103724024431e-06,
+      "loss": 0.9773,
+      "step": 434
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.85590741993093,
+      "learning_rate": 9.508332393437845e-06,
+      "loss": 0.517,
+      "step": 435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.619091928605533,
+      "learning_rate": 9.505620472704721e-06,
+      "loss": 0.596,
+      "step": 436
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.355282837351004,
+      "learning_rate": 9.502901482300155e-06,
+      "loss": 0.9104,
+      "step": 437
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.459137780353942,
+      "learning_rate": 9.500175426490455e-06,
+      "loss": 0.6515,
+      "step": 438
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.619150633766477,
+      "learning_rate": 9.497442309553017e-06,
+      "loss": 0.6243,
+      "step": 439
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.98730102335432,
+      "learning_rate": 9.494702135776315e-06,
+      "loss": 0.7172,
+      "step": 440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.530984794831632,
+      "learning_rate": 9.491954909459895e-06,
+      "loss": 0.3573,
+      "step": 441
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.034027726323064,
+      "learning_rate": 9.489200634914373e-06,
+      "loss": 0.7398,
+      "step": 442
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.564991697105127,
+      "learning_rate": 9.48643931646142e-06,
+      "loss": 0.6976,
+      "step": 443
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.639022395053267,
+      "learning_rate": 9.48367095843376e-06,
+      "loss": 0.4131,
+      "step": 444
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.87102094667738,
+      "learning_rate": 9.480895565175166e-06,
+      "loss": 0.505,
+      "step": 445
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.53426767788405,
+      "learning_rate": 9.478113141040444e-06,
+      "loss": 0.4349,
+      "step": 446
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.436641544590042,
+      "learning_rate": 9.475323690395439e-06,
+      "loss": 0.3993,
+      "step": 447
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.061807693482951,
+      "learning_rate": 9.472527217617016e-06,
+      "loss": 0.6407,
+      "step": 448
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.500711111613098,
+      "learning_rate": 9.469723727093061e-06,
+      "loss": 0.4371,
+      "step": 449
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.88608132428978,
+      "learning_rate": 9.466913223222467e-06,
+      "loss": 0.7292,
+      "step": 450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.642349204612678,
+      "learning_rate": 9.464095710415138e-06,
+      "loss": 0.621,
+      "step": 451
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.754985251557086,
+      "learning_rate": 9.461271193091971e-06,
+      "loss": 0.4804,
+      "step": 452
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.3117914343663,
+      "learning_rate": 9.458439675684854e-06,
+      "loss": 0.6733,
+      "step": 453
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.004304307659222,
+      "learning_rate": 9.455601162636662e-06,
+      "loss": 0.5732,
+      "step": 454
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.76222066862556,
+      "learning_rate": 9.452755658401243e-06,
+      "loss": 0.6024,
+      "step": 455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.383026360942381,
+      "learning_rate": 9.449903167443415e-06,
+      "loss": 0.9344,
+      "step": 456
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.245161999321558,
+      "learning_rate": 9.447043694238961e-06,
+      "loss": 0.4995,
+      "step": 457
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.24255373817042,
+      "learning_rate": 9.444177243274619e-06,
+      "loss": 0.5681,
+      "step": 458
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.980835481694355,
+      "learning_rate": 9.441303819048073e-06,
+      "loss": 0.3731,
+      "step": 459
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.153609852326255,
+      "learning_rate": 9.438423426067953e-06,
+      "loss": 0.6279,
+      "step": 460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.422728169566005,
+      "learning_rate": 9.435536068853819e-06,
+      "loss": 0.6803,
+      "step": 461
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.02872691667991,
+      "learning_rate": 9.432641751936162e-06,
+      "loss": 0.5868,
+      "step": 462
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.65944577308325,
+      "learning_rate": 9.42974047985639e-06,
+      "loss": 0.6751,
+      "step": 463
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.65521827829809,
+      "learning_rate": 9.426832257166831e-06,
+      "loss": 0.2612,
+      "step": 464
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.997597309806288,
+      "learning_rate": 9.42391708843071e-06,
+      "loss": 0.8091,
+      "step": 465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.930839808432063,
+      "learning_rate": 9.420994978222156e-06,
+      "loss": 0.4072,
+      "step": 466
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.272039923356747,
+      "learning_rate": 9.418065931126188e-06,
+      "loss": 0.8452,
+      "step": 467
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.89395745423679,
+      "learning_rate": 9.415129951738713e-06,
+      "loss": 0.8293,
+      "step": 468
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.604517862593061,
+      "learning_rate": 9.41218704466651e-06,
+      "loss": 0.666,
+      "step": 469
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.266987837745313,
+      "learning_rate": 9.40923721452723e-06,
+      "loss": 0.7083,
+      "step": 470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.444328200291517,
+      "learning_rate": 9.406280465949394e-06,
+      "loss": 0.391,
+      "step": 471
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.155156735056769,
+      "learning_rate": 9.403316803572363e-06,
+      "loss": 0.6709,
+      "step": 472
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.834291989080315,
+      "learning_rate": 9.400346232046361e-06,
+      "loss": 0.9306,
+      "step": 473
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.460436054078947,
+      "learning_rate": 9.397368756032445e-06,
+      "loss": 0.5094,
+      "step": 474
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.273531664967878,
+      "learning_rate": 9.394384380202511e-06,
+      "loss": 0.6264,
+      "step": 475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.506217820654758,
+      "learning_rate": 9.391393109239277e-06,
+      "loss": 0.4063,
+      "step": 476
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.464097735609588,
+      "learning_rate": 9.388394947836278e-06,
+      "loss": 0.5183,
+      "step": 477
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.706983315798956,
+      "learning_rate": 9.38538990069787e-06,
+      "loss": 0.3586,
+      "step": 478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.43256204057705,
+      "learning_rate": 9.382377972539202e-06,
+      "loss": 0.3957,
+      "step": 479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.9303093019676,
+      "learning_rate": 9.379359168086231e-06,
+      "loss": 0.8572,
+      "step": 480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.535753096591879,
+      "learning_rate": 9.376333492075692e-06,
+      "loss": 0.7307,
+      "step": 481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.03060889655151,
+      "learning_rate": 9.373300949255112e-06,
+      "loss": 0.8868,
+      "step": 482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.968973989980158,
+      "learning_rate": 9.370261544382784e-06,
+      "loss": 0.5413,
+      "step": 483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.182847882387073,
+      "learning_rate": 9.367215282227775e-06,
+      "loss": 0.3212,
+      "step": 484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.200530512165116,
+      "learning_rate": 9.364162167569907e-06,
+      "loss": 0.5619,
+      "step": 485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.927493078975143,
+      "learning_rate": 9.361102205199762e-06,
+      "loss": 0.5113,
+      "step": 486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.546488613744044,
+      "learning_rate": 9.358035399918652e-06,
+      "loss": 0.4361,
+      "step": 487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.724332076672832,
+      "learning_rate": 9.35496175653864e-06,
+      "loss": 0.7661,
+      "step": 488
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.349579912784495,
+      "learning_rate": 9.351881279882512e-06,
+      "loss": 0.4996,
+      "step": 489
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.98771230302597,
+      "learning_rate": 9.348793974783778e-06,
+      "loss": 0.647,
+      "step": 490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.465561912625265,
+      "learning_rate": 9.34569984608666e-06,
+      "loss": 0.6449,
+      "step": 491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.506298884530864,
+      "learning_rate": 9.34259889864609e-06,
+      "loss": 0.3472,
+      "step": 492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.29272193699987,
+      "learning_rate": 9.339491137327696e-06,
+      "loss": 0.4154,
+      "step": 493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.89338242662377,
+      "learning_rate": 9.336376567007799e-06,
+      "loss": 0.3833,
+      "step": 494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.601782077551498,
+      "learning_rate": 9.333255192573404e-06,
+      "loss": 0.4909,
+      "step": 495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.095297384985853,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.8617,
+      "step": 496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.057930504872592,
+      "learning_rate": 9.326992050962515e-06,
+      "loss": 0.4936,
+      "step": 497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.01103050895663,
+      "learning_rate": 9.32385029361338e-06,
+      "loss": 0.5267,
+      "step": 498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.031931642135508,
+      "learning_rate": 9.320701751804451e-06,
+      "loss": 0.2436,
+      "step": 499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.165708225556706,
+      "learning_rate": 9.317546430476036e-06,
+      "loss": 0.3686,
+      "step": 500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.283718687920384,
+      "learning_rate": 9.314384334579085e-06,
+      "loss": 0.8664,
+      "step": 501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.456167638386002,
+      "learning_rate": 9.311215469075168e-06,
+      "loss": 0.6548,
+      "step": 502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.977122974793179,
+      "learning_rate": 9.30803983893649e-06,
+      "loss": 0.6606,
+      "step": 503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.3134222375462,
+      "learning_rate": 9.304857449145858e-06,
+      "loss": 0.3312,
+      "step": 504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.74755197828282,
+      "learning_rate": 9.301668304696694e-06,
+      "loss": 0.6495,
+      "step": 505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.898259198089015,
+      "learning_rate": 9.298472410593013e-06,
+      "loss": 0.4129,
+      "step": 506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.020073352141033,
+      "learning_rate": 9.295269771849426e-06,
+      "loss": 0.5805,
+      "step": 507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.924108494332405,
+      "learning_rate": 9.292060393491122e-06,
+      "loss": 0.5003,
+      "step": 508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.377999680081595,
+      "learning_rate": 9.288844280553869e-06,
+      "loss": 0.2397,
+      "step": 509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.740780334135453,
+      "learning_rate": 9.285621438083997e-06,
+      "loss": 0.7581,
+      "step": 510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.247228930505976,
+      "learning_rate": 9.282391871138404e-06,
+      "loss": 0.6239,
+      "step": 511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.387435640862735,
+      "learning_rate": 9.279155584784524e-06,
+      "loss": 0.4702,
+      "step": 512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.05787993239625,
+      "learning_rate": 9.275912584100353e-06,
+      "loss": 0.727,
+      "step": 513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.553698611357072,
+      "learning_rate": 9.27266287417441e-06,
+      "loss": 0.4184,
+      "step": 514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.286834178183035,
+      "learning_rate": 9.269406460105742e-06,
+      "loss": 0.6174,
+      "step": 515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.764903360638112,
+      "learning_rate": 9.26614334700392e-06,
+      "loss": 0.8556,
+      "step": 516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.371262164703925,
+      "learning_rate": 9.262873539989023e-06,
+      "loss": 0.6173,
+      "step": 517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.417609913145478,
+      "learning_rate": 9.259597044191635e-06,
+      "loss": 0.294,
+      "step": 518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.2302917295438,
+      "learning_rate": 9.256313864752838e-06,
+      "loss": 0.4982,
+      "step": 519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.585659153214129,
+      "learning_rate": 9.25302400682419e-06,
+      "loss": 0.5701,
+      "step": 520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.614600667347043,
+      "learning_rate": 9.249727475567742e-06,
+      "loss": 0.2716,
+      "step": 521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.130331504868614,
+      "learning_rate": 9.246424276156008e-06,
+      "loss": 0.4188,
+      "step": 522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.750855368989637,
+      "learning_rate": 9.243114413771966e-06,
+      "loss": 0.6,
+      "step": 523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.028865042851002,
+      "learning_rate": 9.23979789360905e-06,
+      "loss": 0.5556,
+      "step": 524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.84417177396771,
+      "learning_rate": 9.23647472087114e-06,
+      "loss": 0.7345,
+      "step": 525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.936328478415838,
+      "learning_rate": 9.233144900772553e-06,
+      "loss": 0.5943,
+      "step": 526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.815830590413693,
+      "learning_rate": 9.22980843853804e-06,
+      "loss": 0.5175,
+      "step": 527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.056942621828679,
+      "learning_rate": 9.226465339402768e-06,
+      "loss": 0.7441,
+      "step": 528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.77806438559145,
+      "learning_rate": 9.223115608612325e-06,
+      "loss": 0.2269,
+      "step": 529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.319379133468905,
+      "learning_rate": 9.2197592514227e-06,
+      "loss": 0.5984,
+      "step": 530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.739926033224762,
+      "learning_rate": 9.21639627310028e-06,
+      "loss": 0.7032,
+      "step": 531
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.758104323600268,
+      "learning_rate": 9.21302667892184e-06,
+      "loss": 0.2339,
+      "step": 532
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.690386042495781,
+      "learning_rate": 9.209650474174539e-06,
+      "loss": 0.4934,
+      "step": 533
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.921167492917776,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 0.4383,
+      "step": 534
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.92173716451255,
+      "learning_rate": 9.202878254173836e-06,
+      "loss": 0.7122,
+      "step": 535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.74101064650462,
+      "learning_rate": 9.199482249546577e-06,
+      "loss": 0.3149,
+      "step": 536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.07530459316679,
+      "learning_rate": 9.196079655602728e-06,
+      "loss": 0.2923,
+      "step": 537
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.85474922380344,
+      "learning_rate": 9.192670477681224e-06,
+      "loss": 0.5423,
+      "step": 538
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.225697314403845,
+      "learning_rate": 9.189254721131333e-06,
+      "loss": 0.799,
+      "step": 539
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.357241546802442,
+      "learning_rate": 9.185832391312644e-06,
+      "loss": 0.3545,
+      "step": 540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.803557845591799,
+      "learning_rate": 9.18240349359506e-06,
+      "loss": 0.3972,
+      "step": 541
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.329680404423371,
+      "learning_rate": 9.178968033358792e-06,
+      "loss": 0.6466,
+      "step": 542
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.081851463681208,
+      "learning_rate": 9.175526015994345e-06,
+      "loss": 0.4528,
+      "step": 543
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.720809663311215,
+      "learning_rate": 9.172077446902515e-06,
+      "loss": 0.4866,
+      "step": 544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.76058348099088,
+      "learning_rate": 9.168622331494375e-06,
+      "loss": 0.8663,
+      "step": 545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.314824794859003,
+      "learning_rate": 9.165160675191272e-06,
+      "loss": 0.4532,
+      "step": 546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.85687338357031,
+      "learning_rate": 9.161692483424817e-06,
+      "loss": 0.5076,
+      "step": 547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.379308204143138,
+      "learning_rate": 9.158217761636876e-06,
+      "loss": 0.3777,
+      "step": 548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.094030544849165,
+      "learning_rate": 9.154736515279557e-06,
+      "loss": 0.5039,
+      "step": 549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.4892633191494,
+      "learning_rate": 9.151248749815208e-06,
+      "loss": 0.3817,
+      "step": 550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.859687519689212,
+      "learning_rate": 9.147754470716407e-06,
+      "loss": 0.5304,
+      "step": 551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.54630614297978,
+      "learning_rate": 9.144253683465953e-06,
+      "loss": 0.5502,
+      "step": 552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.39887951492467,
+      "learning_rate": 9.140746393556853e-06,
+      "loss": 0.4445,
+      "step": 553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.680083600328116,
+      "learning_rate": 9.137232606492323e-06,
+      "loss": 0.2557,
+      "step": 554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.813174707068232,
+      "learning_rate": 9.133712327785769e-06,
+      "loss": 0.345,
+      "step": 555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.33312067267694,
+      "learning_rate": 9.13018556296078e-06,
+      "loss": 0.5196,
+      "step": 556
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.493751546793533,
+      "learning_rate": 9.12665231755113e-06,
+      "loss": 0.5146,
+      "step": 557
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.587826151614221,
+      "learning_rate": 9.123112597100759e-06,
+      "loss": 0.5736,
+      "step": 558
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.4469922725591,
+      "learning_rate": 9.11956640716376e-06,
+      "loss": 0.7266,
+      "step": 559
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.690741887797087,
+      "learning_rate": 9.11601375330439e-06,
+      "loss": 0.4957,
+      "step": 560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.808366580770018,
+      "learning_rate": 9.112454641097037e-06,
+      "loss": 0.8112,
+      "step": 561
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.97042354353918,
+      "learning_rate": 9.108889076126226e-06,
+      "loss": 0.3945,
+      "step": 562
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 32.56759368512608,
+      "learning_rate": 9.105317063986613e-06,
+      "loss": 0.6129,
+      "step": 563
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.431803574555289,
+      "learning_rate": 9.101738610282956e-06,
+      "loss": 0.4552,
+      "step": 564
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.65581111625762,
+      "learning_rate": 9.098153720630138e-06,
+      "loss": 0.4391,
+      "step": 565
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.758459000128104,
+      "learning_rate": 9.094562400653127e-06,
+      "loss": 0.4518,
+      "step": 566
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.029791629666333,
+      "learning_rate": 9.090964655986985e-06,
+      "loss": 0.4558,
+      "step": 567
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.711313263192542,
+      "learning_rate": 9.087360492276858e-06,
+      "loss": 0.5378,
+      "step": 568
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.216158737350378,
+      "learning_rate": 9.083749915177959e-06,
+      "loss": 0.6341,
+      "step": 569
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.62739850028013,
+      "learning_rate": 9.080132930355567e-06,
+      "loss": 0.8432,
+      "step": 570
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.129776180006205,
+      "learning_rate": 9.076509543485014e-06,
+      "loss": 0.5227,
+      "step": 571
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.565074736316749,
+      "learning_rate": 9.07287976025168e-06,
+      "loss": 0.4422,
+      "step": 572
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.506858089551722,
+      "learning_rate": 9.069243586350976e-06,
+      "loss": 0.329,
+      "step": 573
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.641279189201857,
+      "learning_rate": 9.065601027488345e-06,
+      "loss": 0.5581,
+      "step": 574
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.08954162165749,
+      "learning_rate": 9.061952089379248e-06,
+      "loss": 0.3581,
+      "step": 575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.252643106435224,
+      "learning_rate": 9.058296777749154e-06,
+      "loss": 0.2821,
+      "step": 576
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.639897247288893,
+      "learning_rate": 9.054635098333532e-06,
+      "loss": 0.4321,
+      "step": 577
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.125845534492296,
+      "learning_rate": 9.050967056877846e-06,
+      "loss": 0.5631,
+      "step": 578
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.119582943283758,
+      "learning_rate": 9.047292659137542e-06,
+      "loss": 0.4503,
+      "step": 579
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.890007230135126,
+      "learning_rate": 9.043611910878031e-06,
+      "loss": 0.3434,
+      "step": 580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.517972878960965,
+      "learning_rate": 9.0399248178747e-06,
+      "loss": 0.4457,
+      "step": 581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.910940209913173,
+      "learning_rate": 9.03623138591289e-06,
+      "loss": 0.2195,
+      "step": 582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.361398086412844,
+      "learning_rate": 9.032531620787879e-06,
+      "loss": 0.5493,
+      "step": 583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.426198443446308,
+      "learning_rate": 9.028825528304892e-06,
+      "loss": 0.2497,
+      "step": 584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.242386991455056,
+      "learning_rate": 9.025113114279076e-06,
+      "loss": 0.5753,
+      "step": 585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.4516023563780465,
+      "learning_rate": 9.0213943845355e-06,
+      "loss": 0.3517,
+      "step": 586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.800589529548143,
+      "learning_rate": 9.017669344909143e-06,
+      "loss": 0.4984,
+      "step": 587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.143048580178721,
+      "learning_rate": 9.013938001244885e-06,
+      "loss": 0.2151,
+      "step": 588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.081074259824566,
+      "learning_rate": 9.010200359397495e-06,
+      "loss": 0.8493,
+      "step": 589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.145245837687522,
+      "learning_rate": 9.006456425231624e-06,
+      "loss": 0.5276,
+      "step": 590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.026933303639067,
+      "learning_rate": 9.002706204621802e-06,
+      "loss": 0.1736,
+      "step": 591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.563719063755459,
+      "learning_rate": 8.998949703452418e-06,
+      "loss": 0.8722,
+      "step": 592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.360309163278158,
+      "learning_rate": 8.995186927617714e-06,
+      "loss": 0.792,
+      "step": 593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.726353999625424,
+      "learning_rate": 8.99141788302178e-06,
+      "loss": 0.6533,
+      "step": 594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.593473498891232,
+      "learning_rate": 8.987642575578546e-06,
+      "loss": 0.3823,
+      "step": 595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.1748733302172,
+      "learning_rate": 8.98386101121176e-06,
+      "loss": 0.4289,
+      "step": 596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.107103086720384,
+      "learning_rate": 8.980073195854999e-06,
+      "loss": 0.4058,
+      "step": 597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.252593700242617,
+      "learning_rate": 8.976279135451636e-06,
+      "loss": 0.554,
+      "step": 598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.230905980751361,
+      "learning_rate": 8.972478835954851e-06,
+      "loss": 0.5091,
+      "step": 599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.481908291684853,
+      "learning_rate": 8.968672303327614e-06,
+      "loss": 0.5156,
+      "step": 600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.466250265576518,
+      "learning_rate": 8.96485954354267e-06,
+      "loss": 0.462,
+      "step": 601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.4862042570593,
+      "learning_rate": 8.96104056258254e-06,
+      "loss": 0.6422,
+      "step": 602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.835993681541376,
+      "learning_rate": 8.9572153664395e-06,
+      "loss": 0.5118,
+      "step": 603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.091888403648705,
+      "learning_rate": 8.953383961115586e-06,
+      "loss": 0.4578,
+      "step": 604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.775676272236673,
+      "learning_rate": 8.949546352622573e-06,
+      "loss": 0.4302,
+      "step": 605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.358146041246965,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 0.6041,
+      "step": 606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.237645386523313,
+      "learning_rate": 8.941852550225005e-06,
+      "loss": 0.4676,
+      "step": 607
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.306407896237637,
+      "learning_rate": 8.937996368392629e-06,
+      "loss": 0.293,
+      "step": 608
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.86254990672111,
+      "learning_rate": 8.93413400753549e-06,
+      "loss": 0.6665,
+      "step": 609
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.259892389637482,
+      "learning_rate": 8.930265473713939e-06,
+      "loss": 0.5623,
+      "step": 610
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.20221620309941,
+      "learning_rate": 8.926390772998004e-06,
+      "loss": 0.5099,
+      "step": 611
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.201517913642562,
+      "learning_rate": 8.922509911467395e-06,
+      "loss": 0.716,
+      "step": 612
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.517591063140472,
+      "learning_rate": 8.91862289521149e-06,
+      "loss": 0.6078,
+      "step": 613
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.716922135408002,
+      "learning_rate": 8.914729730329321e-06,
+      "loss": 0.7424,
+      "step": 614
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.143238841169707,
+      "learning_rate": 8.910830422929566e-06,
+      "loss": 0.1517,
+      "step": 615
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.672726956304915,
+      "learning_rate": 8.906924979130548e-06,
+      "loss": 0.24,
+      "step": 616
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.334740352519516,
+      "learning_rate": 8.903013405060212e-06,
+      "loss": 0.2895,
+      "step": 617
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.646502694932076,
+      "learning_rate": 8.899095706856122e-06,
+      "loss": 0.7394,
+      "step": 618
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.875638442638373,
+      "learning_rate": 8.895171890665457e-06,
+      "loss": 0.4234,
+      "step": 619
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.05064412208733,
+      "learning_rate": 8.891241962644992e-06,
+      "loss": 0.5487,
+      "step": 620
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.968304781614375,
+      "learning_rate": 8.887305928961088e-06,
+      "loss": 0.4145,
+      "step": 621
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.317206091789457,
+      "learning_rate": 8.883363795789694e-06,
+      "loss": 0.5696,
+      "step": 622
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.879977218875943,
+      "learning_rate": 8.879415569316324e-06,
+      "loss": 0.4713,
+      "step": 623
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.62358283480348,
+      "learning_rate": 8.875461255736055e-06,
+      "loss": 0.7608,
+      "step": 624
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.13770361867563,
+      "learning_rate": 8.871500861253515e-06,
+      "loss": 0.8526,
+      "step": 625
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.495697890547111,
+      "learning_rate": 8.867534392082873e-06,
+      "loss": 0.3285,
+      "step": 626
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.27625891040776,
+      "learning_rate": 8.863561854447829e-06,
+      "loss": 0.5125,
+      "step": 627
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.085900790114003,
+      "learning_rate": 8.859583254581604e-06,
+      "loss": 0.2408,
+      "step": 628
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.799690394679768,
+      "learning_rate": 8.85559859872694e-06,
+      "loss": 0.871,
+      "step": 629
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.067134198168679,
+      "learning_rate": 8.851607893136065e-06,
+      "loss": 0.7519,
+      "step": 630
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.189673287583272,
+      "learning_rate": 8.847611144070716e-06,
+      "loss": 0.6726,
+      "step": 631
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.240406611244559,
+      "learning_rate": 8.8436083578021e-06,
+      "loss": 0.5327,
+      "step": 632
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.62533465108437,
+      "learning_rate": 8.839599540610906e-06,
+      "loss": 0.4613,
+      "step": 633
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.71818559432506,
+      "learning_rate": 8.83558469878728e-06,
+      "loss": 0.3022,
+      "step": 634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.923236493342815,
+      "learning_rate": 8.831563838630825e-06,
+      "loss": 0.5919,
+      "step": 635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.087915684518228,
+      "learning_rate": 8.827536966450584e-06,
+      "loss": 0.6275,
+      "step": 636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.017958781270215,
+      "learning_rate": 8.823504088565035e-06,
+      "loss": 0.2969,
+      "step": 637
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.836112833908901,
+      "learning_rate": 8.819465211302081e-06,
+      "loss": 0.6994,
+      "step": 638
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.797868389201499,
+      "learning_rate": 8.815420340999034e-06,
+      "loss": 0.4284,
+      "step": 639
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.396599076084124,
+      "learning_rate": 8.811369484002614e-06,
+      "loss": 0.5129,
+      "step": 640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.514466827238826,
+      "learning_rate": 8.807312646668932e-06,
+      "loss": 0.5501,
+      "step": 641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.595454633092654,
+      "learning_rate": 8.803249835363486e-06,
+      "loss": 0.2807,
+      "step": 642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.690353494842244,
+      "learning_rate": 8.799181056461143e-06,
+      "loss": 0.4608,
+      "step": 643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.683405211830714,
+      "learning_rate": 8.795106316346136e-06,
+      "loss": 0.4761,
+      "step": 644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.130234354377658,
+      "learning_rate": 8.791025621412052e-06,
+      "loss": 0.3058,
+      "step": 645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.449076663682968,
+      "learning_rate": 8.78693897806182e-06,
+      "loss": 0.5182,
+      "step": 646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.826684813499433,
+      "learning_rate": 8.782846392707704e-06,
+      "loss": 0.4704,
+      "step": 647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.187672535550071,
+      "learning_rate": 8.778747871771293e-06,
+      "loss": 0.4475,
+      "step": 648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.414765025622366,
+      "learning_rate": 8.774643421683484e-06,
+      "loss": 0.4408,
+      "step": 649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.924135705612454,
+      "learning_rate": 8.770533048884483e-06,
+      "loss": 0.4204,
+      "step": 650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.892129709759855,
+      "learning_rate": 8.766416759823785e-06,
+      "loss": 0.4873,
+      "step": 651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.198131815794714,
+      "learning_rate": 8.762294560960173e-06,
+      "loss": 0.1706,
+      "step": 652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.089178286077011,
+      "learning_rate": 8.758166458761695e-06,
+      "loss": 0.5733,
+      "step": 653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.115307458701734,
+      "learning_rate": 8.754032459705672e-06,
+      "loss": 0.4085,
+      "step": 654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.164590981197422,
+      "learning_rate": 8.74989257027867e-06,
+      "loss": 0.3715,
+      "step": 655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.708208157124274,
+      "learning_rate": 8.7457467969765e-06,
+      "loss": 0.2427,
+      "step": 656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.813264749009186,
+      "learning_rate": 8.741595146304202e-06,
+      "loss": 0.5565,
+      "step": 657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.44187172708592,
+      "learning_rate": 8.737437624776047e-06,
+      "loss": 0.2967,
+      "step": 658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.943011818212373,
+      "learning_rate": 8.733274238915508e-06,
+      "loss": 0.677,
+      "step": 659
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.574816640321803,
+      "learning_rate": 8.729104995255265e-06,
+      "loss": 0.2817,
+      "step": 660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.402322285304788,
+      "learning_rate": 8.724929900337186e-06,
+      "loss": 0.3979,
+      "step": 661
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.526541028039155,
+      "learning_rate": 8.720748960712323e-06,
+      "loss": 0.7882,
+      "step": 662
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.831474531761591,
+      "learning_rate": 8.7165621829409e-06,
+      "loss": 0.3234,
+      "step": 663
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.255495736600963,
+      "learning_rate": 8.712369573592296e-06,
+      "loss": 0.4869,
+      "step": 664
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.09892752028399,
+      "learning_rate": 8.708171139245045e-06,
+      "loss": 0.3652,
+      "step": 665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.77002043481429,
+      "learning_rate": 8.703966886486819e-06,
+      "loss": 0.5169,
+      "step": 666
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.005847840054802,
+      "learning_rate": 8.69975682191442e-06,
+      "loss": 0.7055,
+      "step": 667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.489295846320399,
+      "learning_rate": 8.695540952133769e-06,
+      "loss": 0.3733,
+      "step": 668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.21544305149708,
+      "learning_rate": 8.691319283759896e-06,
+      "loss": 0.6678,
+      "step": 669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.535169383036115,
+      "learning_rate": 8.68709182341693e-06,
+      "loss": 0.5525,
+      "step": 670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.980510917527248,
+      "learning_rate": 8.682858577738086e-06,
+      "loss": 0.6075,
+      "step": 671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.468808118527777,
+      "learning_rate": 8.67861955336566e-06,
+      "loss": 0.568,
+      "step": 672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.049978599306831,
+      "learning_rate": 8.67437475695101e-06,
+      "loss": 0.516,
+      "step": 673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.66040721063192,
+      "learning_rate": 8.670124195154557e-06,
+      "loss": 0.3759,
+      "step": 674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.659977682784623,
+      "learning_rate": 8.665867874645767e-06,
+      "loss": 0.481,
+      "step": 675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.28285233149463,
+      "learning_rate": 8.661605802103134e-06,
+      "loss": 0.8561,
+      "step": 676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.509974476618318,
+      "learning_rate": 8.657337984214189e-06,
+      "loss": 0.3294,
+      "step": 677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.995743145473082,
+      "learning_rate": 8.65306442767547e-06,
+      "loss": 0.3342,
+      "step": 678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.944417587367974,
+      "learning_rate": 8.648785139192526e-06,
+      "loss": 0.5559,
+      "step": 679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.908924167149655,
+      "learning_rate": 8.644500125479891e-06,
+      "loss": 0.5753,
+      "step": 680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.29059105817483,
+      "learning_rate": 8.640209393261087e-06,
+      "loss": 0.5358,
+      "step": 681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.288539435926277,
+      "learning_rate": 8.635912949268614e-06,
+      "loss": 0.7426,
+      "step": 682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.86295873251741,
+      "learning_rate": 8.631610800243926e-06,
+      "loss": 0.5026,
+      "step": 683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.665327843523478,
+      "learning_rate": 8.627302952937431e-06,
+      "loss": 0.2277,
+      "step": 684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.170493069795594,
+      "learning_rate": 8.622989414108479e-06,
+      "loss": 0.4337,
+      "step": 685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.73535039397476,
+      "learning_rate": 8.61867019052535e-06,
+      "loss": 0.4019,
+      "step": 686
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.730761756171253,
+      "learning_rate": 8.614345288965249e-06,
+      "loss": 0.7042,
+      "step": 687
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.880536010676554,
+      "learning_rate": 8.61001471621428e-06,
+      "loss": 0.3713,
+      "step": 688
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.77605111269693,
+      "learning_rate": 8.60567847906745e-06,
+      "loss": 0.405,
+      "step": 689
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.629809718991295,
+      "learning_rate": 8.601336584328659e-06,
+      "loss": 0.8516,
+      "step": 690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.23657893856505,
+      "learning_rate": 8.596989038810678e-06,
+      "loss": 0.6202,
+      "step": 691
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.550770023655387,
+      "learning_rate": 8.592635849335148e-06,
+      "loss": 0.46,
+      "step": 692
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.971687364327797,
+      "learning_rate": 8.58827702273256e-06,
+      "loss": 0.4755,
+      "step": 693
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.987613121982326,
+      "learning_rate": 8.583912565842258e-06,
+      "loss": 0.5697,
+      "step": 694
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 15.11549044382133,
+      "learning_rate": 8.579542485512416e-06,
+      "loss": 0.559,
+      "step": 695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.611495056815679,
+      "learning_rate": 8.575166788600031e-06,
+      "loss": 0.5114,
+      "step": 696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.174070976106588,
+      "learning_rate": 8.570785481970915e-06,
+      "loss": 0.4149,
+      "step": 697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.908135841800428,
+      "learning_rate": 8.566398572499685e-06,
+      "loss": 0.5218,
+      "step": 698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.770969060215375,
+      "learning_rate": 8.56200606706974e-06,
+      "loss": 0.5418,
+      "step": 699
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.452896180514323,
+      "learning_rate": 8.557607972573267e-06,
+      "loss": 0.4333,
+      "step": 700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.79965423014118,
+      "learning_rate": 8.553204295911222e-06,
+      "loss": 0.4277,
+      "step": 701
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.335853551489523,
+      "learning_rate": 8.548795043993316e-06,
+      "loss": 0.4034,
+      "step": 702
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.852331858408727,
+      "learning_rate": 8.544380223738015e-06,
+      "loss": 0.4692,
+      "step": 703
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.37887611622151,
+      "learning_rate": 8.539959842072513e-06,
+      "loss": 0.438,
+      "step": 704
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.665145785643587,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.602,
+      "step": 705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.41766594175798,
+      "learning_rate": 8.53110242226333e-06,
+      "loss": 0.3049,
+      "step": 706
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.794514806118022,
+      "learning_rate": 8.526665398017633e-06,
+      "loss": 0.459,
+      "step": 707
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.154157263104445,
+      "learning_rate": 8.522222840157687e-06,
+      "loss": 0.5222,
+      "step": 708
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.89444104464997,
+      "learning_rate": 8.517774755654212e-06,
+      "loss": 0.6427,
+      "step": 709
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.54454637120731,
+      "learning_rate": 8.513321151486602e-06,
+      "loss": 0.4045,
+      "step": 710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.353232286868508,
+      "learning_rate": 8.508862034642909e-06,
+      "loss": 0.4999,
+      "step": 711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.56862839975549,
+      "learning_rate": 8.504397412119838e-06,
+      "loss": 0.5392,
+      "step": 712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.48768823565412,
+      "learning_rate": 8.499927290922734e-06,
+      "loss": 0.414,
+      "step": 713
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.667436419488624,
+      "learning_rate": 8.495451678065563e-06,
+      "loss": 0.4344,
+      "step": 714
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.239022049301735,
+      "learning_rate": 8.490970580570912e-06,
+      "loss": 0.4169,
+      "step": 715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.48052955649948,
+      "learning_rate": 8.486484005469977e-06,
+      "loss": 0.4273,
+      "step": 716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.236036167402592,
+      "learning_rate": 8.481991959802546e-06,
+      "loss": 0.5575,
+      "step": 717
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.364048282155869,
+      "learning_rate": 8.477494450616988e-06,
+      "loss": 0.3966,
+      "step": 718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.568552132253508,
+      "learning_rate": 8.472991484970247e-06,
+      "loss": 0.4265,
+      "step": 719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.404111198693425,
+      "learning_rate": 8.468483069927832e-06,
+      "loss": 0.6073,
+      "step": 720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.125608192068935,
+      "learning_rate": 8.463969212563796e-06,
+      "loss": 0.5328,
+      "step": 721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.605097479501142,
+      "learning_rate": 8.459449919960737e-06,
+      "loss": 0.5642,
+      "step": 722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.730105306634389,
+      "learning_rate": 8.454925199209778e-06,
+      "loss": 0.5612,
+      "step": 723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.624841628566994,
+      "learning_rate": 8.450395057410561e-06,
+      "loss": 0.3602,
+      "step": 724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.521688599159493,
+      "learning_rate": 8.445859501671232e-06,
+      "loss": 0.4536,
+      "step": 725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.392516675905032,
+      "learning_rate": 8.441318539108433e-06,
+      "loss": 0.2624,
+      "step": 726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.514292983934698,
+      "learning_rate": 8.436772176847295e-06,
+      "loss": 0.5886,
+      "step": 727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.534247744419414,
+      "learning_rate": 8.432220422021408e-06,
+      "loss": 0.5777,
+      "step": 728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.206499893653131,
+      "learning_rate": 8.42766328177284e-06,
+      "loss": 0.271,
+      "step": 729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.610775066862617,
+      "learning_rate": 8.423100763252094e-06,
+      "loss": 0.365,
+      "step": 730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.446443559906214,
+      "learning_rate": 8.418532873618125e-06,
+      "loss": 0.4242,
+      "step": 731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.768759355697657,
+      "learning_rate": 8.413959620038306e-06,
+      "loss": 0.6524,
+      "step": 732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.747531975573759,
+      "learning_rate": 8.409381009688431e-06,
+      "loss": 0.4945,
+      "step": 733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.028021949265852,
+      "learning_rate": 8.404797049752697e-06,
+      "loss": 0.6248,
+      "step": 734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.418225127450622,
+      "learning_rate": 8.4002077474237e-06,
+      "loss": 0.4466,
+      "step": 735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.084753246613275,
+      "learning_rate": 8.39561310990241e-06,
+      "loss": 0.6234,
+      "step": 736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.26221417403431,
+      "learning_rate": 8.39101314439818e-06,
+      "loss": 0.9455,
+      "step": 737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.259665158813199,
+      "learning_rate": 8.386407858128707e-06,
+      "loss": 0.4322,
+      "step": 738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.22562360059738,
+      "learning_rate": 8.38179725832005e-06,
+      "loss": 0.4796,
+      "step": 739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.239467412147198,
+      "learning_rate": 8.377181352206604e-06,
+      "loss": 0.5253,
+      "step": 740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.434537552403878,
+      "learning_rate": 8.372560147031087e-06,
+      "loss": 0.5122,
+      "step": 741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.304604279037665,
+      "learning_rate": 8.367933650044526e-06,
+      "loss": 0.3589,
+      "step": 742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.196080897749898,
+      "learning_rate": 8.363301868506264e-06,
+      "loss": 0.4136,
+      "step": 743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.080080916994074,
+      "learning_rate": 8.358664809683926e-06,
+      "loss": 0.6745,
+      "step": 744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.178355170237014,
+      "learning_rate": 8.354022480853418e-06,
+      "loss": 0.574,
+      "step": 745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.672330884844124,
+      "learning_rate": 8.349374889298923e-06,
+      "loss": 0.2227,
+      "step": 746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.673317568226688,
+      "learning_rate": 8.344722042312872e-06,
+      "loss": 0.5035,
+      "step": 747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.601452804011776,
+      "learning_rate": 8.340063947195947e-06,
+      "loss": 0.3122,
+      "step": 748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.612577534646004,
+      "learning_rate": 8.335400611257067e-06,
+      "loss": 0.5286,
+      "step": 749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.514815776925913,
+      "learning_rate": 8.330732041813367e-06,
+      "loss": 0.6596,
+      "step": 750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.710195849532534,
+      "learning_rate": 8.326058246190202e-06,
+      "loss": 0.4108,
+      "step": 751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.285350523407853,
+      "learning_rate": 8.321379231721123e-06,
+      "loss": 0.4211,
+      "step": 752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.869370349357416,
+      "learning_rate": 8.316695005747866e-06,
+      "loss": 0.6561,
+      "step": 753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.15583382112573,
+      "learning_rate": 8.312005575620355e-06,
+      "loss": 0.4642,
+      "step": 754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.455451610421562,
+      "learning_rate": 8.307310948696667e-06,
+      "loss": 0.4986,
+      "step": 755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.57470776263406,
+      "learning_rate": 8.302611132343042e-06,
+      "loss": 0.432,
+      "step": 756
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.07397844116237,
+      "learning_rate": 8.297906133933861e-06,
+      "loss": 0.4036,
+      "step": 757
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.910230095875521,
+      "learning_rate": 8.293195960851634e-06,
+      "loss": 0.6326,
+      "step": 758
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.256678638075714,
+      "learning_rate": 8.288480620486991e-06,
+      "loss": 0.7539,
+      "step": 759
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.9398865780561785,
+      "learning_rate": 8.283760120238672e-06,
+      "loss": 0.1337,
+      "step": 760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.276865530470275,
+      "learning_rate": 8.27903446751351e-06,
+      "loss": 0.3263,
+      "step": 761
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.882800910953,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 0.4415,
+      "step": 762
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.801104393382156,
+      "learning_rate": 8.26956773430041e-06,
+      "loss": 0.4179,
+      "step": 763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.664986313222611,
+      "learning_rate": 8.264826668666516e-06,
+      "loss": 0.472,
+      "step": 764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.520502923088777,
+      "learning_rate": 8.26008048026385e-06,
+      "loss": 0.2778,
+      "step": 765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.951120225641205,
+      "learning_rate": 8.255329176539552e-06,
+      "loss": 0.4354,
+      "step": 766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.322023633370632,
+      "learning_rate": 8.250572764948787e-06,
+      "loss": 0.6368,
+      "step": 767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.405204519865634,
+      "learning_rate": 8.245811252954741e-06,
+      "loss": 0.5628,
+      "step": 768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.211819804228533,
+      "learning_rate": 8.241044648028597e-06,
+      "loss": 0.6515,
+      "step": 769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.633085414240254,
+      "learning_rate": 8.236272957649534e-06,
+      "loss": 0.2599,
+      "step": 770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.803403857724263,
+      "learning_rate": 8.231496189304704e-06,
+      "loss": 0.2248,
+      "step": 771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.317303767754435,
+      "learning_rate": 8.226714350489235e-06,
+      "loss": 0.555,
+      "step": 772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.009664465426518,
+      "learning_rate": 8.221927448706204e-06,
+      "loss": 0.6578,
+      "step": 773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.199045294032182,
+      "learning_rate": 8.217135491466636e-06,
+      "loss": 0.5188,
+      "step": 774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.032952928422251,
+      "learning_rate": 8.212338486289486e-06,
+      "loss": 0.5249,
+      "step": 775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.320273834473651,
+      "learning_rate": 8.207536440701633e-06,
+      "loss": 0.5385,
+      "step": 776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.300560907402463,
+      "learning_rate": 8.20272936223786e-06,
+      "loss": 0.5994,
+      "step": 777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.394654960675332,
+      "learning_rate": 8.197917258440851e-06,
+      "loss": 0.4833,
+      "step": 778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.726270676697984,
+      "learning_rate": 8.193100136861174e-06,
+      "loss": 0.2452,
+      "step": 779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.005883878827241,
+      "learning_rate": 8.18827800505727e-06,
+      "loss": 0.4999,
+      "step": 780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.230329489876919,
+      "learning_rate": 8.183450870595443e-06,
+      "loss": 0.4914,
+      "step": 781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.365224048873309,
+      "learning_rate": 8.178618741049841e-06,
+      "loss": 0.1677,
+      "step": 782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.506982225512495,
+      "learning_rate": 8.173781624002456e-06,
+      "loss": 0.355,
+      "step": 783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.776690326193792,
+      "learning_rate": 8.168939527043104e-06,
+      "loss": 0.298,
+      "step": 784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.276852805619795,
+      "learning_rate": 8.164092457769415e-06,
+      "loss": 0.3634,
+      "step": 785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.691430576807921,
+      "learning_rate": 8.15924042378682e-06,
+      "loss": 0.5947,
+      "step": 786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.838608299501098,
+      "learning_rate": 8.15438343270854e-06,
+      "loss": 0.3637,
+      "step": 787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.795669601902038,
+      "learning_rate": 8.149521492155573e-06,
+      "loss": 0.5435,
+      "step": 788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.597826518560188,
+      "learning_rate": 8.144654609756685e-06,
+      "loss": 0.4146,
+      "step": 789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.08483414180766,
+      "learning_rate": 8.1397827931484e-06,
+      "loss": 0.5755,
+      "step": 790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.767836760691509,
+      "learning_rate": 8.134906049974975e-06,
+      "loss": 0.5221,
+      "step": 791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.082039160046525,
+      "learning_rate": 8.130024387888402e-06,
+      "loss": 0.3184,
+      "step": 792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.647781769097618,
+      "learning_rate": 8.125137814548394e-06,
+      "loss": 0.6682,
+      "step": 793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.287869012963238,
+      "learning_rate": 8.120246337622364e-06,
+      "loss": 0.4346,
+      "step": 794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.222356287706546,
+      "learning_rate": 8.115349964785425e-06,
+      "loss": 0.252,
+      "step": 795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.672114372250452,
+      "learning_rate": 8.110448703720372e-06,
+      "loss": 0.5223,
+      "step": 796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.568658654238156,
+      "learning_rate": 8.105542562117663e-06,
+      "loss": 0.4691,
+      "step": 797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.66777911656101,
+      "learning_rate": 8.100631547675417e-06,
+      "loss": 0.3635,
+      "step": 798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.557322065714295,
+      "learning_rate": 8.095715668099406e-06,
+      "loss": 0.3429,
+      "step": 799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.764516026682733,
+      "learning_rate": 8.090794931103026e-06,
+      "loss": 0.461,
+      "step": 800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.020285678202645,
+      "learning_rate": 8.085869344407301e-06,
+      "loss": 0.3331,
+      "step": 801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.422975741347221,
+      "learning_rate": 8.080938915740863e-06,
+      "loss": 0.4973,
+      "step": 802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.310696245972512,
+      "learning_rate": 8.076003652839936e-06,
+      "loss": 0.4187,
+      "step": 803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.147109288724854,
+      "learning_rate": 8.071063563448341e-06,
+      "loss": 0.3145,
+      "step": 804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.248827497186834,
+      "learning_rate": 8.066118655317458e-06,
+      "loss": 0.3018,
+      "step": 805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.756398464057858,
+      "learning_rate": 8.06116893620624e-06,
+      "loss": 0.8368,
+      "step": 806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.52974129421263,
+      "learning_rate": 8.056214413881183e-06,
+      "loss": 0.6172,
+      "step": 807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.363249307001285,
+      "learning_rate": 8.051255096116322e-06,
+      "loss": 0.4903,
+      "step": 808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.898049285268048,
+      "learning_rate": 8.04629099069321e-06,
+      "loss": 0.5318,
+      "step": 809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.78748029075573,
+      "learning_rate": 8.041322105400923e-06,
+      "loss": 0.6577,
+      "step": 810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.808051751526834,
+      "learning_rate": 8.036348448036029e-06,
+      "loss": 0.686,
+      "step": 811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.896958076968208,
+      "learning_rate": 8.031370026402585e-06,
+      "loss": 0.4541,
+      "step": 812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.266805886990289,
+      "learning_rate": 8.026386848312125e-06,
+      "loss": 0.4783,
+      "step": 813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.864448546069593,
+      "learning_rate": 8.021398921583644e-06,
+      "loss": 0.2306,
+      "step": 814
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.035033849450723,
+      "learning_rate": 8.016406254043595e-06,
+      "loss": 0.3472,
+      "step": 815
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.206269336181785,
+      "learning_rate": 8.01140885352586e-06,
+      "loss": 0.339,
+      "step": 816
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.158672155177223,
+      "learning_rate": 8.006406727871754e-06,
+      "loss": 0.4701,
+      "step": 817
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.268309197205209,
+      "learning_rate": 8.001399884930004e-06,
+      "loss": 0.7795,
+      "step": 818
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.685883229283693,
+      "learning_rate": 7.996388332556735e-06,
+      "loss": 0.2457,
+      "step": 819
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.685893990222208,
+      "learning_rate": 7.991372078615469e-06,
+      "loss": 0.3796,
+      "step": 820
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.791188772768352,
+      "learning_rate": 7.9863511309771e-06,
+      "loss": 0.2602,
+      "step": 821
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.985786895734577,
+      "learning_rate": 7.981325497519892e-06,
+      "loss": 0.4492,
+      "step": 822
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.728818527736347,
+      "learning_rate": 7.97629518612945e-06,
+      "loss": 0.6093,
+      "step": 823
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.637945636826913,
+      "learning_rate": 7.971260204698732e-06,
+      "loss": 0.5082,
+      "step": 824
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.223576231538596,
+      "learning_rate": 7.966220561128018e-06,
+      "loss": 0.3259,
+      "step": 825
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.601166099162938,
+      "learning_rate": 7.961176263324902e-06,
+      "loss": 0.7753,
+      "step": 826
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.296139601564292,
+      "learning_rate": 7.95612731920428e-06,
+      "loss": 0.4444,
+      "step": 827
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.822706280656217,
+      "learning_rate": 7.951073736688348e-06,
+      "loss": 0.7398,
+      "step": 828
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.972677586412486,
+      "learning_rate": 7.946015523706566e-06,
+      "loss": 0.3458,
+      "step": 829
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.425879843644708,
+      "learning_rate": 7.940952688195668e-06,
+      "loss": 0.7229,
+      "step": 830
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.390753617140634,
+      "learning_rate": 7.93588523809964e-06,
+      "loss": 0.6431,
+      "step": 831
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.989526637712588,
+      "learning_rate": 7.930813181369713e-06,
+      "loss": 0.923,
+      "step": 832
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.612825767381356,
+      "learning_rate": 7.925736525964332e-06,
+      "loss": 0.3986,
+      "step": 833
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.421828067703114,
+      "learning_rate": 7.920655279849173e-06,
+      "loss": 0.4148,
+      "step": 834
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.052486214672628,
+      "learning_rate": 7.915569450997107e-06,
+      "loss": 0.5048,
+      "step": 835
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.713428863648017,
+      "learning_rate": 7.9104790473882e-06,
+      "loss": 0.5814,
+      "step": 836
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.475388170320711,
+      "learning_rate": 7.905384077009693e-06,
+      "loss": 0.2815,
+      "step": 837
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.008027395608794,
+      "learning_rate": 7.900284547855992e-06,
+      "loss": 0.2443,
+      "step": 838
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.113219858896548,
+      "learning_rate": 7.895180467928658e-06,
+      "loss": 0.7257,
+      "step": 839
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.266770238239666,
+      "learning_rate": 7.890071845236395e-06,
+      "loss": 0.9316,
+      "step": 840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.479197813261954,
+      "learning_rate": 7.88495868779503e-06,
+      "loss": 0.4266,
+      "step": 841
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.206888987163342,
+      "learning_rate": 7.87984100362751e-06,
+      "loss": 0.684,
+      "step": 842
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.415497278759084,
+      "learning_rate": 7.874718800763876e-06,
+      "loss": 0.3689,
+      "step": 843
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.13965533639083,
+      "learning_rate": 7.869592087241271e-06,
+      "loss": 0.9247,
+      "step": 844
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.1756540266641,
+      "learning_rate": 7.86446087110391e-06,
+      "loss": 0.3632,
+      "step": 845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.871009944544214,
+      "learning_rate": 7.859325160403073e-06,
+      "loss": 0.3478,
+      "step": 846
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.58581841750387,
+      "learning_rate": 7.854184963197088e-06,
+      "loss": 0.6887,
+      "step": 847
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.115272882881465,
+      "learning_rate": 7.849040287551331e-06,
+      "loss": 0.4568,
+      "step": 848
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.420494654807841,
+      "learning_rate": 7.843891141538201e-06,
+      "loss": 0.5524,
+      "step": 849
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.171318241974317,
+      "learning_rate": 7.838737533237111e-06,
+      "loss": 0.5518,
+      "step": 850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.49779281969797,
+      "learning_rate": 7.833579470734476e-06,
+      "loss": 0.4524,
+      "step": 851
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.879753949769127,
+      "learning_rate": 7.8284169621237e-06,
+      "loss": 0.2376,
+      "step": 852
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.620716549175587,
+      "learning_rate": 7.823250015505162e-06,
+      "loss": 0.5439,
+      "step": 853
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.289347940460143,
+      "learning_rate": 7.818078638986208e-06,
+      "loss": 0.4147,
+      "step": 854
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.603915975794902,
+      "learning_rate": 7.812902840681132e-06,
+      "loss": 0.2551,
+      "step": 855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.942958586562655,
+      "learning_rate": 7.807722628711167e-06,
+      "loss": 0.5247,
+      "step": 856
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.047747347222428,
+      "learning_rate": 7.80253801120447e-06,
+      "loss": 0.1714,
+      "step": 857
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.931433738654142,
+      "learning_rate": 7.797348996296116e-06,
+      "loss": 0.535,
+      "step": 858
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.000442816212573,
+      "learning_rate": 7.792155592128072e-06,
+      "loss": 0.5252,
+      "step": 859
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.830508667965475,
+      "learning_rate": 7.786957806849197e-06,
+      "loss": 0.5393,
+      "step": 860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.480387441226519,
+      "learning_rate": 7.781755648615223e-06,
+      "loss": 0.1581,
+      "step": 861
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.280662026827981,
+      "learning_rate": 7.776549125588743e-06,
+      "loss": 0.4115,
+      "step": 862
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.097689927493748,
+      "learning_rate": 7.771338245939205e-06,
+      "loss": 0.5427,
+      "step": 863
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.427182438892707,
+      "learning_rate": 7.766123017842877e-06,
+      "loss": 0.6986,
+      "step": 864
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.36077748552378,
+      "learning_rate": 7.76090344948287e-06,
+      "loss": 0.5655,
+      "step": 865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.52636830181742,
+      "learning_rate": 7.755679549049093e-06,
+      "loss": 0.5787,
+      "step": 866
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.419247857310507,
+      "learning_rate": 7.75045132473825e-06,
+      "loss": 0.1722,
+      "step": 867
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.010495324275274,
+      "learning_rate": 7.745218784753841e-06,
+      "loss": 0.2146,
+      "step": 868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.592558343629278,
+      "learning_rate": 7.739981937306127e-06,
+      "loss": 0.664,
+      "step": 869
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.755589366573375,
+      "learning_rate": 7.734740790612137e-06,
+      "loss": 0.5451,
+      "step": 870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.868439287985701,
+      "learning_rate": 7.729495352895633e-06,
+      "loss": 0.4656,
+      "step": 871
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.398297835691569,
+      "learning_rate": 7.724245632387122e-06,
+      "loss": 0.5265,
+      "step": 872
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.693842719006854,
+      "learning_rate": 7.718991637323828e-06,
+      "loss": 0.6125,
+      "step": 873
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.316814165137444,
+      "learning_rate": 7.713733375949677e-06,
+      "loss": 0.4774,
+      "step": 874
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.73428984456718,
+      "learning_rate": 7.708470856515295e-06,
+      "loss": 0.2462,
+      "step": 875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.433788748024245,
+      "learning_rate": 7.703204087277989e-06,
+      "loss": 0.4002,
+      "step": 876
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.867599005552648,
+      "learning_rate": 7.69793307650173e-06,
+      "loss": 0.476,
+      "step": 877
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.041041857843114,
+      "learning_rate": 7.692657832457146e-06,
+      "loss": 0.584,
+      "step": 878
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.099197229816003,
+      "learning_rate": 7.687378363421512e-06,
+      "loss": 0.4436,
+      "step": 879
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.837154761820468,
+      "learning_rate": 7.682094677678726e-06,
+      "loss": 0.5006,
+      "step": 880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.989861120898263,
+      "learning_rate": 7.676806783519304e-06,
+      "loss": 0.5088,
+      "step": 881
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.840051728533584,
+      "learning_rate": 7.671514689240366e-06,
+      "loss": 0.4533,
+      "step": 882
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.863218145601783,
+      "learning_rate": 7.666218403145625e-06,
+      "loss": 0.4614,
+      "step": 883
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.959719858781327,
+      "learning_rate": 7.660917933545367e-06,
+      "loss": 0.3868,
+      "step": 884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.023458897587899,
+      "learning_rate": 7.655613288756443e-06,
+      "loss": 0.2898,
+      "step": 885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.9095393828131515,
+      "learning_rate": 7.650304477102258e-06,
+      "loss": 0.3415,
+      "step": 886
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.674791212464568,
+      "learning_rate": 7.644991506912753e-06,
+      "loss": 0.2355,
+      "step": 887
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.159665018456524,
+      "learning_rate": 7.639674386524395e-06,
+      "loss": 0.5951,
+      "step": 888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.447304386091854,
+      "learning_rate": 7.634353124280161e-06,
+      "loss": 0.3985,
+      "step": 889
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.470496142422707,
+      "learning_rate": 7.629027728529527e-06,
+      "loss": 0.2365,
+      "step": 890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.789925041141158,
+      "learning_rate": 7.623698207628458e-06,
+      "loss": 0.4334,
+      "step": 891
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.127678441778919,
+      "learning_rate": 7.61836456993939e-06,
+      "loss": 0.3849,
+      "step": 892
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.491904055112274,
+      "learning_rate": 7.613026823831217e-06,
+      "loss": 0.491,
+      "step": 893
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.438390596344213,
+      "learning_rate": 7.607684977679284e-06,
+      "loss": 0.1831,
+      "step": 894
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.499709731549995,
+      "learning_rate": 7.602339039865362e-06,
+      "loss": 0.5286,
+      "step": 895
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.818996224502545,
+      "learning_rate": 7.5969890187776474e-06,
+      "loss": 0.3384,
+      "step": 896
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.622391126483762,
+      "learning_rate": 7.591634922810744e-06,
+      "loss": 0.2655,
+      "step": 897
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.214573683173224,
+      "learning_rate": 7.586276760365645e-06,
+      "loss": 0.3957,
+      "step": 898
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.572179825722651,
+      "learning_rate": 7.580914539849731e-06,
+      "loss": 0.5931,
+      "step": 899
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.599864838395836,
+      "learning_rate": 7.575548269676741e-06,
+      "loss": 0.3567,
+      "step": 900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.519335138049291,
+      "learning_rate": 7.570177958266775e-06,
+      "loss": 0.5207,
+      "step": 901
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.223999573203354,
+      "learning_rate": 7.564803614046276e-06,
+      "loss": 0.582,
+      "step": 902
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.934961770868002,
+      "learning_rate": 7.559425245448006e-06,
+      "loss": 0.3583,
+      "step": 903
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.936806076191598,
+      "learning_rate": 7.554042860911049e-06,
+      "loss": 0.4139,
+      "step": 904
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.084819559622526,
+      "learning_rate": 7.548656468880788e-06,
+      "loss": 0.3148,
+      "step": 905
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.047827588094748,
+      "learning_rate": 7.543266077808893e-06,
+      "loss": 0.4209,
+      "step": 906
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.186947229984948,
+      "learning_rate": 7.53787169615331e-06,
+      "loss": 0.1805,
+      "step": 907
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.859519080877572,
+      "learning_rate": 7.5324733323782465e-06,
+      "loss": 0.4813,
+      "step": 908
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.628152400061126,
+      "learning_rate": 7.52707099495416e-06,
+      "loss": 0.6457,
+      "step": 909
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.30108117503498,
+      "learning_rate": 7.521664692357737e-06,
+      "loss": 0.6157,
+      "step": 910
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.272286880946232,
+      "learning_rate": 7.516254433071894e-06,
+      "loss": 0.312,
+      "step": 911
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.871431690530288,
+      "learning_rate": 7.510840225585749e-06,
+      "loss": 0.176,
+      "step": 912
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.796324932980957,
+      "learning_rate": 7.50542207839462e-06,
+      "loss": 0.4475,
+      "step": 913
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.510599111025023,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.6926,
+      "step": 914
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.783146747017177,
+      "learning_rate": 7.49457399890956e-06,
+      "loss": 0.2763,
+      "step": 915
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.765423618662405,
+      "learning_rate": 7.489144083637117e-06,
+      "loss": 0.788,
+      "step": 916
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.18866604867512,
+      "learning_rate": 7.483710262702635e-06,
+      "loss": 0.3985,
+      "step": 917
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.418174023002933,
+      "learning_rate": 7.478272544632204e-06,
+      "loss": 0.3426,
+      "step": 918
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.877926957826821,
+      "learning_rate": 7.472830937958029e-06,
+      "loss": 0.3326,
+      "step": 919
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.804577282671747,
+      "learning_rate": 7.467385451218418e-06,
+      "loss": 0.4033,
+      "step": 920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.182541980696305,
+      "learning_rate": 7.461936092957767e-06,
+      "loss": 0.4971,
+      "step": 921
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.400725860964071,
+      "learning_rate": 7.456482871726545e-06,
+      "loss": 0.6724,
+      "step": 922
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.986330211916101,
+      "learning_rate": 7.4510257960812824e-06,
+      "loss": 0.6901,
+      "step": 923
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.126633799225575,
+      "learning_rate": 7.44556487458456e-06,
+      "loss": 0.173,
+      "step": 924
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.548104592149162,
+      "learning_rate": 7.440100115804991e-06,
+      "loss": 0.402,
+      "step": 925
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.880873280777033,
+      "learning_rate": 7.434631528317209e-06,
+      "loss": 0.5247,
+      "step": 926
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.332702299884552,
+      "learning_rate": 7.4291591207018565e-06,
+      "loss": 0.4582,
+      "step": 927
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.30930581915382,
+      "learning_rate": 7.4236829015455725e-06,
+      "loss": 0.297,
+      "step": 928
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.651868108801631,
+      "learning_rate": 7.418202879440969e-06,
+      "loss": 0.5303,
+      "step": 929
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.611249541223199,
+      "learning_rate": 7.412719062986632e-06,
+      "loss": 0.6683,
+      "step": 930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.506407187152776,
+      "learning_rate": 7.407231460787099e-06,
+      "loss": 0.4605,
+      "step": 931
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.8075664126980815,
+      "learning_rate": 7.401740081452848e-06,
+      "loss": 0.6384,
+      "step": 932
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.013226101108797,
+      "learning_rate": 7.396244933600285e-06,
+      "loss": 0.6264,
+      "step": 933
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.483913847537034,
+      "learning_rate": 7.390746025851725e-06,
+      "loss": 0.3354,
+      "step": 934
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.770406053687868,
+      "learning_rate": 7.385243366835385e-06,
+      "loss": 0.5484,
+      "step": 935
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.270990276770199,
+      "learning_rate": 7.379736965185369e-06,
+      "loss": 0.3281,
+      "step": 936
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.461776957288274,
+      "learning_rate": 7.374226829541652e-06,
+      "loss": 0.5999,
+      "step": 937
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.531996261823085,
+      "learning_rate": 7.368712968550068e-06,
+      "loss": 0.4971,
+      "step": 938
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.86852836903852,
+      "learning_rate": 7.363195390862298e-06,
+      "loss": 0.4002,
+      "step": 939
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.390810237314641,
+      "learning_rate": 7.3576741051358525e-06,
+      "loss": 0.5065,
+      "step": 940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.273609874743478,
+      "learning_rate": 7.352149120034062e-06,
+      "loss": 0.2733,
+      "step": 941
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.271646280078986,
+      "learning_rate": 7.3466204442260605e-06,
+      "loss": 0.424,
+      "step": 942
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.184285594241308,
+      "learning_rate": 7.341088086386775e-06,
+      "loss": 0.3493,
+      "step": 943
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.339277899733212,
+      "learning_rate": 7.3355520551969055e-06,
+      "loss": 0.4544,
+      "step": 944
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.834333851789557,
+      "learning_rate": 7.330012359342919e-06,
+      "loss": 0.3474,
+      "step": 945
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.319046339004926,
+      "learning_rate": 7.324469007517035e-06,
+      "loss": 0.6389,
+      "step": 946
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.75922380675413,
+      "learning_rate": 7.318922008417203e-06,
+      "loss": 0.6654,
+      "step": 947
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.354457928437528,
+      "learning_rate": 7.313371370747104e-06,
+      "loss": 0.5256,
+      "step": 948
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.129065637535593,
+      "learning_rate": 7.3078171032161175e-06,
+      "loss": 0.3024,
+      "step": 949
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.308795759170206,
+      "learning_rate": 7.302259214539327e-06,
+      "loss": 0.6016,
+      "step": 950
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.589808780872742,
+      "learning_rate": 7.296697713437495e-06,
+      "loss": 0.4572,
+      "step": 951
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.10792181058575,
+      "learning_rate": 7.291132608637053e-06,
+      "loss": 0.7745,
+      "step": 952
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.663966118207991,
+      "learning_rate": 7.285563908870086e-06,
+      "loss": 0.2932,
+      "step": 953
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.79786550987496,
+      "learning_rate": 7.279991622874319e-06,
+      "loss": 0.2123,
+      "step": 954
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.021558279123923,
+      "learning_rate": 7.274415759393107e-06,
+      "loss": 0.6763,
+      "step": 955
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.300239595612451,
+      "learning_rate": 7.268836327175417e-06,
+      "loss": 0.5467,
+      "step": 956
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.574903430633393,
+      "learning_rate": 7.263253334975812e-06,
+      "loss": 0.2903,
+      "step": 957
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.699072868019169,
+      "learning_rate": 7.257666791554448e-06,
+      "loss": 0.3367,
+      "step": 958
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.755639906043921,
+      "learning_rate": 7.252076705677046e-06,
+      "loss": 0.4042,
+      "step": 959
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.500518482284736,
+      "learning_rate": 7.24648308611489e-06,
+      "loss": 0.4898,
+      "step": 960
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.982244138931426,
+      "learning_rate": 7.2408859416448065e-06,
+      "loss": 0.4858,
+      "step": 961
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.553015877874401,
+      "learning_rate": 7.235285281049154e-06,
+      "loss": 0.208,
+      "step": 962
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.951166720872019,
+      "learning_rate": 7.22968111311581e-06,
+      "loss": 0.7537,
+      "step": 963
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.942136949654715,
+      "learning_rate": 7.224073446638149e-06,
+      "loss": 0.4766,
+      "step": 964
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.022259573397292,
+      "learning_rate": 7.21846229041504e-06,
+      "loss": 0.4937,
+      "step": 965
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.297513356992832,
+      "learning_rate": 7.212847653250828e-06,
+      "loss": 0.7728,
+      "step": 966
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.799507520677372,
+      "learning_rate": 7.207229543955319e-06,
+      "loss": 0.3441,
+      "step": 967
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.958094643147433,
+      "learning_rate": 7.201607971343763e-06,
+      "loss": 0.3181,
+      "step": 968
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.704164421436593,
+      "learning_rate": 7.195982944236853e-06,
+      "loss": 0.3022,
+      "step": 969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.818606114653967,
+      "learning_rate": 7.190354471460692e-06,
+      "loss": 0.2873,
+      "step": 970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.34106315283732,
+      "learning_rate": 7.1847225618467975e-06,
+      "loss": 0.4328,
+      "step": 971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.334643632122907,
+      "learning_rate": 7.1790872242320775e-06,
+      "loss": 0.1775,
+      "step": 972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.809965303622986,
+      "learning_rate": 7.173448467458817e-06,
+      "loss": 0.4367,
+      "step": 973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.075449332715628,
+      "learning_rate": 7.167806300374665e-06,
+      "loss": 0.6589,
+      "step": 974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.144663149877118,
+      "learning_rate": 7.162160731832627e-06,
+      "loss": 0.3243,
+      "step": 975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.822550977534975,
+      "learning_rate": 7.15651177069104e-06,
+      "loss": 0.7383,
+      "step": 976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.4880035185404035,
+      "learning_rate": 7.1508594258135685e-06,
+      "loss": 0.4679,
+      "step": 977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.712726394513295,
+      "learning_rate": 7.145203706069183e-06,
+      "loss": 0.393,
+      "step": 978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.86599163408303,
+      "learning_rate": 7.139544620332151e-06,
+      "loss": 0.2996,
+      "step": 979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.455728076415062,
+      "learning_rate": 7.133882177482019e-06,
+      "loss": 0.5095,
+      "step": 980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.366561820948556,
+      "learning_rate": 7.128216386403608e-06,
+      "loss": 0.5252,
+      "step": 981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.841303528980102,
+      "learning_rate": 7.122547255986985e-06,
+      "loss": 0.4737,
+      "step": 982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.24797798028234,
+      "learning_rate": 7.1168747951274596e-06,
+      "loss": 0.2891,
+      "step": 983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.157263496152308,
+      "learning_rate": 7.1111990127255684e-06,
+      "loss": 0.5676,
+      "step": 984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.255735066178092,
+      "learning_rate": 7.105519917687058e-06,
+      "loss": 0.8031,
+      "step": 985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.088622193882017,
+      "learning_rate": 7.099837518922873e-06,
+      "loss": 0.2461,
+      "step": 986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.583589723619978,
+      "learning_rate": 7.094151825349145e-06,
+      "loss": 0.3652,
+      "step": 987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.539673977440312,
+      "learning_rate": 7.088462845887168e-06,
+      "loss": 0.4904,
+      "step": 988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.194189225822065,
+      "learning_rate": 7.082770589463398e-06,
+      "loss": 0.6225,
+      "step": 989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.67359221737759,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 0.4783,
+      "step": 990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.083168800719198,
+      "learning_rate": 7.071376281461994e-06,
+      "loss": 0.512,
+      "step": 991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.224407504008749,
+      "learning_rate": 7.065674247762924e-06,
+      "loss": 0.6191,
+      "step": 992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.934202927192099,
+      "learning_rate": 7.059968972859155e-06,
+      "loss": 0.4955,
+      "step": 993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.471045525649119,
+      "learning_rate": 7.054260465702712e-06,
+      "loss": 0.4877,
+      "step": 994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.098476455026248,
+      "learning_rate": 7.04854873525069e-06,
+      "loss": 0.4333,
+      "step": 995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.06314694285291,
+      "learning_rate": 7.042833790465241e-06,
+      "loss": 0.613,
+      "step": 996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.284446290448441,
+      "learning_rate": 7.037115640313563e-06,
+      "loss": 0.6735,
+      "step": 997
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.35697168242451,
+      "learning_rate": 7.031394293767879e-06,
+      "loss": 0.4054,
+      "step": 998
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.139374059742304,
+      "learning_rate": 7.025669759805431e-06,
+      "loss": 0.3726,
+      "step": 999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.546727011052494,
+      "learning_rate": 7.019942047408461e-06,
+      "loss": 0.2654,
+      "step": 1000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.403364526484237,
+      "learning_rate": 7.0142111655642e-06,
+      "loss": 0.3275,
+      "step": 1001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.428479003494727,
+      "learning_rate": 7.008477123264849e-06,
+      "loss": 0.3167,
+      "step": 1002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.929596342102183,
+      "learning_rate": 7.002739929507569e-06,
+      "loss": 0.2365,
+      "step": 1003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.605442181467153,
+      "learning_rate": 6.996999593294466e-06,
+      "loss": 0.3164,
+      "step": 1004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.419001246853925,
+      "learning_rate": 6.991256123632577e-06,
+      "loss": 0.8182,
+      "step": 1005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.1813510880891,
+      "learning_rate": 6.985509529533859e-06,
+      "loss": 0.602,
+      "step": 1006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.349330429494252,
+      "learning_rate": 6.979759820015166e-06,
+      "loss": 0.5265,
+      "step": 1007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.878079401249398,
+      "learning_rate": 6.974007004098243e-06,
+      "loss": 0.4883,
+      "step": 1008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.303858539733948,
+      "learning_rate": 6.968251090809708e-06,
+      "loss": 0.3041,
+      "step": 1009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.6840817368078485,
+      "learning_rate": 6.96249208918104e-06,
+      "loss": 0.258,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.82669566991317,
+      "learning_rate": 6.956730008248565e-06,
+      "loss": 0.3709,
+      "step": 1011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.453548005337307,
+      "learning_rate": 6.950964857053436e-06,
+      "loss": 0.3207,
+      "step": 1012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.468648437368625,
+      "learning_rate": 6.945196644641631e-06,
+      "loss": 0.4046,
+      "step": 1013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.8990173126418455,
+      "learning_rate": 6.939425380063924e-06,
+      "loss": 0.5832,
+      "step": 1014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.319641770298794,
+      "learning_rate": 6.933651072375884e-06,
+      "loss": 0.3706,
+      "step": 1015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.748764130196562,
+      "learning_rate": 6.927873730637849e-06,
+      "loss": 0.3406,
+      "step": 1016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.73780806040182,
+      "learning_rate": 6.922093363914923e-06,
+      "loss": 0.4664,
+      "step": 1017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.06897179175169,
+      "learning_rate": 6.916309981276954e-06,
+      "loss": 0.5925,
+      "step": 1018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.014569568395036,
+      "learning_rate": 6.9105235917985215e-06,
+      "loss": 0.4542,
+      "step": 1019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.813799108639909,
+      "learning_rate": 6.9047342045589224e-06,
+      "loss": 0.6067,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.78809573613665,
+      "learning_rate": 6.898941828642163e-06,
+      "loss": 0.4096,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.556652996205933,
+      "learning_rate": 6.89314647313693e-06,
+      "loss": 0.4418,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.92096296506994,
+      "learning_rate": 6.887348147136592e-06,
+      "loss": 0.2658,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.358867402856694,
+      "learning_rate": 6.8815468597391785e-06,
+      "loss": 0.4462,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.991751568999254,
+      "learning_rate": 6.875742620047361e-06,
+      "loss": 0.5793,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.854698872477309,
+      "learning_rate": 6.869935437168449e-06,
+      "loss": 0.4027,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.70443834170101,
+      "learning_rate": 6.864125320214364e-06,
+      "loss": 0.3892,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.277247741799264,
+      "learning_rate": 6.858312278301638e-06,
+      "loss": 0.4912,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.289920996270459,
+      "learning_rate": 6.852496320551387e-06,
+      "loss": 0.3457,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.477658125718834,
+      "learning_rate": 6.846677456089305e-06,
+      "loss": 0.8159,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.992801286233085,
+      "learning_rate": 6.840855694045647e-06,
+      "loss": 0.3422,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.978277719533185,
+      "learning_rate": 6.835031043555211e-06,
+      "loss": 0.5569,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.685638460436984,
+      "learning_rate": 6.829203513757332e-06,
+      "loss": 0.6372,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.915005832607024,
+      "learning_rate": 6.82337311379586e-06,
+      "loss": 0.474,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.004214238516166,
+      "learning_rate": 6.817539852819149e-06,
+      "loss": 0.2791,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.780108837831955,
+      "learning_rate": 6.811703739980045e-06,
+      "loss": 0.4929,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.421900114621625,
+      "learning_rate": 6.8058647844358625e-06,
+      "loss": 0.5301,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.86780278187002,
+      "learning_rate": 6.800022995348381e-06,
+      "loss": 0.293,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.857708210409456,
+      "learning_rate": 6.794178381883827e-06,
+      "loss": 0.356,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.05756180887477,
+      "learning_rate": 6.788330953212854e-06,
+      "loss": 0.4752,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.943346135264479,
+      "learning_rate": 6.782480718510538e-06,
+      "loss": 0.6487,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.036150440376149,
+      "learning_rate": 6.776627686956354e-06,
+      "loss": 0.5505,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.682255675214314,
+      "learning_rate": 6.770771867734169e-06,
+      "loss": 0.7012,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.146947519096852,
+      "learning_rate": 6.76491327003222e-06,
+      "loss": 0.7761,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.096781436068701,
+      "learning_rate": 6.7590519030431054e-06,
+      "loss": 0.538,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.789247826814453,
+      "learning_rate": 6.753187775963773e-06,
+      "loss": 0.4735,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.710350218282558,
+      "learning_rate": 6.747320897995493e-06,
+      "loss": 0.581,
+      "step": 1047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.16976808843988,
+      "learning_rate": 6.7414512783438575e-06,
+      "loss": 0.5197,
+      "step": 1048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.113215409077169,
+      "learning_rate": 6.73557892621876e-06,
+      "loss": 0.3605,
+      "step": 1049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.141515338242623,
+      "learning_rate": 6.729703850834381e-06,
+      "loss": 0.4848,
+      "step": 1050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.037791116015522,
+      "learning_rate": 6.723826061409176e-06,
+      "loss": 0.456,
+      "step": 1051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.073982168023747,
+      "learning_rate": 6.717945567165854e-06,
+      "loss": 0.5784,
+      "step": 1052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.157098574363703,
+      "learning_rate": 6.712062377331371e-06,
+      "loss": 0.6205,
+      "step": 1053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.52845942754326,
+      "learning_rate": 6.706176501136914e-06,
+      "loss": 0.4625,
+      "step": 1054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.169593058744603,
+      "learning_rate": 6.700287947817885e-06,
+      "loss": 0.3226,
+      "step": 1055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.874370173662523,
+      "learning_rate": 6.694396726613883e-06,
+      "loss": 0.2981,
+      "step": 1056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.479595833741454,
+      "learning_rate": 6.688502846768697e-06,
+      "loss": 0.6385,
+      "step": 1057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.425590651293703,
+      "learning_rate": 6.682606317530284e-06,
+      "loss": 0.3045,
+      "step": 1058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.071523410918754,
+      "learning_rate": 6.676707148150763e-06,
+      "loss": 0.4712,
+      "step": 1059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.644201207421403,
+      "learning_rate": 6.670805347886392e-06,
+      "loss": 0.5459,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.49346576536051,
+      "learning_rate": 6.6649009259975585e-06,
+      "loss": 0.5596,
+      "step": 1061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.715129272878258,
+      "learning_rate": 6.65899389174876e-06,
+      "loss": 0.3236,
+      "step": 1062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.439265387478165,
+      "learning_rate": 6.653084254408599e-06,
+      "loss": 0.5067,
+      "step": 1063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.826668778035177,
+      "learning_rate": 6.647172023249758e-06,
+      "loss": 0.662,
+      "step": 1064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.137795566095344,
+      "learning_rate": 6.641257207548992e-06,
+      "loss": 0.4822,
+      "step": 1065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.897250874358917,
+      "learning_rate": 6.635339816587109e-06,
+      "loss": 0.2078,
+      "step": 1066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.884037994296202,
+      "learning_rate": 6.629419859648959e-06,
+      "loss": 0.3881,
+      "step": 1067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.463398231120273,
+      "learning_rate": 6.6234973460234184e-06,
+      "loss": 0.7975,
+      "step": 1068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.334062979459183,
+      "learning_rate": 6.6175722850033755e-06,
+      "loss": 0.3141,
+      "step": 1069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.724724816769472,
+      "learning_rate": 6.611644685885713e-06,
+      "loss": 0.5606,
+      "step": 1070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.838423800913443,
+      "learning_rate": 6.6057145579713015e-06,
+      "loss": 0.4222,
+      "step": 1071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.279715746541136,
+      "learning_rate": 6.599781910564973e-06,
+      "loss": 0.5569,
+      "step": 1072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.08173081497578,
+      "learning_rate": 6.593846752975518e-06,
+      "loss": 0.3743,
+      "step": 1073
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.181417946899355,
+      "learning_rate": 6.587909094515663e-06,
+      "loss": 0.4677,
+      "step": 1074
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.615560751479444,
+      "learning_rate": 6.58196894450206e-06,
+      "loss": 0.1893,
+      "step": 1075
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.198662659102776,
+      "learning_rate": 6.576026312255268e-06,
+      "loss": 0.5692,
+      "step": 1076
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.680927222092437,
+      "learning_rate": 6.570081207099744e-06,
+      "loss": 0.6401,
+      "step": 1077
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.854537490531515,
+      "learning_rate": 6.564133638363823e-06,
+      "loss": 0.2684,
+      "step": 1078
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.615350714541232,
+      "learning_rate": 6.558183615379708e-06,
+      "loss": 0.3606,
+      "step": 1079
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.77756185542429,
+      "learning_rate": 6.552231147483448e-06,
+      "loss": 0.6708,
+      "step": 1080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.552454749767117,
+      "learning_rate": 6.546276244014933e-06,
+      "loss": 0.4379,
+      "step": 1081
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.06482377350888,
+      "learning_rate": 6.5403189143178725e-06,
+      "loss": 0.3777,
+      "step": 1082
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.012027354495712,
+      "learning_rate": 6.534359167739784e-06,
+      "loss": 0.5679,
+      "step": 1083
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.483201761594449,
+      "learning_rate": 6.528397013631975e-06,
+      "loss": 0.3148,
+      "step": 1084
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.692824835030049,
+      "learning_rate": 6.522432461349536e-06,
+      "loss": 0.3654,
+      "step": 1085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.306996192326768,
+      "learning_rate": 6.5164655202513135e-06,
+      "loss": 0.1494,
+      "step": 1086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.615178123383902,
+      "learning_rate": 6.510496199699906e-06,
+      "loss": 0.5501,
+      "step": 1087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.688777636086809,
+      "learning_rate": 6.504524509061646e-06,
+      "loss": 0.2488,
+      "step": 1088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.029407334151884,
+      "learning_rate": 6.498550457706584e-06,
+      "loss": 0.3055,
+      "step": 1089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.184285907173871,
+      "learning_rate": 6.492574055008474e-06,
+      "loss": 0.4781,
+      "step": 1090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.638653067981326,
+      "learning_rate": 6.4865953103447585e-06,
+      "loss": 0.2311,
+      "step": 1091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.528374242654134,
+      "learning_rate": 6.480614233096558e-06,
+      "loss": 0.3467,
+      "step": 1092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.2276386539033375,
+      "learning_rate": 6.474630832648651e-06,
+      "loss": 0.1556,
+      "step": 1093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.928072192946448,
+      "learning_rate": 6.4686451183894604e-06,
+      "loss": 0.374,
+      "step": 1094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.086676680696224,
+      "learning_rate": 6.462657099711044e-06,
+      "loss": 0.3722,
+      "step": 1095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.118391943457603,
+      "learning_rate": 6.4566667860090674e-06,
+      "loss": 0.4345,
+      "step": 1096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.63133039600865,
+      "learning_rate": 6.4506741866828035e-06,
+      "loss": 0.5976,
+      "step": 1097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.8736865135665,
+      "learning_rate": 6.444679311135112e-06,
+      "loss": 0.4395,
+      "step": 1098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.436745722431171,
+      "learning_rate": 6.438682168772421e-06,
+      "loss": 0.519,
+      "step": 1099
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.813616202655819,
+      "learning_rate": 6.432682769004717e-06,
+      "loss": 0.4721,
+      "step": 1100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.65683248386648,
+      "learning_rate": 6.426681121245527e-06,
+      "loss": 0.1414,
+      "step": 1101
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.810592852919331,
+      "learning_rate": 6.420677234911908e-06,
+      "loss": 0.3543,
+      "step": 1102
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.871636827802535,
+      "learning_rate": 6.414671119424426e-06,
+      "loss": 0.3333,
+      "step": 1103
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.219403866250927,
+      "learning_rate": 6.408662784207149e-06,
+      "loss": 0.5444,
+      "step": 1104
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.213208585386463,
+      "learning_rate": 6.402652238687624e-06,
+      "loss": 0.4673,
+      "step": 1105
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.200097560863402,
+      "learning_rate": 6.396639492296868e-06,
+      "loss": 0.1665,
+      "step": 1106
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.858715276033205,
+      "learning_rate": 6.390624554469351e-06,
+      "loss": 0.3212,
+      "step": 1107
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.577832526260126,
+      "learning_rate": 6.384607434642982e-06,
+      "loss": 0.6823,
+      "step": 1108
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.552462293713745,
+      "learning_rate": 6.378588142259094e-06,
+      "loss": 0.56,
+      "step": 1109
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.4312868922468995,
+      "learning_rate": 6.372566686762427e-06,
+      "loss": 0.2374,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.40033230657682,
+      "learning_rate": 6.366543077601116e-06,
+      "loss": 0.41,
+      "step": 1111
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.877374877227865,
+      "learning_rate": 6.360517324226676e-06,
+      "loss": 0.3511,
+      "step": 1112
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.258467975280134,
+      "learning_rate": 6.354489436093987e-06,
+      "loss": 0.4155,
+      "step": 1113
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.651038485038006,
+      "learning_rate": 6.348459422661276e-06,
+      "loss": 0.6254,
+      "step": 1114
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.630117166701437,
+      "learning_rate": 6.342427293390108e-06,
+      "loss": 0.5273,
+      "step": 1115
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.732310413537979,
+      "learning_rate": 6.336393057745365e-06,
+      "loss": 0.6226,
+      "step": 1116
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.82795483403045,
+      "learning_rate": 6.330356725195237e-06,
+      "loss": 0.5151,
+      "step": 1117
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.322312719576825,
+      "learning_rate": 6.324318305211201e-06,
+      "loss": 0.2244,
+      "step": 1118
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.030228395271608,
+      "learning_rate": 6.318277807268013e-06,
+      "loss": 0.5275,
+      "step": 1119
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.178368181913658,
+      "learning_rate": 6.312235240843685e-06,
+      "loss": 0.6185,
+      "step": 1120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.48394652019683,
+      "learning_rate": 6.306190615419476e-06,
+      "loss": 0.2911,
+      "step": 1121
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.74893087178822,
+      "learning_rate": 6.300143940479881e-06,
+      "loss": 0.6686,
+      "step": 1122
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.857271003219474,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.3192,
+      "step": 1123
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.599291642714053,
+      "learning_rate": 6.2880444800085535e-06,
+      "loss": 0.2214,
+      "step": 1124
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.964637098763905,
+      "learning_rate": 6.281991713461823e-06,
+      "loss": 0.3988,
+      "step": 1125
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.595984798351758,
+      "learning_rate": 6.275936935369675e-06,
+      "loss": 0.4644,
+      "step": 1126
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.167995824527482,
+      "learning_rate": 6.269880155232534e-06,
+      "loss": 0.3259,
+      "step": 1127
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.504596252435835,
+      "learning_rate": 6.2638213825539595e-06,
+      "loss": 0.3723,
+      "step": 1128
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.8661092229650675,
+      "learning_rate": 6.257760626840644e-06,
+      "loss": 0.4059,
+      "step": 1129
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.002474208246099,
+      "learning_rate": 6.251697897602384e-06,
+      "loss": 0.6143,
+      "step": 1130
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.477239058382684,
+      "learning_rate": 6.245633204352078e-06,
+      "loss": 0.3003,
+      "step": 1131
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.674224036950802,
+      "learning_rate": 6.239566556605706e-06,
+      "loss": 0.5414,
+      "step": 1132
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.169527314388807,
+      "learning_rate": 6.233497963882314e-06,
+      "loss": 0.5818,
+      "step": 1133
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.89263688760969,
+      "learning_rate": 6.227427435703997e-06,
+      "loss": 0.6487,
+      "step": 1134
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.965017659189552,
+      "learning_rate": 6.22135498159589e-06,
+      "loss": 0.5422,
+      "step": 1135
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.616313284002228,
+      "learning_rate": 6.215280611086149e-06,
+      "loss": 0.5534,
+      "step": 1136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.434447174899788,
+      "learning_rate": 6.209204333705937e-06,
+      "loss": 0.5571,
+      "step": 1137
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.009414245276776,
+      "learning_rate": 6.203126158989411e-06,
+      "loss": 0.4274,
+      "step": 1138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.37466391829999,
+      "learning_rate": 6.197046096473701e-06,
+      "loss": 0.4804,
+      "step": 1139
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.484648760368527,
+      "learning_rate": 6.190964155698903e-06,
+      "loss": 0.5939,
+      "step": 1140
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.800745529956103,
+      "learning_rate": 6.184880346208056e-06,
+      "loss": 0.4335,
+      "step": 1141
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.04749655728886,
+      "learning_rate": 6.178794677547138e-06,
+      "loss": 0.3971,
+      "step": 1142
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.335380890900943,
+      "learning_rate": 6.1727071592650345e-06,
+      "loss": 0.293,
+      "step": 1143
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.575649132651304,
+      "learning_rate": 6.1666178009135426e-06,
+      "loss": 0.547,
+      "step": 1144
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.739283149536773,
+      "learning_rate": 6.160526612047339e-06,
+      "loss": 0.4998,
+      "step": 1145
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.440917002226175,
+      "learning_rate": 6.154433602223979e-06,
+      "loss": 0.474,
+      "step": 1146
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.22981323014915,
+      "learning_rate": 6.148338781003873e-06,
+      "loss": 0.383,
+      "step": 1147
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.722600048045983,
+      "learning_rate": 6.14224215795027e-06,
+      "loss": 0.3879,
+      "step": 1148
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.74109123299632,
+      "learning_rate": 6.136143742629252e-06,
+      "loss": 0.4696,
+      "step": 1149
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.448850796308825,
+      "learning_rate": 6.130043544609707e-06,
+      "loss": 0.5078,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.537001974139077,
+      "learning_rate": 6.123941573463327e-06,
+      "loss": 0.4033,
+      "step": 1151
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.743177551585715,
+      "learning_rate": 6.117837838764579e-06,
+      "loss": 0.6524,
+      "step": 1152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.339617110187138,
+      "learning_rate": 6.111732350090703e-06,
+      "loss": 0.4396,
+      "step": 1153
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.03556768021378,
+      "learning_rate": 6.105625117021692e-06,
+      "loss": 0.5616,
+      "step": 1154
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.974875408099988,
+      "learning_rate": 6.099516149140267e-06,
+      "loss": 0.2488,
+      "step": 1155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.051169887459823,
+      "learning_rate": 6.09340545603188e-06,
+      "loss": 0.5384,
+      "step": 1156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.235042321152614,
+      "learning_rate": 6.087293047284687e-06,
+      "loss": 0.5062,
+      "step": 1157
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.426479471110495,
+      "learning_rate": 6.0811789324895365e-06,
+      "loss": 0.6083,
+      "step": 1158
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.038619979547812,
+      "learning_rate": 6.075063121239954e-06,
+      "loss": 0.5478,
+      "step": 1159
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.760261451533111,
+      "learning_rate": 6.068945623132125e-06,
+      "loss": 0.3805,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.9499417878891325,
+      "learning_rate": 6.062826447764883e-06,
+      "loss": 0.2467,
+      "step": 1161
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.824355982825876,
+      "learning_rate": 6.056705604739696e-06,
+      "loss": 0.3895,
+      "step": 1162
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.38175821325658,
+      "learning_rate": 6.050583103660643e-06,
+      "loss": 0.4532,
+      "step": 1163
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.804679946435447,
+      "learning_rate": 6.044458954134411e-06,
+      "loss": 0.4877,
+      "step": 1164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.842685872982434,
+      "learning_rate": 6.038333165770267e-06,
+      "loss": 0.2787,
+      "step": 1165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.085439150729528,
+      "learning_rate": 6.032205748180054e-06,
+      "loss": 0.4214,
+      "step": 1166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.699019760566678,
+      "learning_rate": 6.026076710978172e-06,
+      "loss": 0.516,
+      "step": 1167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.8103397331057,
+      "learning_rate": 6.019946063781559e-06,
+      "loss": 0.251,
+      "step": 1168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.00224900830546,
+      "learning_rate": 6.01381381620968e-06,
+      "loss": 0.4877,
+      "step": 1169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.26492928026321,
+      "learning_rate": 6.0076799778845105e-06,
+      "loss": 0.4188,
+      "step": 1170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.170068766096856,
+      "learning_rate": 6.001544558430527e-06,
+      "loss": 0.2704,
+      "step": 1171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.996006667110514,
+      "learning_rate": 5.99540756747468e-06,
+      "loss": 0.3676,
+      "step": 1172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.526651943045511,
+      "learning_rate": 5.989269014646393e-06,
+      "loss": 0.4119,
+      "step": 1173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.302971110996765,
+      "learning_rate": 5.983128909577532e-06,
+      "loss": 0.2118,
+      "step": 1174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.340076062711805,
+      "learning_rate": 5.976987261902403e-06,
+      "loss": 0.3156,
+      "step": 1175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.038289822041413,
+      "learning_rate": 5.970844081257734e-06,
+      "loss": 0.7092,
+      "step": 1176
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.37046586268542,
+      "learning_rate": 5.964699377282657e-06,
+      "loss": 0.4943,
+      "step": 1177
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.209621345721736,
+      "learning_rate": 5.958553159618693e-06,
+      "loss": 0.589,
+      "step": 1178
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.177236923934109,
+      "learning_rate": 5.952405437909738e-06,
+      "loss": 0.645,
+      "step": 1179
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.47753548307882,
+      "learning_rate": 5.946256221802052e-06,
+      "loss": 0.3826,
+      "step": 1180
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.146697402330108,
+      "learning_rate": 5.940105520944232e-06,
+      "loss": 0.5965,
+      "step": 1181
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.796155642342194,
+      "learning_rate": 5.933953344987215e-06,
+      "loss": 0.6581,
+      "step": 1182
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.47759602355781,
+      "learning_rate": 5.927799703584242e-06,
+      "loss": 0.5336,
+      "step": 1183
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.860894592059796,
+      "learning_rate": 5.921644606390862e-06,
+      "loss": 0.4119,
+      "step": 1184
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.356905372269031,
+      "learning_rate": 5.915488063064902e-06,
+      "loss": 0.2983,
+      "step": 1185
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.417043733670143,
+      "learning_rate": 5.9093300832664625e-06,
+      "loss": 0.5041,
+      "step": 1186
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.822020919477652,
+      "learning_rate": 5.903170676657894e-06,
+      "loss": 0.2479,
+      "step": 1187
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.374583631545006,
+      "learning_rate": 5.897009852903792e-06,
+      "loss": 0.2803,
+      "step": 1188
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.571334519863566,
+      "learning_rate": 5.890847621670966e-06,
+      "loss": 0.2534,
+      "step": 1189
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.621938830760458,
+      "learning_rate": 5.8846839926284435e-06,
+      "loss": 0.7321,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.695059588501781,
+      "learning_rate": 5.878518975447439e-06,
+      "loss": 0.5626,
+      "step": 1191
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.234023307782367,
+      "learning_rate": 5.872352579801351e-06,
+      "loss": 0.3993,
+      "step": 1192
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.352093581170527,
+      "learning_rate": 5.866184815365733e-06,
+      "loss": 0.2633,
+      "step": 1193
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.484265002473958,
+      "learning_rate": 5.860015691818292e-06,
+      "loss": 0.5873,
+      "step": 1194
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.769212095313652,
+      "learning_rate": 5.853845218838868e-06,
+      "loss": 0.1913,
+      "step": 1195
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.4205228881191,
+      "learning_rate": 5.847673406109413e-06,
+      "loss": 0.2847,
+      "step": 1196
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.44302447068203,
+      "learning_rate": 5.841500263313987e-06,
+      "loss": 0.6709,
+      "step": 1197
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.034414839139728,
+      "learning_rate": 5.835325800138736e-06,
+      "loss": 0.6089,
+      "step": 1198
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.249018529601605,
+      "learning_rate": 5.829150026271871e-06,
+      "loss": 0.1997,
+      "step": 1199
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.021064112498987,
+      "learning_rate": 5.82297295140367e-06,
+      "loss": 0.6152,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.522898695611273,
+      "learning_rate": 5.816794585226445e-06,
+      "loss": 0.4864,
+      "step": 1201
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.959209372666291,
+      "learning_rate": 5.810614937434537e-06,
+      "loss": 0.3457,
+      "step": 1202
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.888722477946267,
+      "learning_rate": 5.8044340177242995e-06,
+      "loss": 0.517,
+      "step": 1203
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.321662898990467,
+      "learning_rate": 5.7982518357940755e-06,
+      "loss": 0.4676,
+      "step": 1204
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.351181820208073,
+      "learning_rate": 5.792068401344195e-06,
+      "loss": 0.3999,
+      "step": 1205
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.993966364130094,
+      "learning_rate": 5.78588372407695e-06,
+      "loss": 0.596,
+      "step": 1206
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.915908203229893,
+      "learning_rate": 5.779697813696587e-06,
+      "loss": 0.5144,
+      "step": 1207
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.959337922232766,
+      "learning_rate": 5.77351067990928e-06,
+      "loss": 0.3569,
+      "step": 1208
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.257590822861815,
+      "learning_rate": 5.767322332423128e-06,
+      "loss": 0.4133,
+      "step": 1209
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.985090005009248,
+      "learning_rate": 5.761132780948132e-06,
+      "loss": 0.5451,
+      "step": 1210
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.640775562873882,
+      "learning_rate": 5.7549420351961845e-06,
+      "loss": 0.4622,
+      "step": 1211
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.06794569298899,
+      "learning_rate": 5.748750104881051e-06,
+      "loss": 0.1301,
+      "step": 1212
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.521076614521114,
+      "learning_rate": 5.742556999718353e-06,
+      "loss": 0.5656,
+      "step": 1213
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.2198728228844615,
+      "learning_rate": 5.736362729425558e-06,
+      "loss": 0.5135,
+      "step": 1214
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.672473202385675,
+      "learning_rate": 5.730167303721963e-06,
+      "loss": 0.2372,
+      "step": 1215
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.763355291568327,
+      "learning_rate": 5.723970732328675e-06,
+      "loss": 0.4862,
+      "step": 1216
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.33219354503798,
+      "learning_rate": 5.717773024968602e-06,
+      "loss": 0.3935,
+      "step": 1217
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.012049903590807,
+      "learning_rate": 5.711574191366427e-06,
+      "loss": 0.3742,
+      "step": 1218
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.866813699194147,
+      "learning_rate": 5.70537424124861e-06,
+      "loss": 0.5683,
+      "step": 1219
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.914177724569267,
+      "learning_rate": 5.699173184343357e-06,
+      "loss": 0.4148,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.138756255362503,
+      "learning_rate": 5.692971030380613e-06,
+      "loss": 0.6938,
+      "step": 1221
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.542026305235765,
+      "learning_rate": 5.686767789092041e-06,
+      "loss": 0.4856,
+      "step": 1222
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.285901649423664,
+      "learning_rate": 5.6805634702110155e-06,
+      "loss": 0.5205,
+      "step": 1223
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.709467011197498,
+      "learning_rate": 5.674358083472598e-06,
+      "loss": 0.2515,
+      "step": 1224
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.465130829885597,
+      "learning_rate": 5.668151638613524e-06,
+      "loss": 0.1838,
+      "step": 1225
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.43542708343574,
+      "learning_rate": 5.661944145372193e-06,
+      "loss": 0.5795,
+      "step": 1226
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.90662048359304,
+      "learning_rate": 5.655735613488651e-06,
+      "loss": 0.8311,
+      "step": 1227
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.285708539289056,
+      "learning_rate": 5.649526052704567e-06,
+      "loss": 0.3083,
+      "step": 1228
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.88399400818973,
+      "learning_rate": 5.643315472763229e-06,
+      "loss": 0.7807,
+      "step": 1229
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.131182570441121,
+      "learning_rate": 5.637103883409525e-06,
+      "loss": 0.3478,
+      "step": 1230
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.028681665190737,
+      "learning_rate": 5.630891294389923e-06,
+      "loss": 0.1729,
+      "step": 1231
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.094234686982382,
+      "learning_rate": 5.624677715452465e-06,
+      "loss": 0.3123,
+      "step": 1232
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.075392623098997,
+      "learning_rate": 5.61846315634674e-06,
+      "loss": 0.4252,
+      "step": 1233
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.75575204652137,
+      "learning_rate": 5.612247626823878e-06,
+      "loss": 0.3681,
+      "step": 1234
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.226587666971936,
+      "learning_rate": 5.606031136636534e-06,
+      "loss": 0.4348,
+      "step": 1235
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.36630800342092,
+      "learning_rate": 5.599813695538866e-06,
+      "loss": 0.2487,
+      "step": 1236
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.394604044680419,
+      "learning_rate": 5.593595313286526e-06,
+      "loss": 0.4684,
+      "step": 1237
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.211609688619564,
+      "learning_rate": 5.587375999636645e-06,
+      "loss": 0.7363,
+      "step": 1238
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.345492060841666,
+      "learning_rate": 5.581155764347812e-06,
+      "loss": 0.4719,
+      "step": 1239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.1642539085747,
+      "learning_rate": 5.574934617180063e-06,
+      "loss": 0.7433,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.309283606364493,
+      "learning_rate": 5.568712567894866e-06,
+      "loss": 0.3528,
+      "step": 1241
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.890546676334331,
+      "learning_rate": 5.562489626255104e-06,
+      "loss": 0.1785,
+      "step": 1242
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.397279087590594,
+      "learning_rate": 5.5562658020250585e-06,
+      "loss": 0.3245,
+      "step": 1243
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.955598534874809,
+      "learning_rate": 5.550041104970398e-06,
+      "loss": 0.3532,
+      "step": 1244
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.416181959342918,
+      "learning_rate": 5.5438155448581585e-06,
+      "loss": 0.6445,
+      "step": 1245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.298121259550467,
+      "learning_rate": 5.5375891314567335e-06,
+      "loss": 0.5635,
+      "step": 1246
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.076200608362347,
+      "learning_rate": 5.531361874535853e-06,
+      "loss": 0.6507,
+      "step": 1247
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.447359700096728,
+      "learning_rate": 5.52513378386657e-06,
+      "loss": 0.3575,
+      "step": 1248
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.044235082811221,
+      "learning_rate": 5.518904869221245e-06,
+      "loss": 0.636,
+      "step": 1249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.492250460029549,
+      "learning_rate": 5.512675140373537e-06,
+      "loss": 0.479,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.158826711720739,
+      "learning_rate": 5.506444607098375e-06,
+      "loss": 0.4428,
+      "step": 1251
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.617571798449948,
+      "learning_rate": 5.500213279171955e-06,
+      "loss": 0.3688,
+      "step": 1252
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.6385930682075,
+      "learning_rate": 5.49398116637172e-06,
+      "loss": 0.6141,
+      "step": 1253
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.250563848503393,
+      "learning_rate": 5.487748278476342e-06,
+      "loss": 0.4146,
+      "step": 1254
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.461092897390852,
+      "learning_rate": 5.481514625265709e-06,
+      "loss": 0.41,
+      "step": 1255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.739870819877801,
+      "learning_rate": 5.475280216520913e-06,
+      "loss": 0.2799,
+      "step": 1256
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.544946570448424,
+      "learning_rate": 5.469045062024231e-06,
+      "loss": 0.4943,
+      "step": 1257
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.820867127087581,
+      "learning_rate": 5.462809171559104e-06,
+      "loss": 0.5693,
+      "step": 1258
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.987651313238938,
+      "learning_rate": 5.456572554910137e-06,
+      "loss": 0.3383,
+      "step": 1259
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.482458363140129,
+      "learning_rate": 5.450335221863068e-06,
+      "loss": 0.4527,
+      "step": 1260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.863947134438925,
+      "learning_rate": 5.444097182204762e-06,
+      "loss": 0.3829,
+      "step": 1261
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.496103876889583,
+      "learning_rate": 5.437858445723191e-06,
+      "loss": 0.6383,
+      "step": 1262
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.223504160958149,
+      "learning_rate": 5.431619022207422e-06,
+      "loss": 0.1831,
+      "step": 1263
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.721955695443633,
+      "learning_rate": 5.4253789214475975e-06,
+      "loss": 0.5118,
+      "step": 1264
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.6498419373644975,
+      "learning_rate": 5.4191381532349264e-06,
+      "loss": 0.0979,
+      "step": 1265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.366694795759103,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.3488,
+      "step": 1266
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.039340600197683,
+      "learning_rate": 5.406654653621092e-06,
+      "loss": 0.5968,
+      "step": 1267
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.238666755498793,
+      "learning_rate": 5.400411941807516e-06,
+      "loss": 0.6587,
+      "step": 1268
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.62370459318669,
+      "learning_rate": 5.39416860171624e-06,
+      "loss": 0.4017,
+      "step": 1269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.36464449024479,
+      "learning_rate": 5.387924643143553e-06,
+      "loss": 0.2771,
+      "step": 1270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.238466541903357,
+      "learning_rate": 5.381680075886716e-06,
+      "loss": 0.4496,
+      "step": 1271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.267710588304521,
+      "learning_rate": 5.375434909743942e-06,
+      "loss": 0.1127,
+      "step": 1272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.389431132199778,
+      "learning_rate": 5.369189154514388e-06,
+      "loss": 0.1629,
+      "step": 1273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.202237907650858,
+      "learning_rate": 5.362942819998131e-06,
+      "loss": 0.4412,
+      "step": 1274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.386538671981402,
+      "learning_rate": 5.356695915996162e-06,
+      "loss": 0.4716,
+      "step": 1275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.178281491264384,
+      "learning_rate": 5.350448452310362e-06,
+      "loss": 0.441,
+      "step": 1276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.271508230274065,
+      "learning_rate": 5.344200438743489e-06,
+      "loss": 0.6347,
+      "step": 1277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.65792286115022,
+      "learning_rate": 5.337951885099167e-06,
+      "loss": 0.7768,
+      "step": 1278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.354850823934536,
+      "learning_rate": 5.3317028011818635e-06,
+      "loss": 0.1652,
+      "step": 1279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.368976105319669,
+      "learning_rate": 5.3254531967968845e-06,
+      "loss": 0.3494,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.939425240209033,
+      "learning_rate": 5.319203081750348e-06,
+      "loss": 0.366,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.389015924868149,
+      "learning_rate": 5.312952465849173e-06,
+      "loss": 0.6361,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.012163452908762,
+      "learning_rate": 5.306701358901066e-06,
+      "loss": 0.3682,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.057978543733551,
+      "learning_rate": 5.300449770714502e-06,
+      "loss": 0.2205,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.735396366467914,
+      "learning_rate": 5.294197711098716e-06,
+      "loss": 0.1571,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.677917168605147,
+      "learning_rate": 5.287945189863676e-06,
+      "loss": 0.2863,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.71840978925768,
+      "learning_rate": 5.281692216820078e-06,
+      "loss": 0.483,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.963330073247526,
+      "learning_rate": 5.275438801779328e-06,
+      "loss": 0.5816,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.013837117247888,
+      "learning_rate": 5.269184954553522e-06,
+      "loss": 0.2884,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.391382078873157,
+      "learning_rate": 5.262930684955439e-06,
+      "loss": 0.2595,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.692694724933263,
+      "learning_rate": 5.2566760027985166e-06,
+      "loss": 0.4216,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.245523397906988,
+      "learning_rate": 5.2504209178968395e-06,
+      "loss": 0.5905,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.712516117740153,
+      "learning_rate": 5.2441654400651255e-06,
+      "loss": 0.3077,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.868413071559322,
+      "learning_rate": 5.237909579118713e-06,
+      "loss": 0.4202,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.460178044413867,
+      "learning_rate": 5.231653344873534e-06,
+      "loss": 0.5532,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.477536638516883,
+      "learning_rate": 5.225396747146112e-06,
+      "loss": 0.5092,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.552884627282079,
+      "learning_rate": 5.219139795753539e-06,
+      "loss": 0.3273,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.079469328621949,
+      "learning_rate": 5.212882500513462e-06,
+      "loss": 0.5054,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.656740994792022,
+      "learning_rate": 5.206624871244066e-06,
+      "loss": 0.5172,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.232204113348715,
+      "learning_rate": 5.200366917764062e-06,
+      "loss": 0.5453,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.63869785160946,
+      "learning_rate": 5.194108649892672e-06,
+      "loss": 0.5634,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.763872001324867,
+      "learning_rate": 5.187850077449604e-06,
+      "loss": 0.3341,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.50543284223751,
+      "learning_rate": 5.181591210255051e-06,
+      "loss": 0.3248,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.139040734575019,
+      "learning_rate": 5.175332058129664e-06,
+      "loss": 0.5665,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.806911567115078,
+      "learning_rate": 5.169072630894545e-06,
+      "loss": 0.6251,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.664621771690076,
+      "learning_rate": 5.162812938371226e-06,
+      "loss": 0.4744,
+      "step": 1306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.120293544781397,
+      "learning_rate": 5.1565529903816525e-06,
+      "loss": 0.3341,
+      "step": 1307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.698448137158701,
+      "learning_rate": 5.150292796748174e-06,
+      "loss": 0.6075,
+      "step": 1308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.602424003742613,
+      "learning_rate": 5.144032367293525e-06,
+      "loss": 0.3492,
+      "step": 1309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.897013639650924,
+      "learning_rate": 5.137771711840811e-06,
+      "loss": 0.2273,
+      "step": 1310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.636431255684014,
+      "learning_rate": 5.131510840213488e-06,
+      "loss": 0.2054,
+      "step": 1311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.906507067220765,
+      "learning_rate": 5.125249762235357e-06,
+      "loss": 0.4446,
+      "step": 1312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.228086785082699,
+      "learning_rate": 5.118988487730537e-06,
+      "loss": 0.5642,
+      "step": 1313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.96588636894568,
+      "learning_rate": 5.112727026523461e-06,
+      "loss": 0.5011,
+      "step": 1314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.984540501747077,
+      "learning_rate": 5.1064653884388515e-06,
+      "loss": 0.5288,
+      "step": 1315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.106963370473345,
+      "learning_rate": 5.100203583301706e-06,
+      "loss": 0.4026,
+      "step": 1316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.365776994722431,
+      "learning_rate": 5.0939416209372905e-06,
+      "loss": 0.421,
+      "step": 1317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.60662604302703,
+      "learning_rate": 5.087679511171113e-06,
+      "loss": 0.2446,
+      "step": 1318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.136410044136117,
+      "learning_rate": 5.081417263828914e-06,
+      "loss": 0.4327,
+      "step": 1319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.362206318848802,
+      "learning_rate": 5.075154888736653e-06,
+      "loss": 0.6581,
+      "step": 1320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.818972172497633,
+      "learning_rate": 5.068892395720482e-06,
+      "loss": 0.4422,
+      "step": 1321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.77983864607761,
+      "learning_rate": 5.062629794606748e-06,
+      "loss": 0.5796,
+      "step": 1322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.553194061788364,
+      "learning_rate": 5.056367095221959e-06,
+      "loss": 0.4939,
+      "step": 1323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.36098445187415,
+      "learning_rate": 5.050104307392783e-06,
+      "loss": 0.4407,
+      "step": 1324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.607830211646954,
+      "learning_rate": 5.0438414409460234e-06,
+      "loss": 0.2125,
+      "step": 1325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.414172668597068,
+      "learning_rate": 5.03757850570861e-06,
+      "loss": 0.6797,
+      "step": 1326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.961927403808314,
+      "learning_rate": 5.031315511507577e-06,
+      "loss": 0.4754,
+      "step": 1327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.411631192842114,
+      "learning_rate": 5.025052468170054e-06,
+      "loss": 0.3394,
+      "step": 1328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.505833456509811,
+      "learning_rate": 5.018789385523245e-06,
+      "loss": 0.436,
+      "step": 1329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.079623769474123,
+      "learning_rate": 5.01252627339442e-06,
+      "loss": 0.2952,
+      "step": 1330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.89750232469141,
+      "learning_rate": 5.006263141610891e-06,
+      "loss": 0.3637,
+      "step": 1331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.835155720933331,
+      "learning_rate": 5e-06,
+      "loss": 0.4798,
+      "step": 1332
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.726858792895466,
+      "learning_rate": 4.993736858389111e-06,
+      "loss": 0.4039,
+      "step": 1333
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.8902300094478655,
+      "learning_rate": 4.987473726605581e-06,
+      "loss": 0.6062,
+      "step": 1334
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.468065976656604,
+      "learning_rate": 4.981210614476755e-06,
+      "loss": 0.4849,
+      "step": 1335
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.583057121903037,
+      "learning_rate": 4.974947531829947e-06,
+      "loss": 0.6544,
+      "step": 1336
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.520773274294958,
+      "learning_rate": 4.968684488492424e-06,
+      "loss": 0.5929,
+      "step": 1337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.214698428159136,
+      "learning_rate": 4.9624214942913916e-06,
+      "loss": 0.4177,
+      "step": 1338
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.661385349481618,
+      "learning_rate": 4.956158559053977e-06,
+      "loss": 0.3109,
+      "step": 1339
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.830726148539838,
+      "learning_rate": 4.949895692607218e-06,
+      "loss": 0.2069,
+      "step": 1340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.198631573400142,
+      "learning_rate": 4.9436329047780424e-06,
+      "loss": 0.3312,
+      "step": 1341
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.12339407097875,
+      "learning_rate": 4.9373702053932534e-06,
+      "loss": 0.3737,
+      "step": 1342
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.322406724732632,
+      "learning_rate": 4.9311076042795185e-06,
+      "loss": 0.5057,
+      "step": 1343
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.997249659848713,
+      "learning_rate": 4.924845111263349e-06,
+      "loss": 0.222,
+      "step": 1344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.710878830961932,
+      "learning_rate": 4.918582736171086e-06,
+      "loss": 0.3953,
+      "step": 1345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.939963626072583,
+      "learning_rate": 4.912320488828887e-06,
+      "loss": 0.3267,
+      "step": 1346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.713095050199021,
+      "learning_rate": 4.906058379062712e-06,
+      "loss": 0.5043,
+      "step": 1347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.432700899161436,
+      "learning_rate": 4.899796416698296e-06,
+      "loss": 0.6199,
+      "step": 1348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.952335100699747,
+      "learning_rate": 4.893534611561152e-06,
+      "loss": 0.5902,
+      "step": 1349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.422743293699009,
+      "learning_rate": 4.88727297347654e-06,
+      "loss": 0.6464,
+      "step": 1350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.060333160306979,
+      "learning_rate": 4.881011512269464e-06,
+      "loss": 0.3729,
+      "step": 1351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.697857322670755,
+      "learning_rate": 4.874750237764645e-06,
+      "loss": 0.2836,
+      "step": 1352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.807338243519762,
+      "learning_rate": 4.868489159786513e-06,
+      "loss": 0.5494,
+      "step": 1353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.46682851548217,
+      "learning_rate": 4.862228288159191e-06,
+      "loss": 0.5312,
+      "step": 1354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.254989059687242,
+      "learning_rate": 4.8559676327064755e-06,
+      "loss": 0.4864,
+      "step": 1355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.879027519057562,
+      "learning_rate": 4.8497072032518274e-06,
+      "loss": 0.3772,
+      "step": 1356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.694463499844037,
+      "learning_rate": 4.843447009618351e-06,
+      "loss": 0.448,
+      "step": 1357
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.494623850477497,
+      "learning_rate": 4.837187061628777e-06,
+      "loss": 0.2836,
+      "step": 1358
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.380615983036806,
+      "learning_rate": 4.830927369105457e-06,
+      "loss": 0.4596,
+      "step": 1359
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.497293513071835,
+      "learning_rate": 4.824667941870338e-06,
+      "loss": 0.4786,
+      "step": 1360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.37998809392398,
+      "learning_rate": 4.818408789744951e-06,
+      "loss": 0.4061,
+      "step": 1361
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.187407823555922,
+      "learning_rate": 4.8121499225503974e-06,
+      "loss": 0.4638,
+      "step": 1362
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.3404928517062595,
+      "learning_rate": 4.8058913501073294e-06,
+      "loss": 0.4104,
+      "step": 1363
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.947718777980727,
+      "learning_rate": 4.799633082235938e-06,
+      "loss": 0.4093,
+      "step": 1364
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.275181956978114,
+      "learning_rate": 4.793375128755934e-06,
+      "loss": 0.3508,
+      "step": 1365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.310500761600029,
+      "learning_rate": 4.787117499486539e-06,
+      "loss": 0.3985,
+      "step": 1366
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.179850997227886,
+      "learning_rate": 4.7808602042464615e-06,
+      "loss": 0.7093,
+      "step": 1367
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.8676356288193965,
+      "learning_rate": 4.774603252853889e-06,
+      "loss": 0.2136,
+      "step": 1368
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.239105648353771,
+      "learning_rate": 4.768346655126467e-06,
+      "loss": 0.4479,
+      "step": 1369
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.300208465054627,
+      "learning_rate": 4.762090420881289e-06,
+      "loss": 0.2191,
+      "step": 1370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.815865406365322,
+      "learning_rate": 4.755834559934875e-06,
+      "loss": 0.3375,
+      "step": 1371
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.08674782233264,
+      "learning_rate": 4.749579082103163e-06,
+      "loss": 0.3087,
+      "step": 1372
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.329198110859558,
+      "learning_rate": 4.743323997201485e-06,
+      "loss": 0.2046,
+      "step": 1373
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.328173755473545,
+      "learning_rate": 4.737069315044562e-06,
+      "loss": 0.2113,
+      "step": 1374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.414329260860022,
+      "learning_rate": 4.730815045446478e-06,
+      "loss": 0.4238,
+      "step": 1375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.858656287244083,
+      "learning_rate": 4.724561198220672e-06,
+      "loss": 0.5151,
+      "step": 1376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.247506138950083,
+      "learning_rate": 4.718307783179924e-06,
+      "loss": 0.4116,
+      "step": 1377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.37094737742119,
+      "learning_rate": 4.712054810136327e-06,
+      "loss": 0.165,
+      "step": 1378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.834870417106893,
+      "learning_rate": 4.705802288901286e-06,
+      "loss": 0.3921,
+      "step": 1379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.581550362096307,
+      "learning_rate": 4.699550229285499e-06,
+      "loss": 0.7116,
+      "step": 1380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.8823336297334,
+      "learning_rate": 4.693298641098935e-06,
+      "loss": 0.2384,
+      "step": 1381
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.850368862454253,
+      "learning_rate": 4.687047534150829e-06,
+      "loss": 0.3441,
+      "step": 1382
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.1561746966756,
+      "learning_rate": 4.680796918249653e-06,
+      "loss": 0.5618,
+      "step": 1383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.465367107061454,
+      "learning_rate": 4.674546803203116e-06,
+      "loss": 0.2665,
+      "step": 1384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.24036670367002,
+      "learning_rate": 4.668297198818137e-06,
+      "loss": 0.2254,
+      "step": 1385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.888749649350341,
+      "learning_rate": 4.662048114900837e-06,
+      "loss": 0.2381,
+      "step": 1386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.005739747248711,
+      "learning_rate": 4.6557995612565146e-06,
+      "loss": 0.3735,
+      "step": 1387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.927065345896947,
+      "learning_rate": 4.649551547689641e-06,
+      "loss": 0.3318,
+      "step": 1388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.4782731329058,
+      "learning_rate": 4.643304084003839e-06,
+      "loss": 0.4302,
+      "step": 1389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.227956789949822,
+      "learning_rate": 4.6370571800018695e-06,
+      "loss": 0.2982,
+      "step": 1390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.465849439452189,
+      "learning_rate": 4.630810845485613e-06,
+      "loss": 0.1982,
+      "step": 1391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.052897124409321,
+      "learning_rate": 4.624565090256059e-06,
+      "loss": 0.3195,
+      "step": 1392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.104392197303472,
+      "learning_rate": 4.618319924113286e-06,
+      "loss": 0.3024,
+      "step": 1393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.65457325778261,
+      "learning_rate": 4.612075356856447e-06,
+      "loss": 0.5679,
+      "step": 1394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.903561280850305,
+      "learning_rate": 4.60583139828376e-06,
+      "loss": 0.5545,
+      "step": 1395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.164389605098885,
+      "learning_rate": 4.599588058192487e-06,
+      "loss": 0.395,
+      "step": 1396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.787310555553622,
+      "learning_rate": 4.5933453463789105e-06,
+      "loss": 0.684,
+      "step": 1397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.814309071976997,
+      "learning_rate": 4.587103272638339e-06,
+      "loss": 0.4558,
+      "step": 1398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.968936265671517,
+      "learning_rate": 4.580861846765075e-06,
+      "loss": 0.5723,
+      "step": 1399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.481536732814968,
+      "learning_rate": 4.574621078552403e-06,
+      "loss": 0.6593,
+      "step": 1400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.554008974231142,
+      "learning_rate": 4.568380977792581e-06,
+      "loss": 0.5459,
+      "step": 1401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.5218594658441,
+      "learning_rate": 4.562141554276811e-06,
+      "loss": 0.3774,
+      "step": 1402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.299481744261914,
+      "learning_rate": 4.555902817795239e-06,
+      "loss": 0.5704,
+      "step": 1403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.520098309453494,
+      "learning_rate": 4.549664778136933e-06,
+      "loss": 0.3212,
+      "step": 1404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.253954240207406,
+      "learning_rate": 4.543427445089863e-06,
+      "loss": 0.2249,
+      "step": 1405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.80838004891652,
+      "learning_rate": 4.537190828440898e-06,
+      "loss": 0.5372,
+      "step": 1406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.013976154289539,
+      "learning_rate": 4.5309549379757724e-06,
+      "loss": 0.2614,
+      "step": 1407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.864188158354864,
+      "learning_rate": 4.524719783479088e-06,
+      "loss": 0.4668,
+      "step": 1408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.441090609632564,
+      "learning_rate": 4.5184853747342926e-06,
+      "loss": 0.4086,
+      "step": 1409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.270906571202492,
+      "learning_rate": 4.512251721523659e-06,
+      "loss": 0.4493,
+      "step": 1410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.83374667510523,
+      "learning_rate": 4.506018833628281e-06,
+      "loss": 0.174,
+      "step": 1411
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.203927053647208,
+      "learning_rate": 4.499786720828046e-06,
+      "loss": 0.5114,
+      "step": 1412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.956120952229297,
+      "learning_rate": 4.493555392901626e-06,
+      "loss": 0.3631,
+      "step": 1413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.308813570956035,
+      "learning_rate": 4.487324859626465e-06,
+      "loss": 0.3789,
+      "step": 1414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.957319669364114,
+      "learning_rate": 4.481095130778756e-06,
+      "loss": 0.6516,
+      "step": 1415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.032086552804873,
+      "learning_rate": 4.4748662161334335e-06,
+      "loss": 0.3174,
+      "step": 1416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.217213628254339,
+      "learning_rate": 4.4686381254641495e-06,
+      "loss": 0.1802,
+      "step": 1417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.99563701045662,
+      "learning_rate": 4.462410868543268e-06,
+      "loss": 0.4842,
+      "step": 1418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.461106254786237,
+      "learning_rate": 4.456184455141843e-06,
+      "loss": 0.1724,
+      "step": 1419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.490898471511453,
+      "learning_rate": 4.449958895029604e-06,
+      "loss": 0.217,
+      "step": 1420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.507959836820286,
+      "learning_rate": 4.443734197974943e-06,
+      "loss": 0.5924,
+      "step": 1421
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.370459328246875,
+      "learning_rate": 4.437510373744897e-06,
+      "loss": 0.4636,
+      "step": 1422
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.652808890928566,
+      "learning_rate": 4.431287432105134e-06,
+      "loss": 0.4587,
+      "step": 1423
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.219959059730439,
+      "learning_rate": 4.4250653828199366e-06,
+      "loss": 0.5638,
+      "step": 1424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.949657051043806,
+      "learning_rate": 4.418844235652188e-06,
+      "loss": 0.2335,
+      "step": 1425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.565174038152582,
+      "learning_rate": 4.4126240003633565e-06,
+      "loss": 0.5902,
+      "step": 1426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.231857674459197,
+      "learning_rate": 4.4064046867134755e-06,
+      "loss": 0.5303,
+      "step": 1427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.369422919161507,
+      "learning_rate": 4.400186304461136e-06,
+      "loss": 0.3509,
+      "step": 1428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.112650998020145,
+      "learning_rate": 4.393968863363468e-06,
+      "loss": 0.567,
+      "step": 1429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.26157175929687,
+      "learning_rate": 4.387752373176123e-06,
+      "loss": 0.6461,
+      "step": 1430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.578089480155204,
+      "learning_rate": 4.381536843653262e-06,
+      "loss": 0.3201,
+      "step": 1431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.027937985764357,
+      "learning_rate": 4.375322284547536e-06,
+      "loss": 0.5734,
+      "step": 1432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.138502054747653,
+      "learning_rate": 4.369108705610077e-06,
+      "loss": 0.347,
+      "step": 1433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.023147104849702,
+      "learning_rate": 4.362896116590475e-06,
+      "loss": 0.4312,
+      "step": 1434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.483389454402405,
+      "learning_rate": 4.356684527236773e-06,
+      "loss": 0.2847,
+      "step": 1435
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.603180820098618,
+      "learning_rate": 4.3504739472954355e-06,
+      "loss": 0.3755,
+      "step": 1436
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.953785368924668,
+      "learning_rate": 4.344264386511351e-06,
+      "loss": 0.1066,
+      "step": 1437
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.105682536526328,
+      "learning_rate": 4.3380558546278075e-06,
+      "loss": 0.3103,
+      "step": 1438
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.615791358620706,
+      "learning_rate": 4.331848361386478e-06,
+      "loss": 0.6774,
+      "step": 1439
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.547940804233967,
+      "learning_rate": 4.325641916527405e-06,
+      "loss": 0.3297,
+      "step": 1440
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.707168109729409,
+      "learning_rate": 4.319436529788985e-06,
+      "loss": 0.3985,
+      "step": 1441
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.035881917549933,
+      "learning_rate": 4.313232210907959e-06,
+      "loss": 0.267,
+      "step": 1442
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.48709811548802,
+      "learning_rate": 4.3070289696193886e-06,
+      "loss": 0.4499,
+      "step": 1443
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.750092977644064,
+      "learning_rate": 4.3008268156566444e-06,
+      "loss": 0.5444,
+      "step": 1444
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.772602000077848,
+      "learning_rate": 4.294625758751392e-06,
+      "loss": 0.4672,
+      "step": 1445
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.922487061456263,
+      "learning_rate": 4.2884258086335755e-06,
+      "loss": 0.3861,
+      "step": 1446
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.422604897096992,
+      "learning_rate": 4.282226975031402e-06,
+      "loss": 0.5389,
+      "step": 1447
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.836255667655832,
+      "learning_rate": 4.276029267671327e-06,
+      "loss": 0.6166,
+      "step": 1448
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.16883034783219,
+      "learning_rate": 4.269832696278038e-06,
+      "loss": 0.2187,
+      "step": 1449
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.233150478006507,
+      "learning_rate": 4.2636372705744425e-06,
+      "loss": 0.5722,
+      "step": 1450
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.776438995627759,
+      "learning_rate": 4.257443000281648e-06,
+      "loss": 0.1762,
+      "step": 1451
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.388537661288398,
+      "learning_rate": 4.25124989511895e-06,
+      "loss": 0.3017,
+      "step": 1452
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.900170428247057,
+      "learning_rate": 4.245057964803815e-06,
+      "loss": 0.488,
+      "step": 1453
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.572741204487025,
+      "learning_rate": 4.238867219051868e-06,
+      "loss": 0.2977,
+      "step": 1454
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.03214067816434,
+      "learning_rate": 4.232677667576874e-06,
+      "loss": 0.3281,
+      "step": 1455
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.797733458671555,
+      "learning_rate": 4.226489320090723e-06,
+      "loss": 0.296,
+      "step": 1456
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.67114247201264,
+      "learning_rate": 4.220302186303416e-06,
+      "loss": 0.4469,
+      "step": 1457
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.722884956765451,
+      "learning_rate": 4.214116275923051e-06,
+      "loss": 0.3533,
+      "step": 1458
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.23759969750145,
+      "learning_rate": 4.2079315986558056e-06,
+      "loss": 0.306,
+      "step": 1459
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.240877461181231,
+      "learning_rate": 4.201748164205925e-06,
+      "loss": 0.3752,
+      "step": 1460
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.533934466308994,
+      "learning_rate": 4.195565982275701e-06,
+      "loss": 0.3983,
+      "step": 1461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.359922469282836,
+      "learning_rate": 4.1893850625654626e-06,
+      "loss": 0.5183,
+      "step": 1462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.826206707934249,
+      "learning_rate": 4.183205414773555e-06,
+      "loss": 0.1719,
+      "step": 1463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.79453212510032,
+      "learning_rate": 4.17702704859633e-06,
+      "loss": 0.3448,
+      "step": 1464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.052840709558113,
+      "learning_rate": 4.1708499737281305e-06,
+      "loss": 0.2802,
+      "step": 1465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.06894162722573,
+      "learning_rate": 4.1646741998612676e-06,
+      "loss": 0.3188,
+      "step": 1466
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.143184778807555,
+      "learning_rate": 4.1584997366860145e-06,
+      "loss": 0.4284,
+      "step": 1467
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.31205364032092,
+      "learning_rate": 4.1523265938905885e-06,
+      "loss": 0.2171,
+      "step": 1468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.65883192736996,
+      "learning_rate": 4.146154781161133e-06,
+      "loss": 0.5011,
+      "step": 1469
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.853954326842883,
+      "learning_rate": 4.1399843081817085e-06,
+      "loss": 0.078,
+      "step": 1470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.165104577884692,
+      "learning_rate": 4.133815184634269e-06,
+      "loss": 0.4974,
+      "step": 1471
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.55978007080667,
+      "learning_rate": 4.12764742019865e-06,
+      "loss": 0.3668,
+      "step": 1472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.375241538227273,
+      "learning_rate": 4.121481024552561e-06,
+      "loss": 0.605,
+      "step": 1473
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.108593044289524,
+      "learning_rate": 4.115316007371557e-06,
+      "loss": 0.4421,
+      "step": 1474
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.46808693583527,
+      "learning_rate": 4.109152378329036e-06,
+      "loss": 0.4242,
+      "step": 1475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.019880172300454,
+      "learning_rate": 4.1029901470962105e-06,
+      "loss": 0.3076,
+      "step": 1476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.8259720568784905,
+      "learning_rate": 4.0968293233421065e-06,
+      "loss": 0.2156,
+      "step": 1477
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.085211686897715,
+      "learning_rate": 4.090669916733539e-06,
+      "loss": 0.3175,
+      "step": 1478
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.190554524226714,
+      "learning_rate": 4.0845119369350995e-06,
+      "loss": 0.4506,
+      "step": 1479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.7355418886817,
+      "learning_rate": 4.078355393609139e-06,
+      "loss": 0.4576,
+      "step": 1480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.424056340708123,
+      "learning_rate": 4.0722002964157585e-06,
+      "loss": 0.411,
+      "step": 1481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.240987874419247,
+      "learning_rate": 4.066046655012786e-06,
+      "loss": 0.2843,
+      "step": 1482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.631980906253238,
+      "learning_rate": 4.059894479055767e-06,
+      "loss": 0.3539,
+      "step": 1483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.792641896876604,
+      "learning_rate": 4.053743778197951e-06,
+      "loss": 0.2322,
+      "step": 1484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.09443173888019,
+      "learning_rate": 4.047594562090264e-06,
+      "loss": 0.2493,
+      "step": 1485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.071455445281066,
+      "learning_rate": 4.041446840381309e-06,
+      "loss": 0.4879,
+      "step": 1486
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.53809797532003,
+      "learning_rate": 4.0353006227173455e-06,
+      "loss": 0.295,
+      "step": 1487
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.774343045127807,
+      "learning_rate": 4.029155918742268e-06,
+      "loss": 0.3777,
+      "step": 1488
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.76654045378418,
+      "learning_rate": 4.023012738097598e-06,
+      "loss": 0.5351,
+      "step": 1489
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.50974587340954,
+      "learning_rate": 4.016871090422471e-06,
+      "loss": 0.3448,
+      "step": 1490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.17544656573508,
+      "learning_rate": 4.010730985353609e-06,
+      "loss": 0.5041,
+      "step": 1491
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.302002899664561,
+      "learning_rate": 4.00459243252532e-06,
+      "loss": 0.3991,
+      "step": 1492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.839563884319015,
+      "learning_rate": 3.998455441569473e-06,
+      "loss": 0.5282,
+      "step": 1493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.196359711003553,
+      "learning_rate": 3.992320022115492e-06,
+      "loss": 0.4598,
+      "step": 1494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.793138773314729,
+      "learning_rate": 3.986186183790323e-06,
+      "loss": 0.3006,
+      "step": 1495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.567965732427867,
+      "learning_rate": 3.980053936218444e-06,
+      "loss": 0.3216,
+      "step": 1496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.965367724714955,
+      "learning_rate": 3.973923289021829e-06,
+      "loss": 0.4594,
+      "step": 1497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.100136351390054,
+      "learning_rate": 3.9677942518199465e-06,
+      "loss": 0.4697,
+      "step": 1498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.394205898472151,
+      "learning_rate": 3.961666834229734e-06,
+      "loss": 0.4152,
+      "step": 1499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.777668618012456,
+      "learning_rate": 3.955541045865591e-06,
+      "loss": 0.3972,
+      "step": 1500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.670197496213881,
+      "learning_rate": 3.949416896339357e-06,
+      "loss": 0.3973,
+      "step": 1501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.460873873972286,
+      "learning_rate": 3.943294395260305e-06,
+      "loss": 0.2055,
+      "step": 1502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.246746515497265,
+      "learning_rate": 3.937173552235117e-06,
+      "loss": 0.4135,
+      "step": 1503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.809158567158315,
+      "learning_rate": 3.931054376867877e-06,
+      "loss": 0.4903,
+      "step": 1504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.444135843793509,
+      "learning_rate": 3.924936878760048e-06,
+      "loss": 0.3891,
+      "step": 1505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.55083440198172,
+      "learning_rate": 3.918821067510464e-06,
+      "loss": 0.1857,
+      "step": 1506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.619827346507286,
+      "learning_rate": 3.912706952715314e-06,
+      "loss": 0.3826,
+      "step": 1507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.862576928371584,
+      "learning_rate": 3.906594543968122e-06,
+      "loss": 0.2016,
+      "step": 1508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.257976302927651,
+      "learning_rate": 3.900483850859735e-06,
+      "loss": 0.1452,
+      "step": 1509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.932684623140828,
+      "learning_rate": 3.89437488297831e-06,
+      "loss": 0.5148,
+      "step": 1510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.199340324759861,
+      "learning_rate": 3.888267649909296e-06,
+      "loss": 0.4178,
+      "step": 1511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.002406952247696,
+      "learning_rate": 3.882162161235421e-06,
+      "loss": 0.2315,
+      "step": 1512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.949353177261028,
+      "learning_rate": 3.876058426536674e-06,
+      "loss": 0.565,
+      "step": 1513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.366304965727169,
+      "learning_rate": 3.869956455390295e-06,
+      "loss": 0.5659,
+      "step": 1514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.980697830623451,
+      "learning_rate": 3.86385625737075e-06,
+      "loss": 0.3547,
+      "step": 1515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.185294538060226,
+      "learning_rate": 3.857757842049732e-06,
+      "loss": 0.2116,
+      "step": 1516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.159558352582156,
+      "learning_rate": 3.851661218996129e-06,
+      "loss": 0.5369,
+      "step": 1517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.351451968191936,
+      "learning_rate": 3.845566397776022e-06,
+      "loss": 0.3477,
+      "step": 1518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.755016674015226,
+      "learning_rate": 3.839473387952662e-06,
+      "loss": 0.3331,
+      "step": 1519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.972161300362636,
+      "learning_rate": 3.833382199086459e-06,
+      "loss": 0.6077,
+      "step": 1520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.21673525054098,
+      "learning_rate": 3.827292840734966e-06,
+      "loss": 0.3345,
+      "step": 1521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.298327268072868,
+      "learning_rate": 3.821205322452863e-06,
+      "loss": 0.2459,
+      "step": 1522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.101886992877931,
+      "learning_rate": 3.815119653791943e-06,
+      "loss": 0.3275,
+      "step": 1523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.545711364537286,
+      "learning_rate": 3.8090358443010993e-06,
+      "loss": 0.4035,
+      "step": 1524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.205937249863844,
+      "learning_rate": 3.802953903526301e-06,
+      "loss": 0.2847,
+      "step": 1525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.893491311533815,
+      "learning_rate": 3.796873841010591e-06,
+      "loss": 0.3367,
+      "step": 1526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.091163447432562,
+      "learning_rate": 3.7907956662940644e-06,
+      "loss": 0.1285,
+      "step": 1527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.415142056640292,
+      "learning_rate": 3.784719388913853e-06,
+      "loss": 0.3809,
+      "step": 1528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.405644020799329,
+      "learning_rate": 3.778645018404112e-06,
+      "loss": 0.5295,
+      "step": 1529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.305341847771698,
+      "learning_rate": 3.7725725642960047e-06,
+      "loss": 0.5006,
+      "step": 1530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.733778502031718,
+      "learning_rate": 3.7665020361176872e-06,
+      "loss": 0.3133,
+      "step": 1531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.270239019133196,
+      "learning_rate": 3.760433443394294e-06,
+      "loss": 0.5234,
+      "step": 1532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.302232062700625,
+      "learning_rate": 3.754366795647923e-06,
+      "loss": 0.4634,
+      "step": 1533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.756772289379626,
+      "learning_rate": 3.748302102397618e-06,
+      "loss": 0.5173,
+      "step": 1534
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.286816249754958,
+      "learning_rate": 3.7422393731593586e-06,
+      "loss": 0.2918,
+      "step": 1535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.320213893995408,
+      "learning_rate": 3.7361786174460414e-06,
+      "loss": 0.3342,
+      "step": 1536
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.58588199023222,
+      "learning_rate": 3.730119844767468e-06,
+      "loss": 0.2265,
+      "step": 1537
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.230826940959176,
+      "learning_rate": 3.7240630646303262e-06,
+      "loss": 0.5076,
+      "step": 1538
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.149741851614046,
+      "learning_rate": 3.718008286538179e-06,
+      "loss": 0.4518,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.337433557278371,
+      "learning_rate": 3.711955519991447e-06,
+      "loss": 0.2621,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.12425161465035,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.2856,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.265869506151661,
+      "learning_rate": 3.6998560595201188e-06,
+      "loss": 0.2602,
+      "step": 1542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.921833060783499,
+      "learning_rate": 3.6938093845805257e-06,
+      "loss": 0.6169,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.91010677893322,
+      "learning_rate": 3.687764759156318e-06,
+      "loss": 0.4572,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.793991473233143,
+      "learning_rate": 3.68172219273199e-06,
+      "loss": 0.3478,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.248907484227386,
+      "learning_rate": 3.675681694788801e-06,
+      "loss": 0.4446,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.027999015037292,
+      "learning_rate": 3.669643274804765e-06,
+      "loss": 0.319,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.783079025661976,
+      "learning_rate": 3.6636069422546363e-06,
+      "loss": 0.4412,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.804403348206463,
+      "learning_rate": 3.6575727066098936e-06,
+      "loss": 0.3425,
+      "step": 1549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.041023739321213,
+      "learning_rate": 3.6515405773387257e-06,
+      "loss": 0.4684,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.356848022284,
+      "learning_rate": 3.645510563906014e-06,
+      "loss": 0.2753,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.18019850984098,
+      "learning_rate": 3.639482675773324e-06,
+      "loss": 0.2346,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.977807755199736,
+      "learning_rate": 3.633456922398887e-06,
+      "loss": 0.4696,
+      "step": 1553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.09503479463202,
+      "learning_rate": 3.627433313237576e-06,
+      "loss": 0.5991,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.857931461117108,
+      "learning_rate": 3.621411857740908e-06,
+      "loss": 0.2329,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.335624927411668,
+      "learning_rate": 3.6153925653570186e-06,
+      "loss": 0.4718,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.116727658967495,
+      "learning_rate": 3.6093754455306495e-06,
+      "loss": 0.2129,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.943729329202197,
+      "learning_rate": 3.603360507703133e-06,
+      "loss": 0.5613,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.401357877814739,
+      "learning_rate": 3.597347761312377e-06,
+      "loss": 0.6705,
+      "step": 1559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.186950179813447,
+      "learning_rate": 3.5913372157928515e-06,
+      "loss": 0.4844,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.043592937565245,
+      "learning_rate": 3.585328880575574e-06,
+      "loss": 0.5602,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.393854848270676,
+      "learning_rate": 3.5793227650880928e-06,
+      "loss": 0.6851,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.124505739613044,
+      "learning_rate": 3.573318878754475e-06,
+      "loss": 0.4741,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.936558743988889,
+      "learning_rate": 3.5673172309952846e-06,
+      "loss": 0.4563,
+      "step": 1564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.948154076656229,
+      "learning_rate": 3.5613178312275796e-06,
+      "loss": 0.339,
+      "step": 1565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.831809212203451,
+      "learning_rate": 3.555320688864889e-06,
+      "loss": 0.4929,
+      "step": 1566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.830123104784514,
+      "learning_rate": 3.549325813317197e-06,
+      "loss": 0.3182,
+      "step": 1567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.965317689032,
+      "learning_rate": 3.5433332139909342e-06,
+      "loss": 0.2309,
+      "step": 1568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.337455273809522,
+      "learning_rate": 3.5373429002889583e-06,
+      "loss": 0.442,
+      "step": 1569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.587308372194604,
+      "learning_rate": 3.531354881610539e-06,
+      "loss": 0.301,
+      "step": 1570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.975705928321097,
+      "learning_rate": 3.525369167351349e-06,
+      "loss": 0.3334,
+      "step": 1571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.104802327956442,
+      "learning_rate": 3.519385766903442e-06,
+      "loss": 0.3899,
+      "step": 1572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 17.502606281894145,
+      "learning_rate": 3.5134046896552436e-06,
+      "loss": 0.5094,
+      "step": 1573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.138612858122334,
+      "learning_rate": 3.507425944991529e-06,
+      "loss": 0.5474,
+      "step": 1574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.259431713241304,
+      "learning_rate": 3.501449542293418e-06,
+      "loss": 0.3834,
+      "step": 1575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.649132007733036,
+      "learning_rate": 3.495475490938355e-06,
+      "loss": 0.2045,
+      "step": 1576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.043397832848587,
+      "learning_rate": 3.4895038003000953e-06,
+      "loss": 0.6478,
+      "step": 1577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.087415831950558,
+      "learning_rate": 3.483534479748688e-06,
+      "loss": 0.4931,
+      "step": 1578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.629527976370307,
+      "learning_rate": 3.477567538650466e-06,
+      "loss": 0.5256,
+      "step": 1579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.052862182130642,
+      "learning_rate": 3.4716029863680256e-06,
+      "loss": 0.3948,
+      "step": 1580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.969731427538665,
+      "learning_rate": 3.4656408322602168e-06,
+      "loss": 0.3415,
+      "step": 1581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.461442625969225,
+      "learning_rate": 3.4596810856821304e-06,
+      "loss": 0.3641,
+      "step": 1582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.407047791308006,
+      "learning_rate": 3.45372375598507e-06,
+      "loss": 0.3922,
+      "step": 1583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.182291428869348,
+      "learning_rate": 3.447768852516554e-06,
+      "loss": 0.3186,
+      "step": 1584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.191859371936374,
+      "learning_rate": 3.4418163846202945e-06,
+      "loss": 0.2983,
+      "step": 1585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.811295645116578,
+      "learning_rate": 3.4358663616361775e-06,
+      "loss": 0.6622,
+      "step": 1586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.245353921715301,
+      "learning_rate": 3.429918792900257e-06,
+      "loss": 0.5403,
+      "step": 1587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.34172195184723,
+      "learning_rate": 3.4239736877447327e-06,
+      "loss": 0.3337,
+      "step": 1588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.413007888198166,
+      "learning_rate": 3.4180310554979413e-06,
+      "loss": 0.4306,
+      "step": 1589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.578314454047488,
+      "learning_rate": 3.4120909054843375e-06,
+      "loss": 0.345,
+      "step": 1590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.990773666904295,
+      "learning_rate": 3.406153247024483e-06,
+      "loss": 0.2356,
+      "step": 1591
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.244024209956159,
+      "learning_rate": 3.400218089435029e-06,
+      "loss": 0.2931,
+      "step": 1592
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.915325298807911,
+      "learning_rate": 3.3942854420287006e-06,
+      "loss": 0.7604,
+      "step": 1593
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.179140723436092,
+      "learning_rate": 3.3883553141142884e-06,
+      "loss": 0.4793,
+      "step": 1594
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.955494308264566,
+      "learning_rate": 3.3824277149966265e-06,
+      "loss": 0.2023,
+      "step": 1595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.443732349429494,
+      "learning_rate": 3.3765026539765832e-06,
+      "loss": 0.4244,
+      "step": 1596
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.464585456982883,
+      "learning_rate": 3.3705801403510417e-06,
+      "loss": 0.2999,
+      "step": 1597
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.087301989984667,
+      "learning_rate": 3.3646601834128924e-06,
+      "loss": 0.5977,
+      "step": 1598
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.197112375750976,
+      "learning_rate": 3.3587427924510086e-06,
+      "loss": 0.2904,
+      "step": 1599
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.659122380046885,
+      "learning_rate": 3.352827976750242e-06,
+      "loss": 0.7971,
+      "step": 1600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.074025227252707,
+      "learning_rate": 3.3469157455914013e-06,
+      "loss": 0.3897,
+      "step": 1601
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.722937627733435,
+      "learning_rate": 3.3410061082512422e-06,
+      "loss": 0.338,
+      "step": 1602
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.769774461039551,
+      "learning_rate": 3.335099074002445e-06,
+      "loss": 0.258,
+      "step": 1603
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.426527834447489,
+      "learning_rate": 3.32919465211361e-06,
+      "loss": 0.3433,
+      "step": 1604
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.575049082254282,
+      "learning_rate": 3.323292851849238e-06,
+      "loss": 0.4009,
+      "step": 1605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.744625360656718,
+      "learning_rate": 3.3173936824697174e-06,
+      "loss": 0.2564,
+      "step": 1606
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.472736135137952,
+      "learning_rate": 3.3114971532313058e-06,
+      "loss": 0.6311,
+      "step": 1607
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.870304315651554,
+      "learning_rate": 3.3056032733861188e-06,
+      "loss": 0.566,
+      "step": 1608
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.908317533943925,
+      "learning_rate": 3.2997120521821168e-06,
+      "loss": 0.4134,
+      "step": 1609
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.6700815786929875,
+      "learning_rate": 3.293823498863087e-06,
+      "loss": 0.1716,
+      "step": 1610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.431206141138786,
+      "learning_rate": 3.28793762266863e-06,
+      "loss": 0.245,
+      "step": 1611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.697316420917996,
+      "learning_rate": 3.2820544328341485e-06,
+      "loss": 0.3127,
+      "step": 1612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.721946094041615,
+      "learning_rate": 3.2761739385908264e-06,
+      "loss": 0.1812,
+      "step": 1613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.284809104665317,
+      "learning_rate": 3.2702961491656197e-06,
+      "loss": 0.1136,
+      "step": 1614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.4671180616593,
+      "learning_rate": 3.264421073781241e-06,
+      "loss": 0.3169,
+      "step": 1615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.2007809438331725,
+      "learning_rate": 3.258548721656144e-06,
+      "loss": 0.1195,
+      "step": 1616
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.495468914926477,
+      "learning_rate": 3.252679102004509e-06,
+      "loss": 0.4468,
+      "step": 1617
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.473848853690537,
+      "learning_rate": 3.2468122240362287e-06,
+      "loss": 0.5377,
+      "step": 1618
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.632406421275894,
+      "learning_rate": 3.240948096956894e-06,
+      "loss": 0.3693,
+      "step": 1619
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.717966512389236,
+      "learning_rate": 3.2350867299677802e-06,
+      "loss": 0.5329,
+      "step": 1620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.695028997568485,
+      "learning_rate": 3.2292281322658315e-06,
+      "loss": 0.7197,
+      "step": 1621
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.268382284693438,
+      "learning_rate": 3.223372313043647e-06,
+      "loss": 0.4927,
+      "step": 1622
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.108047030135138,
+      "learning_rate": 3.2175192814894627e-06,
+      "loss": 0.414,
+      "step": 1623
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.453573947018524,
+      "learning_rate": 3.211669046787147e-06,
+      "loss": 0.1539,
+      "step": 1624
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.394423554512505,
+      "learning_rate": 3.205821618116175e-06,
+      "loss": 0.4331,
+      "step": 1625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.083439020680638,
+      "learning_rate": 3.1999770046516198e-06,
+      "loss": 0.3819,
+      "step": 1626
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.408213304555627,
+      "learning_rate": 3.194135215564139e-06,
+      "loss": 0.2884,
+      "step": 1627
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.692629450093069,
+      "learning_rate": 3.188296260019956e-06,
+      "loss": 0.5875,
+      "step": 1628
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.166271048927568,
+      "learning_rate": 3.1824601471808504e-06,
+      "loss": 0.2616,
+      "step": 1629
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.641979517378259,
+      "learning_rate": 3.1766268862041406e-06,
+      "loss": 0.255,
+      "step": 1630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.960285388550341,
+      "learning_rate": 3.17079648624267e-06,
+      "loss": 0.1974,
+      "step": 1631
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.928137697793344,
+      "learning_rate": 3.164968956444791e-06,
+      "loss": 0.2478,
+      "step": 1632
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.517186195307268,
+      "learning_rate": 3.159144305954356e-06,
+      "loss": 0.3773,
+      "step": 1633
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.6161500529936,
+      "learning_rate": 3.1533225439106965e-06,
+      "loss": 0.1728,
+      "step": 1634
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.294433649466768,
+      "learning_rate": 3.1475036794486147e-06,
+      "loss": 0.4297,
+      "step": 1635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.246257502885474,
+      "learning_rate": 3.141687721698363e-06,
+      "loss": 0.1822,
+      "step": 1636
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.666885137693384,
+      "learning_rate": 3.1358746797856367e-06,
+      "loss": 0.5841,
+      "step": 1637
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.702152982858767,
+      "learning_rate": 3.130064562831553e-06,
+      "loss": 0.2529,
+      "step": 1638
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.840343916218031,
+      "learning_rate": 3.1242573799526397e-06,
+      "loss": 0.4089,
+      "step": 1639
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.622747304377013,
+      "learning_rate": 3.118453140260823e-06,
+      "loss": 0.3432,
+      "step": 1640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.52812433271162,
+      "learning_rate": 3.1126518528634096e-06,
+      "loss": 0.6193,
+      "step": 1641
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.335152238857662,
+      "learning_rate": 3.106853526863073e-06,
+      "loss": 0.4281,
+      "step": 1642
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.194563247433896,
+      "learning_rate": 3.1010581713578403e-06,
+      "loss": 0.3576,
+      "step": 1643
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.233668378188844,
+      "learning_rate": 3.0952657954410792e-06,
+      "loss": 0.4275,
+      "step": 1644
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.618414394131295,
+      "learning_rate": 3.0894764082014805e-06,
+      "loss": 0.1185,
+      "step": 1645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.61362579452708,
+      "learning_rate": 3.0836900187230475e-06,
+      "loss": 0.4859,
+      "step": 1646
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.041858342544652,
+      "learning_rate": 3.0779066360850774e-06,
+      "loss": 0.5073,
+      "step": 1647
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.85333330495757,
+      "learning_rate": 3.072126269362151e-06,
+      "loss": 0.4011,
+      "step": 1648
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.96517673296251,
+      "learning_rate": 3.0663489276241166e-06,
+      "loss": 0.6625,
+      "step": 1649
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.957799731386073,
+      "learning_rate": 3.0605746199360755e-06,
+      "loss": 0.457,
+      "step": 1650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.789601295416406,
+      "learning_rate": 3.0548033553583707e-06,
+      "loss": 0.4608,
+      "step": 1651
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.514906657360216,
+      "learning_rate": 3.049035142946565e-06,
+      "loss": 0.2823,
+      "step": 1652
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.963480493966543,
+      "learning_rate": 3.0432699917514375e-06,
+      "loss": 0.2616,
+      "step": 1653
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.054577287711652,
+      "learning_rate": 3.0375079108189613e-06,
+      "loss": 0.1916,
+      "step": 1654
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.900918236170881,
+      "learning_rate": 3.0317489091902936e-06,
+      "loss": 0.3597,
+      "step": 1655
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.66376583988614,
+      "learning_rate": 3.0259929959017585e-06,
+      "loss": 0.0981,
+      "step": 1656
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.103308096919323,
+      "learning_rate": 3.0202401799848347e-06,
+      "loss": 0.5412,
+      "step": 1657
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.682283321442002,
+      "learning_rate": 3.0144904704661413e-06,
+      "loss": 0.1853,
+      "step": 1658
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.639228380677377,
+      "learning_rate": 3.0087438763674226e-06,
+      "loss": 0.4018,
+      "step": 1659
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.298058766584662,
+      "learning_rate": 3.003000406705535e-06,
+      "loss": 0.2208,
+      "step": 1660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.459254325676522,
+      "learning_rate": 2.9972600704924348e-06,
+      "loss": 0.4575,
+      "step": 1661
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.282914391241533,
+      "learning_rate": 2.991522876735154e-06,
+      "loss": 0.4605,
+      "step": 1662
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.690718392665303,
+      "learning_rate": 2.985788834435802e-06,
+      "loss": 0.6466,
+      "step": 1663
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.966338650496088,
+      "learning_rate": 2.9800579525915394e-06,
+      "loss": 0.6171,
+      "step": 1664
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.614214411187506,
+      "learning_rate": 2.9743302401945707e-06,
+      "loss": 0.4833,
+      "step": 1665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.611882006333737,
+      "learning_rate": 2.9686057062321226e-06,
+      "loss": 0.3275,
+      "step": 1666
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.489573951446634,
+      "learning_rate": 2.9628843596864386e-06,
+      "loss": 0.3562,
+      "step": 1667
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.571218692477077,
+      "learning_rate": 2.95716620953476e-06,
+      "loss": 0.4977,
+      "step": 1668
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.41431386115959,
+      "learning_rate": 2.9514512647493123e-06,
+      "loss": 0.2026,
+      "step": 1669
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.2513828973760575,
+      "learning_rate": 2.9457395342972904e-06,
+      "loss": 0.3042,
+      "step": 1670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.323609023840471,
+      "learning_rate": 2.940031027140848e-06,
+      "loss": 0.2988,
+      "step": 1671
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.287655108668993,
+      "learning_rate": 2.9343257522370784e-06,
+      "loss": 0.3218,
+      "step": 1672
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.939183034829941,
+      "learning_rate": 2.928623718538006e-06,
+      "loss": 0.3103,
+      "step": 1673
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.091487080478307,
+      "learning_rate": 2.9229249349905686e-06,
+      "loss": 0.5218,
+      "step": 1674
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.159964049004142,
+      "learning_rate": 2.917229410536604e-06,
+      "loss": 0.5938,
+      "step": 1675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.540020475432147,
+      "learning_rate": 2.9115371541128344e-06,
+      "loss": 0.4675,
+      "step": 1676
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.904924382153865,
+      "learning_rate": 2.905848174650856e-06,
+      "loss": 0.2307,
+      "step": 1677
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.901576147824272,
+      "learning_rate": 2.900162481077126e-06,
+      "loss": 0.3957,
+      "step": 1678
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.109070088247321,
+      "learning_rate": 2.8944800823129413e-06,
+      "loss": 0.3086,
+      "step": 1679
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.583488942734046,
+      "learning_rate": 2.8888009872744332e-06,
+      "loss": 0.5503,
+      "step": 1680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.090375880645535,
+      "learning_rate": 2.883125204872542e-06,
+      "loss": 0.3447,
+      "step": 1681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.020180125441893,
+      "learning_rate": 2.8774527440130173e-06,
+      "loss": 0.2616,
+      "step": 1682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.52212909238047,
+      "learning_rate": 2.8717836135963937e-06,
+      "loss": 0.6226,
+      "step": 1683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.446725605740701,
+      "learning_rate": 2.866117822517982e-06,
+      "loss": 0.5342,
+      "step": 1684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.691842162513392,
+      "learning_rate": 2.860455379667851e-06,
+      "loss": 0.1593,
+      "step": 1685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.256324967659848,
+      "learning_rate": 2.8547962939308187e-06,
+      "loss": 0.5537,
+      "step": 1686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.67606120901613,
+      "learning_rate": 2.8491405741864323e-06,
+      "loss": 0.4101,
+      "step": 1687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.651359497463278,
+      "learning_rate": 2.8434882293089607e-06,
+      "loss": 0.3817,
+      "step": 1688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.903851470810701,
+      "learning_rate": 2.837839268167373e-06,
+      "loss": 0.2284,
+      "step": 1689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.5976503384416985,
+      "learning_rate": 2.8321936996253368e-06,
+      "loss": 0.4001,
+      "step": 1690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.226389935475833,
+      "learning_rate": 2.8265515325411856e-06,
+      "loss": 0.4865,
+      "step": 1691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.87208109505243,
+      "learning_rate": 2.8209127757679246e-06,
+      "loss": 0.1994,
+      "step": 1692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.08400673021425,
+      "learning_rate": 2.8152774381532033e-06,
+      "loss": 0.5544,
+      "step": 1693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.215335249308133,
+      "learning_rate": 2.8096455285393094e-06,
+      "loss": 0.2516,
+      "step": 1694
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.064354693557135,
+      "learning_rate": 2.804017055763149e-06,
+      "loss": 0.5684,
+      "step": 1695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.75353610297087,
+      "learning_rate": 2.798392028656237e-06,
+      "loss": 0.3961,
+      "step": 1696
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.551481866519772,
+      "learning_rate": 2.792770456044683e-06,
+      "loss": 0.5098,
+      "step": 1697
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.719371556083118,
+      "learning_rate": 2.787152346749173e-06,
+      "loss": 0.3072,
+      "step": 1698
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.623744637789311,
+      "learning_rate": 2.7815377095849612e-06,
+      "loss": 0.1459,
+      "step": 1699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.946144734078107,
+      "learning_rate": 2.775926553361855e-06,
+      "loss": 0.5876,
+      "step": 1700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.719721317483218,
+      "learning_rate": 2.7703188868841936e-06,
+      "loss": 0.5188,
+      "step": 1701
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.871624232294797,
+      "learning_rate": 2.7647147189508485e-06,
+      "loss": 0.2969,
+      "step": 1702
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.598797471970249,
+      "learning_rate": 2.7591140583551956e-06,
+      "loss": 0.4102,
+      "step": 1703
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.343006398353749,
+      "learning_rate": 2.7535169138851124e-06,
+      "loss": 0.1691,
+      "step": 1704
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.608714238845527,
+      "learning_rate": 2.7479232943229567e-06,
+      "loss": 0.2172,
+      "step": 1705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.44229405798066,
+      "learning_rate": 2.7423332084455543e-06,
+      "loss": 0.2869,
+      "step": 1706
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.153404915884616,
+      "learning_rate": 2.7367466650241868e-06,
+      "loss": 0.582,
+      "step": 1707
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.010059416597672,
+      "learning_rate": 2.731163672824583e-06,
+      "loss": 0.2708,
+      "step": 1708
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.970824107912996,
+      "learning_rate": 2.7255842406068917e-06,
+      "loss": 0.6092,
+      "step": 1709
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.810520879525855,
+      "learning_rate": 2.720008377125682e-06,
+      "loss": 0.4213,
+      "step": 1710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.856824107801078,
+      "learning_rate": 2.7144360911299158e-06,
+      "loss": 0.29,
+      "step": 1711
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.068889628103712,
+      "learning_rate": 2.708867391362948e-06,
+      "loss": 0.429,
+      "step": 1712
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.948929626931451,
+      "learning_rate": 2.703302286562506e-06,
+      "loss": 0.2868,
+      "step": 1713
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.549804557942514,
+      "learning_rate": 2.697740785460675e-06,
+      "loss": 0.3541,
+      "step": 1714
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.105426346557273,
+      "learning_rate": 2.692182896783885e-06,
+      "loss": 0.6247,
+      "step": 1715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.816871600686308,
+      "learning_rate": 2.686628629252899e-06,
+      "loss": 0.2881,
+      "step": 1716
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.844380598405643,
+      "learning_rate": 2.681077991582797e-06,
+      "loss": 0.4043,
+      "step": 1717
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.852315796446668,
+      "learning_rate": 2.6755309924829657e-06,
+      "loss": 0.245,
+      "step": 1718
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.598332983212499,
+      "learning_rate": 2.6699876406570823e-06,
+      "loss": 0.2594,
+      "step": 1719
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.504100651406315,
+      "learning_rate": 2.664447944803097e-06,
+      "loss": 0.2613,
+      "step": 1720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.778141444066705,
+      "learning_rate": 2.6589119136132273e-06,
+      "loss": 0.646,
+      "step": 1721
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.548171840811458,
+      "learning_rate": 2.6533795557739407e-06,
+      "loss": 0.514,
+      "step": 1722
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.085186585897736,
+      "learning_rate": 2.6478508799659393e-06,
+      "loss": 0.4211,
+      "step": 1723
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.780172557327576,
+      "learning_rate": 2.6423258948641483e-06,
+      "loss": 0.4869,
+      "step": 1724
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.069436493976264,
+      "learning_rate": 2.636804609137703e-06,
+      "loss": 0.3859,
+      "step": 1725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.00911941519549,
+      "learning_rate": 2.6312870314499335e-06,
+      "loss": 0.3792,
+      "step": 1726
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.79971225852881,
+      "learning_rate": 2.62577317045835e-06,
+      "loss": 0.657,
+      "step": 1727
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.003731155053738,
+      "learning_rate": 2.6202630348146323e-06,
+      "loss": 0.4915,
+      "step": 1728
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.320920928099869,
+      "learning_rate": 2.614756633164618e-06,
+      "loss": 0.256,
+      "step": 1729
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.764098673464994,
+      "learning_rate": 2.609253974148278e-06,
+      "loss": 0.2432,
+      "step": 1730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.193433738199325,
+      "learning_rate": 2.603755066399718e-06,
+      "loss": 0.4881,
+      "step": 1731
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.008193308842452,
+      "learning_rate": 2.5982599185471535e-06,
+      "loss": 0.4131,
+      "step": 1732
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.568370511682497,
+      "learning_rate": 2.5927685392129033e-06,
+      "loss": 0.2901,
+      "step": 1733
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.035846053444035,
+      "learning_rate": 2.5872809370133704e-06,
+      "loss": 0.2902,
+      "step": 1734
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.646009712954543,
+      "learning_rate": 2.5817971205590343e-06,
+      "loss": 0.4581,
+      "step": 1735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.619488023781592,
+      "learning_rate": 2.5763170984544304e-06,
+      "loss": 0.5306,
+      "step": 1736
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.71691827477011,
+      "learning_rate": 2.5708408792981443e-06,
+      "loss": 0.3576,
+      "step": 1737
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.308830270714704,
+      "learning_rate": 2.5653684716827904e-06,
+      "loss": 0.5276,
+      "step": 1738
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.972639108132628,
+      "learning_rate": 2.5598998841950105e-06,
+      "loss": 0.5737,
+      "step": 1739
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.19596156352142,
+      "learning_rate": 2.5544351254154407e-06,
+      "loss": 0.2414,
+      "step": 1740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.058113594367686,
+      "learning_rate": 2.5489742039187184e-06,
+      "loss": 0.3432,
+      "step": 1741
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.044073937349259,
+      "learning_rate": 2.5435171282734563e-06,
+      "loss": 0.5406,
+      "step": 1742
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.062007020173741,
+      "learning_rate": 2.5380639070422343e-06,
+      "loss": 0.1837,
+      "step": 1743
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.212983039168492,
+      "learning_rate": 2.5326145487815822e-06,
+      "loss": 0.2607,
+      "step": 1744
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.20596641395671,
+      "learning_rate": 2.527169062041972e-06,
+      "loss": 0.3355,
+      "step": 1745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.985160973854269,
+      "learning_rate": 2.5217274553677975e-06,
+      "loss": 0.374,
+      "step": 1746
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.609530540238346,
+      "learning_rate": 2.516289737297366e-06,
+      "loss": 0.9643,
+      "step": 1747
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.848902676875302,
+      "learning_rate": 2.510855916362884e-06,
+      "loss": 0.572,
+      "step": 1748
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.270230708616433,
+      "learning_rate": 2.5054260010904423e-06,
+      "loss": 0.4742,
+      "step": 1749
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.170980658272782,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.4185,
+      "step": 1750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.755920158041095,
+      "learning_rate": 2.494577921605382e-06,
+      "loss": 0.3944,
+      "step": 1751
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.3881840689155345,
+      "learning_rate": 2.489159774414252e-06,
+      "loss": 0.2457,
+      "step": 1752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.737663635521137,
+      "learning_rate": 2.4837455669281075e-06,
+      "loss": 0.4313,
+      "step": 1753
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.013750306170142,
+      "learning_rate": 2.478335307642264e-06,
+      "loss": 0.2296,
+      "step": 1754
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.890817835741133,
+      "learning_rate": 2.472929005045842e-06,
+      "loss": 0.4483,
+      "step": 1755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.401553772538874,
+      "learning_rate": 2.467526667621754e-06,
+      "loss": 0.6412,
+      "step": 1756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.907563865465455,
+      "learning_rate": 2.4621283038466916e-06,
+      "loss": 0.4491,
+      "step": 1757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.840774623636135,
+      "learning_rate": 2.4567339221911086e-06,
+      "loss": 0.3618,
+      "step": 1758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.353595119176774,
+      "learning_rate": 2.451343531119215e-06,
+      "loss": 0.2685,
+      "step": 1759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.750494873437042,
+      "learning_rate": 2.4459571390889537e-06,
+      "loss": 0.44,
+      "step": 1760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.9524105909762,
+      "learning_rate": 2.4405747545519966e-06,
+      "loss": 0.2839,
+      "step": 1761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.303924202881106,
+      "learning_rate": 2.435196385953727e-06,
+      "loss": 0.3459,
+      "step": 1762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.838253367184715,
+      "learning_rate": 2.4298220417332263e-06,
+      "loss": 0.2987,
+      "step": 1763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.387220913242405,
+      "learning_rate": 2.424451730323261e-06,
+      "loss": 0.5289,
+      "step": 1764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.892397834405033,
+      "learning_rate": 2.4190854601502717e-06,
+      "loss": 0.1964,
+      "step": 1765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.044272135121322,
+      "learning_rate": 2.413723239634356e-06,
+      "loss": 0.4826,
+      "step": 1766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.969186104074804,
+      "learning_rate": 2.4083650771892576e-06,
+      "loss": 0.1205,
+      "step": 1767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.430646673527988,
+      "learning_rate": 2.4030109812223534e-06,
+      "loss": 0.4761,
+      "step": 1768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.407699982884676,
+      "learning_rate": 2.3976609601346395e-06,
+      "loss": 0.2815,
+      "step": 1769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.207511873270246,
+      "learning_rate": 2.3923150223207176e-06,
+      "loss": 0.3378,
+      "step": 1770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.439859290392759,
+      "learning_rate": 2.386973176168783e-06,
+      "loss": 0.2423,
+      "step": 1771
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.456235468359118,
+      "learning_rate": 2.381635430060611e-06,
+      "loss": 0.3815,
+      "step": 1772
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.75950224973206,
+      "learning_rate": 2.376301792371543e-06,
+      "loss": 0.3709,
+      "step": 1773
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.011483328071696,
+      "learning_rate": 2.370972271470475e-06,
+      "loss": 0.3266,
+      "step": 1774
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.635769454177208,
+      "learning_rate": 2.3656468757198414e-06,
+      "loss": 0.3938,
+      "step": 1775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.910342373679152,
+      "learning_rate": 2.3603256134756066e-06,
+      "loss": 0.4842,
+      "step": 1776
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.1495802540475815,
+      "learning_rate": 2.3550084930872475e-06,
+      "loss": 0.159,
+      "step": 1777
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.834503691813105,
+      "learning_rate": 2.3496955228977437e-06,
+      "loss": 0.3577,
+      "step": 1778
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.260847115920575,
+      "learning_rate": 2.3443867112435585e-06,
+      "loss": 0.2029,
+      "step": 1779
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.243631758357619,
+      "learning_rate": 2.3390820664546353e-06,
+      "loss": 0.5201,
+      "step": 1780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.299854136668625,
+      "learning_rate": 2.333781596854377e-06,
+      "loss": 0.2527,
+      "step": 1781
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.268234625796824,
+      "learning_rate": 2.328485310759635e-06,
+      "loss": 0.5393,
+      "step": 1782
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.46839457025244,
+      "learning_rate": 2.323193216480698e-06,
+      "loss": 0.2202,
+      "step": 1783
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.50499985509579,
+      "learning_rate": 2.317905322321276e-06,
+      "loss": 0.3886,
+      "step": 1784
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.398558881916781,
+      "learning_rate": 2.3126216365784894e-06,
+      "loss": 0.4537,
+      "step": 1785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.839040659780117,
+      "learning_rate": 2.307342167542854e-06,
+      "loss": 0.5095,
+      "step": 1786
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.823004728334942,
+      "learning_rate": 2.3020669234982713e-06,
+      "loss": 0.1676,
+      "step": 1787
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.972406099455176,
+      "learning_rate": 2.296795912722014e-06,
+      "loss": 0.3169,
+      "step": 1788
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.9433982224156,
+      "learning_rate": 2.291529143484707e-06,
+      "loss": 0.2783,
+      "step": 1789
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.417844501320433,
+      "learning_rate": 2.286266624050326e-06,
+      "loss": 0.2969,
+      "step": 1790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.5880562134144,
+      "learning_rate": 2.2810083626761757e-06,
+      "loss": 0.2777,
+      "step": 1791
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.379278243615193,
+      "learning_rate": 2.275754367612881e-06,
+      "loss": 0.5929,
+      "step": 1792
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.49629133497484,
+      "learning_rate": 2.2705046471043703e-06,
+      "loss": 0.3696,
+      "step": 1793
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.864106686745443,
+      "learning_rate": 2.265259209387867e-06,
+      "loss": 0.3432,
+      "step": 1794
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.297815721703165,
+      "learning_rate": 2.2600180626938738e-06,
+      "loss": 0.541,
+      "step": 1795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.590813941137464,
+      "learning_rate": 2.25478121524616e-06,
+      "loss": 0.4072,
+      "step": 1796
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.212134878370419,
+      "learning_rate": 2.249548675261751e-06,
+      "loss": 0.2207,
+      "step": 1797
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.916591977876784,
+      "learning_rate": 2.2443204509509094e-06,
+      "loss": 0.5801,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.392154320728896,
+      "learning_rate": 2.2390965505171304e-06,
+      "loss": 0.2051,
+      "step": 1799
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.83332488558745,
+      "learning_rate": 2.2338769821571225e-06,
+      "loss": 0.2007,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.918050249594435,
+      "learning_rate": 2.2286617540607974e-06,
+      "loss": 0.375,
+      "step": 1801
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.718713314500566,
+      "learning_rate": 2.2234508744112564e-06,
+      "loss": 0.2718,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.976246847621736,
+      "learning_rate": 2.2182443513847777e-06,
+      "loss": 0.5644,
+      "step": 1803
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.898338927683716,
+      "learning_rate": 2.213042193150804e-06,
+      "loss": 0.2608,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.300402667403876,
+      "learning_rate": 2.207844407871929e-06,
+      "loss": 0.4014,
+      "step": 1805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.533130014909624,
+      "learning_rate": 2.202651003703885e-06,
+      "loss": 0.4439,
+      "step": 1806
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.821165737835408,
+      "learning_rate": 2.1974619887955294e-06,
+      "loss": 0.2563,
+      "step": 1807
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.24756031971457,
+      "learning_rate": 2.1922773712888356e-06,
+      "loss": 0.5882,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.23707815576226,
+      "learning_rate": 2.1870971593188704e-06,
+      "loss": 0.3455,
+      "step": 1809
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.995453736563052,
+      "learning_rate": 2.181921361013794e-06,
+      "loss": 0.2977,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.251518509398479,
+      "learning_rate": 2.17674998449484e-06,
+      "loss": 0.4404,
+      "step": 1811
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.347557516906262,
+      "learning_rate": 2.1715830378763025e-06,
+      "loss": 0.4171,
+      "step": 1812
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.310559405266783,
+      "learning_rate": 2.1664205292655253e-06,
+      "loss": 0.2501,
+      "step": 1813
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.71094160637401,
+      "learning_rate": 2.16126246676289e-06,
+      "loss": 0.5285,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.663694870278285,
+      "learning_rate": 2.1561088584617995e-06,
+      "loss": 0.529,
+      "step": 1815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.27516416637791,
+      "learning_rate": 2.1509597124486693e-06,
+      "loss": 0.5302,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.964102333471719,
+      "learning_rate": 2.1458150368029146e-06,
+      "loss": 0.4772,
+      "step": 1817
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.871951638083622,
+      "learning_rate": 2.140674839596931e-06,
+      "loss": 0.4823,
+      "step": 1818
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.40002975652929,
+      "learning_rate": 2.135539128896092e-06,
+      "loss": 0.3241,
+      "step": 1819
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.723571749285462,
+      "learning_rate": 2.13040791275873e-06,
+      "loss": 0.4364,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.00894576026266,
+      "learning_rate": 2.1252811992361256e-06,
+      "loss": 0.2686,
+      "step": 1821
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.971689133232392,
+      "learning_rate": 2.1201589963724933e-06,
+      "loss": 0.4137,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.431056651695,
+      "learning_rate": 2.1150413122049714e-06,
+      "loss": 0.7301,
+      "step": 1823
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.326828376087281,
+      "learning_rate": 2.109928154763606e-06,
+      "loss": 0.5025,
+      "step": 1824
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.707570804589606,
+      "learning_rate": 2.104819532071343e-06,
+      "loss": 0.5308,
+      "step": 1825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.47157283078443,
+      "learning_rate": 2.09971545214401e-06,
+      "loss": 0.4805,
+      "step": 1826
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.862190291580385,
+      "learning_rate": 2.094615922990309e-06,
+      "loss": 0.389,
+      "step": 1827
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.042715598973858,
+      "learning_rate": 2.0895209526118016e-06,
+      "loss": 0.356,
+      "step": 1828
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.264566537745118,
+      "learning_rate": 2.084430549002894e-06,
+      "loss": 0.1413,
+      "step": 1829
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.141591811937029,
+      "learning_rate": 2.0793447201508288e-06,
+      "loss": 0.3337,
+      "step": 1830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.412478790119353,
+      "learning_rate": 2.0742634740356695e-06,
+      "loss": 0.3503,
+      "step": 1831
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.62820320531682,
+      "learning_rate": 2.0691868186302898e-06,
+      "loss": 0.1648,
+      "step": 1832
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.689169663762316,
+      "learning_rate": 2.064114761900359e-06,
+      "loss": 0.2484,
+      "step": 1833
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.233364353049398,
+      "learning_rate": 2.0590473118043326e-06,
+      "loss": 0.206,
+      "step": 1834
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.311285795143208,
+      "learning_rate": 2.0539844762934353e-06,
+      "loss": 0.5897,
+      "step": 1835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.880035395838437,
+      "learning_rate": 2.0489262633116536e-06,
+      "loss": 0.5051,
+      "step": 1836
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.554003402715308,
+      "learning_rate": 2.043872680795721e-06,
+      "loss": 0.3117,
+      "step": 1837
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.6566345189428295,
+      "learning_rate": 2.0388237366751005e-06,
+      "loss": 0.1001,
+      "step": 1838
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.765324924201001,
+      "learning_rate": 2.0337794388719845e-06,
+      "loss": 0.2093,
+      "step": 1839
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.224531168085388,
+      "learning_rate": 2.0287397953012686e-06,
+      "loss": 0.274,
+      "step": 1840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.84891745740123,
+      "learning_rate": 2.023704813870551e-06,
+      "loss": 0.2629,
+      "step": 1841
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.913922835693434,
+      "learning_rate": 2.01867450248011e-06,
+      "loss": 0.3178,
+      "step": 1842
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.930213903438952,
+      "learning_rate": 2.0136488690228993e-06,
+      "loss": 0.2699,
+      "step": 1843
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.872479640699487,
+      "learning_rate": 2.008627921384531e-06,
+      "loss": 0.483,
+      "step": 1844
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.42826680416287,
+      "learning_rate": 2.0036116674432653e-06,
+      "loss": 0.5331,
+      "step": 1845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.341887854627865,
+      "learning_rate": 1.998600115069998e-06,
+      "loss": 0.4019,
+      "step": 1846
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.988485873352383,
+      "learning_rate": 1.993593272128248e-06,
+      "loss": 0.6748,
+      "step": 1847
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.659038281832737,
+      "learning_rate": 1.9885911464741413e-06,
+      "loss": 0.3546,
+      "step": 1848
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.541447930265141,
+      "learning_rate": 1.9835937459564065e-06,
+      "loss": 0.2579,
+      "step": 1849
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.51140307594916,
+      "learning_rate": 1.978601078416357e-06,
+      "loss": 0.2877,
+      "step": 1850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.615366485546753,
+      "learning_rate": 1.9736131516878776e-06,
+      "loss": 0.5875,
+      "step": 1851
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.720635148685817,
+      "learning_rate": 1.9686299735974177e-06,
+      "loss": 0.5699,
+      "step": 1852
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.420712180345845,
+      "learning_rate": 1.9636515519639735e-06,
+      "loss": 0.2722,
+      "step": 1853
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.018736187816623,
+      "learning_rate": 1.9586778945990785e-06,
+      "loss": 0.4747,
+      "step": 1854
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.310964597971315,
+      "learning_rate": 1.9537090093067907e-06,
+      "loss": 0.2781,
+      "step": 1855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.244127116333557,
+      "learning_rate": 1.9487449038836803e-06,
+      "loss": 0.3643,
+      "step": 1856
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.98527654678097,
+      "learning_rate": 1.9437855861188175e-06,
+      "loss": 0.3275,
+      "step": 1857
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.675315784912037,
+      "learning_rate": 1.9388310637937606e-06,
+      "loss": 0.4033,
+      "step": 1858
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.0699913471024,
+      "learning_rate": 1.933881344682543e-06,
+      "loss": 0.5364,
+      "step": 1859
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.825078920824781,
+      "learning_rate": 1.928936436551661e-06,
+      "loss": 0.4441,
+      "step": 1860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.42220859713335,
+      "learning_rate": 1.9239963471600635e-06,
+      "loss": 0.2852,
+      "step": 1861
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.506097273577362,
+      "learning_rate": 1.9190610842591386e-06,
+      "loss": 0.3361,
+      "step": 1862
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.548126876855557,
+      "learning_rate": 1.9141306555926985e-06,
+      "loss": 0.2638,
+      "step": 1863
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.1869241035339,
+      "learning_rate": 1.9092050688969736e-06,
+      "loss": 0.2701,
+      "step": 1864
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.511312580581761,
+      "learning_rate": 1.9042843319005944e-06,
+      "loss": 0.2689,
+      "step": 1865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.497510485491633,
+      "learning_rate": 1.8993684523245842e-06,
+      "loss": 0.4596,
+      "step": 1866
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.237334279005607,
+      "learning_rate": 1.8944574378823406e-06,
+      "loss": 0.5006,
+      "step": 1867
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.533754739162614,
+      "learning_rate": 1.8895512962796304e-06,
+      "loss": 0.3731,
+      "step": 1868
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.980405716987704,
+      "learning_rate": 1.8846500352145753e-06,
+      "loss": 0.2758,
+      "step": 1869
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.022299653603152,
+      "learning_rate": 1.879753662377637e-06,
+      "loss": 0.3631,
+      "step": 1870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.522418951891844,
+      "learning_rate": 1.874862185451608e-06,
+      "loss": 0.638,
+      "step": 1871
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.280478830130102,
+      "learning_rate": 1.8699756121115997e-06,
+      "loss": 0.2535,
+      "step": 1872
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.683671277723825,
+      "learning_rate": 1.865093950025027e-06,
+      "loss": 0.4897,
+      "step": 1873
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.304843243173872,
+      "learning_rate": 1.8602172068516011e-06,
+      "loss": 0.1059,
+      "step": 1874
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.033623375817482,
+      "learning_rate": 1.8553453902433144e-06,
+      "loss": 0.5618,
+      "step": 1875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.5939431646745,
+      "learning_rate": 1.8504785078444293e-06,
+      "loss": 0.3372,
+      "step": 1876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.23384403215519,
+      "learning_rate": 1.8456165672914628e-06,
+      "loss": 0.4985,
+      "step": 1877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.106223502434033,
+      "learning_rate": 1.8407595762131814e-06,
+      "loss": 0.1022,
+      "step": 1878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.896961477456104,
+      "learning_rate": 1.8359075422305856e-06,
+      "loss": 0.3514,
+      "step": 1879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.309464662421144,
+      "learning_rate": 1.8310604729568964e-06,
+      "loss": 0.4591,
+      "step": 1880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.758898070501253,
+      "learning_rate": 1.826218375997545e-06,
+      "loss": 0.5803,
+      "step": 1881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.350239622015723,
+      "learning_rate": 1.8213812589501611e-06,
+      "loss": 0.2573,
+      "step": 1882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.441248507228043,
+      "learning_rate": 1.8165491294045596e-06,
+      "loss": 0.1713,
+      "step": 1883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.50422939825495,
+      "learning_rate": 1.811721994942731e-06,
+      "loss": 0.3443,
+      "step": 1884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.101908118514416,
+      "learning_rate": 1.8068998631388268e-06,
+      "loss": 0.5267,
+      "step": 1885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.334813902751531,
+      "learning_rate": 1.8020827415591496e-06,
+      "loss": 0.3043,
+      "step": 1886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.306361105254098,
+      "learning_rate": 1.7972706377621412e-06,
+      "loss": 0.3153,
+      "step": 1887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.814299956395664,
+      "learning_rate": 1.7924635592983687e-06,
+      "loss": 0.3427,
+      "step": 1888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.252866694501568,
+      "learning_rate": 1.7876615137105142e-06,
+      "loss": 0.5449,
+      "step": 1889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.685420399282587,
+      "learning_rate": 1.7828645085333645e-06,
+      "loss": 0.2254,
+      "step": 1890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.671346409858044,
+      "learning_rate": 1.7780725512937964e-06,
+      "loss": 0.407,
+      "step": 1891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.62440073651654,
+      "learning_rate": 1.7732856495107648e-06,
+      "loss": 0.4502,
+      "step": 1892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.89421800045958,
+      "learning_rate": 1.7685038106952952e-06,
+      "loss": 0.4865,
+      "step": 1893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.432100648423548,
+      "learning_rate": 1.7637270423504664e-06,
+      "loss": 0.1669,
+      "step": 1894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.747742256185598,
+      "learning_rate": 1.7589553519714019e-06,
+      "loss": 0.4338,
+      "step": 1895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.47906877944777,
+      "learning_rate": 1.7541887470452606e-06,
+      "loss": 0.2621,
+      "step": 1896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.779049873193055,
+      "learning_rate": 1.7494272350512137e-06,
+      "loss": 0.4691,
+      "step": 1897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.153144240724517,
+      "learning_rate": 1.7446708234604498e-06,
+      "loss": 0.4464,
+      "step": 1898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.276155508326061,
+      "learning_rate": 1.7399195197361507e-06,
+      "loss": 0.526,
+      "step": 1899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.9614120451124,
+      "learning_rate": 1.735173331333484e-06,
+      "loss": 0.2416,
+      "step": 1900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.581879543104444,
+      "learning_rate": 1.7304322656995908e-06,
+      "loss": 0.5122,
+      "step": 1901
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.202512173655756,
+      "learning_rate": 1.7256963302735752e-06,
+      "loss": 0.1715,
+      "step": 1902
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.968609316730051,
+      "learning_rate": 1.7209655324864898e-06,
+      "loss": 0.3412,
+      "step": 1903
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.412411283299877,
+      "learning_rate": 1.7162398797613284e-06,
+      "loss": 0.2769,
+      "step": 1904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.517401449384888,
+      "learning_rate": 1.7115193795130086e-06,
+      "loss": 0.22,
+      "step": 1905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.90002468054369,
+      "learning_rate": 1.7068040391483676e-06,
+      "loss": 0.3432,
+      "step": 1906
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.97914654135758,
+      "learning_rate": 1.7020938660661407e-06,
+      "loss": 0.369,
+      "step": 1907
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.963811204559624,
+      "learning_rate": 1.6973888676569594e-06,
+      "loss": 0.5781,
+      "step": 1908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.931426455717453,
+      "learning_rate": 1.692689051303335e-06,
+      "loss": 0.3827,
+      "step": 1909
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.549418669422327,
+      "learning_rate": 1.6879944243796477e-06,
+      "loss": 0.4392,
+      "step": 1910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.272726581986518,
+      "learning_rate": 1.6833049942521346e-06,
+      "loss": 0.348,
+      "step": 1911
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.201230645875788,
+      "learning_rate": 1.678620768278879e-06,
+      "loss": 0.4333,
+      "step": 1912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.898370283973975,
+      "learning_rate": 1.6739417538097986e-06,
+      "loss": 0.2492,
+      "step": 1913
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.367523786076633,
+      "learning_rate": 1.6692679581866334e-06,
+      "loss": 0.4715,
+      "step": 1914
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.671372870591533,
+      "learning_rate": 1.6645993887429345e-06,
+      "loss": 0.2452,
+      "step": 1915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.701701872383955,
+      "learning_rate": 1.6599360528040538e-06,
+      "loss": 0.4337,
+      "step": 1916
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.724388296237122,
+      "learning_rate": 1.6552779576871297e-06,
+      "loss": 0.144,
+      "step": 1917
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.266750564812154,
+      "learning_rate": 1.650625110701079e-06,
+      "loss": 0.4407,
+      "step": 1918
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.765742276288465,
+      "learning_rate": 1.6459775191465827e-06,
+      "loss": 0.3225,
+      "step": 1919
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.843861460197981,
+      "learning_rate": 1.6413351903160763e-06,
+      "loss": 0.3189,
+      "step": 1920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.3486267913220775,
+      "learning_rate": 1.6366981314937374e-06,
+      "loss": 0.1043,
+      "step": 1921
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.032342880820678,
+      "learning_rate": 1.632066349955474e-06,
+      "loss": 0.5717,
+      "step": 1922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.718076105916003,
+      "learning_rate": 1.6274398529689144e-06,
+      "loss": 0.6106,
+      "step": 1923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.23485913578442,
+      "learning_rate": 1.6228186477933956e-06,
+      "loss": 0.3876,
+      "step": 1924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.409293268389307,
+      "learning_rate": 1.6182027416799505e-06,
+      "loss": 0.5478,
+      "step": 1925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.169465427541791,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.3324,
+      "step": 1926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.64417873623585,
+      "learning_rate": 1.608986855601824e-06,
+      "loss": 0.3091,
+      "step": 1927
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.692126209697866,
+      "learning_rate": 1.60438689009759e-06,
+      "loss": 0.2202,
+      "step": 1928
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.71090675432607,
+      "learning_rate": 1.5997922525763015e-06,
+      "loss": 0.4824,
+      "step": 1929
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.011093029778856,
+      "learning_rate": 1.5952029502473032e-06,
+      "loss": 0.2269,
+      "step": 1930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.009057757867563,
+      "learning_rate": 1.5906189903115704e-06,
+      "loss": 0.2128,
+      "step": 1931
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.88423513625171,
+      "learning_rate": 1.5860403799616951e-06,
+      "loss": 0.2764,
+      "step": 1932
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.99702404702807,
+      "learning_rate": 1.581467126381876e-06,
+      "loss": 0.5297,
+      "step": 1933
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.659761623242984,
+      "learning_rate": 1.5768992367479058e-06,
+      "loss": 0.3731,
+      "step": 1934
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.050651405912584,
+      "learning_rate": 1.5723367182271632e-06,
+      "loss": 0.242,
+      "step": 1935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.449483999528156,
+      "learning_rate": 1.5677795779785932e-06,
+      "loss": 0.1417,
+      "step": 1936
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.97564516670086,
+      "learning_rate": 1.5632278231527081e-06,
+      "loss": 0.2724,
+      "step": 1937
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.145513264496499,
+      "learning_rate": 1.5586814608915673e-06,
+      "loss": 0.3757,
+      "step": 1938
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.960883413600718,
+      "learning_rate": 1.55414049832877e-06,
+      "loss": 0.1361,
+      "step": 1939
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.013926051437153,
+      "learning_rate": 1.549604942589441e-06,
+      "loss": 0.5566,
+      "step": 1940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.669692253067031,
+      "learning_rate": 1.5450748007902234e-06,
+      "loss": 0.4586,
+      "step": 1941
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.97288770227816,
+      "learning_rate": 1.5405500800392643e-06,
+      "loss": 0.2556,
+      "step": 1942
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.087639175498431,
+      "learning_rate": 1.5360307874362052e-06,
+      "loss": 0.6266,
+      "step": 1943
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.325270008374133,
+      "learning_rate": 1.5315169300721694e-06,
+      "loss": 0.5787,
+      "step": 1944
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.719636955209152,
+      "learning_rate": 1.5270085150297538e-06,
+      "loss": 0.2736,
+      "step": 1945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.75355878743217,
+      "learning_rate": 1.5225055493830132e-06,
+      "loss": 0.7731,
+      "step": 1946
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.161161451482341,
+      "learning_rate": 1.518008040197455e-06,
+      "loss": 0.3752,
+      "step": 1947
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.59203164956869,
+      "learning_rate": 1.5135159945300232e-06,
+      "loss": 0.515,
+      "step": 1948
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.688055033119843,
+      "learning_rate": 1.5090294194290882e-06,
+      "loss": 0.2017,
+      "step": 1949
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.507389088577387,
+      "learning_rate": 1.5045483219344387e-06,
+      "loss": 0.6103,
+      "step": 1950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.158250974846569,
+      "learning_rate": 1.500072709077267e-06,
+      "loss": 0.3077,
+      "step": 1951
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.205085488251324,
+      "learning_rate": 1.4956025878801611e-06,
+      "loss": 0.1465,
+      "step": 1952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.631624850161025,
+      "learning_rate": 1.4911379653570913e-06,
+      "loss": 0.5097,
+      "step": 1953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.385994591734201,
+      "learning_rate": 1.4866788485133988e-06,
+      "loss": 0.4302,
+      "step": 1954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.3987336816877605,
+      "learning_rate": 1.4822252443457896e-06,
+      "loss": 0.2105,
+      "step": 1955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.484801203345954,
+      "learning_rate": 1.4777771598423147e-06,
+      "loss": 0.5486,
+      "step": 1956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.446111496004212,
+      "learning_rate": 1.473334601982368e-06,
+      "loss": 0.3271,
+      "step": 1957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.897426929061089,
+      "learning_rate": 1.4688975777366716e-06,
+      "loss": 0.6724,
+      "step": 1958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.181943808831333,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.518,
+      "step": 1959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.772677668142961,
+      "learning_rate": 1.4600401579274876e-06,
+      "loss": 0.2698,
+      "step": 1960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.335641409198365,
+      "learning_rate": 1.455619776261986e-06,
+      "loss": 0.4614,
+      "step": 1961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.471843031615622,
+      "learning_rate": 1.4512049560066837e-06,
+      "loss": 0.6414,
+      "step": 1962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.178698778945709,
+      "learning_rate": 1.4467957040887792e-06,
+      "loss": 0.5407,
+      "step": 1963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.562920542074915,
+      "learning_rate": 1.442392027426735e-06,
+      "loss": 0.2723,
+      "step": 1964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.347474512265087,
+      "learning_rate": 1.4379939329302629e-06,
+      "loss": 0.2116,
+      "step": 1965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.4966250372119125,
+      "learning_rate": 1.433601427500318e-06,
+      "loss": 0.451,
+      "step": 1966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.703764633193813,
+      "learning_rate": 1.4292145180290856e-06,
+      "loss": 0.7158,
+      "step": 1967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.680113157958742,
+      "learning_rate": 1.4248332113999708e-06,
+      "loss": 0.6276,
+      "step": 1968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.150235042863534,
+      "learning_rate": 1.4204575144875871e-06,
+      "loss": 0.3421,
+      "step": 1969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.65653140300182,
+      "learning_rate": 1.4160874341577447e-06,
+      "loss": 0.2975,
+      "step": 1970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.449824412345174,
+      "learning_rate": 1.4117229772674424e-06,
+      "loss": 0.4939,
+      "step": 1971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.25722782410132,
+      "learning_rate": 1.4073641506648555e-06,
+      "loss": 0.2844,
+      "step": 1972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.695726849944576,
+      "learning_rate": 1.4030109611893233e-06,
+      "loss": 0.5272,
+      "step": 1973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.4938052234458965,
+      "learning_rate": 1.3986634156713418e-06,
+      "loss": 0.4215,
+      "step": 1974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.196557724660567,
+      "learning_rate": 1.3943215209325505e-06,
+      "loss": 0.2899,
+      "step": 1975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.722748047433916,
+      "learning_rate": 1.3899852837857219e-06,
+      "loss": 0.4493,
+      "step": 1976
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.641607673304358,
+      "learning_rate": 1.3856547110347523e-06,
+      "loss": 0.3957,
+      "step": 1977
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.81709650280468,
+      "learning_rate": 1.3813298094746491e-06,
+      "loss": 0.3119,
+      "step": 1978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.919933062590274,
+      "learning_rate": 1.3770105858915217e-06,
+      "loss": 0.1832,
+      "step": 1979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.936493618659469,
+      "learning_rate": 1.3726970470625705e-06,
+      "loss": 0.2864,
+      "step": 1980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.458866551430926,
+      "learning_rate": 1.368389199756075e-06,
+      "loss": 0.4431,
+      "step": 1981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.582565379275631,
+      "learning_rate": 1.3640870507313859e-06,
+      "loss": 0.5262,
+      "step": 1982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.254674413810957,
+      "learning_rate": 1.3597906067389116e-06,
+      "loss": 0.3247,
+      "step": 1983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.904664240174014,
+      "learning_rate": 1.3554998745201114e-06,
+      "loss": 0.3265,
+      "step": 1984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.271199796073606,
+      "learning_rate": 1.351214860807476e-06,
+      "loss": 0.5013,
+      "step": 1985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.616321141596117,
+      "learning_rate": 1.3469355723245303e-06,
+      "loss": 0.2956,
+      "step": 1986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.120239952492419,
+      "learning_rate": 1.3426620157858122e-06,
+      "loss": 0.3693,
+      "step": 1987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.919224999687563,
+      "learning_rate": 1.3383941978968673e-06,
+      "loss": 0.3128,
+      "step": 1988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.841366021156138,
+      "learning_rate": 1.334132125354236e-06,
+      "loss": 0.1482,
+      "step": 1989
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.580145531792827,
+      "learning_rate": 1.3298758048454436e-06,
+      "loss": 0.4396,
+      "step": 1990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.205385618184557,
+      "learning_rate": 1.3256252430489907e-06,
+      "loss": 0.418,
+      "step": 1991
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.7944032722828,
+      "learning_rate": 1.321380446634342e-06,
+      "loss": 0.4996,
+      "step": 1992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.24056675986539,
+      "learning_rate": 1.317141422261915e-06,
+      "loss": 0.2738,
+      "step": 1993
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.38358108544251,
+      "learning_rate": 1.3129081765830725e-06,
+      "loss": 0.4784,
+      "step": 1994
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.870362293560709,
+      "learning_rate": 1.308680716240106e-06,
+      "loss": 0.3931,
+      "step": 1995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.747621981035228,
+      "learning_rate": 1.3044590478662328e-06,
+      "loss": 0.2737,
+      "step": 1996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.278735062980735,
+      "learning_rate": 1.3002431780855817e-06,
+      "loss": 0.534,
+      "step": 1997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.829365834131863,
+      "learning_rate": 1.2960331135131826e-06,
+      "loss": 0.6222,
+      "step": 1998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.958418555313173,
+      "learning_rate": 1.2918288607549568e-06,
+      "loss": 0.3556,
+      "step": 1999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.130962333590347,
+      "learning_rate": 1.2876304264077056e-06,
+      "loss": 0.3209,
+      "step": 2000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.773107998455587,
+      "learning_rate": 1.2834378170591017e-06,
+      "loss": 0.49,
+      "step": 2001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.049806839878281,
+      "learning_rate": 1.2792510392876777e-06,
+      "loss": 0.4418,
+      "step": 2002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.317054523122534,
+      "learning_rate": 1.275070099662815e-06,
+      "loss": 0.475,
+      "step": 2003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.193440185409925,
+      "learning_rate": 1.270895004744737e-06,
+      "loss": 0.4668,
+      "step": 2004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.618368715603136,
+      "learning_rate": 1.266725761084493e-06,
+      "loss": 0.2647,
+      "step": 2005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.428089390961345,
+      "learning_rate": 1.262562375223954e-06,
+      "loss": 0.1814,
+      "step": 2006
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.57193031520462,
+      "learning_rate": 1.2584048536957982e-06,
+      "loss": 0.4061,
+      "step": 2007
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.379159392315348,
+      "learning_rate": 1.2542532030235022e-06,
+      "loss": 0.2166,
+      "step": 2008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.70996964253962,
+      "learning_rate": 1.2501074297213312e-06,
+      "loss": 0.2136,
+      "step": 2009
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.026090805308447,
+      "learning_rate": 1.245967540294329e-06,
+      "loss": 0.4429,
+      "step": 2010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.067787686699544,
+      "learning_rate": 1.2418335412383048e-06,
+      "loss": 0.242,
+      "step": 2011
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.14411104076579,
+      "learning_rate": 1.2377054390398285e-06,
+      "loss": 0.1427,
+      "step": 2012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.170653506882074,
+      "learning_rate": 1.233583240176217e-06,
+      "loss": 0.1369,
+      "step": 2013
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.83590005681006,
+      "learning_rate": 1.2294669511155193e-06,
+      "loss": 0.6124,
+      "step": 2014
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.606342445093137,
+      "learning_rate": 1.2253565783165178e-06,
+      "loss": 0.2153,
+      "step": 2015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.413616498514486,
+      "learning_rate": 1.2212521282287093e-06,
+      "loss": 0.4126,
+      "step": 2016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.15844812471716,
+      "learning_rate": 1.2171536072922968e-06,
+      "loss": 0.5583,
+      "step": 2017
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.695194050170219,
+      "learning_rate": 1.2130610219381811e-06,
+      "loss": 0.4052,
+      "step": 2018
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.894906332968972,
+      "learning_rate": 1.2089743785879493e-06,
+      "loss": 0.2352,
+      "step": 2019
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.521733275523475,
+      "learning_rate": 1.204893683653865e-06,
+      "loss": 0.231,
+      "step": 2020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.336998748689888,
+      "learning_rate": 1.2008189435388578e-06,
+      "loss": 0.4688,
+      "step": 2021
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.4878471016339185,
+      "learning_rate": 1.1967501646365147e-06,
+      "loss": 0.1422,
+      "step": 2022
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.697144565568127,
+      "learning_rate": 1.1926873533310696e-06,
+      "loss": 0.2638,
+      "step": 2023
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.768888249028299,
+      "learning_rate": 1.1886305159973883e-06,
+      "loss": 0.1602,
+      "step": 2024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.951329345812272,
+      "learning_rate": 1.1845796590009684e-06,
+      "loss": 0.3115,
+      "step": 2025
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.72638292814094,
+      "learning_rate": 1.1805347886979219e-06,
+      "loss": 0.6219,
+      "step": 2026
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.1763709926677395,
+      "learning_rate": 1.1764959114349666e-06,
+      "loss": 0.1569,
+      "step": 2027
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.7586701592484815,
+      "learning_rate": 1.172463033549418e-06,
+      "loss": 0.2343,
+      "step": 2028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.399853987612607,
+      "learning_rate": 1.1684361613691769e-06,
+      "loss": 0.326,
+      "step": 2029
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.230157879386526,
+      "learning_rate": 1.1644153012127208e-06,
+      "loss": 0.2889,
+      "step": 2030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.813664337603972,
+      "learning_rate": 1.160400459389095e-06,
+      "loss": 0.3923,
+      "step": 2031
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.602094857851897,
+      "learning_rate": 1.1563916421979004e-06,
+      "loss": 0.2798,
+      "step": 2032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.832048260815114,
+      "learning_rate": 1.1523888559292857e-06,
+      "loss": 0.1796,
+      "step": 2033
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.607099150964159,
+      "learning_rate": 1.1483921068639353e-06,
+      "loss": 0.2686,
+      "step": 2034
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.760222238860768,
+      "learning_rate": 1.144401401273062e-06,
+      "loss": 0.3863,
+      "step": 2035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.543244487494375,
+      "learning_rate": 1.1404167454183957e-06,
+      "loss": 0.3158,
+      "step": 2036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.536014615568172,
+      "learning_rate": 1.1364381455521728e-06,
+      "loss": 0.3481,
+      "step": 2037
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.627328599487617,
+      "learning_rate": 1.1324656079171288e-06,
+      "loss": 0.2444,
+      "step": 2038
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.735795456288443,
+      "learning_rate": 1.128499138746486e-06,
+      "loss": 0.4788,
+      "step": 2039
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.529855418036066,
+      "learning_rate": 1.1245387442639456e-06,
+      "loss": 0.3265,
+      "step": 2040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.64856289564184,
+      "learning_rate": 1.120584430683676e-06,
+      "loss": 0.6871,
+      "step": 2041
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.40012925339637,
+      "learning_rate": 1.1166362042103056e-06,
+      "loss": 0.3472,
+      "step": 2042
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.568861675494704,
+      "learning_rate": 1.1126940710389128e-06,
+      "loss": 0.4774,
+      "step": 2043
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.341203605182917,
+      "learning_rate": 1.10875803735501e-06,
+      "loss": 0.4419,
+      "step": 2044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.866549088487773,
+      "learning_rate": 1.1048281093345437e-06,
+      "loss": 0.342,
+      "step": 2045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.482738799520188,
+      "learning_rate": 1.1009042931438784e-06,
+      "loss": 0.3826,
+      "step": 2046
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.6716676083365,
+      "learning_rate": 1.0969865949397902e-06,
+      "loss": 0.3207,
+      "step": 2047
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.80038893883207,
+      "learning_rate": 1.0930750208694535e-06,
+      "loss": 0.2988,
+      "step": 2048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.04984792735308,
+      "learning_rate": 1.0891695770704341e-06,
+      "loss": 0.1684,
+      "step": 2049
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.059887326147047,
+      "learning_rate": 1.0852702696706807e-06,
+      "loss": 0.5772,
+      "step": 2050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.832609594795821,
+      "learning_rate": 1.0813771047885107e-06,
+      "loss": 0.3995,
+      "step": 2051
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.876915479599741,
+      "learning_rate": 1.077490088532605e-06,
+      "loss": 0.2341,
+      "step": 2052
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.507994557809173,
+      "learning_rate": 1.0736092270019987e-06,
+      "loss": 0.3447,
+      "step": 2053
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.002387753370753,
+      "learning_rate": 1.0697345262860638e-06,
+      "loss": 0.2343,
+      "step": 2054
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.90269907943997,
+      "learning_rate": 1.0658659924645109e-06,
+      "loss": 0.409,
+      "step": 2055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.427399122241896,
+      "learning_rate": 1.062003631607373e-06,
+      "loss": 0.3891,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.672846990519414,
+      "learning_rate": 1.0581474497749965e-06,
+      "loss": 0.3062,
+      "step": 2057
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.20622638517937,
+      "learning_rate": 1.0542974530180327e-06,
+      "loss": 0.5652,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.43648166467133,
+      "learning_rate": 1.050453647377428e-06,
+      "loss": 0.3358,
+      "step": 2059
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.38310387794567,
+      "learning_rate": 1.0466160388844149e-06,
+      "loss": 0.4098,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.263807752317014,
+      "learning_rate": 1.0427846335605012e-06,
+      "loss": 0.3908,
+      "step": 2061
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.151822253143494,
+      "learning_rate": 1.0389594374174628e-06,
+      "loss": 0.3012,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.479902205981245,
+      "learning_rate": 1.0351404564573314e-06,
+      "loss": 0.4855,
+      "step": 2063
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.623856792971001,
+      "learning_rate": 1.0313276966723867e-06,
+      "loss": 0.5528,
+      "step": 2064
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.074459489327232,
+      "learning_rate": 1.0275211640451488e-06,
+      "loss": 0.2075,
+      "step": 2065
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.25623047203499,
+      "learning_rate": 1.0237208645483648e-06,
+      "loss": 0.4742,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.681721895898064,
+      "learning_rate": 1.0199268041450023e-06,
+      "loss": 0.4358,
+      "step": 2067
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.611128518564977,
+      "learning_rate": 1.0161389887882394e-06,
+      "loss": 0.2586,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.9472809093074455,
+      "learning_rate": 1.0123574244214552e-06,
+      "loss": 0.4083,
+      "step": 2069
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.305467112776571,
+      "learning_rate": 1.00858211697822e-06,
+      "loss": 0.4624,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.155379884992342,
+      "learning_rate": 1.0048130723822874e-06,
+      "loss": 0.3418,
+      "step": 2071
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.716461852198851,
+      "learning_rate": 1.0010502965475843e-06,
+      "loss": 0.2078,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.502562159109315,
+      "learning_rate": 9.972937953781985e-07,
+      "loss": 0.356,
+      "step": 2073
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.7805910230055835,
+      "learning_rate": 9.935435747683758e-07,
+      "loss": 0.3513,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.465498993546472,
+      "learning_rate": 9.897996406025062e-07,
+      "loss": 0.2176,
+      "step": 2075
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.014889036612077,
+      "learning_rate": 9.860619987551157e-07,
+      "loss": 0.1896,
+      "step": 2076
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.540891574066125,
+      "learning_rate": 9.823306550908568e-07,
+      "loss": 0.4673,
+      "step": 2077
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.100401686886848,
+      "learning_rate": 9.786056154645001e-07,
+      "loss": 0.1845,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.4802083915681115,
+      "learning_rate": 9.74886885720925e-07,
+      "loss": 0.4524,
+      "step": 2079
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.090559610655696,
+      "learning_rate": 9.711744716951093e-07,
+      "loss": 0.2427,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.57905355219723,
+      "learning_rate": 9.674683792121214e-07,
+      "loss": 0.3393,
+      "step": 2081
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.324035302904516,
+      "learning_rate": 9.637686140871121e-07,
+      "loss": 0.4936,
+      "step": 2082
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.331063499723545,
+      "learning_rate": 9.600751821252996e-07,
+      "loss": 0.1537,
+      "step": 2083
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.398338320636404,
+      "learning_rate": 9.563880891219702e-07,
+      "loss": 0.275,
+      "step": 2084
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.997968658296778,
+      "learning_rate": 9.52707340862461e-07,
+      "loss": 0.2726,
+      "step": 2085
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.861685049905185,
+      "learning_rate": 9.490329431221545e-07,
+      "loss": 0.4141,
+      "step": 2086
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.702266450053961,
+      "learning_rate": 9.45364901666469e-07,
+      "loss": 0.4666,
+      "step": 2087
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.373941006523895,
+      "learning_rate": 9.417032222508476e-07,
+      "loss": 0.3967,
+      "step": 2088
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.935709002954415,
+      "learning_rate": 9.380479106207535e-07,
+      "loss": 0.8082,
+      "step": 2089
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.244417756346,
+      "learning_rate": 9.34398972511656e-07,
+      "loss": 0.4678,
+      "step": 2090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.569049989046036,
+      "learning_rate": 9.307564136490255e-07,
+      "loss": 0.5085,
+      "step": 2091
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.43813959059892,
+      "learning_rate": 9.271202397483214e-07,
+      "loss": 0.2136,
+      "step": 2092
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.290245910218419,
+      "learning_rate": 9.234904565149866e-07,
+      "loss": 0.2487,
+      "step": 2093
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.632858368774757,
+      "learning_rate": 9.198670696444339e-07,
+      "loss": 0.6421,
+      "step": 2094
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.91297124560374,
+      "learning_rate": 9.162500848220418e-07,
+      "loss": 0.4586,
+      "step": 2095
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.579249355369184,
+      "learning_rate": 9.126395077231432e-07,
+      "loss": 0.6269,
+      "step": 2096
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.845659648310441,
+      "learning_rate": 9.090353440130156e-07,
+      "loss": 0.4374,
+      "step": 2097
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.979180005784649,
+      "learning_rate": 9.054375993468745e-07,
+      "loss": 0.3077,
+      "step": 2098
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.8242969013660515,
+      "learning_rate": 9.018462793698629e-07,
+      "loss": 0.2807,
+      "step": 2099
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.215445768144914,
+      "learning_rate": 8.982613897170439e-07,
+      "loss": 0.3299,
+      "step": 2100
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.737495402816952,
+      "learning_rate": 8.946829360133891e-07,
+      "loss": 0.1816,
+      "step": 2101
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.556277190694833,
+      "learning_rate": 8.911109238737748e-07,
+      "loss": 0.309,
+      "step": 2102
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.550857873774058,
+      "learning_rate": 8.875453589029648e-07,
+      "loss": 0.2651,
+      "step": 2103
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.669047869598055,
+      "learning_rate": 8.839862466956112e-07,
+      "loss": 0.438,
+      "step": 2104
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.154853333896177,
+      "learning_rate": 8.804335928362401e-07,
+      "loss": 0.422,
+      "step": 2105
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.752143094257852,
+      "learning_rate": 8.768874028992431e-07,
+      "loss": 0.2728,
+      "step": 2106
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.655750084534217,
+      "learning_rate": 8.733476824488707e-07,
+      "loss": 0.4169,
+      "step": 2107
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.145153821056764,
+      "learning_rate": 8.698144370392209e-07,
+      "loss": 0.3689,
+      "step": 2108
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.091185193634063,
+      "learning_rate": 8.662876722142327e-07,
+      "loss": 0.3401,
+      "step": 2109
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.509966533399151,
+      "learning_rate": 8.627673935076769e-07,
+      "loss": 0.3038,
+      "step": 2110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.586810405057567,
+      "learning_rate": 8.592536064431467e-07,
+      "loss": 0.3947,
+      "step": 2111
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.142147738602933,
+      "learning_rate": 8.557463165340479e-07,
+      "loss": 0.2445,
+      "step": 2112
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.819309312524908,
+      "learning_rate": 8.522455292835935e-07,
+      "loss": 0.1627,
+      "step": 2113
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.358358189773751,
+      "learning_rate": 8.487512501847933e-07,
+      "loss": 0.4867,
+      "step": 2114
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.836896044055315,
+      "learning_rate": 8.452634847204449e-07,
+      "loss": 0.4304,
+      "step": 2115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.294186296832018,
+      "learning_rate": 8.417822383631258e-07,
+      "loss": 0.4188,
+      "step": 2116
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.321449057949357,
+      "learning_rate": 8.38307516575183e-07,
+      "loss": 0.5107,
+      "step": 2117
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.83137443594219,
+      "learning_rate": 8.348393248087289e-07,
+      "loss": 0.3317,
+      "step": 2118
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.809632282951776,
+      "learning_rate": 8.313776685056262e-07,
+      "loss": 0.4563,
+      "step": 2119
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.672341674640126,
+      "learning_rate": 8.279225530974866e-07,
+      "loss": 0.2686,
+      "step": 2120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.39500017653997,
+      "learning_rate": 8.244739840056554e-07,
+      "loss": 0.2596,
+      "step": 2121
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.11792984604299,
+      "learning_rate": 8.210319666412087e-07,
+      "loss": 0.4698,
+      "step": 2122
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.206233772378962,
+      "learning_rate": 8.175965064049401e-07,
+      "loss": 0.3789,
+      "step": 2123
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.545593424384634,
+      "learning_rate": 8.141676086873574e-07,
+      "loss": 0.3809,
+      "step": 2124
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.672090496531137,
+      "learning_rate": 8.107452788686681e-07,
+      "loss": 0.4571,
+      "step": 2125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.0618969904223805,
+      "learning_rate": 8.073295223187766e-07,
+      "loss": 0.1252,
+      "step": 2126
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.346445495558898,
+      "learning_rate": 8.039203443972726e-07,
+      "loss": 0.1729,
+      "step": 2127
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.004824208025743,
+      "learning_rate": 8.005177504534228e-07,
+      "loss": 0.1385,
+      "step": 2128
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.146810239767541,
+      "learning_rate": 7.971217458261643e-07,
+      "loss": 0.4395,
+      "step": 2129
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.716083780478664,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.1633,
+      "step": 2130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.87890800688291,
+      "learning_rate": 7.903495258254624e-07,
+      "loss": 0.4777,
+      "step": 2131
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.135267881860537,
+      "learning_rate": 7.869733210781611e-07,
+      "loss": 0.4623,
+      "step": 2132
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.958385880053874,
+      "learning_rate": 7.836037268997221e-07,
+      "loss": 0.1121,
+      "step": 2133
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.445986833889874,
+      "learning_rate": 7.802407485773011e-07,
+      "loss": 0.3134,
+      "step": 2134
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.80204385855301,
+      "learning_rate": 7.768843913876756e-07,
+      "loss": 0.12,
+      "step": 2135
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.4307517759581945,
+      "learning_rate": 7.735346605972322e-07,
+      "loss": 0.3969,
+      "step": 2136
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.379066939703955,
+      "learning_rate": 7.701915614619615e-07,
+      "loss": 0.2102,
+      "step": 2137
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.659034756057517,
+      "learning_rate": 7.668550992274476e-07,
+      "loss": 0.5392,
+      "step": 2138
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.066081053005432,
+      "learning_rate": 7.635252791288611e-07,
+      "loss": 0.3499,
+      "step": 2139
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.385910897277725,
+      "learning_rate": 7.60202106390951e-07,
+      "loss": 0.2682,
+      "step": 2140
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.420515879200593,
+      "learning_rate": 7.568855862280356e-07,
+      "loss": 0.3364,
+      "step": 2141
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.449179692162263,
+      "learning_rate": 7.535757238439939e-07,
+      "loss": 0.4218,
+      "step": 2142
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.853118934664078,
+      "learning_rate": 7.502725244322596e-07,
+      "loss": 0.4658,
+      "step": 2143
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.048957877888816,
+      "learning_rate": 7.469759931758109e-07,
+      "loss": 0.1484,
+      "step": 2144
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.853664110386113,
+      "learning_rate": 7.436861352471647e-07,
+      "loss": 0.2908,
+      "step": 2145
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.687515809812147,
+      "learning_rate": 7.404029558083653e-07,
+      "loss": 0.1989,
+      "step": 2146
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.644212221966484,
+      "learning_rate": 7.371264600109779e-07,
+      "loss": 0.5863,
+      "step": 2147
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.47565836499741,
+      "learning_rate": 7.338566529960817e-07,
+      "loss": 0.2385,
+      "step": 2148
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.176760506918876,
+      "learning_rate": 7.305935398942598e-07,
+      "loss": 0.2667,
+      "step": 2149
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.476455675902974,
+      "learning_rate": 7.273371258255923e-07,
+      "loss": 0.3686,
+      "step": 2150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.882803617055647,
+      "learning_rate": 7.240874158996475e-07,
+      "loss": 0.6826,
+      "step": 2151
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.483889548353686,
+      "learning_rate": 7.208444152154759e-07,
+      "loss": 0.236,
+      "step": 2152
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.552678865326154,
+      "learning_rate": 7.176081288615983e-07,
+      "loss": 0.2578,
+      "step": 2153
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.288585213403897,
+      "learning_rate": 7.143785619160026e-07,
+      "loss": 0.3394,
+      "step": 2154
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.7280597296083595,
+      "learning_rate": 7.11155719446131e-07,
+      "loss": 0.5916,
+      "step": 2155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.934730149795459,
+      "learning_rate": 7.079396065088773e-07,
+      "loss": 0.4483,
+      "step": 2156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.065082053881767,
+      "learning_rate": 7.047302281505735e-07,
+      "loss": 0.457,
+      "step": 2157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.472630107362965,
+      "learning_rate": 7.015275894069862e-07,
+      "loss": 0.1901,
+      "step": 2158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.4480065438226895,
+      "learning_rate": 6.983316953033064e-07,
+      "loss": 0.1435,
+      "step": 2159
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.786809906973199,
+      "learning_rate": 6.951425508541432e-07,
+      "loss": 0.4211,
+      "step": 2160
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.70420091728909,
+      "learning_rate": 6.919601610635118e-07,
+      "loss": 0.2668,
+      "step": 2161
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.720652599316019,
+      "learning_rate": 6.887845309248326e-07,
+      "loss": 0.1755,
+      "step": 2162
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.012228411347202,
+      "learning_rate": 6.856156654209173e-07,
+      "loss": 0.2028,
+      "step": 2163
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.967341785577482,
+      "learning_rate": 6.824535695239643e-07,
+      "loss": 0.2347,
+      "step": 2164
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.843437187148792,
+      "learning_rate": 6.792982481955502e-07,
+      "loss": 0.423,
+      "step": 2165
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.796633967823919,
+      "learning_rate": 6.761497063866207e-07,
+      "loss": 0.3383,
+      "step": 2166
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.274608052885561,
+      "learning_rate": 6.730079490374852e-07,
+      "loss": 0.3553,
+      "step": 2167
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.909826850109335,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.5032,
+      "step": 2168
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.069598409052672,
+      "learning_rate": 6.667448074265954e-07,
+      "loss": 0.5856,
+      "step": 2169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.366387175767343,
+      "learning_rate": 6.63623432992202e-07,
+      "loss": 0.5625,
+      "step": 2170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.732733903160536,
+      "learning_rate": 6.605088626723055e-07,
+      "loss": 0.1697,
+      "step": 2171
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.716580845070284,
+      "learning_rate": 6.574011013539111e-07,
+      "loss": 0.3249,
+      "step": 2172
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.60034408174801,
+      "learning_rate": 6.543001539133409e-07,
+      "loss": 0.4623,
+      "step": 2173
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.913230825805475,
+      "learning_rate": 6.512060252162228e-07,
+      "loss": 0.2645,
+      "step": 2174
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.224601741883596,
+      "learning_rate": 6.481187201174888e-07,
+      "loss": 0.3929,
+      "step": 2175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.832857617305742,
+      "learning_rate": 6.450382434613612e-07,
+      "loss": 0.2291,
+      "step": 2176
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.38978247378926,
+      "learning_rate": 6.41964600081349e-07,
+      "loss": 0.2986,
+      "step": 2177
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.03720768322119,
+      "learning_rate": 6.388977948002406e-07,
+      "loss": 0.3343,
+      "step": 2178
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.948341302440195,
+      "learning_rate": 6.358378324300929e-07,
+      "loss": 0.1974,
+      "step": 2179
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.074787593165627,
+      "learning_rate": 6.327847177722274e-07,
+      "loss": 0.2432,
+      "step": 2180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.799407387712136,
+      "learning_rate": 6.297384556172176e-07,
+      "loss": 0.2261,
+      "step": 2181
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.322696660779327,
+      "learning_rate": 6.2669905074489e-07,
+      "loss": 0.393,
+      "step": 2182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.714401976244911,
+      "learning_rate": 6.236665079243087e-07,
+      "loss": 0.3251,
+      "step": 2183
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.219496818079271,
+      "learning_rate": 6.206408319137703e-07,
+      "loss": 0.6093,
+      "step": 2184
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.112316117039112,
+      "learning_rate": 6.176220274607975e-07,
+      "loss": 0.511,
+      "step": 2185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.629939983216484,
+      "learning_rate": 6.146100993021308e-07,
+      "loss": 0.1958,
+      "step": 2186
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.375401045869026,
+      "learning_rate": 6.116050521637218e-07,
+      "loss": 0.3178,
+      "step": 2187
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.752810054536492,
+      "learning_rate": 6.086068907607245e-07,
+      "loss": 0.1722,
+      "step": 2188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.153151427778937,
+      "learning_rate": 6.056156197974888e-07,
+      "loss": 0.3283,
+      "step": 2189
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.024489421066452,
+      "learning_rate": 6.026312439675553e-07,
+      "loss": 0.4152,
+      "step": 2190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.319748795445489,
+      "learning_rate": 5.996537679536401e-07,
+      "loss": 0.3597,
+      "step": 2191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.460874181418555,
+      "learning_rate": 5.966831964276376e-07,
+      "loss": 0.4824,
+      "step": 2192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.079933500139125,
+      "learning_rate": 5.93719534050608e-07,
+      "loss": 0.2057,
+      "step": 2193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.63177631544456,
+      "learning_rate": 5.907627854727688e-07,
+      "loss": 0.5678,
+      "step": 2194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.164325789910379,
+      "learning_rate": 5.878129553334905e-07,
+      "loss": 0.5206,
+      "step": 2195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.781653065173279,
+      "learning_rate": 5.848700482612873e-07,
+      "loss": 0.1525,
+      "step": 2196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.781805639827291,
+      "learning_rate": 5.819340688738118e-07,
+      "loss": 0.4226,
+      "step": 2197
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.119410686087441,
+      "learning_rate": 5.790050217778442e-07,
+      "loss": 0.3823,
+      "step": 2198
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.430755960571812,
+      "learning_rate": 5.760829115692907e-07,
+      "loss": 0.2282,
+      "step": 2199
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.68202433905999,
+      "learning_rate": 5.7316774283317e-07,
+      "loss": 0.3809,
+      "step": 2200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.33192310518455,
+      "learning_rate": 5.702595201436101e-07,
+      "loss": 0.399,
+      "step": 2201
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.353158096812638,
+      "learning_rate": 5.673582480638395e-07,
+      "loss": 0.512,
+      "step": 2202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.909144573537679,
+      "learning_rate": 5.64463931146183e-07,
+      "loss": 0.285,
+      "step": 2203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.805619233786315,
+      "learning_rate": 5.615765739320494e-07,
+      "loss": 0.2083,
+      "step": 2204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.674669280797438,
+      "learning_rate": 5.586961809519287e-07,
+      "loss": 0.3803,
+      "step": 2205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.711563796421697,
+      "learning_rate": 5.558227567253832e-07,
+      "loss": 0.1039,
+      "step": 2206
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.947597931967001,
+      "learning_rate": 5.529563057610399e-07,
+      "loss": 0.2938,
+      "step": 2207
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.885598223093634,
+      "learning_rate": 5.500968325565859e-07,
+      "loss": 0.4255,
+      "step": 2208
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.010213220109245,
+      "learning_rate": 5.472443415987594e-07,
+      "loss": 0.5701,
+      "step": 2209
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.40239812459973,
+      "learning_rate": 5.443988373633397e-07,
+      "loss": 0.4649,
+      "step": 2210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.329232798637705,
+      "learning_rate": 5.415603243151469e-07,
+      "loss": 0.1554,
+      "step": 2211
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.743789381307879,
+      "learning_rate": 5.387288069080298e-07,
+      "loss": 0.4166,
+      "step": 2212
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.586896366889075,
+      "learning_rate": 5.359042895848626e-07,
+      "loss": 0.4419,
+      "step": 2213
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.734718428233174,
+      "learning_rate": 5.330867767775333e-07,
+      "loss": 0.2971,
+      "step": 2214
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.454990985120913,
+      "learning_rate": 5.302762729069399e-07,
+      "loss": 0.1374,
+      "step": 2215
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.72099788527704,
+      "learning_rate": 5.274727823829839e-07,
+      "loss": 0.1031,
+      "step": 2216
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.901339681417603,
+      "learning_rate": 5.246763096045604e-07,
+      "loss": 0.39,
+      "step": 2217
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.532971786394633,
+      "learning_rate": 5.218868589595555e-07,
+      "loss": 0.3668,
+      "step": 2218
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.295102723638404,
+      "learning_rate": 5.191044348248358e-07,
+      "loss": 0.5066,
+      "step": 2219
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.055599979887792,
+      "learning_rate": 5.163290415662408e-07,
+      "loss": 0.1469,
+      "step": 2220
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.789760233029883,
+      "learning_rate": 5.13560683538582e-07,
+      "loss": 0.3269,
+      "step": 2221
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.466706035952504,
+      "learning_rate": 5.107993650856285e-07,
+      "loss": 0.2534,
+      "step": 2222
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.082743455813846,
+      "learning_rate": 5.080450905401057e-07,
+      "loss": 0.3978,
+      "step": 2223
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.758863714724203,
+      "learning_rate": 5.052978642236866e-07,
+      "loss": 0.1781,
+      "step": 2224
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.022675273265936,
+      "learning_rate": 5.025576904469842e-07,
+      "loss": 0.5463,
+      "step": 2225
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.164529017248615,
+      "learning_rate": 4.998245735095459e-07,
+      "loss": 0.6199,
+      "step": 2226
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.676864292966128,
+      "learning_rate": 4.970985176998455e-07,
+      "loss": 0.2868,
+      "step": 2227
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.700981589777279,
+      "learning_rate": 4.943795272952795e-07,
+      "loss": 0.3086,
+      "step": 2228
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.633017686021188,
+      "learning_rate": 4.916676065621562e-07,
+      "loss": 0.1348,
+      "step": 2229
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.304565582021766,
+      "learning_rate": 4.889627597556911e-07,
+      "loss": 0.1603,
+      "step": 2230
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.420603360565975,
+      "learning_rate": 4.862649911200007e-07,
+      "loss": 0.1076,
+      "step": 2231
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.2868449158239,
+      "learning_rate": 4.835743048880959e-07,
+      "loss": 0.41,
+      "step": 2232
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.172086978746359,
+      "learning_rate": 4.80890705281874e-07,
+      "loss": 0.4561,
+      "step": 2233
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.531603541748778,
+      "learning_rate": 4.782141965121129e-07,
+      "loss": 0.2504,
+      "step": 2234
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.398600393927087,
+      "learning_rate": 4.7554478277846427e-07,
+      "loss": 0.4294,
+      "step": 2235
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.041840277939394,
+      "learning_rate": 4.7288246826944826e-07,
+      "loss": 0.3293,
+      "step": 2236
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.009555889514079,
+      "learning_rate": 4.7022725716244387e-07,
+      "loss": 0.2089,
+      "step": 2237
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.418363255107296,
+      "learning_rate": 4.6757915362368567e-07,
+      "loss": 0.266,
+      "step": 2238
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.9344817893922555,
+      "learning_rate": 4.64938161808256e-07,
+      "loss": 0.2653,
+      "step": 2239
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.220853532389204,
+      "learning_rate": 4.6230428586007734e-07,
+      "loss": 0.1233,
+      "step": 2240
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.213951206148776,
+      "learning_rate": 4.596775299119066e-07,
+      "loss": 0.2863,
+      "step": 2241
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.655462171441577,
+      "learning_rate": 4.570578980853302e-07,
+      "loss": 0.6278,
+      "step": 2242
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.241864198526376,
+      "learning_rate": 4.5444539449075677e-07,
+      "loss": 0.3513,
+      "step": 2243
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.126371905955002,
+      "learning_rate": 4.5184002322740784e-07,
+      "loss": 0.3307,
+      "step": 2244
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.039911617844154,
+      "learning_rate": 4.4924178838331554e-07,
+      "loss": 0.2231,
+      "step": 2245
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.193299611815197,
+      "learning_rate": 4.466506940353138e-07,
+      "loss": 0.1671,
+      "step": 2246
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.2452118677942305,
+      "learning_rate": 4.4406674424903264e-07,
+      "loss": 0.4532,
+      "step": 2247
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.01601985196186,
+      "learning_rate": 4.414899430788916e-07,
+      "loss": 0.1628,
+      "step": 2248
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.969129194210152,
+      "learning_rate": 4.389202945680943e-07,
+      "loss": 0.4071,
+      "step": 2249
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.146047421724461,
+      "learning_rate": 4.363578027486187e-07,
+      "loss": 0.3177,
+      "step": 2250
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.227965230052387,
+      "learning_rate": 4.338024716412165e-07,
+      "loss": 0.4867,
+      "step": 2251
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.784076021987277,
+      "learning_rate": 4.3125430525540193e-07,
+      "loss": 0.1815,
+      "step": 2252
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.835047678949563,
+      "learning_rate": 4.2871330758944786e-07,
+      "loss": 0.3054,
+      "step": 2253
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.1655630612670995,
+      "learning_rate": 4.261794826303783e-07,
+      "loss": 0.1486,
+      "step": 2254
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.764782193145983,
+      "learning_rate": 4.2365283435396366e-07,
+      "loss": 0.2506,
+      "step": 2255
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.25217832961497,
+      "learning_rate": 4.211333667247125e-07,
+      "loss": 0.2388,
+      "step": 2256
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.16626404400971,
+      "learning_rate": 4.186210836958676e-07,
+      "loss": 0.3203,
+      "step": 2257
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.254058513668818,
+      "learning_rate": 4.16115989209398e-07,
+      "loss": 0.4442,
+      "step": 2258
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.648992155390867,
+      "learning_rate": 4.1361808719599163e-07,
+      "loss": 0.2245,
+      "step": 2259
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.208434941946417,
+      "learning_rate": 4.111273815750527e-07,
+      "loss": 0.3419,
+      "step": 2260
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.700782041266212,
+      "learning_rate": 4.086438762546946e-07,
+      "loss": 0.4484,
+      "step": 2261
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.678844591877666,
+      "learning_rate": 4.0616757513173123e-07,
+      "loss": 0.362,
+      "step": 2262
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.972017275699185,
+      "learning_rate": 4.036984820916723e-07,
+      "loss": 0.3912,
+      "step": 2263
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.345753331393052,
+      "learning_rate": 4.012366010087193e-07,
+      "loss": 0.3038,
+      "step": 2264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.848319359573983,
+      "learning_rate": 3.9878193574575566e-07,
+      "loss": 0.405,
+      "step": 2265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.11873054701319,
+      "learning_rate": 3.963344901543437e-07,
+      "loss": 0.2305,
+      "step": 2266
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.444754265791156,
+      "learning_rate": 3.9389426807471764e-07,
+      "loss": 0.4505,
+      "step": 2267
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.75424550946049,
+      "learning_rate": 3.9146127333577757e-07,
+      "loss": 0.2336,
+      "step": 2268
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.864069624755826,
+      "learning_rate": 3.8903550975508075e-07,
+      "loss": 0.3952,
+      "step": 2269
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.354231283794902,
+      "learning_rate": 3.866169811388415e-07,
+      "loss": 0.2501,
+      "step": 2270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.01532297330184,
+      "learning_rate": 3.8420569128192084e-07,
+      "loss": 0.5062,
+      "step": 2271
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.589362799821323,
+      "learning_rate": 3.8180164396782124e-07,
+      "loss": 0.4001,
+      "step": 2272
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.514726086281862,
+      "learning_rate": 3.794048429686803e-07,
+      "loss": 0.3052,
+      "step": 2273
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.73780436077882,
+      "learning_rate": 3.7701529204526856e-07,
+      "loss": 0.2121,
+      "step": 2274
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.843742914468104,
+      "learning_rate": 3.7463299494697735e-07,
+      "loss": 0.3641,
+      "step": 2275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.11091752911843,
+      "learning_rate": 3.722579554118172e-07,
+      "loss": 0.5781,
+      "step": 2276
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.319935696449804,
+      "learning_rate": 3.6989017716641206e-07,
+      "loss": 0.4553,
+      "step": 2277
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.8091995515441495,
+      "learning_rate": 3.675296639259912e-07,
+      "loss": 0.3394,
+      "step": 2278
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.973030488978303,
+      "learning_rate": 3.6517641939438407e-07,
+      "loss": 0.3673,
+      "step": 2279
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.501093671502694,
+      "learning_rate": 3.6283044726401594e-07,
+      "loss": 0.3592,
+      "step": 2280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.470308488218254,
+      "learning_rate": 3.604917512159012e-07,
+      "loss": 0.2425,
+      "step": 2281
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.575863077556576,
+      "learning_rate": 3.581603349196372e-07,
+      "loss": 0.3471,
+      "step": 2282
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.224523005948054,
+      "learning_rate": 3.5583620203339774e-07,
+      "loss": 0.1751,
+      "step": 2283
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.037179619436259,
+      "learning_rate": 3.5351935620393075e-07,
+      "loss": 0.4413,
+      "step": 2284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.62120284942576,
+      "learning_rate": 3.5120980106654825e-07,
+      "loss": 0.1988,
+      "step": 2285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.091065989770666,
+      "learning_rate": 3.4890754024512254e-07,
+      "loss": 0.2769,
+      "step": 2286
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.361199976461995,
+      "learning_rate": 3.466125773520818e-07,
+      "loss": 0.1901,
+      "step": 2287
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.342339642364207,
+      "learning_rate": 3.443249159884038e-07,
+      "loss": 0.1607,
+      "step": 2288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.20549916545604,
+      "learning_rate": 3.420445597436056e-07,
+      "loss": 0.2196,
+      "step": 2289
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.48172162934531,
+      "learning_rate": 3.397715121957468e-07,
+      "loss": 0.3961,
+      "step": 2290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.879870082898456,
+      "learning_rate": 3.3750577691141596e-07,
+      "loss": 0.3851,
+      "step": 2291
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.363404900547624,
+      "learning_rate": 3.352473574457304e-07,
+      "loss": 0.3929,
+      "step": 2292
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.81429516920969,
+      "learning_rate": 3.329962573423262e-07,
+      "loss": 0.5765,
+      "step": 2293
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.161535940982219,
+      "learning_rate": 3.3075248013335614e-07,
+      "loss": 0.372,
+      "step": 2294
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.901995583258259,
+      "learning_rate": 3.2851602933948236e-07,
+      "loss": 0.3259,
+      "step": 2295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.09828344628486,
+      "learning_rate": 3.262869084698711e-07,
+      "loss": 0.4851,
+      "step": 2296
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.722732958583014,
+      "learning_rate": 3.2406512102218903e-07,
+      "loss": 0.3809,
+      "step": 2297
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.762630080543095,
+      "learning_rate": 3.2185067048259245e-07,
+      "loss": 0.4646,
+      "step": 2298
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.228231205772611,
+      "learning_rate": 3.1964356032572866e-07,
+      "loss": 0.1957,
+      "step": 2299
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.493618214420225,
+      "learning_rate": 3.174437940147268e-07,
+      "loss": 0.4684,
+      "step": 2300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.966863725338754,
+      "learning_rate": 3.1525137500119207e-07,
+      "loss": 0.1647,
+      "step": 2301
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.786265473255352,
+      "learning_rate": 3.1306630672520153e-07,
+      "loss": 0.3288,
+      "step": 2302
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.97974301379834,
+      "learning_rate": 3.108885926152988e-07,
+      "loss": 0.2694,
+      "step": 2303
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.470323360438558,
+      "learning_rate": 3.087182360884872e-07,
+      "loss": 0.416,
+      "step": 2304
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.343281976625893,
+      "learning_rate": 3.06555240550227e-07,
+      "loss": 0.1599,
+      "step": 2305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.876354941455844,
+      "learning_rate": 3.0439960939442794e-07,
+      "loss": 0.477,
+      "step": 2306
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.114809201629245,
+      "learning_rate": 3.0225134600344373e-07,
+      "loss": 0.5139,
+      "step": 2307
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.908990949055784,
+      "learning_rate": 3.001104537480676e-07,
+      "loss": 0.4671,
+      "step": 2308
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.555444521073017,
+      "learning_rate": 2.9797693598752673e-07,
+      "loss": 0.4373,
+      "step": 2309
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.841439181209239,
+      "learning_rate": 2.9585079606947843e-07,
+      "loss": 0.6625,
+      "step": 2310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.193015480693626,
+      "learning_rate": 2.9373203733000234e-07,
+      "loss": 0.2076,
+      "step": 2311
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.60871879453958,
+      "learning_rate": 2.916206630935969e-07,
+      "loss": 0.2682,
+      "step": 2312
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.501750740381153,
+      "learning_rate": 2.895166766731744e-07,
+      "loss": 0.1668,
+      "step": 2313
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.283247584560488,
+      "learning_rate": 2.874200813700534e-07,
+      "loss": 0.1543,
+      "step": 2314
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.125036237775026,
+      "learning_rate": 2.8533088047395627e-07,
+      "loss": 0.3075,
+      "step": 2315
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.644316869779583,
+      "learning_rate": 2.8324907726300366e-07,
+      "loss": 0.253,
+      "step": 2316
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.088321004011907,
+      "learning_rate": 2.8117467500370756e-07,
+      "loss": 0.5588,
+      "step": 2317
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.217697778883812,
+      "learning_rate": 2.7910767695096707e-07,
+      "loss": 0.336,
+      "step": 2318
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.628321735830081,
+      "learning_rate": 2.77048086348064e-07,
+      "loss": 0.4627,
+      "step": 2319
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.045547860261054,
+      "learning_rate": 2.7499590642665773e-07,
+      "loss": 0.2944,
+      "step": 2320
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.132408377066821,
+      "learning_rate": 2.729511404067797e-07,
+      "loss": 0.3024,
+      "step": 2321
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.508384540363743,
+      "learning_rate": 2.7091379149682683e-07,
+      "loss": 0.5847,
+      "step": 2322
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.88036139061838,
+      "learning_rate": 2.688838628935603e-07,
+      "loss": 0.3713,
+      "step": 2323
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.737836907995876,
+      "learning_rate": 2.668613577820961e-07,
+      "loss": 0.6328,
+      "step": 2324
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.161629087531022,
+      "learning_rate": 2.6484627933590414e-07,
+      "loss": 0.4648,
+      "step": 2325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.835885084397102,
+      "learning_rate": 2.628386307167996e-07,
+      "loss": 0.5488,
+      "step": 2326
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.282531784707487,
+      "learning_rate": 2.608384150749416e-07,
+      "loss": 0.3321,
+      "step": 2327
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.571836394993312,
+      "learning_rate": 2.5884563554882336e-07,
+      "loss": 0.5925,
+      "step": 2328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.28418156209356,
+      "learning_rate": 2.5686029526527266e-07,
+      "loss": 0.2304,
+      "step": 2329
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.357512688602833,
+      "learning_rate": 2.548823973394449e-07,
+      "loss": 0.4002,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.367536872840859,
+      "learning_rate": 2.5291194487481575e-07,
+      "loss": 0.2234,
+      "step": 2331
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.71307445359601,
+      "learning_rate": 2.509489409631799e-07,
+      "loss": 0.3953,
+      "step": 2332
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.370551243500142,
+      "learning_rate": 2.4899338868464404e-07,
+      "loss": 0.564,
+      "step": 2333
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.31111127509216,
+      "learning_rate": 2.470452911076227e-07,
+      "loss": 0.4974,
+      "step": 2334
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.299414124533375,
+      "learning_rate": 2.4510465128883387e-07,
+      "loss": 0.1427,
+      "step": 2335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.828282276589536,
+      "learning_rate": 2.431714722732942e-07,
+      "loss": 0.3696,
+      "step": 2336
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.76862125112604,
+      "learning_rate": 2.4124575709431276e-07,
+      "loss": 0.4215,
+      "step": 2337
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.262686210172099,
+      "learning_rate": 2.393275087734864e-07,
+      "loss": 0.3258,
+      "step": 2338
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.503114368114424,
+      "learning_rate": 2.3741673032069757e-07,
+      "loss": 0.6684,
+      "step": 2339
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.515449280377295,
+      "learning_rate": 2.355134247341073e-07,
+      "loss": 0.3909,
+      "step": 2340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.291760993078214,
+      "learning_rate": 2.3361759500015123e-07,
+      "loss": 0.3259,
+      "step": 2341
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.323195170085055,
+      "learning_rate": 2.317292440935348e-07,
+      "loss": 0.4752,
+      "step": 2342
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.738336027621637,
+      "learning_rate": 2.2984837497722844e-07,
+      "loss": 0.5449,
+      "step": 2343
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.029292227347815,
+      "learning_rate": 2.2797499060246253e-07,
+      "loss": 0.5741,
+      "step": 2344
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.969517706993196,
+      "learning_rate": 2.2610909390872516e-07,
+      "loss": 0.1136,
+      "step": 2345
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.969533195125056,
+      "learning_rate": 2.242506878237538e-07,
+      "loss": 0.1302,
+      "step": 2346
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.649067179285142,
+      "learning_rate": 2.2239977526353263e-07,
+      "loss": 0.4121,
+      "step": 2347
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.919795867851366,
+      "learning_rate": 2.2055635913228845e-07,
+      "loss": 0.3795,
+      "step": 2348
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.5079147110404,
+      "learning_rate": 2.1872044232248646e-07,
+      "loss": 0.2683,
+      "step": 2349
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.647684386216921,
+      "learning_rate": 2.1689202771482344e-07,
+      "loss": 0.2675,
+      "step": 2350
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.306244451857543,
+      "learning_rate": 2.1507111817822445e-07,
+      "loss": 0.4101,
+      "step": 2351
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.378750707342817,
+      "learning_rate": 2.1325771656984075e-07,
+      "loss": 0.1607,
+      "step": 2352
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.268045617330754,
+      "learning_rate": 2.1145182573504008e-07,
+      "loss": 0.223,
+      "step": 2353
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.001234788889457,
+      "learning_rate": 2.0965344850740698e-07,
+      "loss": 0.2317,
+      "step": 2354
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.59646116418498,
+      "learning_rate": 2.0786258770873647e-07,
+      "loss": 0.2021,
+      "step": 2355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.445338914071707,
+      "learning_rate": 2.060792461490302e-07,
+      "loss": 0.261,
+      "step": 2356
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.135277940515158,
+      "learning_rate": 2.043034266264887e-07,
+      "loss": 0.2989,
+      "step": 2357
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.0646717698297,
+      "learning_rate": 2.0253513192751374e-07,
+      "loss": 0.4315,
+      "step": 2358
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.875941753538371,
+      "learning_rate": 2.0077436482669687e-07,
+      "loss": 0.3466,
+      "step": 2359
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.464575048962711,
+      "learning_rate": 1.9902112808682094e-07,
+      "loss": 0.3525,
+      "step": 2360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.421098660165775,
+      "learning_rate": 1.972754244588504e-07,
+      "loss": 0.2168,
+      "step": 2361
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.726775334058308,
+      "learning_rate": 1.9553725668193192e-07,
+      "loss": 0.3921,
+      "step": 2362
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.086129037026921,
+      "learning_rate": 1.9380662748338662e-07,
+      "loss": 0.307,
+      "step": 2363
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.908546274561068,
+      "learning_rate": 1.9208353957870684e-07,
+      "loss": 0.355,
+      "step": 2364
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.329616360877543,
+      "learning_rate": 1.903679956715526e-07,
+      "loss": 0.3895,
+      "step": 2365
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.515703520179736,
+      "learning_rate": 1.8865999845374794e-07,
+      "loss": 0.3558,
+      "step": 2366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.725370600215836,
+      "learning_rate": 1.8695955060527292e-07,
+      "loss": 0.2965,
+      "step": 2367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.56110149096519,
+      "learning_rate": 1.8526665479426386e-07,
+      "loss": 0.2339,
+      "step": 2368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.25857239899803,
+      "learning_rate": 1.835813136770065e-07,
+      "loss": 0.3262,
+      "step": 2369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.6255118243057565,
+      "learning_rate": 1.8190352989793325e-07,
+      "loss": 0.5723,
+      "step": 2370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.608709168224491,
+      "learning_rate": 1.8023330608961886e-07,
+      "loss": 0.3137,
+      "step": 2371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.98090413191681,
+      "learning_rate": 1.7857064487277475e-07,
+      "loss": 0.6317,
+      "step": 2372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.480819519620084,
+      "learning_rate": 1.7691554885624628e-07,
+      "loss": 0.3192,
+      "step": 2373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.04393860349002,
+      "learning_rate": 1.7526802063700943e-07,
+      "loss": 0.2073,
+      "step": 2374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.358840999263871,
+      "learning_rate": 1.7362806280016464e-07,
+      "loss": 0.1494,
+      "step": 2375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 12.24812260316486,
+      "learning_rate": 1.7199567791893524e-07,
+      "loss": 0.4999,
+      "step": 2376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.538078019092291,
+      "learning_rate": 1.7037086855465902e-07,
+      "loss": 0.3799,
+      "step": 2377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.394479631759364,
+      "learning_rate": 1.6875363725679052e-07,
+      "loss": 0.3825,
+      "step": 2378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.59165828366685,
+      "learning_rate": 1.6714398656289154e-07,
+      "loss": 0.3581,
+      "step": 2379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.579282316849458,
+      "learning_rate": 1.655419189986307e-07,
+      "loss": 0.6393,
+      "step": 2380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.528886006660832,
+      "learning_rate": 1.6394743707777772e-07,
+      "loss": 0.2437,
+      "step": 2381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.069865034650918,
+      "learning_rate": 1.6236054330219853e-07,
+      "loss": 0.3647,
+      "step": 2382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.560385473409923,
+      "learning_rate": 1.6078124016185525e-07,
+      "loss": 0.4791,
+      "step": 2383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.371547322142325,
+      "learning_rate": 1.592095301347968e-07,
+      "loss": 0.4744,
+      "step": 2384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.9273782615382995,
+      "learning_rate": 1.576454156871604e-07,
+      "loss": 0.3367,
+      "step": 2385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.083030542043729,
+      "learning_rate": 1.5608889927316407e-07,
+      "loss": 0.2184,
+      "step": 2386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.381637099793533,
+      "learning_rate": 1.54539983335103e-07,
+      "loss": 0.2297,
+      "step": 2387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.424801118531285,
+      "learning_rate": 1.5299867030334815e-07,
+      "loss": 0.4632,
+      "step": 2388
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.29917623655788,
+      "learning_rate": 1.5146496259634103e-07,
+      "loss": 0.4473,
+      "step": 2389
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.09682862879758,
+      "learning_rate": 1.4993886262058833e-07,
+      "loss": 0.1119,
+      "step": 2390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.320613704537341,
+      "learning_rate": 1.4842037277066012e-07,
+      "loss": 0.2018,
+      "step": 2391
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.039349558069974,
+      "learning_rate": 1.469094954291872e-07,
+      "loss": 0.4431,
+      "step": 2392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.875962399547143,
+      "learning_rate": 1.4540623296685374e-07,
+      "loss": 0.339,
+      "step": 2393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.380520870213245,
+      "learning_rate": 1.439105877423963e-07,
+      "loss": 0.4402,
+      "step": 2394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.267795993519376,
+      "learning_rate": 1.424225621025993e-07,
+      "loss": 0.3127,
+      "step": 2395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.35291322380306,
+      "learning_rate": 1.4094215838229176e-07,
+      "loss": 0.1738,
+      "step": 2396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.210083370935669,
+      "learning_rate": 1.3946937890434276e-07,
+      "loss": 0.4525,
+      "step": 2397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.174275460178519,
+      "learning_rate": 1.3800422597965935e-07,
+      "loss": 0.4041,
+      "step": 2398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.380976942826189,
+      "learning_rate": 1.3654670190718035e-07,
+      "loss": 0.1498,
+      "step": 2399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.228140978195125,
+      "learning_rate": 1.350968089738758e-07,
+      "loss": 0.2223,
+      "step": 2400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.392863360002925,
+      "learning_rate": 1.3365454945474088e-07,
+      "loss": 0.464,
+      "step": 2401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.075569101719713,
+      "learning_rate": 1.322199256127943e-07,
+      "loss": 0.4562,
+      "step": 2402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.387663862427827,
+      "learning_rate": 1.3079293969907259e-07,
+      "loss": 0.361,
+      "step": 2403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.279577877056235,
+      "learning_rate": 1.293735939526286e-07,
+      "loss": 0.2014,
+      "step": 2404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.038096793735251,
+      "learning_rate": 1.279618906005281e-07,
+      "loss": 0.2967,
+      "step": 2405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.484522916417465,
+      "learning_rate": 1.2655783185784253e-07,
+      "loss": 0.3921,
+      "step": 2406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.338686561525826,
+      "learning_rate": 1.2516141992765074e-07,
+      "loss": 0.5691,
+      "step": 2407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.750955924269938,
+      "learning_rate": 1.2377265700103225e-07,
+      "loss": 0.3516,
+      "step": 2408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.028043730100183,
+      "learning_rate": 1.223915452570651e-07,
+      "loss": 0.2024,
+      "step": 2409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.867593180675415,
+      "learning_rate": 1.210180868628219e-07,
+      "loss": 0.4332,
+      "step": 2410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.091637787484238,
+      "learning_rate": 1.196522839733666e-07,
+      "loss": 0.5474,
+      "step": 2411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.11826508229212,
+      "learning_rate": 1.1829413873174988e-07,
+      "loss": 0.591,
+      "step": 2412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.327075743264023,
+      "learning_rate": 1.1694365326900936e-07,
+      "loss": 0.233,
+      "step": 2413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.9259073891784055,
+      "learning_rate": 1.1560082970416164e-07,
+      "loss": 0.3998,
+      "step": 2414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.297350396193844,
+      "learning_rate": 1.1426567014420297e-07,
+      "loss": 0.4298,
+      "step": 2415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.192193645063606,
+      "learning_rate": 1.1293817668410201e-07,
+      "loss": 0.3553,
+      "step": 2416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.975729938513691,
+      "learning_rate": 1.1161835140680033e-07,
+      "loss": 0.3724,
+      "step": 2417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.787560011235373,
+      "learning_rate": 1.1030619638320805e-07,
+      "loss": 0.2499,
+      "step": 2418
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.573934742010731,
+      "learning_rate": 1.0900171367219826e-07,
+      "loss": 0.284,
+      "step": 2419
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.713407643316737,
+      "learning_rate": 1.0770490532060696e-07,
+      "loss": 0.197,
+      "step": 2420
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.536884086270271,
+      "learning_rate": 1.0641577336322761e-07,
+      "loss": 0.1755,
+      "step": 2421
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.423895864310222,
+      "learning_rate": 1.0513431982280997e-07,
+      "loss": 0.6925,
+      "step": 2422
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.279751071287675,
+      "learning_rate": 1.0386054671005452e-07,
+      "loss": 0.1106,
+      "step": 2423
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.156555577432458,
+      "learning_rate": 1.0259445602361084e-07,
+      "loss": 0.5867,
+      "step": 2424
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.67958138064252,
+      "learning_rate": 1.0133604975007483e-07,
+      "loss": 0.2736,
+      "step": 2425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.76406761920036,
+      "learning_rate": 1.0008532986398422e-07,
+      "loss": 0.2032,
+      "step": 2426
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.266684968538936,
+      "learning_rate": 9.884229832781644e-08,
+      "loss": 0.448,
+      "step": 2427
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.494910530780205,
+      "learning_rate": 9.760695709198575e-08,
+      "loss": 0.0848,
+      "step": 2428
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.44821763398629,
+      "learning_rate": 9.637930809483942e-08,
+      "loss": 0.5413,
+      "step": 2429
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.976075117841106,
+      "learning_rate": 9.51593532626538e-08,
+      "loss": 0.3994,
+      "step": 2430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.47659035829514,
+      "learning_rate": 9.39470945096349e-08,
+      "loss": 0.2452,
+      "step": 2431
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.250596612922149,
+      "learning_rate": 9.274253373791064e-08,
+      "loss": 0.4856,
+      "step": 2432
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.1001316345742,
+      "learning_rate": 9.154567283753135e-08,
+      "loss": 0.2541,
+      "step": 2433
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.544225959784822,
+      "learning_rate": 9.035651368646647e-08,
+      "loss": 0.3304,
+      "step": 2434
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.215758543399613,
+      "learning_rate": 8.917505815059902e-08,
+      "loss": 0.2427,
+      "step": 2435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.034226890297433,
+      "learning_rate": 8.800130808372553e-08,
+      "loss": 0.6167,
+      "step": 2436
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.8841745927716,
+      "learning_rate": 8.683526532755171e-08,
+      "loss": 0.3606,
+      "step": 2437
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.254220963828123,
+      "learning_rate": 8.567693171168956e-08,
+      "loss": 0.3032,
+      "step": 2438
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.864877056104323,
+      "learning_rate": 8.452630905365633e-08,
+      "loss": 0.4056,
+      "step": 2439
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.72114346072783,
+      "learning_rate": 8.338339915886784e-08,
+      "loss": 0.3022,
+      "step": 2440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.943830559855649,
+      "learning_rate": 8.224820382064013e-08,
+      "loss": 0.3475,
+      "step": 2441
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.684647142792592,
+      "learning_rate": 8.11207248201834e-08,
+      "loss": 0.1948,
+      "step": 2442
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.8898993272559395,
+      "learning_rate": 8.000096392660029e-08,
+      "loss": 0.2957,
+      "step": 2443
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.533360063707411,
+      "learning_rate": 7.888892289688366e-08,
+      "loss": 0.3787,
+      "step": 2444
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.030284649230856,
+      "learning_rate": 7.778460347591277e-08,
+      "loss": 0.4885,
+      "step": 2445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.239792974268687,
+      "learning_rate": 7.6688007396451e-08,
+      "loss": 0.309,
+      "step": 2446
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.556095756054404,
+      "learning_rate": 7.559913637914418e-08,
+      "loss": 0.3021,
+      "step": 2447
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.242062260909545,
+      "learning_rate": 7.45179921325162e-08,
+      "loss": 0.4514,
+      "step": 2448
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.009438482746365,
+      "learning_rate": 7.34445763529662e-08,
+      "loss": 0.2966,
+      "step": 2449
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.760339730377637,
+      "learning_rate": 7.237889072476856e-08,
+      "loss": 0.3231,
+      "step": 2450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.275704640413498,
+      "learning_rate": 7.13209369200668e-08,
+      "loss": 0.5443,
+      "step": 2451
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.158488618646793,
+      "learning_rate": 7.027071659887364e-08,
+      "loss": 0.8674,
+      "step": 2452
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.712946021551268,
+      "learning_rate": 6.922823140906754e-08,
+      "loss": 0.5116,
+      "step": 2453
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.910467930943143,
+      "learning_rate": 6.819348298638839e-08,
+      "loss": 0.3022,
+      "step": 2454
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.689658651176243,
+      "learning_rate": 6.716647295443801e-08,
+      "loss": 0.3202,
+      "step": 2455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.180318146358221,
+      "learning_rate": 6.614720292467569e-08,
+      "loss": 0.5038,
+      "step": 2456
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.361361411577573,
+      "learning_rate": 6.5135674496416e-08,
+      "loss": 0.506,
+      "step": 2457
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.634045718611473,
+      "learning_rate": 6.4131889256826e-08,
+      "loss": 0.4904,
+      "step": 2458
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.348711387297282,
+      "learning_rate": 6.313584878092361e-08,
+      "loss": 0.3799,
+      "step": 2459
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.555987149624934,
+      "learning_rate": 6.214755463157417e-08,
+      "loss": 0.3841,
+      "step": 2460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.26020814659951,
+      "learning_rate": 6.116700835948842e-08,
+      "loss": 0.5367,
+      "step": 2461
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.40794120871574,
+      "learning_rate": 6.019421150322114e-08,
+      "loss": 0.2811,
+      "step": 2462
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.563643631841472,
+      "learning_rate": 5.922916558916581e-08,
+      "loss": 0.2731,
+      "step": 2463
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.693866758356494,
+      "learning_rate": 5.8271872131555605e-08,
+      "loss": 0.2242,
+      "step": 2464
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.577651796518569,
+      "learning_rate": 5.7322332632458454e-08,
+      "loss": 0.3075,
+      "step": 2465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.049082628718091,
+      "learning_rate": 5.638054858177644e-08,
+      "loss": 0.0841,
+      "step": 2466
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.84672200324526,
+      "learning_rate": 5.544652145724305e-08,
+      "loss": 0.4397,
+      "step": 2467
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.069919924188549,
+      "learning_rate": 5.45202527244193e-08,
+      "loss": 0.2104,
+      "step": 2468
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.380567003345906,
+      "learning_rate": 5.36017438366937e-08,
+      "loss": 0.2778,
+      "step": 2469
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.105491649105515,
+      "learning_rate": 5.26909962352784e-08,
+      "loss": 0.369,
+      "step": 2470
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.424978710013773,
+      "learning_rate": 5.178801134920808e-08,
+      "loss": 0.2047,
+      "step": 2471
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.05251107352017,
+      "learning_rate": 5.089279059533658e-08,
+      "loss": 0.551,
+      "step": 2472
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.972207663779297,
+      "learning_rate": 5.0005335378335294e-08,
+      "loss": 0.5249,
+      "step": 2473
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.590416970197765,
+      "learning_rate": 4.91256470906909e-08,
+      "loss": 0.4834,
+      "step": 2474
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.830105826213474,
+      "learning_rate": 4.825372711270371e-08,
+      "loss": 0.3437,
+      "step": 2475
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.529825118539277,
+      "learning_rate": 4.73895768124838e-08,
+      "loss": 0.4868,
+      "step": 2476
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.534538344231041,
+      "learning_rate": 4.653319754595098e-08,
+      "loss": 0.4452,
+      "step": 2477
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.7357149267879,
+      "learning_rate": 4.568459065683206e-08,
+      "loss": 0.2659,
+      "step": 2478
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.7176035223235315,
+      "learning_rate": 4.484375747665637e-08,
+      "loss": 0.3461,
+      "step": 2479
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.353065212094222,
+      "learning_rate": 4.401069932475799e-08,
+      "loss": 0.3479,
+      "step": 2480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.072785130064468,
+      "learning_rate": 4.318541750827021e-08,
+      "loss": 0.0721,
+      "step": 2481
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.391075843231557,
+      "learning_rate": 4.236791332212498e-08,
+      "loss": 0.4563,
+      "step": 2482
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.863461261178479,
+      "learning_rate": 4.155818804904954e-08,
+      "loss": 0.1473,
+      "step": 2483
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.645847472380593,
+      "learning_rate": 4.0756242959567596e-08,
+      "loss": 0.6493,
+      "step": 2484
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.767246595365589,
+      "learning_rate": 3.996207931199314e-08,
+      "loss": 0.451,
+      "step": 2485
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.229472266943983,
+      "learning_rate": 3.917569835243107e-08,
+      "loss": 0.2483,
+      "step": 2486
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.692096226237597,
+      "learning_rate": 3.839710131477492e-08,
+      "loss": 0.2085,
+      "step": 2487
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.349246700799483,
+      "learning_rate": 3.762628942070412e-08,
+      "loss": 0.252,
+      "step": 2488
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.63930321273986,
+      "learning_rate": 3.686326387968286e-08,
+      "loss": 0.2188,
+      "step": 2489
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.076393027304495,
+      "learning_rate": 3.610802588895845e-08,
+      "loss": 0.1083,
+      "step": 2490
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.157300757672089,
+      "learning_rate": 3.536057663355852e-08,
+      "loss": 0.2656,
+      "step": 2491
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.534746507804247,
+      "learning_rate": 3.462091728628936e-08,
+      "loss": 0.6799,
+      "step": 2492
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.005062687142324,
+      "learning_rate": 3.388904900773371e-08,
+      "loss": 0.3764,
+      "step": 2493
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.72815723333719,
+      "learning_rate": 3.316497294625132e-08,
+      "loss": 0.7183,
+      "step": 2494
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.480540795517165,
+      "learning_rate": 3.2448690237973366e-08,
+      "loss": 0.447,
+      "step": 2495
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.802752874680262,
+      "learning_rate": 3.1740202006804166e-08,
+      "loss": 0.3461,
+      "step": 2496
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.698264389389353,
+      "learning_rate": 3.1039509364416686e-08,
+      "loss": 0.3917,
+      "step": 2497
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.937878523089127,
+      "learning_rate": 3.034661341025258e-08,
+      "loss": 0.3262,
+      "step": 2498
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.729746120592926,
+      "learning_rate": 2.96615152315205e-08,
+      "loss": 0.4779,
+      "step": 2499
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.187368525344585,
+      "learning_rate": 2.898421590319278e-08,
+      "loss": 0.464,
+      "step": 2500
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.719978221612047,
+      "learning_rate": 2.8314716488004878e-08,
+      "loss": 0.4678,
+      "step": 2501
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.227421707232867,
+      "learning_rate": 2.765301803645426e-08,
+      "loss": 0.3537,
+      "step": 2502
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.2539449287688225,
+      "learning_rate": 2.6999121586797626e-08,
+      "loss": 0.3619,
+      "step": 2503
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.183332528257056,
+      "learning_rate": 2.6353028165049254e-08,
+      "loss": 0.3537,
+      "step": 2504
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.698477128168598,
+      "learning_rate": 2.5714738784980986e-08,
+      "loss": 0.1969,
+      "step": 2505
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.975701635424352,
+      "learning_rate": 2.5084254448117794e-08,
+      "loss": 0.4888,
+      "step": 2506
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.060064344116633,
+      "learning_rate": 2.4461576143740007e-08,
+      "loss": 0.555,
+      "step": 2507
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.093201767971806,
+      "learning_rate": 2.3846704848878298e-08,
+      "loss": 0.2883,
+      "step": 2508
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.377596823694224,
+      "learning_rate": 2.323964152831426e-08,
+      "loss": 0.297,
+      "step": 2509
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.003017768719825,
+      "learning_rate": 2.264038713457706e-08,
+      "loss": 0.2105,
+      "step": 2510
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.5864340750090555,
+      "learning_rate": 2.204894260794399e-08,
+      "loss": 0.2885,
+      "step": 2511
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.949368827740956,
+      "learning_rate": 2.1465308876438275e-08,
+      "loss": 0.2601,
+      "step": 2512
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.166456430020295,
+      "learning_rate": 2.0889486855826812e-08,
+      "loss": 0.31,
+      "step": 2513
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.7817222607407,
+      "learning_rate": 2.0321477449619098e-08,
+      "loss": 0.3655,
+      "step": 2514
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.225691445270876,
+      "learning_rate": 1.9761281549067202e-08,
+      "loss": 0.4109,
+      "step": 2515
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.417667799498685,
+      "learning_rate": 1.9208900033161338e-08,
+      "loss": 0.1203,
+      "step": 2516
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.328469612914402,
+      "learning_rate": 1.8664333768631526e-08,
+      "loss": 0.1901,
+      "step": 2517
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.042731106636099,
+      "learning_rate": 1.8127583609945376e-08,
+      "loss": 0.5061,
+      "step": 2518
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.504472569650513,
+      "learning_rate": 1.7598650399305862e-08,
+      "loss": 0.2258,
+      "step": 2519
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.4627866059388,
+      "learning_rate": 1.7077534966650767e-08,
+      "loss": 0.4181,
+      "step": 2520
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.713367399695205,
+      "learning_rate": 1.6564238129650468e-08,
+      "loss": 0.1826,
+      "step": 2521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.329924205006781,
+      "learning_rate": 1.6058760693708487e-08,
+      "loss": 0.3759,
+      "step": 2522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.008648465069133,
+      "learning_rate": 1.556110345195816e-08,
+      "loss": 0.4669,
+      "step": 2523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.407857803707713,
+      "learning_rate": 1.5071267185262086e-08,
+      "loss": 0.2213,
+      "step": 2524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.485560084932878,
+      "learning_rate": 1.4589252662213227e-08,
+      "loss": 0.3658,
+      "step": 2525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.167617744340484,
+      "learning_rate": 1.411506063912882e-08,
+      "loss": 0.5674,
+      "step": 2526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.209403738732304,
+      "learning_rate": 1.3648691860053686e-08,
+      "loss": 0.2163,
+      "step": 2527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.943822958044267,
+      "learning_rate": 1.3190147056757474e-08,
+      "loss": 0.355,
+      "step": 2528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.962809526995259,
+      "learning_rate": 1.2739426948732426e-08,
+      "loss": 0.3251,
+      "step": 2529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.576551674232872,
+      "learning_rate": 1.2296532243193382e-08,
+      "loss": 0.3891,
+      "step": 2530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.935206720795374,
+      "learning_rate": 1.1861463635077785e-08,
+      "loss": 0.4116,
+      "step": 2531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.4259233932409785,
+      "learning_rate": 1.1434221807041234e-08,
+      "loss": 0.3311,
+      "step": 2532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.468441161916848,
+      "learning_rate": 1.1014807429460262e-08,
+      "loss": 0.1565,
+      "step": 2533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.023384904323844,
+      "learning_rate": 1.0603221160429e-08,
+      "loss": 0.4677,
+      "step": 2534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.89357236217317,
+      "learning_rate": 1.019946364575808e-08,
+      "loss": 0.2254,
+      "step": 2535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.49766531318769,
+      "learning_rate": 9.803535518975171e-09,
+      "loss": 0.2768,
+      "step": 2536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.118349818197863,
+      "learning_rate": 9.41543740132167e-09,
+      "loss": 0.3682,
+      "step": 2537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.633053740309622,
+      "learning_rate": 9.035169901754902e-09,
+      "loss": 0.3193,
+      "step": 2538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.510360833047619,
+      "learning_rate": 8.662733616944253e-09,
+      "loss": 0.5116,
+      "step": 2539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.322925425235645,
+      "learning_rate": 8.298129131270594e-09,
+      "loss": 0.312,
+      "step": 2540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.326845455758031,
+      "learning_rate": 7.941357016827967e-09,
+      "loss": 0.4201,
+      "step": 2541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.400542624456659,
+      "learning_rate": 7.59241783341913e-09,
+      "loss": 0.4893,
+      "step": 2542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.60008975140832,
+      "learning_rate": 7.251312128556675e-09,
+      "loss": 0.4917,
+      "step": 2543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.28489309937508,
+      "learning_rate": 6.918040437463025e-09,
+      "loss": 0.3455,
+      "step": 2544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.510596526835135,
+      "learning_rate": 6.592603283067101e-09,
+      "loss": 0.259,
+      "step": 2545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.00301673553428,
+      "learning_rate": 6.2750011760054355e-09,
+      "loss": 0.2293,
+      "step": 2546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.121063412643829,
+      "learning_rate": 5.965234614620508e-09,
+      "loss": 0.2257,
+      "step": 2547
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.799187321596124,
+      "learning_rate": 5.6633040849601865e-09,
+      "loss": 0.2101,
+      "step": 2548
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.778027013953314,
+      "learning_rate": 5.369210060777175e-09,
+      "loss": 0.1685,
+      "step": 2549
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.532605445765191,
+      "learning_rate": 5.082953003528457e-09,
+      "loss": 0.3268,
+      "step": 2550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.067349139614606,
+      "learning_rate": 4.804533362373076e-09,
+      "loss": 0.5316,
+      "step": 2551
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.715537511964083,
+      "learning_rate": 4.533951574174356e-09,
+      "loss": 0.2482,
+      "step": 2552
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.569486224557908,
+      "learning_rate": 4.2712080634949024e-09,
+      "loss": 0.4025,
+      "step": 2553
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.712944111347317,
+      "learning_rate": 4.016303242600495e-09,
+      "loss": 0.4391,
+      "step": 2554
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.5928711884146125,
+      "learning_rate": 3.769237511457302e-09,
+      "loss": 0.4047,
+      "step": 2555
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.204067384218321,
+      "learning_rate": 3.530011257730226e-09,
+      "loss": 0.2574,
+      "step": 2556
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.668396685403804,
+      "learning_rate": 3.298624856784005e-09,
+      "loss": 0.3433,
+      "step": 2557
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.203275352545349,
+      "learning_rate": 3.075078671682108e-09,
+      "loss": 0.4526,
+      "step": 2558
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.180585801712816,
+      "learning_rate": 2.8593730531861764e-09,
+      "loss": 0.3634,
+      "step": 2559
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.32355839313059,
+      "learning_rate": 2.6515083397549156e-09,
+      "loss": 0.406,
+      "step": 2560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.392972906252533,
+      "learning_rate": 2.4514848575446505e-09,
+      "loss": 0.3,
+      "step": 2561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.273842549071784,
+      "learning_rate": 2.2593029204076578e-09,
+      "loss": 0.5776,
+      "step": 2562
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.687407755734984,
+      "learning_rate": 2.0749628298921688e-09,
+      "loss": 0.3127,
+      "step": 2563
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.145302026014107,
+      "learning_rate": 1.8984648752429222e-09,
+      "loss": 0.3846,
+      "step": 2564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.476775618578085,
+      "learning_rate": 1.7298093333989463e-09,
+      "loss": 0.1876,
+      "step": 2565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.730855759839342,
+      "learning_rate": 1.5689964689935555e-09,
+      "loss": 0.2791,
+      "step": 2566
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.489697372576666,
+      "learning_rate": 1.4160265343549084e-09,
+      "loss": 0.2823,
+      "step": 2567
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.103068337586608,
+      "learning_rate": 1.2708997695043412e-09,
+      "loss": 0.16,
+      "step": 2568
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.713280726536281,
+      "learning_rate": 1.133616402158033e-09,
+      "loss": 0.4758,
+      "step": 2569
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.052407841577594,
+      "learning_rate": 1.004176647724231e-09,
+      "loss": 0.4434,
+      "step": 2570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.065605005525713,
+      "learning_rate": 8.82580709303249e-10,
+      "loss": 0.2168,
+      "step": 2571
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.058152934736846,
+      "learning_rate": 7.688287776896897e-10,
+      "loss": 0.3409,
+      "step": 2572
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.634930264600342,
+      "learning_rate": 6.629210313680023e-10,
+      "loss": 0.1238,
+      "step": 2573
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.404701089401403,
+      "learning_rate": 5.648576365169245e-10,
+      "loss": 0.2974,
+      "step": 2574
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.123203402429388,
+      "learning_rate": 4.746387470044855e-10,
+      "loss": 0.2262,
+      "step": 2575
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.770380948601973,
+      "learning_rate": 3.922645043924478e-10,
+      "loss": 0.4437,
+      "step": 2576
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.309938776634853,
+      "learning_rate": 3.1773503793131043e-10,
+      "loss": 0.4454,
+      "step": 2577
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.795914729171075,
+      "learning_rate": 2.5105046456475047e-10,
+      "loss": 0.4762,
+      "step": 2578
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.229107836827541,
+      "learning_rate": 1.9221088892518169e-10,
+      "loss": 0.5075,
+      "step": 2579
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.283021826868119,
+      "learning_rate": 1.4121640333653042e-10,
+      "loss": 0.5212,
+      "step": 2580
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.90977803587339,
+      "learning_rate": 9.806708781368025e-11,
+      "loss": 0.4331,
+      "step": 2581
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.757252983445635,
+      "learning_rate": 6.276301006080676e-11,
+      "loss": 0.296,
+      "step": 2582
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.855438817022135,
+      "learning_rate": 3.530422547304291e-11,
+      "loss": 0.4407,
+      "step": 2583
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.64674490434668,
+      "learning_rate": 1.5690777135368707e-11,
+      "loss": 0.297,
+      "step": 2584
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.596192615058367,
+      "learning_rate": 3.922695822611289e-12,
+      "loss": 0.2754,
+      "step": 2585
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.510757250139918,
+      "learning_rate": 0.0,
+      "loss": 0.7747,
+      "step": 2586
+    },
+    {
+      "epoch": 1.0,
+      "step": 2586,
+      "total_flos": 0.0,
+      "train_loss": 0.6495711437418584,
+      "train_runtime": 45492.3958,
+      "train_samples_per_second": 14.556,
+      "train_steps_per_second": 0.057
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2586,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 0.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}