diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,35033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 50000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001,
+      "grad_norm": 85.6961898803711,
+      "learning_rate": 9.999e-06,
+      "loss": 0.8687,
+      "step": 10
+    },
+    {
+      "epoch": 0.002,
+      "grad_norm": 83.53839874267578,
+      "learning_rate": 9.997000000000001e-06,
+      "loss": 1.418,
+      "step": 20
+    },
+    {
+      "epoch": 0.003,
+      "grad_norm": 95.85935974121094,
+      "learning_rate": 9.995000000000002e-06,
+      "loss": 1.1484,
+      "step": 30
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 44.594295501708984,
+      "learning_rate": 9.993e-06,
+      "loss": 0.8815,
+      "step": 40
+    },
+    {
+      "epoch": 0.005,
+      "grad_norm": 116.1628189086914,
+      "learning_rate": 9.991000000000001e-06,
+      "loss": 1.2495,
+      "step": 50
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 33.84364318847656,
+      "learning_rate": 9.989e-06,
+      "loss": 0.6394,
+      "step": 60
+    },
+    {
+      "epoch": 0.007,
+      "grad_norm": 25.787431716918945,
+      "learning_rate": 9.987000000000001e-06,
+      "loss": 1.0792,
+      "step": 70
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 43.854469299316406,
+      "learning_rate": 9.985000000000002e-06,
+      "loss": 0.8326,
+      "step": 80
+    },
+    {
+      "epoch": 0.009,
+      "grad_norm": 101.5726547241211,
+      "learning_rate": 9.983e-06,
+      "loss": 1.2902,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 61.49679183959961,
+      "learning_rate": 9.981000000000002e-06,
+      "loss": 0.5976,
+      "step": 100
+    },
+    {
+      "epoch": 0.011,
+      "grad_norm": 83.0029525756836,
+      "learning_rate": 9.979e-06,
+      "loss": 1.0847,
+      "step": 110
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 81.3443374633789,
+      "learning_rate": 9.977000000000001e-06,
+      "loss": 0.9674,
+      "step": 120
+    },
+    {
+      "epoch": 0.013,
+      "grad_norm": 61.71631622314453,
+      "learning_rate": 9.975000000000002e-06,
+      "loss": 0.7648,
+      "step": 130
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 52.06660461425781,
+      "learning_rate": 9.973000000000001e-06,
+      "loss": 0.6408,
+      "step": 140
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 67.05174255371094,
+      "learning_rate": 9.971e-06,
+      "loss": 1.1016,
+      "step": 150
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 70.35164642333984,
+      "learning_rate": 9.969e-06,
+      "loss": 1.2591,
+      "step": 160
+    },
+    {
+      "epoch": 0.017,
+      "grad_norm": 113.0372085571289,
+      "learning_rate": 9.967000000000001e-06,
+      "loss": 0.8569,
+      "step": 170
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 46.71174621582031,
+      "learning_rate": 9.965000000000002e-06,
+      "loss": 1.0649,
+      "step": 180
+    },
+    {
+      "epoch": 0.019,
+      "grad_norm": 49.677127838134766,
+      "learning_rate": 9.963000000000001e-06,
+      "loss": 0.7769,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 44.4288444519043,
+      "learning_rate": 9.961e-06,
+      "loss": 0.8335,
+      "step": 200
+    },
+    {
+      "epoch": 0.021,
+      "grad_norm": 49.02952194213867,
+      "learning_rate": 9.959e-06,
+      "loss": 1.0531,
+      "step": 210
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 42.99825668334961,
+      "learning_rate": 9.957000000000001e-06,
+      "loss": 0.9388,
+      "step": 220
+    },
+    {
+      "epoch": 0.023,
+      "grad_norm": 33.82646560668945,
+      "learning_rate": 9.955000000000002e-06,
+      "loss": 0.8255,
+      "step": 230
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 31.44774627685547,
+      "learning_rate": 9.953000000000001e-06,
+      "loss": 0.8447,
+      "step": 240
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 46.26844024658203,
+      "learning_rate": 9.951e-06,
+      "loss": 0.8758,
+      "step": 250
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 32.29243469238281,
+      "learning_rate": 9.949e-06,
+      "loss": 0.5973,
+      "step": 260
+    },
+    {
+      "epoch": 0.027,
+      "grad_norm": 44.86991882324219,
+      "learning_rate": 9.947000000000001e-06,
+      "loss": 0.8255,
+      "step": 270
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 53.69334411621094,
+      "learning_rate": 9.945e-06,
+      "loss": 0.9522,
+      "step": 280
+    },
+    {
+      "epoch": 0.029,
+      "grad_norm": 65.80093383789062,
+      "learning_rate": 9.943000000000001e-06,
+      "loss": 0.8104,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 29.859413146972656,
+      "learning_rate": 9.941e-06,
+      "loss": 0.6795,
+      "step": 300
+    },
+    {
+      "epoch": 0.031,
+      "grad_norm": 37.98838424682617,
+      "learning_rate": 9.939000000000001e-06,
+      "loss": 0.7588,
+      "step": 310
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 16.02583885192871,
+      "learning_rate": 9.937000000000002e-06,
+      "loss": 0.7691,
+      "step": 320
+    },
+    {
+      "epoch": 0.033,
+      "grad_norm": 30.105602264404297,
+      "learning_rate": 9.935e-06,
+      "loss": 0.8934,
+      "step": 330
+    },
+    {
+      "epoch": 0.034,
+      "grad_norm": 27.63981819152832,
+      "learning_rate": 9.933e-06,
+      "loss": 1.0111,
+      "step": 340
+    },
+    {
+      "epoch": 0.035,
+      "grad_norm": 46.90068435668945,
+      "learning_rate": 9.931e-06,
+      "loss": 0.6812,
+      "step": 350
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 45.553165435791016,
+      "learning_rate": 9.929000000000001e-06,
+      "loss": 0.9058,
+      "step": 360
+    },
+    {
+      "epoch": 0.037,
+      "grad_norm": 30.0932559967041,
+      "learning_rate": 9.927000000000002e-06,
+      "loss": 0.6499,
+      "step": 370
+    },
+    {
+      "epoch": 0.038,
+      "grad_norm": 52.90816879272461,
+      "learning_rate": 9.925e-06,
+      "loss": 0.8783,
+      "step": 380
+    },
+    {
+      "epoch": 0.039,
+      "grad_norm": 31.532318115234375,
+      "learning_rate": 9.923e-06,
+      "loss": 0.8316,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.857837677001953,
+      "learning_rate": 9.921e-06,
+      "loss": 0.7904,
+      "step": 400
+    },
+    {
+      "epoch": 0.041,
+      "grad_norm": 32.60575866699219,
+      "learning_rate": 9.919000000000001e-06,
+      "loss": 0.5657,
+      "step": 410
+    },
+    {
+      "epoch": 0.042,
+      "grad_norm": 26.059589385986328,
+      "learning_rate": 9.917000000000002e-06,
+      "loss": 0.8572,
+      "step": 420
+    },
+    {
+      "epoch": 0.043,
+      "grad_norm": 39.00318908691406,
+      "learning_rate": 9.915e-06,
+      "loss": 0.713,
+      "step": 430
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 5.215452671051025,
+      "learning_rate": 9.913e-06,
+      "loss": 0.6709,
+      "step": 440
+    },
+    {
+      "epoch": 0.045,
+      "grad_norm": 15.460463523864746,
+      "learning_rate": 9.911e-06,
+      "loss": 0.639,
+      "step": 450
+    },
+    {
+      "epoch": 0.046,
+      "grad_norm": 60.24131774902344,
+      "learning_rate": 9.909000000000001e-06,
+      "loss": 0.8114,
+      "step": 460
+    },
+    {
+      "epoch": 0.047,
+      "grad_norm": 28.567232131958008,
+      "learning_rate": 9.907000000000002e-06,
+      "loss": 0.6119,
+      "step": 470
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 90.1940689086914,
+      "learning_rate": 9.905000000000001e-06,
+      "loss": 0.7393,
+      "step": 480
+    },
+    {
+      "epoch": 0.049,
+      "grad_norm": 60.29534912109375,
+      "learning_rate": 9.903e-06,
+      "loss": 0.587,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 29.43375587463379,
+      "learning_rate": 9.901e-06,
+      "loss": 0.8577,
+      "step": 500
+    },
+    {
+      "epoch": 0.051,
+      "grad_norm": 14.599111557006836,
+      "learning_rate": 9.899000000000001e-06,
+      "loss": 0.5109,
+      "step": 510
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 28.859493255615234,
+      "learning_rate": 9.897e-06,
+      "loss": 0.6818,
+      "step": 520
+    },
+    {
+      "epoch": 0.053,
+      "grad_norm": 91.49598693847656,
+      "learning_rate": 9.895000000000001e-06,
+      "loss": 0.7266,
+      "step": 530
+    },
+    {
+      "epoch": 0.054,
+      "grad_norm": 43.01456832885742,
+      "learning_rate": 9.893e-06,
+      "loss": 0.9112,
+      "step": 540
+    },
+    {
+      "epoch": 0.055,
+      "grad_norm": 89.42814636230469,
+      "learning_rate": 9.891e-06,
+      "loss": 0.9735,
+      "step": 550
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 31.47621726989746,
+      "learning_rate": 9.889000000000001e-06,
+      "loss": 0.7226,
+      "step": 560
+    },
+    {
+      "epoch": 0.057,
+      "grad_norm": 41.30959701538086,
+      "learning_rate": 9.887e-06,
+      "loss": 0.7916,
+      "step": 570
+    },
+    {
+      "epoch": 0.058,
+      "grad_norm": 46.82521438598633,
+      "learning_rate": 9.885000000000001e-06,
+      "loss": 0.7079,
+      "step": 580
+    },
+    {
+      "epoch": 0.059,
+      "grad_norm": 44.034061431884766,
+      "learning_rate": 9.8832e-06,
+      "loss": 0.7888,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 35.98831558227539,
+      "learning_rate": 9.881200000000001e-06,
+      "loss": 0.6411,
+      "step": 600
+    },
+    {
+      "epoch": 0.061,
+      "grad_norm": 46.98432540893555,
+      "learning_rate": 9.8792e-06,
+      "loss": 0.8364,
+      "step": 610
+    },
+    {
+      "epoch": 0.062,
+      "grad_norm": 52.23887252807617,
+      "learning_rate": 9.877200000000001e-06,
+      "loss": 0.8968,
+      "step": 620
+    },
+    {
+      "epoch": 0.063,
+      "grad_norm": 61.01598358154297,
+      "learning_rate": 9.8752e-06,
+      "loss": 0.8694,
+      "step": 630
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 37.633644104003906,
+      "learning_rate": 9.8732e-06,
+      "loss": 0.9208,
+      "step": 640
+    },
+    {
+      "epoch": 0.065,
+      "grad_norm": 29.999998092651367,
+      "learning_rate": 9.871200000000001e-06,
+      "loss": 0.6904,
+      "step": 650
+    },
+    {
+      "epoch": 0.066,
+      "grad_norm": 15.523207664489746,
+      "learning_rate": 9.8692e-06,
+      "loss": 0.7705,
+      "step": 660
+    },
+    {
+      "epoch": 0.067,
+      "grad_norm": 34.08747863769531,
+      "learning_rate": 9.867200000000001e-06,
+      "loss": 0.9855,
+      "step": 670
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 8.73726749420166,
+      "learning_rate": 9.8652e-06,
+      "loss": 0.6722,
+      "step": 680
+    },
+    {
+      "epoch": 0.069,
+      "grad_norm": 28.97559928894043,
+      "learning_rate": 9.8632e-06,
+      "loss": 0.7683,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 29.997812271118164,
+      "learning_rate": 9.861200000000001e-06,
+      "loss": 0.6177,
+      "step": 700
+    },
+    {
+      "epoch": 0.071,
+      "grad_norm": 25.937652587890625,
+      "learning_rate": 9.8592e-06,
+      "loss": 0.8951,
+      "step": 710
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 51.03562545776367,
+      "learning_rate": 9.857200000000001e-06,
+      "loss": 0.8368,
+      "step": 720
+    },
+    {
+      "epoch": 0.073,
+      "grad_norm": 16.005319595336914,
+      "learning_rate": 9.8552e-06,
+      "loss": 0.6646,
+      "step": 730
+    },
+    {
+      "epoch": 0.074,
+      "grad_norm": 34.2577018737793,
+      "learning_rate": 9.8532e-06,
+      "loss": 0.9092,
+      "step": 740
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 20.1041316986084,
+      "learning_rate": 9.851200000000001e-06,
+      "loss": 0.7914,
+      "step": 750
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 30.207904815673828,
+      "learning_rate": 9.8492e-06,
+      "loss": 0.9624,
+      "step": 760
+    },
+    {
+      "epoch": 0.077,
+      "grad_norm": 15.775735855102539,
+      "learning_rate": 9.847200000000001e-06,
+      "loss": 0.6117,
+      "step": 770
+    },
+    {
+      "epoch": 0.078,
+      "grad_norm": 18.047714233398438,
+      "learning_rate": 9.8452e-06,
+      "loss": 0.706,
+      "step": 780
+    },
+    {
+      "epoch": 0.079,
+      "grad_norm": 17.006519317626953,
+      "learning_rate": 9.843200000000001e-06,
+      "loss": 0.7123,
+      "step": 790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 28.868335723876953,
+      "learning_rate": 9.841200000000002e-06,
+      "loss": 0.7956,
+      "step": 800
+    },
+    {
+      "epoch": 0.081,
+      "grad_norm": 42.61716842651367,
+      "learning_rate": 9.8392e-06,
+      "loss": 0.7163,
+      "step": 810
+    },
+    {
+      "epoch": 0.082,
+      "grad_norm": 21.97026824951172,
+      "learning_rate": 9.837200000000001e-06,
+      "loss": 0.8151,
+      "step": 820
+    },
+    {
+      "epoch": 0.083,
+      "grad_norm": 14.37778091430664,
+      "learning_rate": 9.8354e-06,
+      "loss": 0.8179,
+      "step": 830
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 24.292551040649414,
+      "learning_rate": 9.833400000000001e-06,
+      "loss": 0.8949,
+      "step": 840
+    },
+    {
+      "epoch": 0.085,
+      "grad_norm": 31.621431350708008,
+      "learning_rate": 9.8314e-06,
+      "loss": 0.9193,
+      "step": 850
+    },
+    {
+      "epoch": 0.086,
+      "grad_norm": 22.210474014282227,
+      "learning_rate": 9.829400000000001e-06,
+      "loss": 0.7189,
+      "step": 860
+    },
+    {
+      "epoch": 0.087,
+      "grad_norm": 22.361305236816406,
+      "learning_rate": 9.8274e-06,
+      "loss": 0.7964,
+      "step": 870
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 20.3325252532959,
+      "learning_rate": 9.8254e-06,
+      "loss": 0.7291,
+      "step": 880
+    },
+    {
+      "epoch": 0.089,
+      "grad_norm": 34.202213287353516,
+      "learning_rate": 9.823400000000002e-06,
+      "loss": 0.7168,
+      "step": 890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 78.82975006103516,
+      "learning_rate": 9.8214e-06,
+      "loss": 0.7371,
+      "step": 900
+    },
+    {
+      "epoch": 0.091,
+      "grad_norm": 27.09836769104004,
+      "learning_rate": 9.819400000000001e-06,
+      "loss": 0.8064,
+      "step": 910
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 31.45117950439453,
+      "learning_rate": 9.8174e-06,
+      "loss": 0.9471,
+      "step": 920
+    },
+    {
+      "epoch": 0.093,
+      "grad_norm": 74.64904022216797,
+      "learning_rate": 9.815400000000001e-06,
+      "loss": 0.8619,
+      "step": 930
+    },
+    {
+      "epoch": 0.094,
+      "grad_norm": 21.551801681518555,
+      "learning_rate": 9.8134e-06,
+      "loss": 0.8292,
+      "step": 940
+    },
+    {
+      "epoch": 0.095,
+      "grad_norm": 12.729266166687012,
+      "learning_rate": 9.8114e-06,
+      "loss": 0.6445,
+      "step": 950
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 17.378259658813477,
+      "learning_rate": 9.809400000000001e-06,
+      "loss": 0.6488,
+      "step": 960
+    },
+    {
+      "epoch": 0.097,
+      "grad_norm": 35.8254508972168,
+      "learning_rate": 9.8074e-06,
+      "loss": 0.9045,
+      "step": 970
+    },
+    {
+      "epoch": 0.098,
+      "grad_norm": 37.96717071533203,
+      "learning_rate": 9.805400000000001e-06,
+      "loss": 0.8012,
+      "step": 980
+    },
+    {
+      "epoch": 0.099,
+      "grad_norm": 25.49050521850586,
+      "learning_rate": 9.8034e-06,
+      "loss": 1.0158,
+      "step": 990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.505203247070312,
+      "learning_rate": 9.8014e-06,
+      "loss": 0.7721,
+      "step": 1000
+    },
+    {
+      "epoch": 0.101,
+      "grad_norm": 13.868611335754395,
+      "learning_rate": 9.799400000000001e-06,
+      "loss": 0.6163,
+      "step": 1010
+    },
+    {
+      "epoch": 0.102,
+      "grad_norm": 23.564525604248047,
+      "learning_rate": 9.797400000000002e-06,
+      "loss": 0.7585,
+      "step": 1020
+    },
+    {
+      "epoch": 0.103,
+      "grad_norm": 23.647708892822266,
+      "learning_rate": 9.795400000000001e-06,
+      "loss": 0.5131,
+      "step": 1030
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 26.601909637451172,
+      "learning_rate": 9.7934e-06,
+      "loss": 0.8254,
+      "step": 1040
+    },
+    {
+      "epoch": 0.105,
+      "grad_norm": 13.147595405578613,
+      "learning_rate": 9.7914e-06,
+      "loss": 0.7593,
+      "step": 1050
+    },
+    {
+      "epoch": 0.106,
+      "grad_norm": 41.67106628417969,
+      "learning_rate": 9.789400000000002e-06,
+      "loss": 0.7138,
+      "step": 1060
+    },
+    {
+      "epoch": 0.107,
+      "grad_norm": 17.47908592224121,
+      "learning_rate": 9.7874e-06,
+      "loss": 0.7506,
+      "step": 1070
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 53.28046798706055,
+      "learning_rate": 9.785400000000001e-06,
+      "loss": 0.7343,
+      "step": 1080
+    },
+    {
+      "epoch": 0.109,
+      "grad_norm": 34.919734954833984,
+      "learning_rate": 9.7834e-06,
+      "loss": 0.8653,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 26.23090171813965,
+      "learning_rate": 9.781400000000001e-06,
+      "loss": 0.717,
+      "step": 1100
+    },
+    {
+      "epoch": 0.111,
+      "grad_norm": 22.868240356445312,
+      "learning_rate": 9.779400000000002e-06,
+      "loss": 0.5845,
+      "step": 1110
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 32.55518341064453,
+      "learning_rate": 9.7774e-06,
+      "loss": 0.8183,
+      "step": 1120
+    },
+    {
+      "epoch": 0.113,
+      "grad_norm": 14.902539253234863,
+      "learning_rate": 9.7754e-06,
+      "loss": 0.7217,
+      "step": 1130
+    },
+    {
+      "epoch": 0.114,
+      "grad_norm": 39.822017669677734,
+      "learning_rate": 9.7734e-06,
+      "loss": 0.6007,
+      "step": 1140
+    },
+    {
+      "epoch": 0.115,
+      "grad_norm": 94.34113311767578,
+      "learning_rate": 9.771400000000001e-06,
+      "loss": 0.6261,
+      "step": 1150
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 46.82630157470703,
+      "learning_rate": 9.769400000000002e-06,
+      "loss": 0.9339,
+      "step": 1160
+    },
+    {
+      "epoch": 0.117,
+      "grad_norm": 29.478790283203125,
+      "learning_rate": 9.7674e-06,
+      "loss": 0.5997,
+      "step": 1170
+    },
+    {
+      "epoch": 0.118,
+      "grad_norm": 24.89198875427246,
+      "learning_rate": 9.7654e-06,
+      "loss": 0.9387,
+      "step": 1180
+    },
+    {
+      "epoch": 0.119,
+      "grad_norm": 41.31000518798828,
+      "learning_rate": 9.7634e-06,
+      "loss": 0.7602,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 22.63460350036621,
+      "learning_rate": 9.761400000000001e-06,
+      "loss": 0.6417,
+      "step": 1200
+    },
+    {
+      "epoch": 0.121,
+      "grad_norm": 30.559040069580078,
+      "learning_rate": 9.759400000000002e-06,
+      "loss": 0.8557,
+      "step": 1210
+    },
+    {
+      "epoch": 0.122,
+      "grad_norm": 35.72277069091797,
+      "learning_rate": 9.7574e-06,
+      "loss": 0.7212,
+      "step": 1220
+    },
+    {
+      "epoch": 0.123,
+      "grad_norm": 19.481369018554688,
+      "learning_rate": 9.7554e-06,
+      "loss": 0.7667,
+      "step": 1230
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 5.373721599578857,
+      "learning_rate": 9.7534e-06,
+      "loss": 0.5497,
+      "step": 1240
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 40.84144592285156,
+      "learning_rate": 9.751400000000001e-06,
+      "loss": 0.717,
+      "step": 1250
+    },
+    {
+      "epoch": 0.126,
+      "grad_norm": 23.940317153930664,
+      "learning_rate": 9.749400000000002e-06,
+      "loss": 0.7161,
+      "step": 1260
+    },
+    {
+      "epoch": 0.127,
+      "grad_norm": 21.63367462158203,
+      "learning_rate": 9.747400000000001e-06,
+      "loss": 0.8505,
+      "step": 1270
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 24.01799774169922,
+      "learning_rate": 9.7454e-06,
+      "loss": 0.6953,
+      "step": 1280
+    },
+    {
+      "epoch": 0.129,
+      "grad_norm": 31.635465621948242,
+      "learning_rate": 9.7434e-06,
+      "loss": 0.7121,
+      "step": 1290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.7454776763916,
+      "learning_rate": 9.741400000000001e-06,
+      "loss": 0.8494,
+      "step": 1300
+    },
+    {
+      "epoch": 0.131,
+      "grad_norm": 25.52398109436035,
+      "learning_rate": 9.7394e-06,
+      "loss": 0.7229,
+      "step": 1310
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 30.034908294677734,
+      "learning_rate": 9.737400000000001e-06,
+      "loss": 0.7787,
+      "step": 1320
+    },
+    {
+      "epoch": 0.133,
+      "grad_norm": 11.845321655273438,
+      "learning_rate": 9.7354e-06,
+      "loss": 0.8125,
+      "step": 1330
+    },
+    {
+      "epoch": 0.134,
+      "grad_norm": 25.581470489501953,
+      "learning_rate": 9.7334e-06,
+      "loss": 0.619,
+      "step": 1340
+    },
+    {
+      "epoch": 0.135,
+      "grad_norm": 37.586360931396484,
+      "learning_rate": 9.731400000000001e-06,
+      "loss": 0.7928,
+      "step": 1350
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 40.47153091430664,
+      "learning_rate": 9.7294e-06,
+      "loss": 0.7839,
+      "step": 1360
+    },
+    {
+      "epoch": 0.137,
+      "grad_norm": 25.44156837463379,
+      "learning_rate": 9.727400000000001e-06,
+      "loss": 0.6141,
+      "step": 1370
+    },
+    {
+      "epoch": 0.138,
+      "grad_norm": 23.60976219177246,
+      "learning_rate": 9.7254e-06,
+      "loss": 0.9117,
+      "step": 1380
+    },
+    {
+      "epoch": 0.139,
+      "grad_norm": 23.967390060424805,
+      "learning_rate": 9.7234e-06,
+      "loss": 0.6394,
+      "step": 1390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.87657928466797,
+      "learning_rate": 9.721400000000001e-06,
+      "loss": 0.6832,
+      "step": 1400
+    },
+    {
+      "epoch": 0.141,
+      "grad_norm": 18.35723114013672,
+      "learning_rate": 9.7194e-06,
+      "loss": 0.6657,
+      "step": 1410
+    },
+    {
+      "epoch": 0.142,
+      "grad_norm": 44.30083465576172,
+      "learning_rate": 9.717400000000001e-06,
+      "loss": 0.7639,
+      "step": 1420
+    },
+    {
+      "epoch": 0.143,
+      "grad_norm": 30.185522079467773,
+      "learning_rate": 9.7154e-06,
+      "loss": 0.8301,
+      "step": 1430
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 16.693023681640625,
+      "learning_rate": 9.713400000000001e-06,
+      "loss": 0.7492,
+      "step": 1440
+    },
+    {
+      "epoch": 0.145,
+      "grad_norm": 46.640113830566406,
+      "learning_rate": 9.711400000000002e-06,
+      "loss": 0.7384,
+      "step": 1450
+    },
+    {
+      "epoch": 0.146,
+      "grad_norm": 14.902393341064453,
+      "learning_rate": 9.7094e-06,
+      "loss": 0.6725,
+      "step": 1460
+    },
+    {
+      "epoch": 0.147,
+      "grad_norm": 24.02662467956543,
+      "learning_rate": 9.7076e-06,
+      "loss": 0.7805,
+      "step": 1470
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 169.85626220703125,
+      "learning_rate": 9.7056e-06,
+      "loss": 0.9443,
+      "step": 1480
+    },
+    {
+      "epoch": 0.149,
+      "grad_norm": 25.04290771484375,
+      "learning_rate": 9.703600000000001e-06,
+      "loss": 0.7477,
+      "step": 1490
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 21.511314392089844,
+      "learning_rate": 9.7016e-06,
+      "loss": 0.6933,
+      "step": 1500
+    },
+    {
+      "epoch": 0.151,
+      "grad_norm": 34.140377044677734,
+      "learning_rate": 9.699600000000001e-06,
+      "loss": 0.7412,
+      "step": 1510
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 29.6060848236084,
+      "learning_rate": 9.6976e-06,
+      "loss": 0.669,
+      "step": 1520
+    },
+    {
+      "epoch": 0.153,
+      "grad_norm": 26.4044246673584,
+      "learning_rate": 9.6956e-06,
+      "loss": 0.7043,
+      "step": 1530
+    },
+    {
+      "epoch": 0.154,
+      "grad_norm": 35.81393814086914,
+      "learning_rate": 9.693600000000002e-06,
+      "loss": 0.6838,
+      "step": 1540
+    },
+    {
+      "epoch": 0.155,
+      "grad_norm": 31.620975494384766,
+      "learning_rate": 9.6916e-06,
+      "loss": 0.5365,
+      "step": 1550
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 25.63037109375,
+      "learning_rate": 9.689600000000001e-06,
+      "loss": 0.5885,
+      "step": 1560
+    },
+    {
+      "epoch": 0.157,
+      "grad_norm": 54.12228012084961,
+      "learning_rate": 9.6876e-06,
+      "loss": 0.8745,
+      "step": 1570
+    },
+    {
+      "epoch": 0.158,
+      "grad_norm": 32.007720947265625,
+      "learning_rate": 9.685600000000001e-06,
+      "loss": 0.8223,
+      "step": 1580
+    },
+    {
+      "epoch": 0.159,
+      "grad_norm": 29.423542022705078,
+      "learning_rate": 9.683600000000002e-06,
+      "loss": 0.8957,
+      "step": 1590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 20.0703125,
+      "learning_rate": 9.6816e-06,
+      "loss": 0.5619,
+      "step": 1600
+    },
+    {
+      "epoch": 0.161,
+      "grad_norm": 24.575084686279297,
+      "learning_rate": 9.679600000000001e-06,
+      "loss": 0.8571,
+      "step": 1610
+    },
+    {
+      "epoch": 0.162,
+      "grad_norm": 18.266128540039062,
+      "learning_rate": 9.6776e-06,
+      "loss": 0.7458,
+      "step": 1620
+    },
+    {
+      "epoch": 0.163,
+      "grad_norm": 16.532207489013672,
+      "learning_rate": 9.675600000000001e-06,
+      "loss": 0.7339,
+      "step": 1630
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 32.43312072753906,
+      "learning_rate": 9.6736e-06,
+      "loss": 0.8568,
+      "step": 1640
+    },
+    {
+      "epoch": 0.165,
+      "grad_norm": 28.956008911132812,
+      "learning_rate": 9.6716e-06,
+      "loss": 0.698,
+      "step": 1650
+    },
+    {
+      "epoch": 0.166,
+      "grad_norm": 22.16632652282715,
+      "learning_rate": 9.669600000000001e-06,
+      "loss": 0.6527,
+      "step": 1660
+    },
+    {
+      "epoch": 0.167,
+      "grad_norm": 38.9682502746582,
+      "learning_rate": 9.6676e-06,
+      "loss": 0.7314,
+      "step": 1670
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 35.68903350830078,
+      "learning_rate": 9.665600000000001e-06,
+      "loss": 1.1168,
+      "step": 1680
+    },
+    {
+      "epoch": 0.169,
+      "grad_norm": 7.560277462005615,
+      "learning_rate": 9.6636e-06,
+      "loss": 0.7461,
+      "step": 1690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.822694778442383,
+      "learning_rate": 9.6616e-06,
+      "loss": 0.7006,
+      "step": 1700
+    },
+    {
+      "epoch": 0.171,
+      "grad_norm": 36.905982971191406,
+      "learning_rate": 9.659600000000002e-06,
+      "loss": 0.5886,
+      "step": 1710
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 32.26405715942383,
+      "learning_rate": 9.6576e-06,
+      "loss": 0.7071,
+      "step": 1720
+    },
+    {
+      "epoch": 0.173,
+      "grad_norm": 27.82394027709961,
+      "learning_rate": 9.655600000000001e-06,
+      "loss": 0.6715,
+      "step": 1730
+    },
+    {
+      "epoch": 0.174,
+      "grad_norm": 30.092735290527344,
+      "learning_rate": 9.6536e-06,
+      "loss": 0.7702,
+      "step": 1740
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 55.14793395996094,
+      "learning_rate": 9.651600000000001e-06,
+      "loss": 0.9611,
+      "step": 1750
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 33.633914947509766,
+      "learning_rate": 9.649600000000002e-06,
+      "loss": 0.8827,
+      "step": 1760
+    },
+    {
+      "epoch": 0.177,
+      "grad_norm": 21.440275192260742,
+      "learning_rate": 9.6476e-06,
+      "loss": 0.6002,
+      "step": 1770
+    },
+    {
+      "epoch": 0.178,
+      "grad_norm": 54.77036666870117,
+      "learning_rate": 9.645600000000001e-06,
+      "loss": 0.6598,
+      "step": 1780
+    },
+    {
+      "epoch": 0.179,
+      "grad_norm": 9.992846488952637,
+      "learning_rate": 9.6436e-06,
+      "loss": 0.7752,
+      "step": 1790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.36663246154785,
+      "learning_rate": 9.641600000000001e-06,
+      "loss": 0.6992,
+      "step": 1800
+    },
+    {
+      "epoch": 0.181,
+      "grad_norm": 22.170259475708008,
+      "learning_rate": 9.639600000000002e-06,
+      "loss": 0.6805,
+      "step": 1810
+    },
+    {
+      "epoch": 0.182,
+      "grad_norm": 38.906578063964844,
+      "learning_rate": 9.6376e-06,
+      "loss": 0.8172,
+      "step": 1820
+    },
+    {
+      "epoch": 0.183,
+      "grad_norm": 54.6015739440918,
+      "learning_rate": 9.6356e-06,
+      "loss": 0.9855,
+      "step": 1830
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 59.32664108276367,
+      "learning_rate": 9.6336e-06,
+      "loss": 0.8278,
+      "step": 1840
+    },
+    {
+      "epoch": 0.185,
+      "grad_norm": 14.2304048538208,
+      "learning_rate": 9.631600000000001e-06,
+      "loss": 0.8763,
+      "step": 1850
+    },
+    {
+      "epoch": 0.186,
+      "grad_norm": 29.78812026977539,
+      "learning_rate": 9.629600000000002e-06,
+      "loss": 0.9618,
+      "step": 1860
+    },
+    {
+      "epoch": 0.187,
+      "grad_norm": 14.675698280334473,
+      "learning_rate": 9.6276e-06,
+      "loss": 0.6545,
+      "step": 1870
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 24.466876983642578,
+      "learning_rate": 9.6256e-06,
+      "loss": 0.7284,
+      "step": 1880
+    },
+    {
+      "epoch": 0.189,
+      "grad_norm": 7.637043476104736,
+      "learning_rate": 9.6236e-06,
+      "loss": 0.723,
+      "step": 1890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.321901321411133,
+      "learning_rate": 9.621600000000001e-06,
+      "loss": 0.8099,
+      "step": 1900
+    },
+    {
+      "epoch": 0.191,
+      "grad_norm": 18.512929916381836,
+      "learning_rate": 9.619600000000002e-06,
+      "loss": 0.7765,
+      "step": 1910
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 16.116796493530273,
+      "learning_rate": 9.617600000000001e-06,
+      "loss": 0.7287,
+      "step": 1920
+    },
+    {
+      "epoch": 0.193,
+      "grad_norm": 29.692922592163086,
+      "learning_rate": 9.6156e-06,
+      "loss": 0.8104,
+      "step": 1930
+    },
+    {
+      "epoch": 0.194,
+      "grad_norm": 34.71347427368164,
+      "learning_rate": 9.6136e-06,
+      "loss": 0.6277,
+      "step": 1940
+    },
+    {
+      "epoch": 0.195,
+      "grad_norm": 8.180668830871582,
+      "learning_rate": 9.611600000000001e-06,
+      "loss": 0.5576,
+      "step": 1950
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 24.448158264160156,
+      "learning_rate": 9.609600000000002e-06,
+      "loss": 0.6852,
+      "step": 1960
+    },
+    {
+      "epoch": 0.197,
+      "grad_norm": 29.8192081451416,
+      "learning_rate": 9.607600000000001e-06,
+      "loss": 0.5809,
+      "step": 1970
+    },
+    {
+      "epoch": 0.198,
+      "grad_norm": 25.80632972717285,
+      "learning_rate": 9.6056e-06,
+      "loss": 0.9618,
+      "step": 1980
+    },
+    {
+      "epoch": 0.199,
+      "grad_norm": 44.71590042114258,
+      "learning_rate": 9.6036e-06,
+      "loss": 0.9214,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 25.80580711364746,
+      "learning_rate": 9.601600000000001e-06,
+      "loss": 0.6859,
+      "step": 2000
+    },
+    {
+      "epoch": 0.201,
+      "grad_norm": 11.477659225463867,
+      "learning_rate": 9.5996e-06,
+      "loss": 0.6822,
+      "step": 2010
+    },
+    {
+      "epoch": 0.202,
+      "grad_norm": 25.939342498779297,
+      "learning_rate": 9.597600000000001e-06,
+      "loss": 0.7485,
+      "step": 2020
+    },
+    {
+      "epoch": 0.203,
+      "grad_norm": 30.334260940551758,
+      "learning_rate": 9.5956e-06,
+      "loss": 0.8371,
+      "step": 2030
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 34.063961029052734,
+      "learning_rate": 9.5936e-06,
+      "loss": 1.0189,
+      "step": 2040
+    },
+    {
+      "epoch": 0.205,
+      "grad_norm": 19.317014694213867,
+      "learning_rate": 9.591600000000001e-06,
+      "loss": 0.6944,
+      "step": 2050
+    },
+    {
+      "epoch": 0.206,
+      "grad_norm": 26.51333999633789,
+      "learning_rate": 9.5896e-06,
+      "loss": 0.6656,
+      "step": 2060
+    },
+    {
+      "epoch": 0.207,
+      "grad_norm": 10.476828575134277,
+      "learning_rate": 9.5876e-06,
+      "loss": 0.6312,
+      "step": 2070
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 27.542985916137695,
+      "learning_rate": 9.5856e-06,
+      "loss": 0.5976,
+      "step": 2080
+    },
+    {
+      "epoch": 0.209,
+      "grad_norm": 31.365577697753906,
+      "learning_rate": 9.583600000000001e-06,
+      "loss": 0.9775,
+      "step": 2090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 40.898616790771484,
+      "learning_rate": 9.581600000000002e-06,
+      "loss": 0.5258,
+      "step": 2100
+    },
+    {
+      "epoch": 0.211,
+      "grad_norm": 15.9881591796875,
+      "learning_rate": 9.5796e-06,
+      "loss": 0.7965,
+      "step": 2110
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 25.101795196533203,
+      "learning_rate": 9.5776e-06,
+      "loss": 0.5959,
+      "step": 2120
+    },
+    {
+      "epoch": 0.213,
+      "grad_norm": 39.432334899902344,
+      "learning_rate": 9.5756e-06,
+      "loss": 0.8365,
+      "step": 2130
+    },
+    {
+      "epoch": 0.214,
+      "grad_norm": 41.93818283081055,
+      "learning_rate": 9.573600000000001e-06,
+      "loss": 0.7073,
+      "step": 2140
+    },
+    {
+      "epoch": 0.215,
+      "grad_norm": 5.648146629333496,
+      "learning_rate": 9.571600000000002e-06,
+      "loss": 0.7014,
+      "step": 2150
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 35.35142517089844,
+      "learning_rate": 9.5696e-06,
+      "loss": 0.7572,
+      "step": 2160
+    },
+    {
+      "epoch": 0.217,
+      "grad_norm": 23.585378646850586,
+      "learning_rate": 9.567600000000001e-06,
+      "loss": 0.7152,
+      "step": 2170
+    },
+    {
+      "epoch": 0.218,
+      "grad_norm": 19.717735290527344,
+      "learning_rate": 9.5656e-06,
+      "loss": 0.6492,
+      "step": 2180
+    },
+    {
+      "epoch": 0.219,
+      "grad_norm": 10.743090629577637,
+      "learning_rate": 9.563600000000001e-06,
+      "loss": 0.7152,
+      "step": 2190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 33.635536193847656,
+      "learning_rate": 9.5616e-06,
+      "loss": 0.6707,
+      "step": 2200
+    },
+    {
+      "epoch": 0.221,
+      "grad_norm": 24.042377471923828,
+      "learning_rate": 9.5596e-06,
+      "loss": 0.5237,
+      "step": 2210
+    },
+    {
+      "epoch": 0.222,
+      "grad_norm": 48.697303771972656,
+      "learning_rate": 9.557600000000001e-06,
+      "loss": 0.7784,
+      "step": 2220
+    },
+    {
+      "epoch": 0.223,
+      "grad_norm": 62.7592658996582,
+      "learning_rate": 9.5556e-06,
+      "loss": 0.5507,
+      "step": 2230
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 38.88966369628906,
+      "learning_rate": 9.553600000000001e-06,
+      "loss": 0.8598,
+      "step": 2240
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 29.976181030273438,
+      "learning_rate": 9.5516e-06,
+      "loss": 0.8038,
+      "step": 2250
+    },
+    {
+      "epoch": 0.226,
+      "grad_norm": 18.87337875366211,
+      "learning_rate": 9.549600000000001e-06,
+      "loss": 0.4878,
+      "step": 2260
+    },
+    {
+      "epoch": 0.227,
+      "grad_norm": 38.70804214477539,
+      "learning_rate": 9.547600000000002e-06,
+      "loss": 0.8455,
+      "step": 2270
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 9.813104629516602,
+      "learning_rate": 9.5456e-06,
+      "loss": 0.4139,
+      "step": 2280
+    },
+    {
+      "epoch": 0.229,
+      "grad_norm": 55.018157958984375,
+      "learning_rate": 9.543600000000001e-06,
+      "loss": 1.0867,
+      "step": 2290
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.863101005554199,
+      "learning_rate": 9.5416e-06,
+      "loss": 0.6808,
+      "step": 2300
+    },
+    {
+      "epoch": 0.231,
+      "grad_norm": 21.807111740112305,
+      "learning_rate": 9.539600000000001e-06,
+      "loss": 0.9384,
+      "step": 2310
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 27.61972999572754,
+      "learning_rate": 9.537600000000002e-06,
+      "loss": 0.7813,
+      "step": 2320
+    },
+    {
+      "epoch": 0.233,
+      "grad_norm": 48.68235397338867,
+      "learning_rate": 9.5356e-06,
+      "loss": 0.81,
+      "step": 2330
+    },
+    {
+      "epoch": 0.234,
+      "grad_norm": 26.039478302001953,
+      "learning_rate": 9.533600000000001e-06,
+      "loss": 0.7981,
+      "step": 2340
+    },
+    {
+      "epoch": 0.235,
+      "grad_norm": 25.755840301513672,
+      "learning_rate": 9.5316e-06,
+      "loss": 0.7751,
+      "step": 2350
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 25.323991775512695,
+      "learning_rate": 9.529600000000001e-06,
+      "loss": 0.8628,
+      "step": 2360
+    },
+    {
+      "epoch": 0.237,
+      "grad_norm": 25.357213973999023,
+      "learning_rate": 9.527600000000002e-06,
+      "loss": 0.6644,
+      "step": 2370
+    },
+    {
+      "epoch": 0.238,
+      "grad_norm": 49.66263198852539,
+      "learning_rate": 9.5256e-06,
+      "loss": 0.6991,
+      "step": 2380
+    },
+    {
+      "epoch": 0.239,
+      "grad_norm": 23.181354522705078,
+      "learning_rate": 9.523600000000001e-06,
+      "loss": 0.9643,
+      "step": 2390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.877561569213867,
+      "learning_rate": 9.5216e-06,
+      "loss": 0.9063,
+      "step": 2400
+    },
+    {
+      "epoch": 0.241,
+      "grad_norm": 8.216836929321289,
+      "learning_rate": 9.519600000000001e-06,
+      "loss": 0.823,
+      "step": 2410
+    },
+    {
+      "epoch": 0.242,
+      "grad_norm": 16.65896987915039,
+      "learning_rate": 9.517600000000002e-06,
+      "loss": 0.587,
+      "step": 2420
+    },
+    {
+      "epoch": 0.243,
+      "grad_norm": 12.24148178100586,
+      "learning_rate": 9.515600000000001e-06,
+      "loss": 0.5954,
+      "step": 2430
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 24.519540786743164,
+      "learning_rate": 9.5136e-06,
+      "loss": 0.862,
+      "step": 2440
+    },
+    {
+      "epoch": 0.245,
+      "grad_norm": 6.640892028808594,
+      "learning_rate": 9.5116e-06,
+      "loss": 0.8153,
+      "step": 2450
+    },
+    {
+      "epoch": 0.246,
+      "grad_norm": 35.18092346191406,
+      "learning_rate": 9.509600000000001e-06,
+      "loss": 0.8742,
+      "step": 2460
+    },
+    {
+      "epoch": 0.247,
+      "grad_norm": 33.762699127197266,
+      "learning_rate": 9.507600000000002e-06,
+      "loss": 1.0229,
+      "step": 2470
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 22.815547943115234,
+      "learning_rate": 9.505600000000001e-06,
+      "loss": 0.7814,
+      "step": 2480
+    },
+    {
+      "epoch": 0.249,
+      "grad_norm": 46.83992385864258,
+      "learning_rate": 9.5036e-06,
+      "loss": 0.8382,
+      "step": 2490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.844913482666016,
+      "learning_rate": 9.5016e-06,
+      "loss": 0.5333,
+      "step": 2500
+    },
+    {
+      "epoch": 0.251,
+      "grad_norm": 18.164127349853516,
+      "learning_rate": 9.499600000000001e-06,
+      "loss": 0.6708,
+      "step": 2510
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 40.63623046875,
+      "learning_rate": 9.497600000000002e-06,
+      "loss": 0.7313,
+      "step": 2520
+    },
+    {
+      "epoch": 0.253,
+      "grad_norm": 32.359256744384766,
+      "learning_rate": 9.495600000000001e-06,
+      "loss": 0.7149,
+      "step": 2530
+    },
+    {
+      "epoch": 0.254,
+      "grad_norm": 22.28809356689453,
+      "learning_rate": 9.4936e-06,
+      "loss": 0.8074,
+      "step": 2540
+    },
+    {
+      "epoch": 0.255,
+      "grad_norm": 19.11900520324707,
+      "learning_rate": 9.4916e-06,
+      "loss": 0.8352,
+      "step": 2550
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 44.57843017578125,
+      "learning_rate": 9.489600000000001e-06,
+      "loss": 0.6453,
+      "step": 2560
+    },
+    {
+      "epoch": 0.257,
+      "grad_norm": 13.637824058532715,
+      "learning_rate": 9.4876e-06,
+      "loss": 0.7488,
+      "step": 2570
+    },
+    {
+      "epoch": 0.258,
+      "grad_norm": 13.40095043182373,
+      "learning_rate": 9.485600000000001e-06,
+      "loss": 0.7969,
+      "step": 2580
+    },
+    {
+      "epoch": 0.259,
+      "grad_norm": 31.110763549804688,
+      "learning_rate": 9.4836e-06,
+      "loss": 0.8985,
+      "step": 2590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 21.71891212463379,
+      "learning_rate": 9.4816e-06,
+      "loss": 0.5378,
+      "step": 2600
+    },
+    {
+      "epoch": 0.261,
+      "grad_norm": 43.970008850097656,
+      "learning_rate": 9.479600000000002e-06,
+      "loss": 0.7833,
+      "step": 2610
+    },
+    {
+      "epoch": 0.262,
+      "grad_norm": 25.333736419677734,
+      "learning_rate": 9.4776e-06,
+      "loss": 0.8815,
+      "step": 2620
+    },
+    {
+      "epoch": 0.263,
+      "grad_norm": 22.61463737487793,
+      "learning_rate": 9.4756e-06,
+      "loss": 0.6996,
+      "step": 2630
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 29.46462059020996,
+      "learning_rate": 9.4736e-06,
+      "loss": 0.7356,
+      "step": 2640
+    },
+    {
+      "epoch": 0.265,
+      "grad_norm": 19.242321014404297,
+      "learning_rate": 9.471600000000001e-06,
+      "loss": 0.9462,
+      "step": 2650
+    },
+    {
+      "epoch": 0.266,
+      "grad_norm": 29.384977340698242,
+      "learning_rate": 9.469600000000002e-06,
+      "loss": 0.8388,
+      "step": 2660
+    },
+    {
+      "epoch": 0.267,
+      "grad_norm": 25.901931762695312,
+      "learning_rate": 9.4676e-06,
+      "loss": 0.5984,
+      "step": 2670
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 18.902681350708008,
+      "learning_rate": 9.4656e-06,
+      "loss": 0.6226,
+      "step": 2680
+    },
+    {
+      "epoch": 0.269,
+      "grad_norm": 38.78340148925781,
+      "learning_rate": 9.4636e-06,
+      "loss": 0.729,
+      "step": 2690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 29.76132583618164,
+      "learning_rate": 9.461600000000001e-06,
+      "loss": 0.8336,
+      "step": 2700
+    },
+    {
+      "epoch": 0.271,
+      "grad_norm": 26.546491622924805,
+      "learning_rate": 9.459600000000002e-06,
+      "loss": 0.7328,
+      "step": 2710
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 14.950640678405762,
+      "learning_rate": 9.4576e-06,
+      "loss": 0.6283,
+      "step": 2720
+    },
+    {
+      "epoch": 0.273,
+      "grad_norm": 23.570337295532227,
+      "learning_rate": 9.4556e-06,
+      "loss": 0.7588,
+      "step": 2730
+    },
+    {
+      "epoch": 0.274,
+      "grad_norm": 33.28166961669922,
+      "learning_rate": 9.4536e-06,
+      "loss": 0.8514,
+      "step": 2740
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 62.68339920043945,
+      "learning_rate": 9.451600000000001e-06,
+      "loss": 0.6574,
+      "step": 2750
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 20.3784236907959,
+      "learning_rate": 9.449600000000002e-06,
+      "loss": 1.0182,
+      "step": 2760
+    },
+    {
+      "epoch": 0.277,
+      "grad_norm": 8.02643871307373,
+      "learning_rate": 9.4476e-06,
+      "loss": 0.8452,
+      "step": 2770
+    },
+    {
+      "epoch": 0.278,
+      "grad_norm": 15.038368225097656,
+      "learning_rate": 9.4456e-06,
+      "loss": 0.7439,
+      "step": 2780
+    },
+    {
+      "epoch": 0.279,
+      "grad_norm": 30.92201042175293,
+      "learning_rate": 9.4436e-06,
+      "loss": 0.9571,
+      "step": 2790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 28.545917510986328,
+      "learning_rate": 9.441600000000001e-06,
+      "loss": 0.734,
+      "step": 2800
+    },
+    {
+      "epoch": 0.281,
+      "grad_norm": 23.49970054626465,
+      "learning_rate": 9.4396e-06,
+      "loss": 0.9253,
+      "step": 2810
+    },
+    {
+      "epoch": 0.282,
+      "grad_norm": 24.050922393798828,
+      "learning_rate": 9.437600000000001e-06,
+      "loss": 0.8319,
+      "step": 2820
+    },
+    {
+      "epoch": 0.283,
+      "grad_norm": 13.687472343444824,
+      "learning_rate": 9.4356e-06,
+      "loss": 0.8918,
+      "step": 2830
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 13.137202262878418,
+      "learning_rate": 9.4336e-06,
+      "loss": 0.9102,
+      "step": 2840
+    },
+    {
+      "epoch": 0.285,
+      "grad_norm": 10.729533195495605,
+      "learning_rate": 9.431600000000001e-06,
+      "loss": 0.7587,
+      "step": 2850
+    },
+    {
+      "epoch": 0.286,
+      "grad_norm": 30.572080612182617,
+      "learning_rate": 9.4296e-06,
+      "loss": 0.8736,
+      "step": 2860
+    },
+    {
+      "epoch": 0.287,
+      "grad_norm": 53.83828353881836,
+      "learning_rate": 9.427600000000001e-06,
+      "loss": 0.8348,
+      "step": 2870
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 14.523385047912598,
+      "learning_rate": 9.4256e-06,
+      "loss": 0.8429,
+      "step": 2880
+    },
+    {
+      "epoch": 0.289,
+      "grad_norm": 22.808746337890625,
+      "learning_rate": 9.4236e-06,
+      "loss": 0.7983,
+      "step": 2890
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.653146743774414,
+      "learning_rate": 9.421600000000001e-06,
+      "loss": 0.7738,
+      "step": 2900
+    },
+    {
+      "epoch": 0.291,
+      "grad_norm": 11.08884334564209,
+      "learning_rate": 9.4196e-06,
+      "loss": 0.8135,
+      "step": 2910
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 11.973413467407227,
+      "learning_rate": 9.417600000000001e-06,
+      "loss": 0.703,
+      "step": 2920
+    },
+    {
+      "epoch": 0.293,
+      "grad_norm": 41.609397888183594,
+      "learning_rate": 9.4156e-06,
+      "loss": 0.5246,
+      "step": 2930
+    },
+    {
+      "epoch": 0.294,
+      "grad_norm": 7.794600963592529,
+      "learning_rate": 9.4136e-06,
+      "loss": 0.6844,
+      "step": 2940
+    },
+    {
+      "epoch": 0.295,
+      "grad_norm": 47.65550231933594,
+      "learning_rate": 9.411600000000002e-06,
+      "loss": 0.7167,
+      "step": 2950
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 28.062644958496094,
+      "learning_rate": 9.4096e-06,
+      "loss": 0.7536,
+      "step": 2960
+    },
+    {
+      "epoch": 0.297,
+      "grad_norm": 11.334671020507812,
+      "learning_rate": 9.407600000000001e-06,
+      "loss": 0.7353,
+      "step": 2970
+    },
+    {
+      "epoch": 0.298,
+      "grad_norm": 38.403076171875,
+      "learning_rate": 9.4056e-06,
+      "loss": 0.8809,
+      "step": 2980
+    },
+    {
+      "epoch": 0.299,
+      "grad_norm": 21.346092224121094,
+      "learning_rate": 9.403600000000001e-06,
+      "loss": 0.5042,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.392056465148926,
+      "learning_rate": 9.4016e-06,
+      "loss": 0.5377,
+      "step": 3000
+    },
+    {
+      "epoch": 0.301,
+      "grad_norm": 28.40464210510254,
+      "learning_rate": 9.3996e-06,
+      "loss": 0.8087,
+      "step": 3010
+    },
+    {
+      "epoch": 0.302,
+      "grad_norm": 38.49671173095703,
+      "learning_rate": 9.397600000000001e-06,
+      "loss": 0.8251,
+      "step": 3020
+    },
+    {
+      "epoch": 0.303,
+      "grad_norm": 22.052934646606445,
+      "learning_rate": 9.3956e-06,
+      "loss": 0.6006,
+      "step": 3030
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 44.930458068847656,
+      "learning_rate": 9.393600000000001e-06,
+      "loss": 0.9245,
+      "step": 3040
+    },
+    {
+      "epoch": 0.305,
+      "grad_norm": 56.81517028808594,
+      "learning_rate": 9.3916e-06,
+      "loss": 0.6957,
+      "step": 3050
+    },
+    {
+      "epoch": 0.306,
+      "grad_norm": 43.52264404296875,
+      "learning_rate": 9.3896e-06,
+      "loss": 0.7816,
+      "step": 3060
+    },
+    {
+      "epoch": 0.307,
+      "grad_norm": 24.788782119750977,
+      "learning_rate": 9.387600000000001e-06,
+      "loss": 0.7707,
+      "step": 3070
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 40.297454833984375,
+      "learning_rate": 9.3856e-06,
+      "loss": 0.6905,
+      "step": 3080
+    },
+    {
+      "epoch": 0.309,
+      "grad_norm": 16.154659271240234,
+      "learning_rate": 9.383600000000001e-06,
+      "loss": 0.6413,
+      "step": 3090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 24.906816482543945,
+      "learning_rate": 9.3816e-06,
+      "loss": 0.6919,
+      "step": 3100
+    },
+    {
+      "epoch": 0.311,
+      "grad_norm": 9.070622444152832,
+      "learning_rate": 9.3796e-06,
+      "loss": 0.7292,
+      "step": 3110
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 25.71718406677246,
+      "learning_rate": 9.377600000000001e-06,
+      "loss": 0.767,
+      "step": 3120
+    },
+    {
+      "epoch": 0.313,
+      "grad_norm": 19.88960075378418,
+      "learning_rate": 9.3756e-06,
+      "loss": 0.8662,
+      "step": 3130
+    },
+    {
+      "epoch": 0.314,
+      "grad_norm": 21.357192993164062,
+      "learning_rate": 9.373600000000001e-06,
+      "loss": 0.5693,
+      "step": 3140
+    },
+    {
+      "epoch": 0.315,
+      "grad_norm": 39.57415771484375,
+      "learning_rate": 9.3716e-06,
+      "loss": 0.8721,
+      "step": 3150
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 24.280065536499023,
+      "learning_rate": 9.369600000000001e-06,
+      "loss": 0.7282,
+      "step": 3160
+    },
+    {
+      "epoch": 0.317,
+      "grad_norm": 34.4137077331543,
+      "learning_rate": 9.367600000000002e-06,
+      "loss": 0.9604,
+      "step": 3170
+    },
+    {
+      "epoch": 0.318,
+      "grad_norm": 28.870506286621094,
+      "learning_rate": 9.3656e-06,
+      "loss": 0.7249,
+      "step": 3180
+    },
+    {
+      "epoch": 0.319,
+      "grad_norm": 31.467092514038086,
+      "learning_rate": 9.363600000000001e-06,
+      "loss": 0.7248,
+      "step": 3190
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 50.64829635620117,
+      "learning_rate": 9.3616e-06,
+      "loss": 0.8032,
+      "step": 3200
+    },
+    {
+      "epoch": 0.321,
+      "grad_norm": 33.74970626831055,
+      "learning_rate": 9.359600000000001e-06,
+      "loss": 0.7636,
+      "step": 3210
+    },
+    {
+      "epoch": 0.322,
+      "grad_norm": 20.4691162109375,
+      "learning_rate": 9.357600000000002e-06,
+      "loss": 0.7254,
+      "step": 3220
+    },
+    {
+      "epoch": 0.323,
+      "grad_norm": 28.088464736938477,
+      "learning_rate": 9.3556e-06,
+      "loss": 0.4999,
+      "step": 3230
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 35.50647735595703,
+      "learning_rate": 9.3536e-06,
+      "loss": 0.5983,
+      "step": 3240
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 26.18625259399414,
+      "learning_rate": 9.3516e-06,
+      "loss": 0.8037,
+      "step": 3250
+    },
+    {
+      "epoch": 0.326,
+      "grad_norm": 33.88105392456055,
+      "learning_rate": 9.349600000000001e-06,
+      "loss": 0.6124,
+      "step": 3260
+    },
+    {
+      "epoch": 0.327,
+      "grad_norm": 28.297372817993164,
+      "learning_rate": 9.347600000000002e-06,
+      "loss": 0.82,
+      "step": 3270
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 26.45084571838379,
+      "learning_rate": 9.3456e-06,
+      "loss": 0.7743,
+      "step": 3280
+    },
+    {
+      "epoch": 0.329,
+      "grad_norm": 45.73817443847656,
+      "learning_rate": 9.3436e-06,
+      "loss": 0.589,
+      "step": 3290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 30.2762508392334,
+      "learning_rate": 9.3416e-06,
+      "loss": 0.6909,
+      "step": 3300
+    },
+    {
+      "epoch": 0.331,
+      "grad_norm": 16.09083366394043,
+      "learning_rate": 9.339600000000001e-06,
+      "loss": 0.5877,
+      "step": 3310
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 28.67357635498047,
+      "learning_rate": 9.337600000000002e-06,
+      "loss": 0.7198,
+      "step": 3320
+    },
+    {
+      "epoch": 0.333,
+      "grad_norm": 33.21657943725586,
+      "learning_rate": 9.335600000000001e-06,
+      "loss": 0.5661,
+      "step": 3330
+    },
+    {
+      "epoch": 0.334,
+      "grad_norm": 30.5037841796875,
+      "learning_rate": 9.3336e-06,
+      "loss": 0.7943,
+      "step": 3340
+    },
+    {
+      "epoch": 0.335,
+      "grad_norm": 34.517642974853516,
+      "learning_rate": 9.3316e-06,
+      "loss": 0.6678,
+      "step": 3350
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 47.0224723815918,
+      "learning_rate": 9.329600000000001e-06,
+      "loss": 0.7275,
+      "step": 3360
+    },
+    {
+      "epoch": 0.337,
+      "grad_norm": 33.63166427612305,
+      "learning_rate": 9.3276e-06,
+      "loss": 0.6178,
+      "step": 3370
+    },
+    {
+      "epoch": 0.338,
+      "grad_norm": 16.433677673339844,
+      "learning_rate": 9.325600000000001e-06,
+      "loss": 0.8244,
+      "step": 3380
+    },
+    {
+      "epoch": 0.339,
+      "grad_norm": 31.32615089416504,
+      "learning_rate": 9.3236e-06,
+      "loss": 0.7691,
+      "step": 3390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.45935821533203,
+      "learning_rate": 9.3216e-06,
+      "loss": 0.5259,
+      "step": 3400
+    },
+    {
+      "epoch": 0.341,
+      "grad_norm": 10.065863609313965,
+      "learning_rate": 9.319600000000001e-06,
+      "loss": 0.6325,
+      "step": 3410
+    },
+    {
+      "epoch": 0.342,
+      "grad_norm": 19.148534774780273,
+      "learning_rate": 9.3176e-06,
+      "loss": 0.5141,
+      "step": 3420
+    },
+    {
+      "epoch": 0.343,
+      "grad_norm": 30.4162654876709,
+      "learning_rate": 9.315600000000001e-06,
+      "loss": 0.7945,
+      "step": 3430
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 53.84823226928711,
+      "learning_rate": 9.3136e-06,
+      "loss": 0.7548,
+      "step": 3440
+    },
+    {
+      "epoch": 0.345,
+      "grad_norm": 3.4882357120513916,
+      "learning_rate": 9.3116e-06,
+      "loss": 0.5881,
+      "step": 3450
+    },
+    {
+      "epoch": 0.346,
+      "grad_norm": 53.241600036621094,
+      "learning_rate": 9.309600000000001e-06,
+      "loss": 0.6289,
+      "step": 3460
+    },
+    {
+      "epoch": 0.347,
+      "grad_norm": 31.7698917388916,
+      "learning_rate": 9.3076e-06,
+      "loss": 0.7232,
+      "step": 3470
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 13.279865264892578,
+      "learning_rate": 9.305600000000001e-06,
+      "loss": 0.5825,
+      "step": 3480
+    },
+    {
+      "epoch": 0.349,
+      "grad_norm": 14.930980682373047,
+      "learning_rate": 9.3036e-06,
+      "loss": 0.3373,
+      "step": 3490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 44.31536865234375,
+      "learning_rate": 9.301600000000001e-06,
+      "loss": 0.9145,
+      "step": 3500
+    },
+    {
+      "epoch": 0.351,
+      "grad_norm": 11.612078666687012,
+      "learning_rate": 9.299600000000002e-06,
+      "loss": 0.7549,
+      "step": 3510
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 19.171342849731445,
+      "learning_rate": 9.2976e-06,
+      "loss": 0.6653,
+      "step": 3520
+    },
+    {
+      "epoch": 0.353,
+      "grad_norm": 13.827188491821289,
+      "learning_rate": 9.295600000000001e-06,
+      "loss": 0.8802,
+      "step": 3530
+    },
+    {
+      "epoch": 0.354,
+      "grad_norm": 17.23624610900879,
+      "learning_rate": 9.2936e-06,
+      "loss": 0.8222,
+      "step": 3540
+    },
+    {
+      "epoch": 0.355,
+      "grad_norm": 17.355684280395508,
+      "learning_rate": 9.291600000000001e-06,
+      "loss": 0.5068,
+      "step": 3550
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 34.479156494140625,
+      "learning_rate": 9.289600000000002e-06,
+      "loss": 0.784,
+      "step": 3560
+    },
+    {
+      "epoch": 0.357,
+      "grad_norm": 29.13589859008789,
+      "learning_rate": 9.2876e-06,
+      "loss": 0.7737,
+      "step": 3570
+    },
+    {
+      "epoch": 0.358,
+      "grad_norm": 33.8214225769043,
+      "learning_rate": 9.285600000000001e-06,
+      "loss": 0.8217,
+      "step": 3580
+    },
+    {
+      "epoch": 0.359,
+      "grad_norm": 41.37405776977539,
+      "learning_rate": 9.2836e-06,
+      "loss": 0.7403,
+      "step": 3590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 31.224672317504883,
+      "learning_rate": 9.281600000000001e-06,
+      "loss": 0.692,
+      "step": 3600
+    },
+    {
+      "epoch": 0.361,
+      "grad_norm": 21.726320266723633,
+      "learning_rate": 9.2796e-06,
+      "loss": 0.6739,
+      "step": 3610
+    },
+    {
+      "epoch": 0.362,
+      "grad_norm": 3.601778030395508,
+      "learning_rate": 9.2776e-06,
+      "loss": 0.5509,
+      "step": 3620
+    },
+    {
+      "epoch": 0.363,
+      "grad_norm": 34.594703674316406,
+      "learning_rate": 9.275600000000001e-06,
+      "loss": 0.9467,
+      "step": 3630
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 14.18492603302002,
+      "learning_rate": 9.2736e-06,
+      "loss": 0.6881,
+      "step": 3640
+    },
+    {
+      "epoch": 0.365,
+      "grad_norm": 10.007046699523926,
+      "learning_rate": 9.271600000000001e-06,
+      "loss": 0.5958,
+      "step": 3650
+    },
+    {
+      "epoch": 0.366,
+      "grad_norm": 26.733074188232422,
+      "learning_rate": 9.2696e-06,
+      "loss": 0.7343,
+      "step": 3660
+    },
+    {
+      "epoch": 0.367,
+      "grad_norm": 14.416747093200684,
+      "learning_rate": 9.267600000000001e-06,
+      "loss": 0.8798,
+      "step": 3670
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 26.770458221435547,
+      "learning_rate": 9.265600000000002e-06,
+      "loss": 0.7193,
+      "step": 3680
+    },
+    {
+      "epoch": 0.369,
+      "grad_norm": 15.376206398010254,
+      "learning_rate": 9.2636e-06,
+      "loss": 0.7509,
+      "step": 3690
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 20.67218589782715,
+      "learning_rate": 9.261600000000001e-06,
+      "loss": 0.6673,
+      "step": 3700
+    },
+    {
+      "epoch": 0.371,
+      "grad_norm": 26.45255470275879,
+      "learning_rate": 9.2596e-06,
+      "loss": 0.8062,
+      "step": 3710
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 29.3276309967041,
+      "learning_rate": 9.257600000000001e-06,
+      "loss": 0.6185,
+      "step": 3720
+    },
+    {
+      "epoch": 0.373,
+      "grad_norm": 24.602161407470703,
+      "learning_rate": 9.255600000000002e-06,
+      "loss": 0.7317,
+      "step": 3730
+    },
+    {
+      "epoch": 0.374,
+      "grad_norm": 29.78669548034668,
+      "learning_rate": 9.2536e-06,
+      "loss": 0.863,
+      "step": 3740
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 16.003917694091797,
+      "learning_rate": 9.251600000000001e-06,
+      "loss": 0.6759,
+      "step": 3750
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 5.657262325286865,
+      "learning_rate": 9.2496e-06,
+      "loss": 0.3055,
+      "step": 3760
+    },
+    {
+      "epoch": 0.377,
+      "grad_norm": 83.36643981933594,
+      "learning_rate": 9.247600000000001e-06,
+      "loss": 0.7421,
+      "step": 3770
+    },
+    {
+      "epoch": 0.378,
+      "grad_norm": 36.70709991455078,
+      "learning_rate": 9.245600000000002e-06,
+      "loss": 0.7514,
+      "step": 3780
+    },
+    {
+      "epoch": 0.379,
+      "grad_norm": 38.573974609375,
+      "learning_rate": 9.2436e-06,
+      "loss": 0.6547,
+      "step": 3790
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 45.923824310302734,
+      "learning_rate": 9.2416e-06,
+      "loss": 0.6894,
+      "step": 3800
+    },
+    {
+      "epoch": 0.381,
+      "grad_norm": 26.007081985473633,
+      "learning_rate": 9.2396e-06,
+      "loss": 0.5406,
+      "step": 3810
+    },
+    {
+      "epoch": 0.382,
+      "grad_norm": 42.57291030883789,
+      "learning_rate": 9.237600000000001e-06,
+      "loss": 0.5942,
+      "step": 3820
+    },
+    {
+      "epoch": 0.383,
+      "grad_norm": 37.315677642822266,
+      "learning_rate": 9.235600000000002e-06,
+      "loss": 0.7275,
+      "step": 3830
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 18.446134567260742,
+      "learning_rate": 9.2336e-06,
+      "loss": 0.6234,
+      "step": 3840
+    },
+    {
+      "epoch": 0.385,
+      "grad_norm": 21.959327697753906,
+      "learning_rate": 9.2316e-06,
+      "loss": 0.9139,
+      "step": 3850
+    },
+    {
+      "epoch": 0.386,
+      "grad_norm": 16.19249725341797,
+      "learning_rate": 9.229800000000001e-06,
+      "loss": 0.9198,
+      "step": 3860
+    },
+    {
+      "epoch": 0.387,
+      "grad_norm": 24.801454544067383,
+      "learning_rate": 9.227800000000002e-06,
+      "loss": 0.8498,
+      "step": 3870
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 20.562063217163086,
+      "learning_rate": 9.2258e-06,
+      "loss": 0.9006,
+      "step": 3880
+    },
+    {
+      "epoch": 0.389,
+      "grad_norm": 27.41887855529785,
+      "learning_rate": 9.223800000000001e-06,
+      "loss": 0.6526,
+      "step": 3890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.00798225402832,
+      "learning_rate": 9.2218e-06,
+      "loss": 0.6432,
+      "step": 3900
+    },
+    {
+      "epoch": 0.391,
+      "grad_norm": 22.514768600463867,
+      "learning_rate": 9.219800000000001e-06,
+      "loss": 0.8321,
+      "step": 3910
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 22.683561325073242,
+      "learning_rate": 9.217800000000002e-06,
+      "loss": 0.6577,
+      "step": 3920
+    },
+    {
+      "epoch": 0.393,
+      "grad_norm": 38.5459098815918,
+      "learning_rate": 9.2158e-06,
+      "loss": 0.7376,
+      "step": 3930
+    },
+    {
+      "epoch": 0.394,
+      "grad_norm": 7.456604480743408,
+      "learning_rate": 9.2138e-06,
+      "loss": 0.5598,
+      "step": 3940
+    },
+    {
+      "epoch": 0.395,
+      "grad_norm": 22.23004913330078,
+      "learning_rate": 9.2118e-06,
+      "loss": 0.7943,
+      "step": 3950
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 14.584146499633789,
+      "learning_rate": 9.209800000000001e-06,
+      "loss": 0.802,
+      "step": 3960
+    },
+    {
+      "epoch": 0.397,
+      "grad_norm": 5.253112316131592,
+      "learning_rate": 9.207800000000002e-06,
+      "loss": 0.698,
+      "step": 3970
+    },
+    {
+      "epoch": 0.398,
+      "grad_norm": 23.153915405273438,
+      "learning_rate": 9.205800000000001e-06,
+      "loss": 0.7036,
+      "step": 3980
+    },
+    {
+      "epoch": 0.399,
+      "grad_norm": 32.066951751708984,
+      "learning_rate": 9.2038e-06,
+      "loss": 0.7933,
+      "step": 3990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.568124771118164,
+      "learning_rate": 9.2018e-06,
+      "loss": 0.5587,
+      "step": 4000
+    },
+    {
+      "epoch": 0.401,
+      "grad_norm": 62.3616943359375,
+      "learning_rate": 9.199800000000001e-06,
+      "loss": 0.8117,
+      "step": 4010
+    },
+    {
+      "epoch": 0.402,
+      "grad_norm": 3.00107479095459,
+      "learning_rate": 9.197800000000002e-06,
+      "loss": 0.6618,
+      "step": 4020
+    },
+    {
+      "epoch": 0.403,
+      "grad_norm": 12.941449165344238,
+      "learning_rate": 9.195800000000001e-06,
+      "loss": 0.5325,
+      "step": 4030
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 39.835296630859375,
+      "learning_rate": 9.1938e-06,
+      "loss": 0.6589,
+      "step": 4040
+    },
+    {
+      "epoch": 0.405,
+      "grad_norm": 44.73147201538086,
+      "learning_rate": 9.1918e-06,
+      "loss": 0.7197,
+      "step": 4050
+    },
+    {
+      "epoch": 0.406,
+      "grad_norm": 29.60512924194336,
+      "learning_rate": 9.189800000000001e-06,
+      "loss": 0.6466,
+      "step": 4060
+    },
+    {
+      "epoch": 0.407,
+      "grad_norm": 11.546348571777344,
+      "learning_rate": 9.1878e-06,
+      "loss": 0.5344,
+      "step": 4070
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 23.700122833251953,
+      "learning_rate": 9.185800000000001e-06,
+      "loss": 0.9021,
+      "step": 4080
+    },
+    {
+      "epoch": 0.409,
+      "grad_norm": 35.56307601928711,
+      "learning_rate": 9.1838e-06,
+      "loss": 0.8117,
+      "step": 4090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 19.759023666381836,
+      "learning_rate": 9.1818e-06,
+      "loss": 0.9188,
+      "step": 4100
+    },
+    {
+      "epoch": 0.411,
+      "grad_norm": 25.926471710205078,
+      "learning_rate": 9.179800000000001e-06,
+      "loss": 0.514,
+      "step": 4110
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 28.19685173034668,
+      "learning_rate": 9.1778e-06,
+      "loss": 0.5714,
+      "step": 4120
+    },
+    {
+      "epoch": 0.413,
+      "grad_norm": 29.041406631469727,
+      "learning_rate": 9.1758e-06,
+      "loss": 0.6469,
+      "step": 4130
+    },
+    {
+      "epoch": 0.414,
+      "grad_norm": 27.198598861694336,
+      "learning_rate": 9.1738e-06,
+      "loss": 0.7159,
+      "step": 4140
+    },
+    {
+      "epoch": 0.415,
+      "grad_norm": 20.31609535217285,
+      "learning_rate": 9.171800000000001e-06,
+      "loss": 0.7121,
+      "step": 4150
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 49.4409065246582,
+      "learning_rate": 9.169800000000002e-06,
+      "loss": 0.818,
+      "step": 4160
+    },
+    {
+      "epoch": 0.417,
+      "grad_norm": 9.885689735412598,
+      "learning_rate": 9.1678e-06,
+      "loss": 0.7141,
+      "step": 4170
+    },
+    {
+      "epoch": 0.418,
+      "grad_norm": 50.130733489990234,
+      "learning_rate": 9.1658e-06,
+      "loss": 0.5292,
+      "step": 4180
+    },
+    {
+      "epoch": 0.419,
+      "grad_norm": 35.08009719848633,
+      "learning_rate": 9.1638e-06,
+      "loss": 0.7885,
+      "step": 4190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 45.62377166748047,
+      "learning_rate": 9.161800000000001e-06,
+      "loss": 0.831,
+      "step": 4200
+    },
+    {
+      "epoch": 0.421,
+      "grad_norm": 25.133146286010742,
+      "learning_rate": 9.159800000000002e-06,
+      "loss": 0.7066,
+      "step": 4210
+    },
+    {
+      "epoch": 0.422,
+      "grad_norm": 36.50223159790039,
+      "learning_rate": 9.1578e-06,
+      "loss": 0.9691,
+      "step": 4220
+    },
+    {
+      "epoch": 0.423,
+      "grad_norm": 45.53746795654297,
+      "learning_rate": 9.1558e-06,
+      "loss": 0.9619,
+      "step": 4230
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 26.474241256713867,
+      "learning_rate": 9.1538e-06,
+      "loss": 0.7747,
+      "step": 4240
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 39.54206848144531,
+      "learning_rate": 9.151800000000001e-06,
+      "loss": 0.8397,
+      "step": 4250
+    },
+    {
+      "epoch": 0.426,
+      "grad_norm": 27.512920379638672,
+      "learning_rate": 9.149800000000002e-06,
+      "loss": 0.6549,
+      "step": 4260
+    },
+    {
+      "epoch": 0.427,
+      "grad_norm": 15.803271293640137,
+      "learning_rate": 9.1478e-06,
+      "loss": 0.8272,
+      "step": 4270
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 22.9871768951416,
+      "learning_rate": 9.1458e-06,
+      "loss": 0.6667,
+      "step": 4280
+    },
+    {
+      "epoch": 0.429,
+      "grad_norm": 20.597959518432617,
+      "learning_rate": 9.1438e-06,
+      "loss": 0.5645,
+      "step": 4290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 39.363590240478516,
+      "learning_rate": 9.141800000000001e-06,
+      "loss": 0.8324,
+      "step": 4300
+    },
+    {
+      "epoch": 0.431,
+      "grad_norm": 34.98888397216797,
+      "learning_rate": 9.1398e-06,
+      "loss": 0.9688,
+      "step": 4310
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 20.55095100402832,
+      "learning_rate": 9.1378e-06,
+      "loss": 0.772,
+      "step": 4320
+    },
+    {
+      "epoch": 0.433,
+      "grad_norm": 30.377758026123047,
+      "learning_rate": 9.1358e-06,
+      "loss": 0.9099,
+      "step": 4330
+    },
+    {
+      "epoch": 0.434,
+      "grad_norm": 31.90020179748535,
+      "learning_rate": 9.1338e-06,
+      "loss": 0.9172,
+      "step": 4340
+    },
+    {
+      "epoch": 0.435,
+      "grad_norm": 28.935474395751953,
+      "learning_rate": 9.131800000000001e-06,
+      "loss": 0.5936,
+      "step": 4350
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 31.516016006469727,
+      "learning_rate": 9.1298e-06,
+      "loss": 0.7803,
+      "step": 4360
+    },
+    {
+      "epoch": 0.437,
+      "grad_norm": 89.58314514160156,
+      "learning_rate": 9.127800000000001e-06,
+      "loss": 1.0418,
+      "step": 4370
+    },
+    {
+      "epoch": 0.438,
+      "grad_norm": 27.75042152404785,
+      "learning_rate": 9.1258e-06,
+      "loss": 0.6251,
+      "step": 4380
+    },
+    {
+      "epoch": 0.439,
+      "grad_norm": 18.389522552490234,
+      "learning_rate": 9.1238e-06,
+      "loss": 0.6817,
+      "step": 4390
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 30.223163604736328,
+      "learning_rate": 9.121800000000001e-06,
+      "loss": 0.741,
+      "step": 4400
+    },
+    {
+      "epoch": 0.441,
+      "grad_norm": 18.69849395751953,
+      "learning_rate": 9.1198e-06,
+      "loss": 0.5822,
+      "step": 4410
+    },
+    {
+      "epoch": 0.442,
+      "grad_norm": 16.272411346435547,
+      "learning_rate": 9.117800000000001e-06,
+      "loss": 0.6329,
+      "step": 4420
+    },
+    {
+      "epoch": 0.443,
+      "grad_norm": 23.2576904296875,
+      "learning_rate": 9.1158e-06,
+      "loss": 0.8683,
+      "step": 4430
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 19.748661041259766,
+      "learning_rate": 9.1138e-06,
+      "loss": 0.8318,
+      "step": 4440
+    },
+    {
+      "epoch": 0.445,
+      "grad_norm": 34.46539306640625,
+      "learning_rate": 9.111800000000001e-06,
+      "loss": 0.6966,
+      "step": 4450
+    },
+    {
+      "epoch": 0.446,
+      "grad_norm": 35.254310607910156,
+      "learning_rate": 9.1098e-06,
+      "loss": 0.7138,
+      "step": 4460
+    },
+    {
+      "epoch": 0.447,
+      "grad_norm": 32.39616775512695,
+      "learning_rate": 9.107800000000001e-06,
+      "loss": 0.6498,
+      "step": 4470
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 23.23666000366211,
+      "learning_rate": 9.1058e-06,
+      "loss": 0.9085,
+      "step": 4480
+    },
+    {
+      "epoch": 0.449,
+      "grad_norm": 28.106037139892578,
+      "learning_rate": 9.1038e-06,
+      "loss": 0.6794,
+      "step": 4490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 37.72524642944336,
+      "learning_rate": 9.1018e-06,
+      "loss": 0.5902,
+      "step": 4500
+    },
+    {
+      "epoch": 0.451,
+      "grad_norm": 22.42662239074707,
+      "learning_rate": 9.0998e-06,
+      "loss": 0.7308,
+      "step": 4510
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 20.135847091674805,
+      "learning_rate": 9.097800000000001e-06,
+      "loss": 0.6052,
+      "step": 4520
+    },
+    {
+      "epoch": 0.453,
+      "grad_norm": 3.0650992393493652,
+      "learning_rate": 9.095800000000002e-06,
+      "loss": 0.6174,
+      "step": 4530
+    },
+    {
+      "epoch": 0.454,
+      "grad_norm": 309.6864013671875,
+      "learning_rate": 9.093800000000001e-06,
+      "loss": 1.0139,
+      "step": 4540
+    },
+    {
+      "epoch": 0.455,
+      "grad_norm": 32.27659606933594,
+      "learning_rate": 9.0918e-06,
+      "loss": 0.8334,
+      "step": 4550
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 4.334067344665527,
+      "learning_rate": 9.0898e-06,
+      "loss": 0.6487,
+      "step": 4560
+    },
+    {
+      "epoch": 0.457,
+      "grad_norm": 7.285878658294678,
+      "learning_rate": 9.087800000000001e-06,
+      "loss": 0.5274,
+      "step": 4570
+    },
+    {
+      "epoch": 0.458,
+      "grad_norm": 68.66569519042969,
+      "learning_rate": 9.085800000000002e-06,
+      "loss": 0.7734,
+      "step": 4580
+    },
+    {
+      "epoch": 0.459,
+      "grad_norm": 36.57508850097656,
+      "learning_rate": 9.083800000000001e-06,
+      "loss": 0.8785,
+      "step": 4590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 49.88356018066406,
+      "learning_rate": 9.0818e-06,
+      "loss": 0.6072,
+      "step": 4600
+    },
+    {
+      "epoch": 0.461,
+      "grad_norm": 28.811969757080078,
+      "learning_rate": 9.0798e-06,
+      "loss": 0.701,
+      "step": 4610
+    },
+    {
+      "epoch": 0.462,
+      "grad_norm": 38.20576477050781,
+      "learning_rate": 9.077800000000001e-06,
+      "loss": 0.6207,
+      "step": 4620
+    },
+    {
+      "epoch": 0.463,
+      "grad_norm": 34.115169525146484,
+      "learning_rate": 9.075800000000002e-06,
+      "loss": 0.7065,
+      "step": 4630
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 16.451030731201172,
+      "learning_rate": 9.073800000000001e-06,
+      "loss": 0.6654,
+      "step": 4640
+    },
+    {
+      "epoch": 0.465,
+      "grad_norm": 50.430564880371094,
+      "learning_rate": 9.0718e-06,
+      "loss": 0.576,
+      "step": 4650
+    },
+    {
+      "epoch": 0.466,
+      "grad_norm": 34.623878479003906,
+      "learning_rate": 9.0698e-06,
+      "loss": 0.4831,
+      "step": 4660
+    },
+    {
+      "epoch": 0.467,
+      "grad_norm": 28.038461685180664,
+      "learning_rate": 9.067800000000002e-06,
+      "loss": 0.6403,
+      "step": 4670
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 18.259531021118164,
+      "learning_rate": 9.0658e-06,
+      "loss": 0.6536,
+      "step": 4680
+    },
+    {
+      "epoch": 0.469,
+      "grad_norm": 27.56315803527832,
+      "learning_rate": 9.063800000000001e-06,
+      "loss": 0.8802,
+      "step": 4690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 24.04572868347168,
+      "learning_rate": 9.0618e-06,
+      "loss": 0.7131,
+      "step": 4700
+    },
+    {
+      "epoch": 0.471,
+      "grad_norm": 34.79553985595703,
+      "learning_rate": 9.059800000000001e-06,
+      "loss": 0.9388,
+      "step": 4710
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 20.2165470123291,
+      "learning_rate": 9.057800000000002e-06,
+      "loss": 0.7212,
+      "step": 4720
+    },
+    {
+      "epoch": 0.473,
+      "grad_norm": 26.301525115966797,
+      "learning_rate": 9.0558e-06,
+      "loss": 0.8441,
+      "step": 4730
+    },
+    {
+      "epoch": 0.474,
+      "grad_norm": 30.426250457763672,
+      "learning_rate": 9.0538e-06,
+      "loss": 0.6735,
+      "step": 4740
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 16.273841857910156,
+      "learning_rate": 9.0518e-06,
+      "loss": 0.7563,
+      "step": 4750
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 17.23615264892578,
+      "learning_rate": 9.049800000000001e-06,
+      "loss": 0.6499,
+      "step": 4760
+    },
+    {
+      "epoch": 0.477,
+      "grad_norm": 13.083768844604492,
+      "learning_rate": 9.047800000000002e-06,
+      "loss": 0.5476,
+      "step": 4770
+    },
+    {
+      "epoch": 0.478,
+      "grad_norm": 26.91759490966797,
+      "learning_rate": 9.0458e-06,
+      "loss": 0.5266,
+      "step": 4780
+    },
+    {
+      "epoch": 0.479,
+      "grad_norm": 98.05184173583984,
+      "learning_rate": 9.0438e-06,
+      "loss": 0.8325,
+      "step": 4790
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 23.273862838745117,
+      "learning_rate": 9.0418e-06,
+      "loss": 0.9052,
+      "step": 4800
+    },
+    {
+      "epoch": 0.481,
+      "grad_norm": 19.2165584564209,
+      "learning_rate": 9.039800000000001e-06,
+      "loss": 0.9031,
+      "step": 4810
+    },
+    {
+      "epoch": 0.482,
+      "grad_norm": 18.614849090576172,
+      "learning_rate": 9.037800000000002e-06,
+      "loss": 0.7544,
+      "step": 4820
+    },
+    {
+      "epoch": 0.483,
+      "grad_norm": 13.22616195678711,
+      "learning_rate": 9.0358e-06,
+      "loss": 0.8122,
+      "step": 4830
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 26.158973693847656,
+      "learning_rate": 9.0338e-06,
+      "loss": 0.8789,
+      "step": 4840
+    },
+    {
+      "epoch": 0.485,
+      "grad_norm": 24.33793830871582,
+      "learning_rate": 9.0318e-06,
+      "loss": 0.7403,
+      "step": 4850
+    },
+    {
+      "epoch": 0.486,
+      "grad_norm": 13.415386199951172,
+      "learning_rate": 9.029800000000001e-06,
+      "loss": 0.6783,
+      "step": 4860
+    },
+    {
+      "epoch": 0.487,
+      "grad_norm": 26.246070861816406,
+      "learning_rate": 9.0278e-06,
+      "loss": 0.8308,
+      "step": 4870
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 25.685985565185547,
+      "learning_rate": 9.025800000000001e-06,
+      "loss": 0.8727,
+      "step": 4880
+    },
+    {
+      "epoch": 0.489,
+      "grad_norm": 27.296411514282227,
+      "learning_rate": 9.0238e-06,
+      "loss": 0.8406,
+      "step": 4890
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 14.517251014709473,
+      "learning_rate": 9.0218e-06,
+      "loss": 0.7182,
+      "step": 4900
+    },
+    {
+      "epoch": 0.491,
+      "grad_norm": 37.61231994628906,
+      "learning_rate": 9.019800000000001e-06,
+      "loss": 0.9733,
+      "step": 4910
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 14.171152114868164,
+      "learning_rate": 9.0178e-06,
+      "loss": 0.7648,
+      "step": 4920
+    },
+    {
+      "epoch": 0.493,
+      "grad_norm": 18.871143341064453,
+      "learning_rate": 9.015800000000001e-06,
+      "loss": 0.8402,
+      "step": 4930
+    },
+    {
+      "epoch": 0.494,
+      "grad_norm": 10.27568531036377,
+      "learning_rate": 9.0138e-06,
+      "loss": 0.6241,
+      "step": 4940
+    },
+    {
+      "epoch": 0.495,
+      "grad_norm": 25.206148147583008,
+      "learning_rate": 9.0118e-06,
+      "loss": 0.7524,
+      "step": 4950
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 16.134567260742188,
+      "learning_rate": 9.009800000000001e-06,
+      "loss": 0.5719,
+      "step": 4960
+    },
+    {
+      "epoch": 0.497,
+      "grad_norm": 27.282073974609375,
+      "learning_rate": 9.0078e-06,
+      "loss": 0.9313,
+      "step": 4970
+    },
+    {
+      "epoch": 0.498,
+      "grad_norm": 31.6767578125,
+      "learning_rate": 9.005800000000001e-06,
+      "loss": 0.7742,
+      "step": 4980
+    },
+    {
+      "epoch": 0.499,
+      "grad_norm": 47.10121536254883,
+      "learning_rate": 9.0038e-06,
+      "loss": 0.726,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 32.248146057128906,
+      "learning_rate": 9.0018e-06,
+      "loss": 0.6758,
+      "step": 5000
+    },
+    {
+      "epoch": 0.501,
+      "grad_norm": 21.67167854309082,
+      "learning_rate": 8.999800000000001e-06,
+      "loss": 0.8787,
+      "step": 5010
+    },
+    {
+      "epoch": 0.502,
+      "grad_norm": 22.77020835876465,
+      "learning_rate": 8.9978e-06,
+      "loss": 0.5766,
+      "step": 5020
+    },
+    {
+      "epoch": 0.503,
+      "grad_norm": 31.036142349243164,
+      "learning_rate": 8.995800000000001e-06,
+      "loss": 0.837,
+      "step": 5030
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 18.027769088745117,
+      "learning_rate": 8.9938e-06,
+      "loss": 0.8028,
+      "step": 5040
+    },
+    {
+      "epoch": 0.505,
+      "grad_norm": 23.31261444091797,
+      "learning_rate": 8.991800000000001e-06,
+      "loss": 0.5677,
+      "step": 5050
+    },
+    {
+      "epoch": 0.506,
+      "grad_norm": 22.90900421142578,
+      "learning_rate": 8.989800000000002e-06,
+      "loss": 0.7327,
+      "step": 5060
+    },
+    {
+      "epoch": 0.507,
+      "grad_norm": 29.72652244567871,
+      "learning_rate": 8.9878e-06,
+      "loss": 0.5881,
+      "step": 5070
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 29.568384170532227,
+      "learning_rate": 8.985800000000001e-06,
+      "loss": 0.6233,
+      "step": 5080
+    },
+    {
+      "epoch": 0.509,
+      "grad_norm": 38.03068923950195,
+      "learning_rate": 8.9838e-06,
+      "loss": 0.7105,
+      "step": 5090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 42.80902862548828,
+      "learning_rate": 8.981800000000001e-06,
+      "loss": 0.727,
+      "step": 5100
+    },
+    {
+      "epoch": 0.511,
+      "grad_norm": 39.31882858276367,
+      "learning_rate": 8.9798e-06,
+      "loss": 0.6516,
+      "step": 5110
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 28.98011589050293,
+      "learning_rate": 8.9778e-06,
+      "loss": 0.6184,
+      "step": 5120
+    },
+    {
+      "epoch": 0.513,
+      "grad_norm": 26.84391212463379,
+      "learning_rate": 8.975800000000001e-06,
+      "loss": 1.0467,
+      "step": 5130
+    },
+    {
+      "epoch": 0.514,
+      "grad_norm": 35.17725372314453,
+      "learning_rate": 8.9738e-06,
+      "loss": 0.9125,
+      "step": 5140
+    },
+    {
+      "epoch": 0.515,
+      "grad_norm": 34.84568405151367,
+      "learning_rate": 8.971800000000001e-06,
+      "loss": 0.5917,
+      "step": 5150
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 23.5303897857666,
+      "learning_rate": 8.9698e-06,
+      "loss": 0.6795,
+      "step": 5160
+    },
+    {
+      "epoch": 0.517,
+      "grad_norm": 48.519927978515625,
+      "learning_rate": 8.9678e-06,
+      "loss": 0.7996,
+      "step": 5170
+    },
+    {
+      "epoch": 0.518,
+      "grad_norm": 9.060611724853516,
+      "learning_rate": 8.965800000000001e-06,
+      "loss": 0.67,
+      "step": 5180
+    },
+    {
+      "epoch": 0.519,
+      "grad_norm": 15.595134735107422,
+      "learning_rate": 8.9638e-06,
+      "loss": 0.9602,
+      "step": 5190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.556856155395508,
+      "learning_rate": 8.961800000000001e-06,
+      "loss": 0.7247,
+      "step": 5200
+    },
+    {
+      "epoch": 0.521,
+      "grad_norm": 21.102563858032227,
+      "learning_rate": 8.9598e-06,
+      "loss": 0.8323,
+      "step": 5210
+    },
+    {
+      "epoch": 0.522,
+      "grad_norm": 18.13496208190918,
+      "learning_rate": 8.957800000000001e-06,
+      "loss": 0.6505,
+      "step": 5220
+    },
+    {
+      "epoch": 0.523,
+      "grad_norm": 5.696938514709473,
+      "learning_rate": 8.955800000000002e-06,
+      "loss": 0.7781,
+      "step": 5230
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 37.51520538330078,
+      "learning_rate": 8.9538e-06,
+      "loss": 1.074,
+      "step": 5240
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 8.744912147521973,
+      "learning_rate": 8.951800000000001e-06,
+      "loss": 0.6568,
+      "step": 5250
+    },
+    {
+      "epoch": 0.526,
+      "grad_norm": 23.32012367248535,
+      "learning_rate": 8.9498e-06,
+      "loss": 0.9314,
+      "step": 5260
+    },
+    {
+      "epoch": 0.527,
+      "grad_norm": 15.782977104187012,
+      "learning_rate": 8.947800000000001e-06,
+      "loss": 0.7599,
+      "step": 5270
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 14.995272636413574,
+      "learning_rate": 8.945800000000002e-06,
+      "loss": 0.6917,
+      "step": 5280
+    },
+    {
+      "epoch": 0.529,
+      "grad_norm": 15.904824256896973,
+      "learning_rate": 8.9438e-06,
+      "loss": 0.7478,
+      "step": 5290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 25.889888763427734,
+      "learning_rate": 8.9418e-06,
+      "loss": 0.6467,
+      "step": 5300
+    },
+    {
+      "epoch": 0.531,
+      "grad_norm": 10.381155014038086,
+      "learning_rate": 8.9398e-06,
+      "loss": 0.6683,
+      "step": 5310
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 21.153766632080078,
+      "learning_rate": 8.937800000000001e-06,
+      "loss": 0.5883,
+      "step": 5320
+    },
+    {
+      "epoch": 0.533,
+      "grad_norm": 35.405887603759766,
+      "learning_rate": 8.935800000000002e-06,
+      "loss": 0.6827,
+      "step": 5330
+    },
+    {
+      "epoch": 0.534,
+      "grad_norm": 34.55176544189453,
+      "learning_rate": 8.9338e-06,
+      "loss": 0.5616,
+      "step": 5340
+    },
+    {
+      "epoch": 0.535,
+      "grad_norm": 48.410457611083984,
+      "learning_rate": 8.9318e-06,
+      "loss": 0.8742,
+      "step": 5350
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 30.154441833496094,
+      "learning_rate": 8.9298e-06,
+      "loss": 0.7361,
+      "step": 5360
+    },
+    {
+      "epoch": 0.537,
+      "grad_norm": 22.52509117126465,
+      "learning_rate": 8.927800000000001e-06,
+      "loss": 0.8061,
+      "step": 5370
+    },
+    {
+      "epoch": 0.538,
+      "grad_norm": 33.107120513916016,
+      "learning_rate": 8.925800000000002e-06,
+      "loss": 0.9955,
+      "step": 5380
+    },
+    {
+      "epoch": 0.539,
+      "grad_norm": 26.009675979614258,
+      "learning_rate": 8.923800000000001e-06,
+      "loss": 0.7532,
+      "step": 5390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.156719207763672,
+      "learning_rate": 8.9218e-06,
+      "loss": 0.6046,
+      "step": 5400
+    },
+    {
+      "epoch": 0.541,
+      "grad_norm": 34.604610443115234,
+      "learning_rate": 8.9198e-06,
+      "loss": 1.0056,
+      "step": 5410
+    },
+    {
+      "epoch": 0.542,
+      "grad_norm": 7.621533393859863,
+      "learning_rate": 8.917800000000001e-06,
+      "loss": 0.6947,
+      "step": 5420
+    },
+    {
+      "epoch": 0.543,
+      "grad_norm": 22.182506561279297,
+      "learning_rate": 8.915800000000002e-06,
+      "loss": 0.7304,
+      "step": 5430
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 17.386693954467773,
+      "learning_rate": 8.913800000000001e-06,
+      "loss": 0.8156,
+      "step": 5440
+    },
+    {
+      "epoch": 0.545,
+      "grad_norm": 28.76013946533203,
+      "learning_rate": 8.9118e-06,
+      "loss": 0.4821,
+      "step": 5450
+    },
+    {
+      "epoch": 0.546,
+      "grad_norm": 42.765296936035156,
+      "learning_rate": 8.9098e-06,
+      "loss": 0.871,
+      "step": 5460
+    },
+    {
+      "epoch": 0.547,
+      "grad_norm": 31.4482479095459,
+      "learning_rate": 8.907800000000001e-06,
+      "loss": 0.79,
+      "step": 5470
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 21.309450149536133,
+      "learning_rate": 8.9058e-06,
+      "loss": 0.5208,
+      "step": 5480
+    },
+    {
+      "epoch": 0.549,
+      "grad_norm": 26.79330062866211,
+      "learning_rate": 8.904e-06,
+      "loss": 0.5881,
+      "step": 5490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.553604125976562,
+      "learning_rate": 8.902e-06,
+      "loss": 0.9493,
+      "step": 5500
+    },
+    {
+      "epoch": 0.551,
+      "grad_norm": 25.66444969177246,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 0.5148,
+      "step": 5510
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 33.9870491027832,
+      "learning_rate": 8.898000000000002e-06,
+      "loss": 0.4237,
+      "step": 5520
+    },
+    {
+      "epoch": 0.553,
+      "grad_norm": 38.512290954589844,
+      "learning_rate": 8.896000000000001e-06,
+      "loss": 0.673,
+      "step": 5530
+    },
+    {
+      "epoch": 0.554,
+      "grad_norm": 35.605525970458984,
+      "learning_rate": 8.894e-06,
+      "loss": 0.767,
+      "step": 5540
+    },
+    {
+      "epoch": 0.555,
+      "grad_norm": 42.295989990234375,
+      "learning_rate": 8.892e-06,
+      "loss": 0.9592,
+      "step": 5550
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 40.66447830200195,
+      "learning_rate": 8.890000000000001e-06,
+      "loss": 0.7368,
+      "step": 5560
+    },
+    {
+      "epoch": 0.557,
+      "grad_norm": 28.294403076171875,
+      "learning_rate": 8.888e-06,
+      "loss": 0.8183,
+      "step": 5570
+    },
+    {
+      "epoch": 0.558,
+      "grad_norm": 25.432598114013672,
+      "learning_rate": 8.886000000000001e-06,
+      "loss": 0.8464,
+      "step": 5580
+    },
+    {
+      "epoch": 0.559,
+      "grad_norm": 33.40468215942383,
+      "learning_rate": 8.884e-06,
+      "loss": 0.6672,
+      "step": 5590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 27.27119255065918,
+      "learning_rate": 8.882e-06,
+      "loss": 0.6753,
+      "step": 5600
+    },
+    {
+      "epoch": 0.561,
+      "grad_norm": 27.768653869628906,
+      "learning_rate": 8.880000000000001e-06,
+      "loss": 0.6132,
+      "step": 5610
+    },
+    {
+      "epoch": 0.562,
+      "grad_norm": 9.56386947631836,
+      "learning_rate": 8.878e-06,
+      "loss": 0.8525,
+      "step": 5620
+    },
+    {
+      "epoch": 0.563,
+      "grad_norm": 39.86705017089844,
+      "learning_rate": 8.876e-06,
+      "loss": 0.9229,
+      "step": 5630
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 31.844165802001953,
+      "learning_rate": 8.874e-06,
+      "loss": 0.9162,
+      "step": 5640
+    },
+    {
+      "epoch": 0.565,
+      "grad_norm": 30.67936897277832,
+      "learning_rate": 8.872e-06,
+      "loss": 0.6348,
+      "step": 5650
+    },
+    {
+      "epoch": 0.566,
+      "grad_norm": 25.196563720703125,
+      "learning_rate": 8.870000000000001e-06,
+      "loss": 0.7222,
+      "step": 5660
+    },
+    {
+      "epoch": 0.567,
+      "grad_norm": 14.649309158325195,
+      "learning_rate": 8.868e-06,
+      "loss": 0.791,
+      "step": 5670
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 23.247085571289062,
+      "learning_rate": 8.866000000000001e-06,
+      "loss": 0.6081,
+      "step": 5680
+    },
+    {
+      "epoch": 0.569,
+      "grad_norm": 23.32643699645996,
+      "learning_rate": 8.864e-06,
+      "loss": 0.605,
+      "step": 5690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 32.223018646240234,
+      "learning_rate": 8.862000000000001e-06,
+      "loss": 0.7627,
+      "step": 5700
+    },
+    {
+      "epoch": 0.571,
+      "grad_norm": 19.816926956176758,
+      "learning_rate": 8.860000000000002e-06,
+      "loss": 0.8573,
+      "step": 5710
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 31.972070693969727,
+      "learning_rate": 8.858e-06,
+      "loss": 0.7598,
+      "step": 5720
+    },
+    {
+      "epoch": 0.573,
+      "grad_norm": 28.69130516052246,
+      "learning_rate": 8.856000000000001e-06,
+      "loss": 0.7115,
+      "step": 5730
+    },
+    {
+      "epoch": 0.574,
+      "grad_norm": 35.84492111206055,
+      "learning_rate": 8.854e-06,
+      "loss": 0.5372,
+      "step": 5740
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 6.977468967437744,
+      "learning_rate": 8.852000000000001e-06,
+      "loss": 0.5373,
+      "step": 5750
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 51.22468566894531,
+      "learning_rate": 8.85e-06,
+      "loss": 0.6888,
+      "step": 5760
+    },
+    {
+      "epoch": 0.577,
+      "grad_norm": 17.154935836791992,
+      "learning_rate": 8.848e-06,
+      "loss": 0.3186,
+      "step": 5770
+    },
+    {
+      "epoch": 0.578,
+      "grad_norm": 30.214448928833008,
+      "learning_rate": 8.846000000000001e-06,
+      "loss": 0.6309,
+      "step": 5780
+    },
+    {
+      "epoch": 0.579,
+      "grad_norm": 54.17653274536133,
+      "learning_rate": 8.844e-06,
+      "loss": 0.7933,
+      "step": 5790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 45.278411865234375,
+      "learning_rate": 8.842000000000001e-06,
+      "loss": 0.709,
+      "step": 5800
+    },
+    {
+      "epoch": 0.581,
+      "grad_norm": 5.875872611999512,
+      "learning_rate": 8.84e-06,
+      "loss": 0.4124,
+      "step": 5810
+    },
+    {
+      "epoch": 0.582,
+      "grad_norm": 54.30577087402344,
+      "learning_rate": 8.838e-06,
+      "loss": 0.6046,
+      "step": 5820
+    },
+    {
+      "epoch": 0.583,
+      "grad_norm": 28.242185592651367,
+      "learning_rate": 8.836000000000001e-06,
+      "loss": 0.8088,
+      "step": 5830
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 17.646381378173828,
+      "learning_rate": 8.834e-06,
+      "loss": 0.6872,
+      "step": 5840
+    },
+    {
+      "epoch": 0.585,
+      "grad_norm": 29.147197723388672,
+      "learning_rate": 8.832000000000001e-06,
+      "loss": 0.5049,
+      "step": 5850
+    },
+    {
+      "epoch": 0.586,
+      "grad_norm": 41.977230072021484,
+      "learning_rate": 8.83e-06,
+      "loss": 0.8069,
+      "step": 5860
+    },
+    {
+      "epoch": 0.587,
+      "grad_norm": 17.308279037475586,
+      "learning_rate": 8.828000000000001e-06,
+      "loss": 0.5329,
+      "step": 5870
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 59.8780632019043,
+      "learning_rate": 8.826000000000002e-06,
+      "loss": 0.8051,
+      "step": 5880
+    },
+    {
+      "epoch": 0.589,
+      "grad_norm": 72.70011138916016,
+      "learning_rate": 8.824e-06,
+      "loss": 1.0038,
+      "step": 5890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.008209228515625,
+      "learning_rate": 8.822000000000001e-06,
+      "loss": 0.442,
+      "step": 5900
+    },
+    {
+      "epoch": 0.591,
+      "grad_norm": 45.6512565612793,
+      "learning_rate": 8.82e-06,
+      "loss": 0.4802,
+      "step": 5910
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 9.083815574645996,
+      "learning_rate": 8.818000000000001e-06,
+      "loss": 0.5706,
+      "step": 5920
+    },
+    {
+      "epoch": 0.593,
+      "grad_norm": 31.648679733276367,
+      "learning_rate": 8.816000000000002e-06,
+      "loss": 0.5389,
+      "step": 5930
+    },
+    {
+      "epoch": 0.594,
+      "grad_norm": 47.933197021484375,
+      "learning_rate": 8.814e-06,
+      "loss": 0.6758,
+      "step": 5940
+    },
+    {
+      "epoch": 0.595,
+      "grad_norm": 44.82785415649414,
+      "learning_rate": 8.812000000000001e-06,
+      "loss": 0.6103,
+      "step": 5950
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 15.51645565032959,
+      "learning_rate": 8.81e-06,
+      "loss": 0.7726,
+      "step": 5960
+    },
+    {
+      "epoch": 0.597,
+      "grad_norm": 31.250015258789062,
+      "learning_rate": 8.808000000000001e-06,
+      "loss": 0.6967,
+      "step": 5970
+    },
+    {
+      "epoch": 0.598,
+      "grad_norm": 39.70091247558594,
+      "learning_rate": 8.806000000000002e-06,
+      "loss": 0.9665,
+      "step": 5980
+    },
+    {
+      "epoch": 0.599,
+      "grad_norm": 17.304738998413086,
+      "learning_rate": 8.804e-06,
+      "loss": 0.6742,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 51.37207794189453,
+      "learning_rate": 8.802e-06,
+      "loss": 0.767,
+      "step": 6000
+    },
+    {
+      "epoch": 0.601,
+      "grad_norm": 21.40468978881836,
+      "learning_rate": 8.8e-06,
+      "loss": 0.6364,
+      "step": 6010
+    },
+    {
+      "epoch": 0.602,
+      "grad_norm": 23.359695434570312,
+      "learning_rate": 8.798000000000001e-06,
+      "loss": 0.6353,
+      "step": 6020
+    },
+    {
+      "epoch": 0.603,
+      "grad_norm": 25.836505889892578,
+      "learning_rate": 8.796000000000002e-06,
+      "loss": 0.8381,
+      "step": 6030
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 37.61728286743164,
+      "learning_rate": 8.794e-06,
+      "loss": 0.6782,
+      "step": 6040
+    },
+    {
+      "epoch": 0.605,
+      "grad_norm": 27.693248748779297,
+      "learning_rate": 8.792e-06,
+      "loss": 0.7325,
+      "step": 6050
+    },
+    {
+      "epoch": 0.606,
+      "grad_norm": 23.111059188842773,
+      "learning_rate": 8.79e-06,
+      "loss": 0.8304,
+      "step": 6060
+    },
+    {
+      "epoch": 0.607,
+      "grad_norm": 34.3321647644043,
+      "learning_rate": 8.788000000000001e-06,
+      "loss": 0.5949,
+      "step": 6070
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 43.452415466308594,
+      "learning_rate": 8.786000000000002e-06,
+      "loss": 0.669,
+      "step": 6080
+    },
+    {
+      "epoch": 0.609,
+      "grad_norm": 12.943218231201172,
+      "learning_rate": 8.784000000000001e-06,
+      "loss": 0.6943,
+      "step": 6090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.950276374816895,
+      "learning_rate": 8.782e-06,
+      "loss": 0.5902,
+      "step": 6100
+    },
+    {
+      "epoch": 0.611,
+      "grad_norm": 9.990435600280762,
+      "learning_rate": 8.78e-06,
+      "loss": 0.5351,
+      "step": 6110
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 33.92399215698242,
+      "learning_rate": 8.778000000000001e-06,
+      "loss": 0.858,
+      "step": 6120
+    },
+    {
+      "epoch": 0.613,
+      "grad_norm": 29.9383544921875,
+      "learning_rate": 8.776e-06,
+      "loss": 0.7106,
+      "step": 6130
+    },
+    {
+      "epoch": 0.614,
+      "grad_norm": 25.9768123626709,
+      "learning_rate": 8.774000000000001e-06,
+      "loss": 0.7565,
+      "step": 6140
+    },
+    {
+      "epoch": 0.615,
+      "grad_norm": 14.911945343017578,
+      "learning_rate": 8.772e-06,
+      "loss": 0.6397,
+      "step": 6150
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 42.49845886230469,
+      "learning_rate": 8.77e-06,
+      "loss": 0.5812,
+      "step": 6160
+    },
+    {
+      "epoch": 0.617,
+      "grad_norm": 24.447677612304688,
+      "learning_rate": 8.768000000000001e-06,
+      "loss": 0.6996,
+      "step": 6170
+    },
+    {
+      "epoch": 0.618,
+      "grad_norm": 22.37032127380371,
+      "learning_rate": 8.766e-06,
+      "loss": 0.7133,
+      "step": 6180
+    },
+    {
+      "epoch": 0.619,
+      "grad_norm": 15.080531120300293,
+      "learning_rate": 8.764e-06,
+      "loss": 0.7305,
+      "step": 6190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.373855113983154,
+      "learning_rate": 8.762e-06,
+      "loss": 0.7308,
+      "step": 6200
+    },
+    {
+      "epoch": 0.621,
+      "grad_norm": 31.23784065246582,
+      "learning_rate": 8.76e-06,
+      "loss": 0.7615,
+      "step": 6210
+    },
+    {
+      "epoch": 0.622,
+      "grad_norm": 38.272438049316406,
+      "learning_rate": 8.758000000000002e-06,
+      "loss": 0.6086,
+      "step": 6220
+    },
+    {
+      "epoch": 0.623,
+      "grad_norm": 23.415599822998047,
+      "learning_rate": 8.756e-06,
+      "loss": 0.6437,
+      "step": 6230
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 27.43158531188965,
+      "learning_rate": 8.754e-06,
+      "loss": 0.7331,
+      "step": 6240
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 20.39682388305664,
+      "learning_rate": 8.752e-06,
+      "loss": 0.5126,
+      "step": 6250
+    },
+    {
+      "epoch": 0.626,
+      "grad_norm": 55.00613784790039,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.7794,
+      "step": 6260
+    },
+    {
+      "epoch": 0.627,
+      "grad_norm": 38.01939010620117,
+      "learning_rate": 8.748000000000002e-06,
+      "loss": 0.617,
+      "step": 6270
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 21.09652328491211,
+      "learning_rate": 8.746e-06,
+      "loss": 0.742,
+      "step": 6280
+    },
+    {
+      "epoch": 0.629,
+      "grad_norm": 14.96897029876709,
+      "learning_rate": 8.744e-06,
+      "loss": 0.3911,
+      "step": 6290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 31.505590438842773,
+      "learning_rate": 8.742e-06,
+      "loss": 1.0415,
+      "step": 6300
+    },
+    {
+      "epoch": 0.631,
+      "grad_norm": 36.91202926635742,
+      "learning_rate": 8.740000000000001e-06,
+      "loss": 0.7963,
+      "step": 6310
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 23.781566619873047,
+      "learning_rate": 8.738000000000002e-06,
+      "loss": 0.4309,
+      "step": 6320
+    },
+    {
+      "epoch": 0.633,
+      "grad_norm": 37.20154571533203,
+      "learning_rate": 8.736e-06,
+      "loss": 0.7661,
+      "step": 6330
+    },
+    {
+      "epoch": 0.634,
+      "grad_norm": 49.68278121948242,
+      "learning_rate": 8.734e-06,
+      "loss": 0.9511,
+      "step": 6340
+    },
+    {
+      "epoch": 0.635,
+      "grad_norm": 36.95859909057617,
+      "learning_rate": 8.732e-06,
+      "loss": 0.7286,
+      "step": 6350
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 42.05426025390625,
+      "learning_rate": 8.730000000000001e-06,
+      "loss": 0.684,
+      "step": 6360
+    },
+    {
+      "epoch": 0.637,
+      "grad_norm": 55.31623840332031,
+      "learning_rate": 8.728e-06,
+      "loss": 1.0039,
+      "step": 6370
+    },
+    {
+      "epoch": 0.638,
+      "grad_norm": 37.254356384277344,
+      "learning_rate": 8.726e-06,
+      "loss": 0.5336,
+      "step": 6380
+    },
+    {
+      "epoch": 0.639,
+      "grad_norm": 9.793030738830566,
+      "learning_rate": 8.724e-06,
+      "loss": 0.6437,
+      "step": 6390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 41.21820831298828,
+      "learning_rate": 8.722e-06,
+      "loss": 0.6156,
+      "step": 6400
+    },
+    {
+      "epoch": 0.641,
+      "grad_norm": 16.490699768066406,
+      "learning_rate": 8.720000000000001e-06,
+      "loss": 0.6881,
+      "step": 6410
+    },
+    {
+      "epoch": 0.642,
+      "grad_norm": 45.798728942871094,
+      "learning_rate": 8.718e-06,
+      "loss": 0.7894,
+      "step": 6420
+    },
+    {
+      "epoch": 0.643,
+      "grad_norm": 36.68819808959961,
+      "learning_rate": 8.716000000000001e-06,
+      "loss": 0.771,
+      "step": 6430
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 9.799590110778809,
+      "learning_rate": 8.714e-06,
+      "loss": 0.6979,
+      "step": 6440
+    },
+    {
+      "epoch": 0.645,
+      "grad_norm": 24.45409393310547,
+      "learning_rate": 8.712e-06,
+      "loss": 0.65,
+      "step": 6450
+    },
+    {
+      "epoch": 0.646,
+      "grad_norm": 17.76416778564453,
+      "learning_rate": 8.710000000000001e-06,
+      "loss": 0.8058,
+      "step": 6460
+    },
+    {
+      "epoch": 0.647,
+      "grad_norm": 16.39788055419922,
+      "learning_rate": 8.708e-06,
+      "loss": 0.6635,
+      "step": 6470
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 18.585445404052734,
+      "learning_rate": 8.706000000000001e-06,
+      "loss": 0.7145,
+      "step": 6480
+    },
+    {
+      "epoch": 0.649,
+      "grad_norm": 31.643970489501953,
+      "learning_rate": 8.704e-06,
+      "loss": 0.5295,
+      "step": 6490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.48748779296875,
+      "learning_rate": 8.702e-06,
+      "loss": 0.5848,
+      "step": 6500
+    },
+    {
+      "epoch": 0.651,
+      "grad_norm": 36.65018081665039,
+      "learning_rate": 8.700000000000001e-06,
+      "loss": 0.7816,
+      "step": 6510
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 29.191734313964844,
+      "learning_rate": 8.698e-06,
+      "loss": 0.6342,
+      "step": 6520
+    },
+    {
+      "epoch": 0.653,
+      "grad_norm": 29.119672775268555,
+      "learning_rate": 8.696000000000001e-06,
+      "loss": 0.6923,
+      "step": 6530
+    },
+    {
+      "epoch": 0.654,
+      "grad_norm": 33.278594970703125,
+      "learning_rate": 8.694e-06,
+      "loss": 0.617,
+      "step": 6540
+    },
+    {
+      "epoch": 0.655,
+      "grad_norm": 36.83278274536133,
+      "learning_rate": 8.692e-06,
+      "loss": 0.7368,
+      "step": 6550
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 25.805545806884766,
+      "learning_rate": 8.690000000000002e-06,
+      "loss": 0.7087,
+      "step": 6560
+    },
+    {
+      "epoch": 0.657,
+      "grad_norm": 22.268075942993164,
+      "learning_rate": 8.688e-06,
+      "loss": 0.6209,
+      "step": 6570
+    },
+    {
+      "epoch": 0.658,
+      "grad_norm": 12.060050964355469,
+      "learning_rate": 8.686000000000001e-06,
+      "loss": 0.4913,
+      "step": 6580
+    },
+    {
+      "epoch": 0.659,
+      "grad_norm": 13.607654571533203,
+      "learning_rate": 8.684e-06,
+      "loss": 0.7358,
+      "step": 6590
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 38.86949920654297,
+      "learning_rate": 8.682000000000001e-06,
+      "loss": 0.7662,
+      "step": 6600
+    },
+    {
+      "epoch": 0.661,
+      "grad_norm": 22.411563873291016,
+      "learning_rate": 8.68e-06,
+      "loss": 0.5944,
+      "step": 6610
+    },
+    {
+      "epoch": 0.662,
+      "grad_norm": 27.0049991607666,
+      "learning_rate": 8.678e-06,
+      "loss": 0.6924,
+      "step": 6620
+    },
+    {
+      "epoch": 0.663,
+      "grad_norm": 16.259170532226562,
+      "learning_rate": 8.676000000000001e-06,
+      "loss": 0.6007,
+      "step": 6630
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 21.972728729248047,
+      "learning_rate": 8.674e-06,
+      "loss": 0.5624,
+      "step": 6640
+    },
+    {
+      "epoch": 0.665,
+      "grad_norm": 24.113021850585938,
+      "learning_rate": 8.672000000000001e-06,
+      "loss": 0.7037,
+      "step": 6650
+    },
+    {
+      "epoch": 0.666,
+      "grad_norm": 24.1397705078125,
+      "learning_rate": 8.67e-06,
+      "loss": 0.6052,
+      "step": 6660
+    },
+    {
+      "epoch": 0.667,
+      "grad_norm": 23.034231185913086,
+      "learning_rate": 8.668e-06,
+      "loss": 0.653,
+      "step": 6670
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 22.749948501586914,
+      "learning_rate": 8.666000000000001e-06,
+      "loss": 0.5321,
+      "step": 6680
+    },
+    {
+      "epoch": 0.669,
+      "grad_norm": 37.435977935791016,
+      "learning_rate": 8.664e-06,
+      "loss": 0.8462,
+      "step": 6690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 19.840953826904297,
+      "learning_rate": 8.662000000000001e-06,
+      "loss": 0.7411,
+      "step": 6700
+    },
+    {
+      "epoch": 0.671,
+      "grad_norm": 28.068639755249023,
+      "learning_rate": 8.66e-06,
+      "loss": 0.513,
+      "step": 6710
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 13.38555908203125,
+      "learning_rate": 8.658e-06,
+      "loss": 0.7307,
+      "step": 6720
+    },
+    {
+      "epoch": 0.673,
+      "grad_norm": 23.34636878967285,
+      "learning_rate": 8.656000000000001e-06,
+      "loss": 0.9183,
+      "step": 6730
+    },
+    {
+      "epoch": 0.674,
+      "grad_norm": 23.878576278686523,
+      "learning_rate": 8.654e-06,
+      "loss": 0.5841,
+      "step": 6740
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 31.957456588745117,
+      "learning_rate": 8.652000000000001e-06,
+      "loss": 0.8774,
+      "step": 6750
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 41.79953384399414,
+      "learning_rate": 8.65e-06,
+      "loss": 0.9604,
+      "step": 6760
+    },
+    {
+      "epoch": 0.677,
+      "grad_norm": 19.412567138671875,
+      "learning_rate": 8.648000000000001e-06,
+      "loss": 0.6639,
+      "step": 6770
+    },
+    {
+      "epoch": 0.678,
+      "grad_norm": 23.539968490600586,
+      "learning_rate": 8.646000000000002e-06,
+      "loss": 0.5529,
+      "step": 6780
+    },
+    {
+      "epoch": 0.679,
+      "grad_norm": 16.99551773071289,
+      "learning_rate": 8.644e-06,
+      "loss": 0.7736,
+      "step": 6790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 33.947628021240234,
+      "learning_rate": 8.642e-06,
+      "loss": 0.7398,
+      "step": 6800
+    },
+    {
+      "epoch": 0.681,
+      "grad_norm": 31.256986618041992,
+      "learning_rate": 8.64e-06,
+      "loss": 0.7431,
+      "step": 6810
+    },
+    {
+      "epoch": 0.682,
+      "grad_norm": 34.16581344604492,
+      "learning_rate": 8.638000000000001e-06,
+      "loss": 0.7627,
+      "step": 6820
+    },
+    {
+      "epoch": 0.683,
+      "grad_norm": 28.380054473876953,
+      "learning_rate": 8.636000000000002e-06,
+      "loss": 0.5163,
+      "step": 6830
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 17.26369285583496,
+      "learning_rate": 8.634e-06,
+      "loss": 0.5183,
+      "step": 6840
+    },
+    {
+      "epoch": 0.685,
+      "grad_norm": 9.074869155883789,
+      "learning_rate": 8.632e-06,
+      "loss": 0.754,
+      "step": 6850
+    },
+    {
+      "epoch": 0.686,
+      "grad_norm": 12.758150100708008,
+      "learning_rate": 8.63e-06,
+      "loss": 0.6343,
+      "step": 6860
+    },
+    {
+      "epoch": 0.687,
+      "grad_norm": 9.85903263092041,
+      "learning_rate": 8.628000000000001e-06,
+      "loss": 0.7374,
+      "step": 6870
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 46.6954231262207,
+      "learning_rate": 8.626000000000002e-06,
+      "loss": 0.6382,
+      "step": 6880
+    },
+    {
+      "epoch": 0.689,
+      "grad_norm": 26.677377700805664,
+      "learning_rate": 8.624e-06,
+      "loss": 1.0401,
+      "step": 6890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 26.68614387512207,
+      "learning_rate": 8.622e-06,
+      "loss": 0.6229,
+      "step": 6900
+    },
+    {
+      "epoch": 0.691,
+      "grad_norm": 10.952954292297363,
+      "learning_rate": 8.62e-06,
+      "loss": 0.8177,
+      "step": 6910
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 32.28446960449219,
+      "learning_rate": 8.618000000000001e-06,
+      "loss": 0.6967,
+      "step": 6920
+    },
+    {
+      "epoch": 0.693,
+      "grad_norm": 32.433128356933594,
+      "learning_rate": 8.616000000000002e-06,
+      "loss": 0.7636,
+      "step": 6930
+    },
+    {
+      "epoch": 0.694,
+      "grad_norm": 39.171165466308594,
+      "learning_rate": 8.614000000000001e-06,
+      "loss": 0.5737,
+      "step": 6940
+    },
+    {
+      "epoch": 0.695,
+      "grad_norm": 32.63961410522461,
+      "learning_rate": 8.612e-06,
+      "loss": 0.7406,
+      "step": 6950
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 30.88587188720703,
+      "learning_rate": 8.61e-06,
+      "loss": 0.9034,
+      "step": 6960
+    },
+    {
+      "epoch": 0.697,
+      "grad_norm": 29.65443992614746,
+      "learning_rate": 8.608000000000001e-06,
+      "loss": 0.4538,
+      "step": 6970
+    },
+    {
+      "epoch": 0.698,
+      "grad_norm": 35.46700668334961,
+      "learning_rate": 8.606e-06,
+      "loss": 0.9041,
+      "step": 6980
+    },
+    {
+      "epoch": 0.699,
+      "grad_norm": 12.977072715759277,
+      "learning_rate": 8.604000000000001e-06,
+      "loss": 0.6177,
+      "step": 6990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.877992630004883,
+      "learning_rate": 8.602e-06,
+      "loss": 0.5468,
+      "step": 7000
+    },
+    {
+      "epoch": 0.701,
+      "grad_norm": 54.49048614501953,
+      "learning_rate": 8.6e-06,
+      "loss": 0.7231,
+      "step": 7010
+    },
+    {
+      "epoch": 0.702,
+      "grad_norm": 51.610084533691406,
+      "learning_rate": 8.598000000000001e-06,
+      "loss": 0.7289,
+      "step": 7020
+    },
+    {
+      "epoch": 0.703,
+      "grad_norm": 38.165809631347656,
+      "learning_rate": 8.596e-06,
+      "loss": 0.9512,
+      "step": 7030
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 26.098329544067383,
+      "learning_rate": 8.594000000000001e-06,
+      "loss": 0.7228,
+      "step": 7040
+    },
+    {
+      "epoch": 0.705,
+      "grad_norm": 46.73612976074219,
+      "learning_rate": 8.592e-06,
+      "loss": 0.7202,
+      "step": 7050
+    },
+    {
+      "epoch": 0.706,
+      "grad_norm": 40.09733200073242,
+      "learning_rate": 8.59e-06,
+      "loss": 0.4938,
+      "step": 7060
+    },
+    {
+      "epoch": 0.707,
+      "grad_norm": 43.058841705322266,
+      "learning_rate": 8.588000000000001e-06,
+      "loss": 0.7683,
+      "step": 7070
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 50.56329345703125,
+      "learning_rate": 8.586e-06,
+      "loss": 1.064,
+      "step": 7080
+    },
+    {
+      "epoch": 0.709,
+      "grad_norm": 39.37093734741211,
+      "learning_rate": 8.584000000000001e-06,
+      "loss": 0.8581,
+      "step": 7090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.572067260742188,
+      "learning_rate": 8.582e-06,
+      "loss": 0.5938,
+      "step": 7100
+    },
+    {
+      "epoch": 0.711,
+      "grad_norm": 16.67902183532715,
+      "learning_rate": 8.580000000000001e-06,
+      "loss": 0.5792,
+      "step": 7110
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 21.606800079345703,
+      "learning_rate": 8.578000000000002e-06,
+      "loss": 0.8628,
+      "step": 7120
+    },
+    {
+      "epoch": 0.713,
+      "grad_norm": 29.96168327331543,
+      "learning_rate": 8.576e-06,
+      "loss": 0.7258,
+      "step": 7130
+    },
+    {
+      "epoch": 0.714,
+      "grad_norm": 17.330432891845703,
+      "learning_rate": 8.574000000000001e-06,
+      "loss": 0.769,
+      "step": 7140
+    },
+    {
+      "epoch": 0.715,
+      "grad_norm": 43.22488021850586,
+      "learning_rate": 8.572e-06,
+      "loss": 0.7662,
+      "step": 7150
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 24.21693992614746,
+      "learning_rate": 8.570000000000001e-06,
+      "loss": 0.8548,
+      "step": 7160
+    },
+    {
+      "epoch": 0.717,
+      "grad_norm": 14.228105545043945,
+      "learning_rate": 8.568e-06,
+      "loss": 0.7917,
+      "step": 7170
+    },
+    {
+      "epoch": 0.718,
+      "grad_norm": 34.96318817138672,
+      "learning_rate": 8.566e-06,
+      "loss": 1.0241,
+      "step": 7180
+    },
+    {
+      "epoch": 0.719,
+      "grad_norm": 23.818344116210938,
+      "learning_rate": 8.564000000000001e-06,
+      "loss": 0.8201,
+      "step": 7190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.386028289794922,
+      "learning_rate": 8.562e-06,
+      "loss": 0.9085,
+      "step": 7200
+    },
+    {
+      "epoch": 0.721,
+      "grad_norm": 11.841687202453613,
+      "learning_rate": 8.560000000000001e-06,
+      "loss": 0.713,
+      "step": 7210
+    },
+    {
+      "epoch": 0.722,
+      "grad_norm": 6.8726725578308105,
+      "learning_rate": 8.558e-06,
+      "loss": 0.6639,
+      "step": 7220
+    },
+    {
+      "epoch": 0.723,
+      "grad_norm": 31.94175910949707,
+      "learning_rate": 8.556e-06,
+      "loss": 0.5847,
+      "step": 7230
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 17.904970169067383,
+      "learning_rate": 8.554000000000001e-06,
+      "loss": 0.6752,
+      "step": 7240
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 10.443678855895996,
+      "learning_rate": 8.552e-06,
+      "loss": 0.9382,
+      "step": 7250
+    },
+    {
+      "epoch": 0.726,
+      "grad_norm": 21.056140899658203,
+      "learning_rate": 8.550000000000001e-06,
+      "loss": 0.7227,
+      "step": 7260
+    },
+    {
+      "epoch": 0.727,
+      "grad_norm": 36.38212585449219,
+      "learning_rate": 8.548e-06,
+      "loss": 0.563,
+      "step": 7270
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 29.130191802978516,
+      "learning_rate": 8.546000000000001e-06,
+      "loss": 0.5688,
+      "step": 7280
+    },
+    {
+      "epoch": 0.729,
+      "grad_norm": 19.342304229736328,
+      "learning_rate": 8.544000000000002e-06,
+      "loss": 0.6482,
+      "step": 7290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 32.37727737426758,
+      "learning_rate": 8.542e-06,
+      "loss": 0.6761,
+      "step": 7300
+    },
+    {
+      "epoch": 0.731,
+      "grad_norm": 33.89266586303711,
+      "learning_rate": 8.540000000000001e-06,
+      "loss": 0.7243,
+      "step": 7310
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 26.32709312438965,
+      "learning_rate": 8.538e-06,
+      "loss": 0.754,
+      "step": 7320
+    },
+    {
+      "epoch": 0.733,
+      "grad_norm": 18.493139266967773,
+      "learning_rate": 8.536000000000001e-06,
+      "loss": 0.3922,
+      "step": 7330
+    },
+    {
+      "epoch": 0.734,
+      "grad_norm": 17.698272705078125,
+      "learning_rate": 8.534000000000002e-06,
+      "loss": 0.8431,
+      "step": 7340
+    },
+    {
+      "epoch": 0.735,
+      "grad_norm": 56.601158142089844,
+      "learning_rate": 8.532e-06,
+      "loss": 0.905,
+      "step": 7350
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 35.621917724609375,
+      "learning_rate": 8.530000000000001e-06,
+      "loss": 0.9357,
+      "step": 7360
+    },
+    {
+      "epoch": 0.737,
+      "grad_norm": 30.305463790893555,
+      "learning_rate": 8.528e-06,
+      "loss": 0.4989,
+      "step": 7370
+    },
+    {
+      "epoch": 0.738,
+      "grad_norm": 35.86700439453125,
+      "learning_rate": 8.526000000000001e-06,
+      "loss": 0.6451,
+      "step": 7380
+    },
+    {
+      "epoch": 0.739,
+      "grad_norm": 35.63044357299805,
+      "learning_rate": 8.524000000000002e-06,
+      "loss": 0.6942,
+      "step": 7390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 21.295764923095703,
+      "learning_rate": 8.522e-06,
+      "loss": 0.8271,
+      "step": 7400
+    },
+    {
+      "epoch": 0.741,
+      "grad_norm": 21.100643157958984,
+      "learning_rate": 8.52e-06,
+      "loss": 0.5248,
+      "step": 7410
+    },
+    {
+      "epoch": 0.742,
+      "grad_norm": 27.021936416625977,
+      "learning_rate": 8.518e-06,
+      "loss": 0.7567,
+      "step": 7420
+    },
+    {
+      "epoch": 0.743,
+      "grad_norm": 38.12123107910156,
+      "learning_rate": 8.516000000000001e-06,
+      "loss": 0.7024,
+      "step": 7430
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 11.784056663513184,
+      "learning_rate": 8.514000000000002e-06,
+      "loss": 0.7275,
+      "step": 7440
+    },
+    {
+      "epoch": 0.745,
+      "grad_norm": 25.664180755615234,
+      "learning_rate": 8.512e-06,
+      "loss": 0.8583,
+      "step": 7450
+    },
+    {
+      "epoch": 0.746,
+      "grad_norm": 23.51327133178711,
+      "learning_rate": 8.51e-06,
+      "loss": 0.9263,
+      "step": 7460
+    },
+    {
+      "epoch": 0.747,
+      "grad_norm": 7.7203593254089355,
+      "learning_rate": 8.508e-06,
+      "loss": 0.6919,
+      "step": 7470
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 24.703340530395508,
+      "learning_rate": 8.506000000000001e-06,
+      "loss": 0.7369,
+      "step": 7480
+    },
+    {
+      "epoch": 0.749,
+      "grad_norm": 31.4427490234375,
+      "learning_rate": 8.504000000000002e-06,
+      "loss": 0.7534,
+      "step": 7490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.99972915649414,
+      "learning_rate": 8.502000000000001e-06,
+      "loss": 0.7025,
+      "step": 7500
+    },
+    {
+      "epoch": 0.751,
+      "grad_norm": 35.3527717590332,
+      "learning_rate": 8.5e-06,
+      "loss": 0.6326,
+      "step": 7510
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 14.60549259185791,
+      "learning_rate": 8.498e-06,
+      "loss": 0.6411,
+      "step": 7520
+    },
+    {
+      "epoch": 0.753,
+      "grad_norm": 11.679787635803223,
+      "learning_rate": 8.496000000000001e-06,
+      "loss": 0.7096,
+      "step": 7530
+    },
+    {
+      "epoch": 0.754,
+      "grad_norm": 16.99401092529297,
+      "learning_rate": 8.494e-06,
+      "loss": 0.6471,
+      "step": 7540
+    },
+    {
+      "epoch": 0.755,
+      "grad_norm": 31.19361114501953,
+      "learning_rate": 8.492000000000001e-06,
+      "loss": 0.7787,
+      "step": 7550
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 20.032594680786133,
+      "learning_rate": 8.49e-06,
+      "loss": 0.7043,
+      "step": 7560
+    },
+    {
+      "epoch": 0.757,
+      "grad_norm": 34.8859748840332,
+      "learning_rate": 8.488e-06,
+      "loss": 0.6349,
+      "step": 7570
+    },
+    {
+      "epoch": 0.758,
+      "grad_norm": 33.36531066894531,
+      "learning_rate": 8.486000000000001e-06,
+      "loss": 0.6744,
+      "step": 7580
+    },
+    {
+      "epoch": 0.759,
+      "grad_norm": 8.187764167785645,
+      "learning_rate": 8.484e-06,
+      "loss": 0.9517,
+      "step": 7590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.946749687194824,
+      "learning_rate": 8.482e-06,
+      "loss": 0.6614,
+      "step": 7600
+    },
+    {
+      "epoch": 0.761,
+      "grad_norm": 27.84326934814453,
+      "learning_rate": 8.48e-06,
+      "loss": 0.6593,
+      "step": 7610
+    },
+    {
+      "epoch": 0.762,
+      "grad_norm": 30.7228946685791,
+      "learning_rate": 8.478e-06,
+      "loss": 0.9224,
+      "step": 7620
+    },
+    {
+      "epoch": 0.763,
+      "grad_norm": 30.333284378051758,
+      "learning_rate": 8.476000000000002e-06,
+      "loss": 0.7381,
+      "step": 7630
+    },
+    {
+      "epoch": 0.764,
+      "grad_norm": 20.903934478759766,
+      "learning_rate": 8.474e-06,
+      "loss": 0.8126,
+      "step": 7640
+    },
+    {
+      "epoch": 0.765,
+      "grad_norm": 19.930513381958008,
+      "learning_rate": 8.472e-06,
+      "loss": 0.4864,
+      "step": 7650
+    },
+    {
+      "epoch": 0.766,
+      "grad_norm": 28.226978302001953,
+      "learning_rate": 8.47e-06,
+      "loss": 0.7474,
+      "step": 7660
+    },
+    {
+      "epoch": 0.767,
+      "grad_norm": 35.864070892333984,
+      "learning_rate": 8.468000000000001e-06,
+      "loss": 0.7979,
+      "step": 7670
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 18.937253952026367,
+      "learning_rate": 8.466000000000002e-06,
+      "loss": 0.6326,
+      "step": 7680
+    },
+    {
+      "epoch": 0.769,
+      "grad_norm": 33.21320724487305,
+      "learning_rate": 8.464e-06,
+      "loss": 0.8367,
+      "step": 7690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.86391258239746,
+      "learning_rate": 8.462e-06,
+      "loss": 0.7428,
+      "step": 7700
+    },
+    {
+      "epoch": 0.771,
+      "grad_norm": 39.599666595458984,
+      "learning_rate": 8.46e-06,
+      "loss": 0.6803,
+      "step": 7710
+    },
+    {
+      "epoch": 0.772,
+      "grad_norm": 31.304195404052734,
+      "learning_rate": 8.458000000000001e-06,
+      "loss": 0.6783,
+      "step": 7720
+    },
+    {
+      "epoch": 0.773,
+      "grad_norm": 21.93630027770996,
+      "learning_rate": 8.456000000000002e-06,
+      "loss": 0.4458,
+      "step": 7730
+    },
+    {
+      "epoch": 0.774,
+      "grad_norm": 10.390119552612305,
+      "learning_rate": 8.454e-06,
+      "loss": 0.5747,
+      "step": 7740
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 29.907333374023438,
+      "learning_rate": 8.452e-06,
+      "loss": 0.6706,
+      "step": 7750
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 14.718128204345703,
+      "learning_rate": 8.45e-06,
+      "loss": 0.8387,
+      "step": 7760
+    },
+    {
+      "epoch": 0.777,
+      "grad_norm": 32.11901092529297,
+      "learning_rate": 8.448000000000001e-06,
+      "loss": 0.7932,
+      "step": 7770
+    },
+    {
+      "epoch": 0.778,
+      "grad_norm": 18.3746280670166,
+      "learning_rate": 8.446e-06,
+      "loss": 0.3266,
+      "step": 7780
+    },
+    {
+      "epoch": 0.779,
+      "grad_norm": 17.5025634765625,
+      "learning_rate": 8.444e-06,
+      "loss": 0.5833,
+      "step": 7790
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 44.1501579284668,
+      "learning_rate": 8.442e-06,
+      "loss": 0.9204,
+      "step": 7800
+    },
+    {
+      "epoch": 0.781,
+      "grad_norm": 4.0904107093811035,
+      "learning_rate": 8.44e-06,
+      "loss": 0.7701,
+      "step": 7810
+    },
+    {
+      "epoch": 0.782,
+      "grad_norm": 14.378214836120605,
+      "learning_rate": 8.438000000000001e-06,
+      "loss": 0.6579,
+      "step": 7820
+    },
+    {
+      "epoch": 0.783,
+      "grad_norm": 31.078834533691406,
+      "learning_rate": 8.436e-06,
+      "loss": 0.7474,
+      "step": 7830
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 32.57761764526367,
+      "learning_rate": 8.434000000000001e-06,
+      "loss": 0.6718,
+      "step": 7840
+    },
+    {
+      "epoch": 0.785,
+      "grad_norm": 29.638538360595703,
+      "learning_rate": 8.432e-06,
+      "loss": 0.4598,
+      "step": 7850
+    },
+    {
+      "epoch": 0.786,
+      "grad_norm": 15.80543041229248,
+      "learning_rate": 8.43e-06,
+      "loss": 0.591,
+      "step": 7860
+    },
+    {
+      "epoch": 0.787,
+      "grad_norm": 22.110992431640625,
+      "learning_rate": 8.428000000000001e-06,
+      "loss": 0.8063,
+      "step": 7870
+    },
+    {
+      "epoch": 0.788,
+      "grad_norm": 20.21088218688965,
+      "learning_rate": 8.426e-06,
+      "loss": 0.7009,
+      "step": 7880
+    },
+    {
+      "epoch": 0.789,
+      "grad_norm": 29.424182891845703,
+      "learning_rate": 8.424000000000001e-06,
+      "loss": 0.6799,
+      "step": 7890
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 22.46828842163086,
+      "learning_rate": 8.422e-06,
+      "loss": 0.6094,
+      "step": 7900
+    },
+    {
+      "epoch": 0.791,
+      "grad_norm": 16.11787986755371,
+      "learning_rate": 8.42e-06,
+      "loss": 0.4994,
+      "step": 7910
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 13.472208023071289,
+      "learning_rate": 8.418000000000001e-06,
+      "loss": 0.6699,
+      "step": 7920
+    },
+    {
+      "epoch": 0.793,
+      "grad_norm": 23.473478317260742,
+      "learning_rate": 8.416e-06,
+      "loss": 0.6749,
+      "step": 7930
+    },
+    {
+      "epoch": 0.794,
+      "grad_norm": 6.8579792976379395,
+      "learning_rate": 8.414000000000001e-06,
+      "loss": 0.4462,
+      "step": 7940
+    },
+    {
+      "epoch": 0.795,
+      "grad_norm": 34.45481491088867,
+      "learning_rate": 8.412e-06,
+      "loss": 0.9276,
+      "step": 7950
+    },
+    {
+      "epoch": 0.796,
+      "grad_norm": 33.5301399230957,
+      "learning_rate": 8.41e-06,
+      "loss": 0.6733,
+      "step": 7960
+    },
+    {
+      "epoch": 0.797,
+      "grad_norm": 57.889190673828125,
+      "learning_rate": 8.408e-06,
+      "loss": 0.9723,
+      "step": 7970
+    },
+    {
+      "epoch": 0.798,
+      "grad_norm": 6.853766441345215,
+      "learning_rate": 8.406e-06,
+      "loss": 0.6781,
+      "step": 7980
+    },
+    {
+      "epoch": 0.799,
+      "grad_norm": 47.17698669433594,
+      "learning_rate": 8.404000000000001e-06,
+      "loss": 0.6879,
+      "step": 7990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.459867477416992,
+      "learning_rate": 8.402e-06,
+      "loss": 0.7155,
+      "step": 8000
+    },
+    {
+      "epoch": 0.801,
+      "grad_norm": 17.590360641479492,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.7116,
+      "step": 8010
+    },
+    {
+      "epoch": 0.802,
+      "grad_norm": 20.620521545410156,
+      "learning_rate": 8.398e-06,
+      "loss": 0.7005,
+      "step": 8020
+    },
+    {
+      "epoch": 0.803,
+      "grad_norm": 44.69348907470703,
+      "learning_rate": 8.396e-06,
+      "loss": 0.7358,
+      "step": 8030
+    },
+    {
+      "epoch": 0.804,
+      "grad_norm": 31.715669631958008,
+      "learning_rate": 8.394000000000001e-06,
+      "loss": 0.8132,
+      "step": 8040
+    },
+    {
+      "epoch": 0.805,
+      "grad_norm": 19.612058639526367,
+      "learning_rate": 8.392e-06,
+      "loss": 0.5062,
+      "step": 8050
+    },
+    {
+      "epoch": 0.806,
+      "grad_norm": 24.114898681640625,
+      "learning_rate": 8.390000000000001e-06,
+      "loss": 0.9493,
+      "step": 8060
+    },
+    {
+      "epoch": 0.807,
+      "grad_norm": 32.89696502685547,
+      "learning_rate": 8.388e-06,
+      "loss": 0.6696,
+      "step": 8070
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 31.07137680053711,
+      "learning_rate": 8.386e-06,
+      "loss": 0.6975,
+      "step": 8080
+    },
+    {
+      "epoch": 0.809,
+      "grad_norm": 31.295612335205078,
+      "learning_rate": 8.384000000000001e-06,
+      "loss": 0.6132,
+      "step": 8090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 25.72049903869629,
+      "learning_rate": 8.382e-06,
+      "loss": 0.4594,
+      "step": 8100
+    },
+    {
+      "epoch": 0.811,
+      "grad_norm": 33.49985885620117,
+      "learning_rate": 8.380000000000001e-06,
+      "loss": 0.7982,
+      "step": 8110
+    },
+    {
+      "epoch": 0.812,
+      "grad_norm": 26.266956329345703,
+      "learning_rate": 8.378e-06,
+      "loss": 0.652,
+      "step": 8120
+    },
+    {
+      "epoch": 0.813,
+      "grad_norm": 8.741480827331543,
+      "learning_rate": 8.376e-06,
+      "loss": 0.6933,
+      "step": 8130
+    },
+    {
+      "epoch": 0.814,
+      "grad_norm": 60.22322082519531,
+      "learning_rate": 8.374000000000001e-06,
+      "loss": 0.9593,
+      "step": 8140
+    },
+    {
+      "epoch": 0.815,
+      "grad_norm": 20.960670471191406,
+      "learning_rate": 8.372e-06,
+      "loss": 0.5567,
+      "step": 8150
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 24.88983726501465,
+      "learning_rate": 8.370000000000001e-06,
+      "loss": 0.5051,
+      "step": 8160
+    },
+    {
+      "epoch": 0.817,
+      "grad_norm": 32.92258834838867,
+      "learning_rate": 8.368e-06,
+      "loss": 0.7988,
+      "step": 8170
+    },
+    {
+      "epoch": 0.818,
+      "grad_norm": 39.90187072753906,
+      "learning_rate": 8.366000000000001e-06,
+      "loss": 0.7172,
+      "step": 8180
+    },
+    {
+      "epoch": 0.819,
+      "grad_norm": 24.0626163482666,
+      "learning_rate": 8.364000000000002e-06,
+      "loss": 0.5431,
+      "step": 8190
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 53.17387771606445,
+      "learning_rate": 8.362e-06,
+      "loss": 0.8249,
+      "step": 8200
+    },
+    {
+      "epoch": 0.821,
+      "grad_norm": 12.743687629699707,
+      "learning_rate": 8.36e-06,
+      "loss": 0.5468,
+      "step": 8210
+    },
+    {
+      "epoch": 0.822,
+      "grad_norm": 23.088817596435547,
+      "learning_rate": 8.358e-06,
+      "loss": 0.648,
+      "step": 8220
+    },
+    {
+      "epoch": 0.823,
+      "grad_norm": 22.222497940063477,
+      "learning_rate": 8.356000000000001e-06,
+      "loss": 0.5813,
+      "step": 8230
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 22.03820037841797,
+      "learning_rate": 8.354000000000002e-06,
+      "loss": 0.8584,
+      "step": 8240
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 7.797426700592041,
+      "learning_rate": 8.352e-06,
+      "loss": 0.8026,
+      "step": 8250
+    },
+    {
+      "epoch": 0.826,
+      "grad_norm": 12.721060752868652,
+      "learning_rate": 8.35e-06,
+      "loss": 0.6768,
+      "step": 8260
+    },
+    {
+      "epoch": 0.827,
+      "grad_norm": 27.198408126831055,
+      "learning_rate": 8.348e-06,
+      "loss": 0.6818,
+      "step": 8270
+    },
+    {
+      "epoch": 0.828,
+      "grad_norm": 22.904878616333008,
+      "learning_rate": 8.346000000000001e-06,
+      "loss": 0.9425,
+      "step": 8280
+    },
+    {
+      "epoch": 0.829,
+      "grad_norm": 22.876659393310547,
+      "learning_rate": 8.344000000000002e-06,
+      "loss": 0.3033,
+      "step": 8290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 35.75187301635742,
+      "learning_rate": 8.342e-06,
+      "loss": 0.9064,
+      "step": 8300
+    },
+    {
+      "epoch": 0.831,
+      "grad_norm": 26.52242660522461,
+      "learning_rate": 8.34e-06,
+      "loss": 0.733,
+      "step": 8310
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.9999659061431885,
+      "learning_rate": 8.338e-06,
+      "loss": 0.7176,
+      "step": 8320
+    },
+    {
+      "epoch": 0.833,
+      "grad_norm": 33.7828369140625,
+      "learning_rate": 8.336000000000001e-06,
+      "loss": 0.8047,
+      "step": 8330
+    },
+    {
+      "epoch": 0.834,
+      "grad_norm": 17.01401138305664,
+      "learning_rate": 8.334e-06,
+      "loss": 0.7676,
+      "step": 8340
+    },
+    {
+      "epoch": 0.835,
+      "grad_norm": 24.290319442749023,
+      "learning_rate": 8.332000000000001e-06,
+      "loss": 0.774,
+      "step": 8350
+    },
+    {
+      "epoch": 0.836,
+      "grad_norm": 24.34035301208496,
+      "learning_rate": 8.33e-06,
+      "loss": 0.7797,
+      "step": 8360
+    },
+    {
+      "epoch": 0.837,
+      "grad_norm": 19.819849014282227,
+      "learning_rate": 8.328e-06,
+      "loss": 0.6335,
+      "step": 8370
+    },
+    {
+      "epoch": 0.838,
+      "grad_norm": 33.246543884277344,
+      "learning_rate": 8.326000000000001e-06,
+      "loss": 0.6919,
+      "step": 8380
+    },
+    {
+      "epoch": 0.839,
+      "grad_norm": 23.2971134185791,
+      "learning_rate": 8.324e-06,
+      "loss": 0.7181,
+      "step": 8390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 24.520105361938477,
+      "learning_rate": 8.322000000000001e-06,
+      "loss": 0.7179,
+      "step": 8400
+    },
+    {
+      "epoch": 0.841,
+      "grad_norm": 11.137072563171387,
+      "learning_rate": 8.32e-06,
+      "loss": 0.6092,
+      "step": 8410
+    },
+    {
+      "epoch": 0.842,
+      "grad_norm": 15.37700366973877,
+      "learning_rate": 8.318e-06,
+      "loss": 0.6143,
+      "step": 8420
+    },
+    {
+      "epoch": 0.843,
+      "grad_norm": 19.72783660888672,
+      "learning_rate": 8.316000000000001e-06,
+      "loss": 0.7324,
+      "step": 8430
+    },
+    {
+      "epoch": 0.844,
+      "grad_norm": 7.460726737976074,
+      "learning_rate": 8.314e-06,
+      "loss": 0.8602,
+      "step": 8440
+    },
+    {
+      "epoch": 0.845,
+      "grad_norm": 18.754709243774414,
+      "learning_rate": 8.312000000000001e-06,
+      "loss": 0.7999,
+      "step": 8450
+    },
+    {
+      "epoch": 0.846,
+      "grad_norm": 19.63807487487793,
+      "learning_rate": 8.31e-06,
+      "loss": 0.9064,
+      "step": 8460
+    },
+    {
+      "epoch": 0.847,
+      "grad_norm": 19.692235946655273,
+      "learning_rate": 8.308e-06,
+      "loss": 0.5737,
+      "step": 8470
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 53.53981018066406,
+      "learning_rate": 8.306000000000001e-06,
+      "loss": 0.7999,
+      "step": 8480
+    },
+    {
+      "epoch": 0.849,
+      "grad_norm": 20.587724685668945,
+      "learning_rate": 8.304e-06,
+      "loss": 0.5465,
+      "step": 8490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 35.91649627685547,
+      "learning_rate": 8.302000000000001e-06,
+      "loss": 0.7199,
+      "step": 8500
+    },
+    {
+      "epoch": 0.851,
+      "grad_norm": 31.062335968017578,
+      "learning_rate": 8.3e-06,
+      "loss": 0.6207,
+      "step": 8510
+    },
+    {
+      "epoch": 0.852,
+      "grad_norm": 28.461896896362305,
+      "learning_rate": 8.298000000000001e-06,
+      "loss": 0.8531,
+      "step": 8520
+    },
+    {
+      "epoch": 0.853,
+      "grad_norm": 42.22334671020508,
+      "learning_rate": 8.296000000000002e-06,
+      "loss": 0.7788,
+      "step": 8530
+    },
+    {
+      "epoch": 0.854,
+      "grad_norm": 27.250551223754883,
+      "learning_rate": 8.294e-06,
+      "loss": 0.499,
+      "step": 8540
+    },
+    {
+      "epoch": 0.855,
+      "grad_norm": 26.948123931884766,
+      "learning_rate": 8.292000000000001e-06,
+      "loss": 0.7913,
+      "step": 8550
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 17.879419326782227,
+      "learning_rate": 8.29e-06,
+      "loss": 0.7423,
+      "step": 8560
+    },
+    {
+      "epoch": 0.857,
+      "grad_norm": 23.53091812133789,
+      "learning_rate": 8.288000000000001e-06,
+      "loss": 0.6809,
+      "step": 8570
+    },
+    {
+      "epoch": 0.858,
+      "grad_norm": 16.258769989013672,
+      "learning_rate": 8.286e-06,
+      "loss": 0.8998,
+      "step": 8580
+    },
+    {
+      "epoch": 0.859,
+      "grad_norm": 23.789386749267578,
+      "learning_rate": 8.284e-06,
+      "loss": 0.6641,
+      "step": 8590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 28.253061294555664,
+      "learning_rate": 8.282000000000001e-06,
+      "loss": 0.8821,
+      "step": 8600
+    },
+    {
+      "epoch": 0.861,
+      "grad_norm": 25.612363815307617,
+      "learning_rate": 8.28e-06,
+      "loss": 0.7803,
+      "step": 8610
+    },
+    {
+      "epoch": 0.862,
+      "grad_norm": 22.466379165649414,
+      "learning_rate": 8.278000000000001e-06,
+      "loss": 0.5856,
+      "step": 8620
+    },
+    {
+      "epoch": 0.863,
+      "grad_norm": 19.716615676879883,
+      "learning_rate": 8.276e-06,
+      "loss": 0.5913,
+      "step": 8630
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 22.004531860351562,
+      "learning_rate": 8.274e-06,
+      "loss": 0.819,
+      "step": 8640
+    },
+    {
+      "epoch": 0.865,
+      "grad_norm": 29.626161575317383,
+      "learning_rate": 8.272000000000001e-06,
+      "loss": 0.47,
+      "step": 8650
+    },
+    {
+      "epoch": 0.866,
+      "grad_norm": 16.204734802246094,
+      "learning_rate": 8.27e-06,
+      "loss": 0.4727,
+      "step": 8660
+    },
+    {
+      "epoch": 0.867,
+      "grad_norm": 37.761837005615234,
+      "learning_rate": 8.268000000000001e-06,
+      "loss": 0.6507,
+      "step": 8670
+    },
+    {
+      "epoch": 0.868,
+      "grad_norm": 40.671695709228516,
+      "learning_rate": 8.266e-06,
+      "loss": 0.7335,
+      "step": 8680
+    },
+    {
+      "epoch": 0.869,
+      "grad_norm": 41.19644546508789,
+      "learning_rate": 8.264e-06,
+      "loss": 0.9103,
+      "step": 8690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 33.284019470214844,
+      "learning_rate": 8.262000000000002e-06,
+      "loss": 0.864,
+      "step": 8700
+    },
+    {
+      "epoch": 0.871,
+      "grad_norm": 20.35594940185547,
+      "learning_rate": 8.26e-06,
+      "loss": 0.7424,
+      "step": 8710
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 17.536413192749023,
+      "learning_rate": 8.258000000000001e-06,
+      "loss": 0.7686,
+      "step": 8720
+    },
+    {
+      "epoch": 0.873,
+      "grad_norm": 25.731748580932617,
+      "learning_rate": 8.256e-06,
+      "loss": 0.7772,
+      "step": 8730
+    },
+    {
+      "epoch": 0.874,
+      "grad_norm": 13.456388473510742,
+      "learning_rate": 8.254000000000001e-06,
+      "loss": 0.6912,
+      "step": 8740
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 24.622709274291992,
+      "learning_rate": 8.252000000000002e-06,
+      "loss": 0.5566,
+      "step": 8750
+    },
+    {
+      "epoch": 0.876,
+      "grad_norm": 29.366680145263672,
+      "learning_rate": 8.25e-06,
+      "loss": 0.8911,
+      "step": 8760
+    },
+    {
+      "epoch": 0.877,
+      "grad_norm": 31.854909896850586,
+      "learning_rate": 8.248e-06,
+      "loss": 0.7445,
+      "step": 8770
+    },
+    {
+      "epoch": 0.878,
+      "grad_norm": 22.391782760620117,
+      "learning_rate": 8.246e-06,
+      "loss": 0.8338,
+      "step": 8780
+    },
+    {
+      "epoch": 0.879,
+      "grad_norm": 23.707935333251953,
+      "learning_rate": 8.244000000000001e-06,
+      "loss": 0.5422,
+      "step": 8790
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 32.04408645629883,
+      "learning_rate": 8.242000000000002e-06,
+      "loss": 0.8426,
+      "step": 8800
+    },
+    {
+      "epoch": 0.881,
+      "grad_norm": 14.735499382019043,
+      "learning_rate": 8.24e-06,
+      "loss": 0.7227,
+      "step": 8810
+    },
+    {
+      "epoch": 0.882,
+      "grad_norm": 27.44671630859375,
+      "learning_rate": 8.238e-06,
+      "loss": 0.7975,
+      "step": 8820
+    },
+    {
+      "epoch": 0.883,
+      "grad_norm": 15.00820255279541,
+      "learning_rate": 8.236e-06,
+      "loss": 0.6293,
+      "step": 8830
+    },
+    {
+      "epoch": 0.884,
+      "grad_norm": 23.110822677612305,
+      "learning_rate": 8.234000000000001e-06,
+      "loss": 0.5946,
+      "step": 8840
+    },
+    {
+      "epoch": 0.885,
+      "grad_norm": 22.80463218688965,
+      "learning_rate": 8.232000000000002e-06,
+      "loss": 0.6666,
+      "step": 8850
+    },
+    {
+      "epoch": 0.886,
+      "grad_norm": 33.523399353027344,
+      "learning_rate": 8.23e-06,
+      "loss": 0.8169,
+      "step": 8860
+    },
+    {
+      "epoch": 0.887,
+      "grad_norm": 19.332923889160156,
+      "learning_rate": 8.228e-06,
+      "loss": 0.6937,
+      "step": 8870
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 12.804621696472168,
+      "learning_rate": 8.226e-06,
+      "loss": 0.6046,
+      "step": 8880
+    },
+    {
+      "epoch": 0.889,
+      "grad_norm": 18.660120010375977,
+      "learning_rate": 8.224000000000001e-06,
+      "loss": 0.7445,
+      "step": 8890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.0673246383667,
+      "learning_rate": 8.222000000000002e-06,
+      "loss": 0.8442,
+      "step": 8900
+    },
+    {
+      "epoch": 0.891,
+      "grad_norm": 22.311397552490234,
+      "learning_rate": 8.220000000000001e-06,
+      "loss": 0.6267,
+      "step": 8910
+    },
+    {
+      "epoch": 0.892,
+      "grad_norm": 35.950958251953125,
+      "learning_rate": 8.218e-06,
+      "loss": 0.6888,
+      "step": 8920
+    },
+    {
+      "epoch": 0.893,
+      "grad_norm": 19.454551696777344,
+      "learning_rate": 8.216e-06,
+      "loss": 0.7747,
+      "step": 8930
+    },
+    {
+      "epoch": 0.894,
+      "grad_norm": 24.78478240966797,
+      "learning_rate": 8.214000000000001e-06,
+      "loss": 0.6721,
+      "step": 8940
+    },
+    {
+      "epoch": 0.895,
+      "grad_norm": 21.27381706237793,
+      "learning_rate": 8.212e-06,
+      "loss": 0.7577,
+      "step": 8950
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 24.68490982055664,
+      "learning_rate": 8.210000000000001e-06,
+      "loss": 0.7143,
+      "step": 8960
+    },
+    {
+      "epoch": 0.897,
+      "grad_norm": 22.952789306640625,
+      "learning_rate": 8.208e-06,
+      "loss": 0.6053,
+      "step": 8970
+    },
+    {
+      "epoch": 0.898,
+      "grad_norm": 29.25204849243164,
+      "learning_rate": 8.206e-06,
+      "loss": 0.5921,
+      "step": 8980
+    },
+    {
+      "epoch": 0.899,
+      "grad_norm": 26.693504333496094,
+      "learning_rate": 8.204000000000001e-06,
+      "loss": 0.6771,
+      "step": 8990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 30.661624908447266,
+      "learning_rate": 8.202e-06,
+      "loss": 0.6005,
+      "step": 9000
+    },
+    {
+      "epoch": 0.901,
+      "grad_norm": 36.014434814453125,
+      "learning_rate": 8.2e-06,
+      "loss": 0.6479,
+      "step": 9010
+    },
+    {
+      "epoch": 0.902,
+      "grad_norm": 27.50763511657715,
+      "learning_rate": 8.198e-06,
+      "loss": 0.6197,
+      "step": 9020
+    },
+    {
+      "epoch": 0.903,
+      "grad_norm": 27.790868759155273,
+      "learning_rate": 8.196e-06,
+      "loss": 0.7401,
+      "step": 9030
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 25.92188262939453,
+      "learning_rate": 8.194000000000002e-06,
+      "loss": 0.6251,
+      "step": 9040
+    },
+    {
+      "epoch": 0.905,
+      "grad_norm": 12.84793758392334,
+      "learning_rate": 8.192e-06,
+      "loss": 0.693,
+      "step": 9050
+    },
+    {
+      "epoch": 0.906,
+      "grad_norm": 26.72030258178711,
+      "learning_rate": 8.19e-06,
+      "loss": 0.6391,
+      "step": 9060
+    },
+    {
+      "epoch": 0.907,
+      "grad_norm": 36.05525207519531,
+      "learning_rate": 8.188e-06,
+      "loss": 0.7277,
+      "step": 9070
+    },
+    {
+      "epoch": 0.908,
+      "grad_norm": 52.11860275268555,
+      "learning_rate": 8.186000000000001e-06,
+      "loss": 1.0863,
+      "step": 9080
+    },
+    {
+      "epoch": 0.909,
+      "grad_norm": 21.379451751708984,
+      "learning_rate": 8.184000000000002e-06,
+      "loss": 0.7588,
+      "step": 9090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 25.653507232666016,
+      "learning_rate": 8.182e-06,
+      "loss": 0.7232,
+      "step": 9100
+    },
+    {
+      "epoch": 0.911,
+      "grad_norm": 84.60053253173828,
+      "learning_rate": 8.18e-06,
+      "loss": 0.8049,
+      "step": 9110
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 15.349387168884277,
+      "learning_rate": 8.178e-06,
+      "loss": 0.7509,
+      "step": 9120
+    },
+    {
+      "epoch": 0.913,
+      "grad_norm": 18.783823013305664,
+      "learning_rate": 8.176000000000001e-06,
+      "loss": 0.6542,
+      "step": 9130
+    },
+    {
+      "epoch": 0.914,
+      "grad_norm": 53.27655029296875,
+      "learning_rate": 8.174e-06,
+      "loss": 0.8797,
+      "step": 9140
+    },
+    {
+      "epoch": 0.915,
+      "grad_norm": 27.371829986572266,
+      "learning_rate": 8.172e-06,
+      "loss": 0.6392,
+      "step": 9150
+    },
+    {
+      "epoch": 0.916,
+      "grad_norm": 17.546588897705078,
+      "learning_rate": 8.17e-06,
+      "loss": 0.8283,
+      "step": 9160
+    },
+    {
+      "epoch": 0.917,
+      "grad_norm": 53.05363464355469,
+      "learning_rate": 8.168e-06,
+      "loss": 0.6169,
+      "step": 9170
+    },
+    {
+      "epoch": 0.918,
+      "grad_norm": 43.26979446411133,
+      "learning_rate": 8.166000000000001e-06,
+      "loss": 0.6253,
+      "step": 9180
+    },
+    {
+      "epoch": 0.919,
+      "grad_norm": 33.546607971191406,
+      "learning_rate": 8.164e-06,
+      "loss": 0.7871,
+      "step": 9190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.680317878723145,
+      "learning_rate": 8.162e-06,
+      "loss": 0.8808,
+      "step": 9200
+    },
+    {
+      "epoch": 0.921,
+      "grad_norm": 53.66688919067383,
+      "learning_rate": 8.1602e-06,
+      "loss": 0.5977,
+      "step": 9210
+    },
+    {
+      "epoch": 0.922,
+      "grad_norm": 14.099869728088379,
+      "learning_rate": 8.158200000000001e-06,
+      "loss": 0.973,
+      "step": 9220
+    },
+    {
+      "epoch": 0.923,
+      "grad_norm": 5.584235668182373,
+      "learning_rate": 8.1562e-06,
+      "loss": 0.464,
+      "step": 9230
+    },
+    {
+      "epoch": 0.924,
+      "grad_norm": 15.951348304748535,
+      "learning_rate": 8.1542e-06,
+      "loss": 0.5346,
+      "step": 9240
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 31.42668342590332,
+      "learning_rate": 8.1522e-06,
+      "loss": 0.619,
+      "step": 9250
+    },
+    {
+      "epoch": 0.926,
+      "grad_norm": 18.869773864746094,
+      "learning_rate": 8.1502e-06,
+      "loss": 0.5971,
+      "step": 9260
+    },
+    {
+      "epoch": 0.927,
+      "grad_norm": 35.41682434082031,
+      "learning_rate": 8.148200000000001e-06,
+      "loss": 0.9642,
+      "step": 9270
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 32.13036346435547,
+      "learning_rate": 8.1462e-06,
+      "loss": 0.7917,
+      "step": 9280
+    },
+    {
+      "epoch": 0.929,
+      "grad_norm": 19.31574249267578,
+      "learning_rate": 8.1442e-06,
+      "loss": 0.7301,
+      "step": 9290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 47.323856353759766,
+      "learning_rate": 8.142200000000001e-06,
+      "loss": 0.7374,
+      "step": 9300
+    },
+    {
+      "epoch": 0.931,
+      "grad_norm": 9.623072624206543,
+      "learning_rate": 8.1402e-06,
+      "loss": 0.5891,
+      "step": 9310
+    },
+    {
+      "epoch": 0.932,
+      "grad_norm": 31.580720901489258,
+      "learning_rate": 8.138200000000001e-06,
+      "loss": 0.7645,
+      "step": 9320
+    },
+    {
+      "epoch": 0.933,
+      "grad_norm": 23.051807403564453,
+      "learning_rate": 8.1362e-06,
+      "loss": 0.7925,
+      "step": 9330
+    },
+    {
+      "epoch": 0.934,
+      "grad_norm": 19.1823787689209,
+      "learning_rate": 8.1342e-06,
+      "loss": 0.8137,
+      "step": 9340
+    },
+    {
+      "epoch": 0.935,
+      "grad_norm": 40.923606872558594,
+      "learning_rate": 8.132200000000002e-06,
+      "loss": 0.7003,
+      "step": 9350
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 18.946191787719727,
+      "learning_rate": 8.1302e-06,
+      "loss": 0.6805,
+      "step": 9360
+    },
+    {
+      "epoch": 0.937,
+      "grad_norm": 29.28422737121582,
+      "learning_rate": 8.128200000000001e-06,
+      "loss": 0.8437,
+      "step": 9370
+    },
+    {
+      "epoch": 0.938,
+      "grad_norm": 16.271589279174805,
+      "learning_rate": 8.1262e-06,
+      "loss": 0.5734,
+      "step": 9380
+    },
+    {
+      "epoch": 0.939,
+      "grad_norm": 31.045360565185547,
+      "learning_rate": 8.124200000000001e-06,
+      "loss": 0.6937,
+      "step": 9390
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 25.518590927124023,
+      "learning_rate": 8.122200000000002e-06,
+      "loss": 0.7622,
+      "step": 9400
+    },
+    {
+      "epoch": 0.941,
+      "grad_norm": 14.280061721801758,
+      "learning_rate": 8.1202e-06,
+      "loss": 0.4491,
+      "step": 9410
+    },
+    {
+      "epoch": 0.942,
+      "grad_norm": 13.236858367919922,
+      "learning_rate": 8.118200000000001e-06,
+      "loss": 0.8063,
+      "step": 9420
+    },
+    {
+      "epoch": 0.943,
+      "grad_norm": 25.479963302612305,
+      "learning_rate": 8.1162e-06,
+      "loss": 0.847,
+      "step": 9430
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 47.634483337402344,
+      "learning_rate": 8.114200000000001e-06,
+      "loss": 0.7497,
+      "step": 9440
+    },
+    {
+      "epoch": 0.945,
+      "grad_norm": 20.367294311523438,
+      "learning_rate": 8.112200000000002e-06,
+      "loss": 0.4785,
+      "step": 9450
+    },
+    {
+      "epoch": 0.946,
+      "grad_norm": 34.85844039916992,
+      "learning_rate": 8.1102e-06,
+      "loss": 0.7972,
+      "step": 9460
+    },
+    {
+      "epoch": 0.947,
+      "grad_norm": 30.255382537841797,
+      "learning_rate": 8.1082e-06,
+      "loss": 0.7756,
+      "step": 9470
+    },
+    {
+      "epoch": 0.948,
+      "grad_norm": 30.46112060546875,
+      "learning_rate": 8.1062e-06,
+      "loss": 0.8008,
+      "step": 9480
+    },
+    {
+      "epoch": 0.949,
+      "grad_norm": 15.27729606628418,
+      "learning_rate": 8.104200000000001e-06,
+      "loss": 0.5169,
+      "step": 9490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 81.02428436279297,
+      "learning_rate": 8.102200000000002e-06,
+      "loss": 0.7141,
+      "step": 9500
+    },
+    {
+      "epoch": 0.951,
+      "grad_norm": 22.240251541137695,
+      "learning_rate": 8.1002e-06,
+      "loss": 0.5889,
+      "step": 9510
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 28.84646987915039,
+      "learning_rate": 8.0982e-06,
+      "loss": 0.6834,
+      "step": 9520
+    },
+    {
+      "epoch": 0.953,
+      "grad_norm": 35.79847717285156,
+      "learning_rate": 8.0962e-06,
+      "loss": 0.7559,
+      "step": 9530
+    },
+    {
+      "epoch": 0.954,
+      "grad_norm": 25.810239791870117,
+      "learning_rate": 8.094200000000001e-06,
+      "loss": 0.7268,
+      "step": 9540
+    },
+    {
+      "epoch": 0.955,
+      "grad_norm": 34.47658157348633,
+      "learning_rate": 8.092200000000002e-06,
+      "loss": 0.6295,
+      "step": 9550
+    },
+    {
+      "epoch": 0.956,
+      "grad_norm": 33.548397064208984,
+      "learning_rate": 8.090200000000001e-06,
+      "loss": 0.7543,
+      "step": 9560
+    },
+    {
+      "epoch": 0.957,
+      "grad_norm": 28.522300720214844,
+      "learning_rate": 8.0882e-06,
+      "loss": 0.6737,
+      "step": 9570
+    },
+    {
+      "epoch": 0.958,
+      "grad_norm": 4.9988226890563965,
+      "learning_rate": 8.0862e-06,
+      "loss": 0.6983,
+      "step": 9580
+    },
+    {
+      "epoch": 0.959,
+      "grad_norm": 4.651895523071289,
+      "learning_rate": 8.084200000000001e-06,
+      "loss": 0.6858,
+      "step": 9590
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 32.256072998046875,
+      "learning_rate": 8.082200000000002e-06,
+      "loss": 0.7514,
+      "step": 9600
+    },
+    {
+      "epoch": 0.961,
+      "grad_norm": 21.96092987060547,
+      "learning_rate": 8.080200000000001e-06,
+      "loss": 0.6136,
+      "step": 9610
+    },
+    {
+      "epoch": 0.962,
+      "grad_norm": 11.350577354431152,
+      "learning_rate": 8.0782e-06,
+      "loss": 0.6035,
+      "step": 9620
+    },
+    {
+      "epoch": 0.963,
+      "grad_norm": 23.908449172973633,
+      "learning_rate": 8.0762e-06,
+      "loss": 0.515,
+      "step": 9630
+    },
+    {
+      "epoch": 0.964,
+      "grad_norm": 20.183002471923828,
+      "learning_rate": 8.074200000000001e-06,
+      "loss": 0.6201,
+      "step": 9640
+    },
+    {
+      "epoch": 0.965,
+      "grad_norm": 48.198543548583984,
+      "learning_rate": 8.0722e-06,
+      "loss": 0.6703,
+      "step": 9650
+    },
+    {
+      "epoch": 0.966,
+      "grad_norm": 57.94309616088867,
+      "learning_rate": 8.070200000000001e-06,
+      "loss": 0.952,
+      "step": 9660
+    },
+    {
+      "epoch": 0.967,
+      "grad_norm": 0.18479055166244507,
+      "learning_rate": 8.0682e-06,
+      "loss": 0.5543,
+      "step": 9670
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 31.414304733276367,
+      "learning_rate": 8.0662e-06,
+      "loss": 0.7087,
+      "step": 9680
+    },
+    {
+      "epoch": 0.969,
+      "grad_norm": 21.823001861572266,
+      "learning_rate": 8.064200000000001e-06,
+      "loss": 0.8891,
+      "step": 9690
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 20.574464797973633,
+      "learning_rate": 8.0622e-06,
+      "loss": 0.7342,
+      "step": 9700
+    },
+    {
+      "epoch": 0.971,
+      "grad_norm": 40.75229263305664,
+      "learning_rate": 8.0602e-06,
+      "loss": 0.6528,
+      "step": 9710
+    },
+    {
+      "epoch": 0.972,
+      "grad_norm": 31.854001998901367,
+      "learning_rate": 8.0582e-06,
+      "loss": 0.7637,
+      "step": 9720
+    },
+    {
+      "epoch": 0.973,
+      "grad_norm": 58.702945709228516,
+      "learning_rate": 8.056200000000001e-06,
+      "loss": 1.0291,
+      "step": 9730
+    },
+    {
+      "epoch": 0.974,
+      "grad_norm": 22.919151306152344,
+      "learning_rate": 8.054200000000002e-06,
+      "loss": 0.6746,
+      "step": 9740
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 13.362497329711914,
+      "learning_rate": 8.0522e-06,
+      "loss": 0.8915,
+      "step": 9750
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 24.370723724365234,
+      "learning_rate": 8.0502e-06,
+      "loss": 0.5496,
+      "step": 9760
+    },
+    {
+      "epoch": 0.977,
+      "grad_norm": 26.763219833374023,
+      "learning_rate": 8.0482e-06,
+      "loss": 0.8504,
+      "step": 9770
+    },
+    {
+      "epoch": 0.978,
+      "grad_norm": 23.564735412597656,
+      "learning_rate": 8.046200000000001e-06,
+      "loss": 0.5827,
+      "step": 9780
+    },
+    {
+      "epoch": 0.979,
+      "grad_norm": 34.036258697509766,
+      "learning_rate": 8.044200000000002e-06,
+      "loss": 0.7833,
+      "step": 9790
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 23.769405364990234,
+      "learning_rate": 8.0422e-06,
+      "loss": 0.8223,
+      "step": 9800
+    },
+    {
+      "epoch": 0.981,
+      "grad_norm": 16.29641342163086,
+      "learning_rate": 8.0402e-06,
+      "loss": 0.8408,
+      "step": 9810
+    },
+    {
+      "epoch": 0.982,
+      "grad_norm": 6.285024642944336,
+      "learning_rate": 8.0382e-06,
+      "loss": 0.6709,
+      "step": 9820
+    },
+    {
+      "epoch": 0.983,
+      "grad_norm": 16.67902946472168,
+      "learning_rate": 8.036200000000001e-06,
+      "loss": 0.8057,
+      "step": 9830
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 15.825904846191406,
+      "learning_rate": 8.0342e-06,
+      "loss": 0.5174,
+      "step": 9840
+    },
+    {
+      "epoch": 0.985,
+      "grad_norm": 22.62651824951172,
+      "learning_rate": 8.0322e-06,
+      "loss": 0.5081,
+      "step": 9850
+    },
+    {
+      "epoch": 0.986,
+      "grad_norm": 22.813600540161133,
+      "learning_rate": 8.0302e-06,
+      "loss": 0.6055,
+      "step": 9860
+    },
+    {
+      "epoch": 0.987,
+      "grad_norm": 39.27797317504883,
+      "learning_rate": 8.0282e-06,
+      "loss": 0.6965,
+      "step": 9870
+    },
+    {
+      "epoch": 0.988,
+      "grad_norm": 21.403615951538086,
+      "learning_rate": 8.026200000000001e-06,
+      "loss": 0.6509,
+      "step": 9880
+    },
+    {
+      "epoch": 0.989,
+      "grad_norm": 38.3160400390625,
+      "learning_rate": 8.0242e-06,
+      "loss": 0.4817,
+      "step": 9890
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.218068599700928,
+      "learning_rate": 8.022200000000001e-06,
+      "loss": 0.8259,
+      "step": 9900
+    },
+    {
+      "epoch": 0.991,
+      "grad_norm": 38.34626388549805,
+      "learning_rate": 8.0202e-06,
+      "loss": 0.7534,
+      "step": 9910
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 17.265363693237305,
+      "learning_rate": 8.0182e-06,
+      "loss": 0.807,
+      "step": 9920
+    },
+    {
+      "epoch": 0.993,
+      "grad_norm": 25.756587982177734,
+      "learning_rate": 8.016200000000001e-06,
+      "loss": 0.7307,
+      "step": 9930
+    },
+    {
+      "epoch": 0.994,
+      "grad_norm": 27.01302719116211,
+      "learning_rate": 8.0142e-06,
+      "loss": 0.7551,
+      "step": 9940
+    },
+    {
+      "epoch": 0.995,
+      "grad_norm": 22.30595588684082,
+      "learning_rate": 8.012200000000001e-06,
+      "loss": 0.6335,
+      "step": 9950
+    },
+    {
+      "epoch": 0.996,
+      "grad_norm": 29.83885955810547,
+      "learning_rate": 8.0102e-06,
+      "loss": 0.6483,
+      "step": 9960
+    },
+    {
+      "epoch": 0.997,
+      "grad_norm": 22.51917839050293,
+      "learning_rate": 8.0082e-06,
+      "loss": 0.7563,
+      "step": 9970
+    },
+    {
+      "epoch": 0.998,
+      "grad_norm": 28.65583038330078,
+      "learning_rate": 8.006200000000001e-06,
+      "loss": 0.5714,
+      "step": 9980
+    },
+    {
+      "epoch": 0.999,
+      "grad_norm": 40.359317779541016,
+      "learning_rate": 8.0042e-06,
+      "loss": 0.615,
+      "step": 9990
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 48.593536376953125,
+      "learning_rate": 8.002200000000001e-06,
+      "loss": 0.6024,
+      "step": 10000
+    },
+    {
+      "epoch": 1.001,
+      "grad_norm": 32.619136810302734,
+      "learning_rate": 8.0002e-06,
+      "loss": 0.7511,
+      "step": 10010
+    },
+    {
+      "epoch": 1.002,
+      "grad_norm": 22.86610221862793,
+      "learning_rate": 7.9982e-06,
+      "loss": 0.7653,
+      "step": 10020
+    },
+    {
+      "epoch": 1.003,
+      "grad_norm": 36.1226692199707,
+      "learning_rate": 7.996200000000001e-06,
+      "loss": 0.7023,
+      "step": 10030
+    },
+    {
+      "epoch": 1.004,
+      "grad_norm": 48.830448150634766,
+      "learning_rate": 7.9942e-06,
+      "loss": 0.6821,
+      "step": 10040
+    },
+    {
+      "epoch": 1.005,
+      "grad_norm": 40.12712097167969,
+      "learning_rate": 7.992200000000001e-06,
+      "loss": 0.9133,
+      "step": 10050
+    },
+    {
+      "epoch": 1.006,
+      "grad_norm": 16.794452667236328,
+      "learning_rate": 7.9902e-06,
+      "loss": 0.6606,
+      "step": 10060
+    },
+    {
+      "epoch": 1.007,
+      "grad_norm": 22.306949615478516,
+      "learning_rate": 7.988200000000001e-06,
+      "loss": 0.5966,
+      "step": 10070
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 10.76391887664795,
+      "learning_rate": 7.9862e-06,
+      "loss": 0.5588,
+      "step": 10080
+    },
+    {
+      "epoch": 1.009,
+      "grad_norm": 30.505229949951172,
+      "learning_rate": 7.9842e-06,
+      "loss": 0.7045,
+      "step": 10090
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 25.643178939819336,
+      "learning_rate": 7.982200000000001e-06,
+      "loss": 0.6035,
+      "step": 10100
+    },
+    {
+      "epoch": 1.011,
+      "grad_norm": 29.607240676879883,
+      "learning_rate": 7.9802e-06,
+      "loss": 0.4267,
+      "step": 10110
+    },
+    {
+      "epoch": 1.012,
+      "grad_norm": 28.369211196899414,
+      "learning_rate": 7.978200000000001e-06,
+      "loss": 0.6734,
+      "step": 10120
+    },
+    {
+      "epoch": 1.013,
+      "grad_norm": 23.7529354095459,
+      "learning_rate": 7.9762e-06,
+      "loss": 0.632,
+      "step": 10130
+    },
+    {
+      "epoch": 1.014,
+      "grad_norm": 38.64054870605469,
+      "learning_rate": 7.9742e-06,
+      "loss": 0.7071,
+      "step": 10140
+    },
+    {
+      "epoch": 1.015,
+      "grad_norm": 24.465662002563477,
+      "learning_rate": 7.972200000000001e-06,
+      "loss": 0.7538,
+      "step": 10150
+    },
+    {
+      "epoch": 1.016,
+      "grad_norm": 40.59769821166992,
+      "learning_rate": 7.9702e-06,
+      "loss": 0.7346,
+      "step": 10160
+    },
+    {
+      "epoch": 1.017,
+      "grad_norm": 4.554698467254639,
+      "learning_rate": 7.968200000000001e-06,
+      "loss": 0.5459,
+      "step": 10170
+    },
+    {
+      "epoch": 1.018,
+      "grad_norm": 10.751768112182617,
+      "learning_rate": 7.9662e-06,
+      "loss": 0.4939,
+      "step": 10180
+    },
+    {
+      "epoch": 1.019,
+      "grad_norm": 2.6401894092559814,
+      "learning_rate": 7.9642e-06,
+      "loss": 0.5954,
+      "step": 10190
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.48724937438965,
+      "learning_rate": 7.962200000000001e-06,
+      "loss": 0.6032,
+      "step": 10200
+    },
+    {
+      "epoch": 1.021,
+      "grad_norm": 27.149873733520508,
+      "learning_rate": 7.9602e-06,
+      "loss": 0.4046,
+      "step": 10210
+    },
+    {
+      "epoch": 1.022,
+      "grad_norm": 25.46242332458496,
+      "learning_rate": 7.958200000000001e-06,
+      "loss": 0.8915,
+      "step": 10220
+    },
+    {
+      "epoch": 1.023,
+      "grad_norm": 50.827213287353516,
+      "learning_rate": 7.9562e-06,
+      "loss": 0.6709,
+      "step": 10230
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 21.56336212158203,
+      "learning_rate": 7.954200000000001e-06,
+      "loss": 0.573,
+      "step": 10240
+    },
+    {
+      "epoch": 1.025,
+      "grad_norm": 20.957317352294922,
+      "learning_rate": 7.952200000000002e-06,
+      "loss": 0.5161,
+      "step": 10250
+    },
+    {
+      "epoch": 1.026,
+      "grad_norm": 46.837120056152344,
+      "learning_rate": 7.9502e-06,
+      "loss": 0.7277,
+      "step": 10260
+    },
+    {
+      "epoch": 1.027,
+      "grad_norm": 75.76361846923828,
+      "learning_rate": 7.9482e-06,
+      "loss": 0.7414,
+      "step": 10270
+    },
+    {
+      "epoch": 1.028,
+      "grad_norm": 53.485206604003906,
+      "learning_rate": 7.9462e-06,
+      "loss": 1.0627,
+      "step": 10280
+    },
+    {
+      "epoch": 1.029,
+      "grad_norm": 21.28236961364746,
+      "learning_rate": 7.944200000000001e-06,
+      "loss": 0.5271,
+      "step": 10290
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.299246788024902,
+      "learning_rate": 7.942200000000002e-06,
+      "loss": 0.6222,
+      "step": 10300
+    },
+    {
+      "epoch": 1.031,
+      "grad_norm": 46.65543746948242,
+      "learning_rate": 7.9402e-06,
+      "loss": 0.8833,
+      "step": 10310
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 28.530805587768555,
+      "learning_rate": 7.9382e-06,
+      "loss": 0.9351,
+      "step": 10320
+    },
+    {
+      "epoch": 1.033,
+      "grad_norm": 21.32673454284668,
+      "learning_rate": 7.9362e-06,
+      "loss": 0.5873,
+      "step": 10330
+    },
+    {
+      "epoch": 1.034,
+      "grad_norm": 33.533451080322266,
+      "learning_rate": 7.934200000000001e-06,
+      "loss": 0.7576,
+      "step": 10340
+    },
+    {
+      "epoch": 1.035,
+      "grad_norm": 36.07699966430664,
+      "learning_rate": 7.932200000000002e-06,
+      "loss": 0.7857,
+      "step": 10350
+    },
+    {
+      "epoch": 1.036,
+      "grad_norm": 22.704744338989258,
+      "learning_rate": 7.9302e-06,
+      "loss": 0.7796,
+      "step": 10360
+    },
+    {
+      "epoch": 1.037,
+      "grad_norm": 21.681188583374023,
+      "learning_rate": 7.9282e-06,
+      "loss": 0.8053,
+      "step": 10370
+    },
+    {
+      "epoch": 1.038,
+      "grad_norm": 25.426740646362305,
+      "learning_rate": 7.9262e-06,
+      "loss": 0.7377,
+      "step": 10380
+    },
+    {
+      "epoch": 1.039,
+      "grad_norm": 34.104248046875,
+      "learning_rate": 7.924200000000001e-06,
+      "loss": 0.6537,
+      "step": 10390
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 14.893959999084473,
+      "learning_rate": 7.922200000000002e-06,
+      "loss": 0.4739,
+      "step": 10400
+    },
+    {
+      "epoch": 1.041,
+      "grad_norm": 64.21956634521484,
+      "learning_rate": 7.9202e-06,
+      "loss": 0.8915,
+      "step": 10410
+    },
+    {
+      "epoch": 1.042,
+      "grad_norm": 35.314414978027344,
+      "learning_rate": 7.9182e-06,
+      "loss": 0.8416,
+      "step": 10420
+    },
+    {
+      "epoch": 1.043,
+      "grad_norm": 9.086963653564453,
+      "learning_rate": 7.9162e-06,
+      "loss": 0.7986,
+      "step": 10430
+    },
+    {
+      "epoch": 1.044,
+      "grad_norm": 20.71559715270996,
+      "learning_rate": 7.914200000000001e-06,
+      "loss": 0.5166,
+      "step": 10440
+    },
+    {
+      "epoch": 1.045,
+      "grad_norm": 32.6236572265625,
+      "learning_rate": 7.9122e-06,
+      "loss": 0.7166,
+      "step": 10450
+    },
+    {
+      "epoch": 1.046,
+      "grad_norm": 34.77704620361328,
+      "learning_rate": 7.910200000000001e-06,
+      "loss": 0.7366,
+      "step": 10460
+    },
+    {
+      "epoch": 1.047,
+      "grad_norm": 30.373966217041016,
+      "learning_rate": 7.9082e-06,
+      "loss": 0.5817,
+      "step": 10470
+    },
+    {
+      "epoch": 1.048,
+      "grad_norm": 21.460689544677734,
+      "learning_rate": 7.9062e-06,
+      "loss": 0.6465,
+      "step": 10480
+    },
+    {
+      "epoch": 1.049,
+      "grad_norm": 21.833818435668945,
+      "learning_rate": 7.904200000000001e-06,
+      "loss": 0.5789,
+      "step": 10490
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 22.837757110595703,
+      "learning_rate": 7.9022e-06,
+      "loss": 0.8323,
+      "step": 10500
+    },
+    {
+      "epoch": 1.051,
+      "grad_norm": 14.228981018066406,
+      "learning_rate": 7.9002e-06,
+      "loss": 0.6262,
+      "step": 10510
+    },
+    {
+      "epoch": 1.052,
+      "grad_norm": 24.556312561035156,
+      "learning_rate": 7.8982e-06,
+      "loss": 0.7267,
+      "step": 10520
+    },
+    {
+      "epoch": 1.053,
+      "grad_norm": 54.67141342163086,
+      "learning_rate": 7.8962e-06,
+      "loss": 0.4909,
+      "step": 10530
+    },
+    {
+      "epoch": 1.054,
+      "grad_norm": 33.58781051635742,
+      "learning_rate": 7.894200000000001e-06,
+      "loss": 0.6022,
+      "step": 10540
+    },
+    {
+      "epoch": 1.055,
+      "grad_norm": 16.152503967285156,
+      "learning_rate": 7.8922e-06,
+      "loss": 0.5752,
+      "step": 10550
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 50.25066375732422,
+      "learning_rate": 7.890200000000001e-06,
+      "loss": 0.7387,
+      "step": 10560
+    },
+    {
+      "epoch": 1.057,
+      "grad_norm": 23.39357566833496,
+      "learning_rate": 7.8882e-06,
+      "loss": 0.5076,
+      "step": 10570
+    },
+    {
+      "epoch": 1.058,
+      "grad_norm": 22.392078399658203,
+      "learning_rate": 7.8862e-06,
+      "loss": 0.961,
+      "step": 10580
+    },
+    {
+      "epoch": 1.059,
+      "grad_norm": 12.517849922180176,
+      "learning_rate": 7.884200000000002e-06,
+      "loss": 0.6681,
+      "step": 10590
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 25.450721740722656,
+      "learning_rate": 7.8822e-06,
+      "loss": 0.7726,
+      "step": 10600
+    },
+    {
+      "epoch": 1.061,
+      "grad_norm": 37.7824821472168,
+      "learning_rate": 7.880200000000001e-06,
+      "loss": 0.4592,
+      "step": 10610
+    },
+    {
+      "epoch": 1.062,
+      "grad_norm": 33.96101379394531,
+      "learning_rate": 7.8782e-06,
+      "loss": 0.9868,
+      "step": 10620
+    },
+    {
+      "epoch": 1.063,
+      "grad_norm": 7.661472320556641,
+      "learning_rate": 7.876200000000001e-06,
+      "loss": 0.6406,
+      "step": 10630
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 32.47237777709961,
+      "learning_rate": 7.8742e-06,
+      "loss": 0.6936,
+      "step": 10640
+    },
+    {
+      "epoch": 1.065,
+      "grad_norm": 54.157474517822266,
+      "learning_rate": 7.8722e-06,
+      "loss": 0.7992,
+      "step": 10650
+    },
+    {
+      "epoch": 1.066,
+      "grad_norm": 31.294652938842773,
+      "learning_rate": 7.870200000000001e-06,
+      "loss": 0.7245,
+      "step": 10660
+    },
+    {
+      "epoch": 1.067,
+      "grad_norm": 41.6014404296875,
+      "learning_rate": 7.8682e-06,
+      "loss": 0.6573,
+      "step": 10670
+    },
+    {
+      "epoch": 1.068,
+      "grad_norm": 19.699817657470703,
+      "learning_rate": 7.866200000000001e-06,
+      "loss": 0.6731,
+      "step": 10680
+    },
+    {
+      "epoch": 1.069,
+      "grad_norm": 41.20882034301758,
+      "learning_rate": 7.8642e-06,
+      "loss": 0.3784,
+      "step": 10690
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 25.886396408081055,
+      "learning_rate": 7.8622e-06,
+      "loss": 0.509,
+      "step": 10700
+    },
+    {
+      "epoch": 1.071,
+      "grad_norm": 40.70294189453125,
+      "learning_rate": 7.860200000000001e-06,
+      "loss": 0.6683,
+      "step": 10710
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 41.22813415527344,
+      "learning_rate": 7.8582e-06,
+      "loss": 0.7252,
+      "step": 10720
+    },
+    {
+      "epoch": 1.073,
+      "grad_norm": 21.7695255279541,
+      "learning_rate": 7.856200000000001e-06,
+      "loss": 0.6357,
+      "step": 10730
+    },
+    {
+      "epoch": 1.074,
+      "grad_norm": 41.75007247924805,
+      "learning_rate": 7.8542e-06,
+      "loss": 0.7398,
+      "step": 10740
+    },
+    {
+      "epoch": 1.075,
+      "grad_norm": 7.795687675476074,
+      "learning_rate": 7.8522e-06,
+      "loss": 0.8729,
+      "step": 10750
+    },
+    {
+      "epoch": 1.076,
+      "grad_norm": 17.92356300354004,
+      "learning_rate": 7.850200000000002e-06,
+      "loss": 0.6661,
+      "step": 10760
+    },
+    {
+      "epoch": 1.077,
+      "grad_norm": 15.715928077697754,
+      "learning_rate": 7.8482e-06,
+      "loss": 0.528,
+      "step": 10770
+    },
+    {
+      "epoch": 1.078,
+      "grad_norm": 27.70500946044922,
+      "learning_rate": 7.846200000000001e-06,
+      "loss": 0.8231,
+      "step": 10780
+    },
+    {
+      "epoch": 1.079,
+      "grad_norm": 16.61052894592285,
+      "learning_rate": 7.8442e-06,
+      "loss": 0.6257,
+      "step": 10790
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.07752513885498,
+      "learning_rate": 7.842200000000001e-06,
+      "loss": 0.7334,
+      "step": 10800
+    },
+    {
+      "epoch": 1.081,
+      "grad_norm": 34.947757720947266,
+      "learning_rate": 7.840200000000002e-06,
+      "loss": 0.6238,
+      "step": 10810
+    },
+    {
+      "epoch": 1.082,
+      "grad_norm": 19.702144622802734,
+      "learning_rate": 7.8382e-06,
+      "loss": 0.5864,
+      "step": 10820
+    },
+    {
+      "epoch": 1.083,
+      "grad_norm": 35.57761764526367,
+      "learning_rate": 7.836200000000001e-06,
+      "loss": 0.747,
+      "step": 10830
+    },
+    {
+      "epoch": 1.084,
+      "grad_norm": 26.503231048583984,
+      "learning_rate": 7.8342e-06,
+      "loss": 0.73,
+      "step": 10840
+    },
+    {
+      "epoch": 1.085,
+      "grad_norm": 37.302093505859375,
+      "learning_rate": 7.832200000000001e-06,
+      "loss": 0.6326,
+      "step": 10850
+    },
+    {
+      "epoch": 1.086,
+      "grad_norm": 24.52855110168457,
+      "learning_rate": 7.830200000000002e-06,
+      "loss": 0.4165,
+      "step": 10860
+    },
+    {
+      "epoch": 1.087,
+      "grad_norm": 17.49752426147461,
+      "learning_rate": 7.8282e-06,
+      "loss": 0.5934,
+      "step": 10870
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 50.52231979370117,
+      "learning_rate": 7.8262e-06,
+      "loss": 0.482,
+      "step": 10880
+    },
+    {
+      "epoch": 1.089,
+      "grad_norm": 23.141277313232422,
+      "learning_rate": 7.8242e-06,
+      "loss": 0.5202,
+      "step": 10890
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 28.306123733520508,
+      "learning_rate": 7.822200000000001e-06,
+      "loss": 0.7558,
+      "step": 10900
+    },
+    {
+      "epoch": 1.091,
+      "grad_norm": 25.076086044311523,
+      "learning_rate": 7.820200000000002e-06,
+      "loss": 0.8782,
+      "step": 10910
+    },
+    {
+      "epoch": 1.092,
+      "grad_norm": 39.36845016479492,
+      "learning_rate": 7.8182e-06,
+      "loss": 0.6682,
+      "step": 10920
+    },
+    {
+      "epoch": 1.093,
+      "grad_norm": 7.6542067527771,
+      "learning_rate": 7.8162e-06,
+      "loss": 0.8071,
+      "step": 10930
+    },
+    {
+      "epoch": 1.094,
+      "grad_norm": 28.47787094116211,
+      "learning_rate": 7.8142e-06,
+      "loss": 0.8684,
+      "step": 10940
+    },
+    {
+      "epoch": 1.095,
+      "grad_norm": 32.25487518310547,
+      "learning_rate": 7.812200000000001e-06,
+      "loss": 0.5147,
+      "step": 10950
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 29.876779556274414,
+      "learning_rate": 7.810200000000002e-06,
+      "loss": 0.7231,
+      "step": 10960
+    },
+    {
+      "epoch": 1.097,
+      "grad_norm": 15.228774070739746,
+      "learning_rate": 7.808200000000001e-06,
+      "loss": 0.7224,
+      "step": 10970
+    },
+    {
+      "epoch": 1.098,
+      "grad_norm": 18.995376586914062,
+      "learning_rate": 7.8062e-06,
+      "loss": 0.5264,
+      "step": 10980
+    },
+    {
+      "epoch": 1.099,
+      "grad_norm": 29.122920989990234,
+      "learning_rate": 7.8042e-06,
+      "loss": 0.5653,
+      "step": 10990
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 47.868804931640625,
+      "learning_rate": 7.802200000000001e-06,
+      "loss": 0.6261,
+      "step": 11000
+    },
+    {
+      "epoch": 1.101,
+      "grad_norm": 30.663339614868164,
+      "learning_rate": 7.8002e-06,
+      "loss": 0.48,
+      "step": 11010
+    },
+    {
+      "epoch": 1.102,
+      "grad_norm": 8.724465370178223,
+      "learning_rate": 7.798200000000001e-06,
+      "loss": 0.4851,
+      "step": 11020
+    },
+    {
+      "epoch": 1.103,
+      "grad_norm": 52.42881393432617,
+      "learning_rate": 7.7962e-06,
+      "loss": 0.6022,
+      "step": 11030
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 47.27696990966797,
+      "learning_rate": 7.7942e-06,
+      "loss": 0.8039,
+      "step": 11040
+    },
+    {
+      "epoch": 1.105,
+      "grad_norm": 51.4878044128418,
+      "learning_rate": 7.792200000000001e-06,
+      "loss": 0.7665,
+      "step": 11050
+    },
+    {
+      "epoch": 1.106,
+      "grad_norm": 35.13274383544922,
+      "learning_rate": 7.7902e-06,
+      "loss": 0.6382,
+      "step": 11060
+    },
+    {
+      "epoch": 1.107,
+      "grad_norm": 61.97047424316406,
+      "learning_rate": 7.7882e-06,
+      "loss": 0.4421,
+      "step": 11070
+    },
+    {
+      "epoch": 1.108,
+      "grad_norm": 54.35153579711914,
+      "learning_rate": 7.7862e-06,
+      "loss": 0.9091,
+      "step": 11080
+    },
+    {
+      "epoch": 1.109,
+      "grad_norm": 5.162842273712158,
+      "learning_rate": 7.7842e-06,
+      "loss": 0.42,
+      "step": 11090
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 29.818588256835938,
+      "learning_rate": 7.782200000000001e-06,
+      "loss": 0.8216,
+      "step": 11100
+    },
+    {
+      "epoch": 1.111,
+      "grad_norm": 46.03116989135742,
+      "learning_rate": 7.7802e-06,
+      "loss": 0.9207,
+      "step": 11110
+    },
+    {
+      "epoch": 1.112,
+      "grad_norm": 40.337379455566406,
+      "learning_rate": 7.7782e-06,
+      "loss": 0.7755,
+      "step": 11120
+    },
+    {
+      "epoch": 1.113,
+      "grad_norm": 27.618555068969727,
+      "learning_rate": 7.7762e-06,
+      "loss": 0.5692,
+      "step": 11130
+    },
+    {
+      "epoch": 1.114,
+      "grad_norm": 39.48080062866211,
+      "learning_rate": 7.774200000000001e-06,
+      "loss": 0.7518,
+      "step": 11140
+    },
+    {
+      "epoch": 1.115,
+      "grad_norm": 26.62726402282715,
+      "learning_rate": 7.772200000000002e-06,
+      "loss": 0.6853,
+      "step": 11150
+    },
+    {
+      "epoch": 1.116,
+      "grad_norm": 20.606592178344727,
+      "learning_rate": 7.7702e-06,
+      "loss": 0.8051,
+      "step": 11160
+    },
+    {
+      "epoch": 1.117,
+      "grad_norm": 38.8145751953125,
+      "learning_rate": 7.7682e-06,
+      "loss": 0.6414,
+      "step": 11170
+    },
+    {
+      "epoch": 1.1179999999999999,
+      "grad_norm": 36.362327575683594,
+      "learning_rate": 7.7662e-06,
+      "loss": 0.6396,
+      "step": 11180
+    },
+    {
+      "epoch": 1.119,
+      "grad_norm": 15.188089370727539,
+      "learning_rate": 7.764200000000001e-06,
+      "loss": 0.7237,
+      "step": 11190
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 43.664649963378906,
+      "learning_rate": 7.762200000000002e-06,
+      "loss": 0.806,
+      "step": 11200
+    },
+    {
+      "epoch": 1.121,
+      "grad_norm": 45.87483596801758,
+      "learning_rate": 7.7602e-06,
+      "loss": 0.6526,
+      "step": 11210
+    },
+    {
+      "epoch": 1.1219999999999999,
+      "grad_norm": 41.26497268676758,
+      "learning_rate": 7.7582e-06,
+      "loss": 0.4962,
+      "step": 11220
+    },
+    {
+      "epoch": 1.123,
+      "grad_norm": 23.14554786682129,
+      "learning_rate": 7.7562e-06,
+      "loss": 0.6414,
+      "step": 11230
+    },
+    {
+      "epoch": 1.124,
+      "grad_norm": 13.66366958618164,
+      "learning_rate": 7.754200000000001e-06,
+      "loss": 0.811,
+      "step": 11240
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 43.1004638671875,
+      "learning_rate": 7.7522e-06,
+      "loss": 0.6546,
+      "step": 11250
+    },
+    {
+      "epoch": 1.126,
+      "grad_norm": 33.18350601196289,
+      "learning_rate": 7.7502e-06,
+      "loss": 0.7879,
+      "step": 11260
+    },
+    {
+      "epoch": 1.127,
+      "grad_norm": 11.210447311401367,
+      "learning_rate": 7.7482e-06,
+      "loss": 0.6429,
+      "step": 11270
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 50.47068405151367,
+      "learning_rate": 7.7462e-06,
+      "loss": 0.4329,
+      "step": 11280
+    },
+    {
+      "epoch": 1.129,
+      "grad_norm": 38.2538948059082,
+      "learning_rate": 7.744200000000001e-06,
+      "loss": 0.7161,
+      "step": 11290
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 22.432058334350586,
+      "learning_rate": 7.7422e-06,
+      "loss": 0.535,
+      "step": 11300
+    },
+    {
+      "epoch": 1.131,
+      "grad_norm": 44.708797454833984,
+      "learning_rate": 7.740200000000001e-06,
+      "loss": 0.7098,
+      "step": 11310
+    },
+    {
+      "epoch": 1.1320000000000001,
+      "grad_norm": 27.494529724121094,
+      "learning_rate": 7.7382e-06,
+      "loss": 0.6044,
+      "step": 11320
+    },
+    {
+      "epoch": 1.133,
+      "grad_norm": 37.533042907714844,
+      "learning_rate": 7.7362e-06,
+      "loss": 0.944,
+      "step": 11330
+    },
+    {
+      "epoch": 1.134,
+      "grad_norm": 24.845661163330078,
+      "learning_rate": 7.734200000000001e-06,
+      "loss": 0.7786,
+      "step": 11340
+    },
+    {
+      "epoch": 1.135,
+      "grad_norm": 36.159950256347656,
+      "learning_rate": 7.7322e-06,
+      "loss": 0.5647,
+      "step": 11350
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 33.843299865722656,
+      "learning_rate": 7.730200000000001e-06,
+      "loss": 0.9112,
+      "step": 11360
+    },
+    {
+      "epoch": 1.137,
+      "grad_norm": 46.498435974121094,
+      "learning_rate": 7.7282e-06,
+      "loss": 1.0631,
+      "step": 11370
+    },
+    {
+      "epoch": 1.138,
+      "grad_norm": 24.42860221862793,
+      "learning_rate": 7.7262e-06,
+      "loss": 0.8096,
+      "step": 11380
+    },
+    {
+      "epoch": 1.139,
+      "grad_norm": 14.397269248962402,
+      "learning_rate": 7.724200000000001e-06,
+      "loss": 0.5918,
+      "step": 11390
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 14.705584526062012,
+      "learning_rate": 7.7222e-06,
+      "loss": 0.5131,
+      "step": 11400
+    },
+    {
+      "epoch": 1.141,
+      "grad_norm": 26.62394142150879,
+      "learning_rate": 7.720200000000001e-06,
+      "loss": 0.6488,
+      "step": 11410
+    },
+    {
+      "epoch": 1.142,
+      "grad_norm": 32.86175537109375,
+      "learning_rate": 7.7182e-06,
+      "loss": 0.5914,
+      "step": 11420
+    },
+    {
+      "epoch": 1.143,
+      "grad_norm": 20.663698196411133,
+      "learning_rate": 7.7162e-06,
+      "loss": 0.6432,
+      "step": 11430
+    },
+    {
+      "epoch": 1.144,
+      "grad_norm": 30.446578979492188,
+      "learning_rate": 7.7142e-06,
+      "loss": 0.7907,
+      "step": 11440
+    },
+    {
+      "epoch": 1.145,
+      "grad_norm": 34.20303726196289,
+      "learning_rate": 7.7122e-06,
+      "loss": 0.7505,
+      "step": 11450
+    },
+    {
+      "epoch": 1.146,
+      "grad_norm": 26.968290328979492,
+      "learning_rate": 7.710200000000001e-06,
+      "loss": 0.8869,
+      "step": 11460
+    },
+    {
+      "epoch": 1.147,
+      "grad_norm": 50.744842529296875,
+      "learning_rate": 7.7082e-06,
+      "loss": 0.87,
+      "step": 11470
+    },
+    {
+      "epoch": 1.148,
+      "grad_norm": 31.850502014160156,
+      "learning_rate": 7.706200000000001e-06,
+      "loss": 0.5349,
+      "step": 11480
+    },
+    {
+      "epoch": 1.149,
+      "grad_norm": 32.834651947021484,
+      "learning_rate": 7.7042e-06,
+      "loss": 0.5677,
+      "step": 11490
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 21.380491256713867,
+      "learning_rate": 7.7022e-06,
+      "loss": 0.6704,
+      "step": 11500
+    },
+    {
+      "epoch": 1.151,
+      "grad_norm": 11.736533164978027,
+      "learning_rate": 7.700200000000001e-06,
+      "loss": 0.7262,
+      "step": 11510
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 52.395469665527344,
+      "learning_rate": 7.6982e-06,
+      "loss": 0.6247,
+      "step": 11520
+    },
+    {
+      "epoch": 1.153,
+      "grad_norm": 39.82683181762695,
+      "learning_rate": 7.696200000000001e-06,
+      "loss": 0.7259,
+      "step": 11530
+    },
+    {
+      "epoch": 1.154,
+      "grad_norm": 23.57517433166504,
+      "learning_rate": 7.6942e-06,
+      "loss": 0.7778,
+      "step": 11540
+    },
+    {
+      "epoch": 1.155,
+      "grad_norm": 8.457917213439941,
+      "learning_rate": 7.6922e-06,
+      "loss": 0.835,
+      "step": 11550
+    },
+    {
+      "epoch": 1.156,
+      "grad_norm": 16.930097579956055,
+      "learning_rate": 7.690200000000001e-06,
+      "loss": 0.6884,
+      "step": 11560
+    },
+    {
+      "epoch": 1.157,
+      "grad_norm": 37.95302963256836,
+      "learning_rate": 7.6882e-06,
+      "loss": 0.8589,
+      "step": 11570
+    },
+    {
+      "epoch": 1.158,
+      "grad_norm": 26.018352508544922,
+      "learning_rate": 7.686200000000001e-06,
+      "loss": 0.6359,
+      "step": 11580
+    },
+    {
+      "epoch": 1.159,
+      "grad_norm": 16.526905059814453,
+      "learning_rate": 7.6842e-06,
+      "loss": 0.5665,
+      "step": 11590
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 39.45627212524414,
+      "learning_rate": 7.6822e-06,
+      "loss": 0.845,
+      "step": 11600
+    },
+    {
+      "epoch": 1.161,
+      "grad_norm": 27.748611450195312,
+      "learning_rate": 7.680200000000001e-06,
+      "loss": 0.5258,
+      "step": 11610
+    },
+    {
+      "epoch": 1.162,
+      "grad_norm": 14.75036907196045,
+      "learning_rate": 7.6782e-06,
+      "loss": 0.7445,
+      "step": 11620
+    },
+    {
+      "epoch": 1.163,
+      "grad_norm": 32.02853775024414,
+      "learning_rate": 7.676200000000001e-06,
+      "loss": 0.5285,
+      "step": 11630
+    },
+    {
+      "epoch": 1.164,
+      "grad_norm": 23.34952163696289,
+      "learning_rate": 7.6742e-06,
+      "loss": 0.7296,
+      "step": 11640
+    },
+    {
+      "epoch": 1.165,
+      "grad_norm": 43.00461959838867,
+      "learning_rate": 7.6722e-06,
+      "loss": 0.7876,
+      "step": 11650
+    },
+    {
+      "epoch": 1.166,
+      "grad_norm": 33.411170959472656,
+      "learning_rate": 7.670200000000002e-06,
+      "loss": 0.5697,
+      "step": 11660
+    },
+    {
+      "epoch": 1.167,
+      "grad_norm": 32.586212158203125,
+      "learning_rate": 7.6682e-06,
+      "loss": 0.5464,
+      "step": 11670
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 9.358975410461426,
+      "learning_rate": 7.6662e-06,
+      "loss": 0.515,
+      "step": 11680
+    },
+    {
+      "epoch": 1.169,
+      "grad_norm": 25.917869567871094,
+      "learning_rate": 7.6642e-06,
+      "loss": 0.8162,
+      "step": 11690
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.3763427734375,
+      "learning_rate": 7.662200000000001e-06,
+      "loss": 0.4068,
+      "step": 11700
+    },
+    {
+      "epoch": 1.171,
+      "grad_norm": 32.39506912231445,
+      "learning_rate": 7.660200000000002e-06,
+      "loss": 0.6225,
+      "step": 11710
+    },
+    {
+      "epoch": 1.172,
+      "grad_norm": 24.119741439819336,
+      "learning_rate": 7.6582e-06,
+      "loss": 0.8509,
+      "step": 11720
+    },
+    {
+      "epoch": 1.173,
+      "grad_norm": 46.87370300292969,
+      "learning_rate": 7.6562e-06,
+      "loss": 0.828,
+      "step": 11730
+    },
+    {
+      "epoch": 1.174,
+      "grad_norm": 28.37071990966797,
+      "learning_rate": 7.6542e-06,
+      "loss": 0.9559,
+      "step": 11740
+    },
+    {
+      "epoch": 1.175,
+      "grad_norm": 38.91966247558594,
+      "learning_rate": 7.652200000000001e-06,
+      "loss": 0.7304,
+      "step": 11750
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 42.96327590942383,
+      "learning_rate": 7.650200000000002e-06,
+      "loss": 0.7791,
+      "step": 11760
+    },
+    {
+      "epoch": 1.177,
+      "grad_norm": 23.778846740722656,
+      "learning_rate": 7.6482e-06,
+      "loss": 0.8178,
+      "step": 11770
+    },
+    {
+      "epoch": 1.178,
+      "grad_norm": 27.186412811279297,
+      "learning_rate": 7.6462e-06,
+      "loss": 0.7396,
+      "step": 11780
+    },
+    {
+      "epoch": 1.179,
+      "grad_norm": 32.69036865234375,
+      "learning_rate": 7.6442e-06,
+      "loss": 0.8204,
+      "step": 11790
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 18.225202560424805,
+      "learning_rate": 7.642200000000001e-06,
+      "loss": 0.6291,
+      "step": 11800
+    },
+    {
+      "epoch": 1.181,
+      "grad_norm": 16.933101654052734,
+      "learning_rate": 7.6402e-06,
+      "loss": 0.6593,
+      "step": 11810
+    },
+    {
+      "epoch": 1.182,
+      "grad_norm": 32.985843658447266,
+      "learning_rate": 7.6382e-06,
+      "loss": 0.836,
+      "step": 11820
+    },
+    {
+      "epoch": 1.183,
+      "grad_norm": 26.52162742614746,
+      "learning_rate": 7.6362e-06,
+      "loss": 0.7246,
+      "step": 11830
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 49.35344314575195,
+      "learning_rate": 7.6342e-06,
+      "loss": 0.6721,
+      "step": 11840
+    },
+    {
+      "epoch": 1.185,
+      "grad_norm": 12.153249740600586,
+      "learning_rate": 7.632200000000001e-06,
+      "loss": 0.8688,
+      "step": 11850
+    },
+    {
+      "epoch": 1.186,
+      "grad_norm": 33.72947311401367,
+      "learning_rate": 7.6302e-06,
+      "loss": 0.6242,
+      "step": 11860
+    },
+    {
+      "epoch": 1.187,
+      "grad_norm": 52.174407958984375,
+      "learning_rate": 7.628200000000001e-06,
+      "loss": 0.7162,
+      "step": 11870
+    },
+    {
+      "epoch": 1.188,
+      "grad_norm": 30.34244728088379,
+      "learning_rate": 7.6262e-06,
+      "loss": 0.7312,
+      "step": 11880
+    },
+    {
+      "epoch": 1.189,
+      "grad_norm": 34.500083923339844,
+      "learning_rate": 7.624200000000001e-06,
+      "loss": 0.7341,
+      "step": 11890
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 29.85059356689453,
+      "learning_rate": 7.6222000000000005e-06,
+      "loss": 0.5486,
+      "step": 11900
+    },
+    {
+      "epoch": 1.191,
+      "grad_norm": 17.79989242553711,
+      "learning_rate": 7.620200000000001e-06,
+      "loss": 0.7282,
+      "step": 11910
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 10.709245681762695,
+      "learning_rate": 7.618200000000001e-06,
+      "loss": 0.5993,
+      "step": 11920
+    },
+    {
+      "epoch": 1.193,
+      "grad_norm": 25.175249099731445,
+      "learning_rate": 7.6162e-06,
+      "loss": 0.7604,
+      "step": 11930
+    },
+    {
+      "epoch": 1.194,
+      "grad_norm": 26.545978546142578,
+      "learning_rate": 7.614200000000001e-06,
+      "loss": 0.6331,
+      "step": 11940
+    },
+    {
+      "epoch": 1.195,
+      "grad_norm": 39.96792984008789,
+      "learning_rate": 7.6122000000000006e-06,
+      "loss": 0.7824,
+      "step": 11950
+    },
+    {
+      "epoch": 1.196,
+      "grad_norm": 38.01614761352539,
+      "learning_rate": 7.610200000000001e-06,
+      "loss": 0.8789,
+      "step": 11960
+    },
+    {
+      "epoch": 1.197,
+      "grad_norm": 5.848873615264893,
+      "learning_rate": 7.608200000000001e-06,
+      "loss": 0.5995,
+      "step": 11970
+    },
+    {
+      "epoch": 1.198,
+      "grad_norm": 78.58563232421875,
+      "learning_rate": 7.6062e-06,
+      "loss": 0.6577,
+      "step": 11980
+    },
+    {
+      "epoch": 1.199,
+      "grad_norm": 18.236574172973633,
+      "learning_rate": 7.604200000000001e-06,
+      "loss": 0.5221,
+      "step": 11990
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 19.23676872253418,
+      "learning_rate": 7.602200000000001e-06,
+      "loss": 0.4422,
+      "step": 12000
+    },
+    {
+      "epoch": 1.201,
+      "grad_norm": 29.161029815673828,
+      "learning_rate": 7.6002000000000005e-06,
+      "loss": 0.6526,
+      "step": 12010
+    },
+    {
+      "epoch": 1.202,
+      "grad_norm": 25.65962791442871,
+      "learning_rate": 7.598200000000001e-06,
+      "loss": 0.7309,
+      "step": 12020
+    },
+    {
+      "epoch": 1.203,
+      "grad_norm": 32.263160705566406,
+      "learning_rate": 7.5962e-06,
+      "loss": 0.7232,
+      "step": 12030
+    },
+    {
+      "epoch": 1.204,
+      "grad_norm": 103.59854125976562,
+      "learning_rate": 7.5942e-06,
+      "loss": 0.7941,
+      "step": 12040
+    },
+    {
+      "epoch": 1.205,
+      "grad_norm": 28.62127685546875,
+      "learning_rate": 7.592200000000001e-06,
+      "loss": 0.5728,
+      "step": 12050
+    },
+    {
+      "epoch": 1.206,
+      "grad_norm": 28.61754035949707,
+      "learning_rate": 7.590200000000001e-06,
+      "loss": 0.5754,
+      "step": 12060
+    },
+    {
+      "epoch": 1.207,
+      "grad_norm": 28.66469955444336,
+      "learning_rate": 7.588200000000001e-06,
+      "loss": 0.6362,
+      "step": 12070
+    },
+    {
+      "epoch": 1.208,
+      "grad_norm": 21.962121963500977,
+      "learning_rate": 7.5862e-06,
+      "loss": 0.5206,
+      "step": 12080
+    },
+    {
+      "epoch": 1.209,
+      "grad_norm": 10.966731071472168,
+      "learning_rate": 7.5842e-06,
+      "loss": 0.8621,
+      "step": 12090
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.3472261428833008,
+      "learning_rate": 7.582200000000001e-06,
+      "loss": 0.6844,
+      "step": 12100
+    },
+    {
+      "epoch": 1.211,
+      "grad_norm": 51.0182991027832,
+      "learning_rate": 7.580200000000001e-06,
+      "loss": 0.6243,
+      "step": 12110
+    },
+    {
+      "epoch": 1.212,
+      "grad_norm": 13.555909156799316,
+      "learning_rate": 7.578200000000001e-06,
+      "loss": 0.5435,
+      "step": 12120
+    },
+    {
+      "epoch": 1.213,
+      "grad_norm": 43.063419342041016,
+      "learning_rate": 7.5762e-06,
+      "loss": 0.72,
+      "step": 12130
+    },
+    {
+      "epoch": 1.214,
+      "grad_norm": 24.662275314331055,
+      "learning_rate": 7.5742e-06,
+      "loss": 0.4902,
+      "step": 12140
+    },
+    {
+      "epoch": 1.215,
+      "grad_norm": 41.610626220703125,
+      "learning_rate": 7.572200000000001e-06,
+      "loss": 0.6616,
+      "step": 12150
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 44.63747787475586,
+      "learning_rate": 7.570200000000001e-06,
+      "loss": 0.8976,
+      "step": 12160
+    },
+    {
+      "epoch": 1.217,
+      "grad_norm": 4.047074794769287,
+      "learning_rate": 7.5682000000000015e-06,
+      "loss": 0.9258,
+      "step": 12170
+    },
+    {
+      "epoch": 1.218,
+      "grad_norm": 52.99892044067383,
+      "learning_rate": 7.5662000000000005e-06,
+      "loss": 0.7729,
+      "step": 12180
+    },
+    {
+      "epoch": 1.219,
+      "grad_norm": 42.149688720703125,
+      "learning_rate": 7.5642e-06,
+      "loss": 0.7279,
+      "step": 12190
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.185766220092773,
+      "learning_rate": 7.562200000000001e-06,
+      "loss": 0.7866,
+      "step": 12200
+    },
+    {
+      "epoch": 1.221,
+      "grad_norm": 10.956640243530273,
+      "learning_rate": 7.560200000000001e-06,
+      "loss": 0.5585,
+      "step": 12210
+    },
+    {
+      "epoch": 1.222,
+      "grad_norm": 19.674837112426758,
+      "learning_rate": 7.558200000000001e-06,
+      "loss": 0.6338,
+      "step": 12220
+    },
+    {
+      "epoch": 1.223,
+      "grad_norm": 10.953132629394531,
+      "learning_rate": 7.556200000000001e-06,
+      "loss": 0.5947,
+      "step": 12230
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 33.46485900878906,
+      "learning_rate": 7.5542000000000005e-06,
+      "loss": 0.7458,
+      "step": 12240
+    },
+    {
+      "epoch": 1.225,
+      "grad_norm": 9.161663055419922,
+      "learning_rate": 7.5522e-06,
+      "loss": 0.5293,
+      "step": 12250
+    },
+    {
+      "epoch": 1.226,
+      "grad_norm": 36.02226257324219,
+      "learning_rate": 7.550200000000001e-06,
+      "loss": 0.6799,
+      "step": 12260
+    },
+    {
+      "epoch": 1.227,
+      "grad_norm": 31.329750061035156,
+      "learning_rate": 7.548200000000001e-06,
+      "loss": 0.57,
+      "step": 12270
+    },
+    {
+      "epoch": 1.228,
+      "grad_norm": 17.420188903808594,
+      "learning_rate": 7.5462e-06,
+      "loss": 0.7771,
+      "step": 12280
+    },
+    {
+      "epoch": 1.229,
+      "grad_norm": 20.56215476989746,
+      "learning_rate": 7.5442000000000005e-06,
+      "loss": 0.743,
+      "step": 12290
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 19.782087326049805,
+      "learning_rate": 7.5422e-06,
+      "loss": 0.7162,
+      "step": 12300
+    },
+    {
+      "epoch": 1.231,
+      "grad_norm": 12.472650527954102,
+      "learning_rate": 7.540200000000001e-06,
+      "loss": 0.583,
+      "step": 12310
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 32.87686538696289,
+      "learning_rate": 7.538200000000001e-06,
+      "loss": 0.6691,
+      "step": 12320
+    },
+    {
+      "epoch": 1.233,
+      "grad_norm": 4.495699405670166,
+      "learning_rate": 7.5362e-06,
+      "loss": 0.4486,
+      "step": 12330
+    },
+    {
+      "epoch": 1.234,
+      "grad_norm": 21.480024337768555,
+      "learning_rate": 7.534200000000001e-06,
+      "loss": 0.7801,
+      "step": 12340
+    },
+    {
+      "epoch": 1.2349999999999999,
+      "grad_norm": 14.798094749450684,
+      "learning_rate": 7.5322000000000005e-06,
+      "loss": 0.5866,
+      "step": 12350
+    },
+    {
+      "epoch": 1.236,
+      "grad_norm": 32.35955810546875,
+      "learning_rate": 7.530200000000001e-06,
+      "loss": 0.743,
+      "step": 12360
+    },
+    {
+      "epoch": 1.237,
+      "grad_norm": 44.85438537597656,
+      "learning_rate": 7.528200000000001e-06,
+      "loss": 0.7529,
+      "step": 12370
+    },
+    {
+      "epoch": 1.238,
+      "grad_norm": 22.723072052001953,
+      "learning_rate": 7.5262e-06,
+      "loss": 0.8058,
+      "step": 12380
+    },
+    {
+      "epoch": 1.2389999999999999,
+      "grad_norm": 34.42487716674805,
+      "learning_rate": 7.524200000000001e-06,
+      "loss": 0.7757,
+      "step": 12390
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 45.39601135253906,
+      "learning_rate": 7.522200000000001e-06,
+      "loss": 0.7571,
+      "step": 12400
+    },
+    {
+      "epoch": 1.241,
+      "grad_norm": 34.43247604370117,
+      "learning_rate": 7.5202000000000004e-06,
+      "loss": 0.7506,
+      "step": 12410
+    },
+    {
+      "epoch": 1.242,
+      "grad_norm": 41.917396545410156,
+      "learning_rate": 7.518200000000001e-06,
+      "loss": 0.8379,
+      "step": 12420
+    },
+    {
+      "epoch": 1.2429999999999999,
+      "grad_norm": 38.676753997802734,
+      "learning_rate": 7.5162e-06,
+      "loss": 0.6781,
+      "step": 12430
+    },
+    {
+      "epoch": 1.244,
+      "grad_norm": 12.698298454284668,
+      "learning_rate": 7.5142e-06,
+      "loss": 0.6937,
+      "step": 12440
+    },
+    {
+      "epoch": 1.245,
+      "grad_norm": 11.609149932861328,
+      "learning_rate": 7.512200000000001e-06,
+      "loss": 0.7535,
+      "step": 12450
+    },
+    {
+      "epoch": 1.246,
+      "grad_norm": 16.510948181152344,
+      "learning_rate": 7.5102000000000005e-06,
+      "loss": 0.7634,
+      "step": 12460
+    },
+    {
+      "epoch": 1.2469999999999999,
+      "grad_norm": 25.841136932373047,
+      "learning_rate": 7.508200000000001e-06,
+      "loss": 0.5595,
+      "step": 12470
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 44.224754333496094,
+      "learning_rate": 7.5062e-06,
+      "loss": 0.9578,
+      "step": 12480
+    },
+    {
+      "epoch": 1.249,
+      "grad_norm": 32.64692687988281,
+      "learning_rate": 7.5042e-06,
+      "loss": 0.5536,
+      "step": 12490
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 19.11054039001465,
+      "learning_rate": 7.502200000000001e-06,
+      "loss": 0.6679,
+      "step": 12500
+    },
+    {
+      "epoch": 1.251,
+      "grad_norm": 20.72027015686035,
+      "learning_rate": 7.500200000000001e-06,
+      "loss": 0.8676,
+      "step": 12510
+    },
+    {
+      "epoch": 1.252,
+      "grad_norm": 10.526891708374023,
+      "learning_rate": 7.498200000000001e-06,
+      "loss": 0.6176,
+      "step": 12520
+    },
+    {
+      "epoch": 1.2530000000000001,
+      "grad_norm": 20.656282424926758,
+      "learning_rate": 7.4962e-06,
+      "loss": 0.7816,
+      "step": 12530
+    },
+    {
+      "epoch": 1.254,
+      "grad_norm": 57.771263122558594,
+      "learning_rate": 7.4942e-06,
+      "loss": 0.5708,
+      "step": 12540
+    },
+    {
+      "epoch": 1.255,
+      "grad_norm": 10.302445411682129,
+      "learning_rate": 7.492200000000001e-06,
+      "loss": 0.7098,
+      "step": 12550
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 70.08512115478516,
+      "learning_rate": 7.490200000000001e-06,
+      "loss": 0.8798,
+      "step": 12560
+    },
+    {
+      "epoch": 1.2570000000000001,
+      "grad_norm": 21.335031509399414,
+      "learning_rate": 7.488200000000001e-06,
+      "loss": 0.6085,
+      "step": 12570
+    },
+    {
+      "epoch": 1.258,
+      "grad_norm": 25.092174530029297,
+      "learning_rate": 7.4862000000000004e-06,
+      "loss": 0.6516,
+      "step": 12580
+    },
+    {
+      "epoch": 1.259,
+      "grad_norm": 11.581510543823242,
+      "learning_rate": 7.4842e-06,
+      "loss": 0.5533,
+      "step": 12590
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.392578125,
+      "learning_rate": 7.482200000000001e-06,
+      "loss": 0.6309,
+      "step": 12600
+    },
+    {
+      "epoch": 1.2610000000000001,
+      "grad_norm": 46.0922737121582,
+      "learning_rate": 7.480200000000001e-06,
+      "loss": 0.7325,
+      "step": 12610
+    },
+    {
+      "epoch": 1.262,
+      "grad_norm": 18.586734771728516,
+      "learning_rate": 7.478200000000001e-06,
+      "loss": 0.7337,
+      "step": 12620
+    },
+    {
+      "epoch": 1.263,
+      "grad_norm": 5.978358745574951,
+      "learning_rate": 7.4762000000000005e-06,
+      "loss": 0.5667,
+      "step": 12630
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 37.30823516845703,
+      "learning_rate": 7.4742e-06,
+      "loss": 0.7755,
+      "step": 12640
+    },
+    {
+      "epoch": 1.2650000000000001,
+      "grad_norm": 24.3037052154541,
+      "learning_rate": 7.4722e-06,
+      "loss": 0.7101,
+      "step": 12650
+    },
+    {
+      "epoch": 1.266,
+      "grad_norm": 27.828401565551758,
+      "learning_rate": 7.470200000000001e-06,
+      "loss": 0.5885,
+      "step": 12660
+    },
+    {
+      "epoch": 1.267,
+      "grad_norm": 23.27979278564453,
+      "learning_rate": 7.468200000000001e-06,
+      "loss": 0.6291,
+      "step": 12670
+    },
+    {
+      "epoch": 1.268,
+      "grad_norm": 22.77558708190918,
+      "learning_rate": 7.4662e-06,
+      "loss": 0.7666,
+      "step": 12680
+    },
+    {
+      "epoch": 1.2690000000000001,
+      "grad_norm": 23.251815795898438,
+      "learning_rate": 7.4642000000000005e-06,
+      "loss": 0.5757,
+      "step": 12690
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 25.934316635131836,
+      "learning_rate": 7.4622e-06,
+      "loss": 0.632,
+      "step": 12700
+    },
+    {
+      "epoch": 1.271,
+      "grad_norm": 25.416095733642578,
+      "learning_rate": 7.460200000000001e-06,
+      "loss": 0.7164,
+      "step": 12710
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 25.40232276916504,
+      "learning_rate": 7.458200000000001e-06,
+      "loss": 0.6782,
+      "step": 12720
+    },
+    {
+      "epoch": 1.2730000000000001,
+      "grad_norm": 48.11226272583008,
+      "learning_rate": 7.4562e-06,
+      "loss": 0.6615,
+      "step": 12730
+    },
+    {
+      "epoch": 1.274,
+      "grad_norm": 8.603801727294922,
+      "learning_rate": 7.4542000000000006e-06,
+      "loss": 0.7099,
+      "step": 12740
+    },
+    {
+      "epoch": 1.275,
+      "grad_norm": 18.267532348632812,
+      "learning_rate": 7.4522e-06,
+      "loss": 0.7334,
+      "step": 12750
+    },
+    {
+      "epoch": 1.276,
+      "grad_norm": 45.3951530456543,
+      "learning_rate": 7.450200000000001e-06,
+      "loss": 0.6886,
+      "step": 12760
+    },
+    {
+      "epoch": 1.2770000000000001,
+      "grad_norm": 235.92100524902344,
+      "learning_rate": 7.448200000000001e-06,
+      "loss": 0.5436,
+      "step": 12770
+    },
+    {
+      "epoch": 1.278,
+      "grad_norm": 15.398147583007812,
+      "learning_rate": 7.4462e-06,
+      "loss": 0.7053,
+      "step": 12780
+    },
+    {
+      "epoch": 1.279,
+      "grad_norm": 48.43658447265625,
+      "learning_rate": 7.444200000000001e-06,
+      "loss": 0.5759,
+      "step": 12790
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 39.65992736816406,
+      "learning_rate": 7.4422000000000005e-06,
+      "loss": 0.699,
+      "step": 12800
+    },
+    {
+      "epoch": 1.2810000000000001,
+      "grad_norm": 20.03291893005371,
+      "learning_rate": 7.4402e-06,
+      "loss": 0.6684,
+      "step": 12810
+    },
+    {
+      "epoch": 1.282,
+      "grad_norm": 15.064152717590332,
+      "learning_rate": 7.438200000000001e-06,
+      "loss": 0.7231,
+      "step": 12820
+    },
+    {
+      "epoch": 1.283,
+      "grad_norm": 35.2615966796875,
+      "learning_rate": 7.4362e-06,
+      "loss": 0.7014,
+      "step": 12830
+    },
+    {
+      "epoch": 1.284,
+      "grad_norm": 19.830982208251953,
+      "learning_rate": 7.4342e-06,
+      "loss": 0.6552,
+      "step": 12840
+    },
+    {
+      "epoch": 1.285,
+      "grad_norm": 9.698526382446289,
+      "learning_rate": 7.432200000000001e-06,
+      "loss": 0.7807,
+      "step": 12850
+    },
+    {
+      "epoch": 1.286,
+      "grad_norm": 25.383045196533203,
+      "learning_rate": 7.4302000000000005e-06,
+      "loss": 0.6524,
+      "step": 12860
+    },
+    {
+      "epoch": 1.287,
+      "grad_norm": 26.713987350463867,
+      "learning_rate": 7.428200000000001e-06,
+      "loss": 0.6857,
+      "step": 12870
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 8.724356651306152,
+      "learning_rate": 7.4262e-06,
+      "loss": 0.4876,
+      "step": 12880
+    },
+    {
+      "epoch": 1.289,
+      "grad_norm": 22.239267349243164,
+      "learning_rate": 7.4242e-06,
+      "loss": 0.6277,
+      "step": 12890
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 28.160232543945312,
+      "learning_rate": 7.422200000000001e-06,
+      "loss": 0.6837,
+      "step": 12900
+    },
+    {
+      "epoch": 1.291,
+      "grad_norm": 30.82021713256836,
+      "learning_rate": 7.4202000000000005e-06,
+      "loss": 0.5708,
+      "step": 12910
+    },
+    {
+      "epoch": 1.292,
+      "grad_norm": 24.322532653808594,
+      "learning_rate": 7.418200000000001e-06,
+      "loss": 0.5548,
+      "step": 12920
+    },
+    {
+      "epoch": 1.293,
+      "grad_norm": 29.292064666748047,
+      "learning_rate": 7.4162e-06,
+      "loss": 0.8325,
+      "step": 12930
+    },
+    {
+      "epoch": 1.294,
+      "grad_norm": 33.295448303222656,
+      "learning_rate": 7.4142e-06,
+      "loss": 0.7986,
+      "step": 12940
+    },
+    {
+      "epoch": 1.295,
+      "grad_norm": 18.731597900390625,
+      "learning_rate": 7.412200000000001e-06,
+      "loss": 1.0747,
+      "step": 12950
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 22.99925994873047,
+      "learning_rate": 7.410200000000001e-06,
+      "loss": 0.6548,
+      "step": 12960
+    },
+    {
+      "epoch": 1.297,
+      "grad_norm": 25.831560134887695,
+      "learning_rate": 7.408200000000001e-06,
+      "loss": 0.7995,
+      "step": 12970
+    },
+    {
+      "epoch": 1.298,
+      "grad_norm": 21.3696231842041,
+      "learning_rate": 7.4062e-06,
+      "loss": 0.8104,
+      "step": 12980
+    },
+    {
+      "epoch": 1.299,
+      "grad_norm": 29.745773315429688,
+      "learning_rate": 7.4042e-06,
+      "loss": 0.6324,
+      "step": 12990
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 25.20359992980957,
+      "learning_rate": 7.402200000000001e-06,
+      "loss": 0.6775,
+      "step": 13000
+    },
+    {
+      "epoch": 1.301,
+      "grad_norm": 22.265050888061523,
+      "learning_rate": 7.400200000000001e-06,
+      "loss": 0.5683,
+      "step": 13010
+    },
+    {
+      "epoch": 1.302,
+      "grad_norm": 26.836889266967773,
+      "learning_rate": 7.398200000000001e-06,
+      "loss": 0.4955,
+      "step": 13020
+    },
+    {
+      "epoch": 1.303,
+      "grad_norm": 20.822649002075195,
+      "learning_rate": 7.396200000000001e-06,
+      "loss": 0.594,
+      "step": 13030
+    },
+    {
+      "epoch": 1.304,
+      "grad_norm": 5.273739814758301,
+      "learning_rate": 7.3942e-06,
+      "loss": 0.4596,
+      "step": 13040
+    },
+    {
+      "epoch": 1.305,
+      "grad_norm": 41.83501052856445,
+      "learning_rate": 7.3922e-06,
+      "loss": 0.5578,
+      "step": 13050
+    },
+    {
+      "epoch": 1.306,
+      "grad_norm": 20.89312171936035,
+      "learning_rate": 7.390200000000001e-06,
+      "loss": 0.5613,
+      "step": 13060
+    },
+    {
+      "epoch": 1.307,
+      "grad_norm": 39.89366149902344,
+      "learning_rate": 7.388200000000001e-06,
+      "loss": 0.9497,
+      "step": 13070
+    },
+    {
+      "epoch": 1.308,
+      "grad_norm": 37.19102478027344,
+      "learning_rate": 7.386200000000001e-06,
+      "loss": 0.5325,
+      "step": 13080
+    },
+    {
+      "epoch": 1.309,
+      "grad_norm": 3.879763603210449,
+      "learning_rate": 7.3842e-06,
+      "loss": 0.52,
+      "step": 13090
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.704940795898438,
+      "learning_rate": 7.3822e-06,
+      "loss": 0.4688,
+      "step": 13100
+    },
+    {
+      "epoch": 1.311,
+      "grad_norm": 23.836936950683594,
+      "learning_rate": 7.380200000000001e-06,
+      "loss": 0.6162,
+      "step": 13110
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 40.60661697387695,
+      "learning_rate": 7.378200000000001e-06,
+      "loss": 0.6149,
+      "step": 13120
+    },
+    {
+      "epoch": 1.313,
+      "grad_norm": 23.078330993652344,
+      "learning_rate": 7.3762000000000015e-06,
+      "loss": 0.6401,
+      "step": 13130
+    },
+    {
+      "epoch": 1.314,
+      "grad_norm": 26.6516170501709,
+      "learning_rate": 7.3742000000000005e-06,
+      "loss": 0.7671,
+      "step": 13140
+    },
+    {
+      "epoch": 1.315,
+      "grad_norm": 35.08804702758789,
+      "learning_rate": 7.3722e-06,
+      "loss": 0.5731,
+      "step": 13150
+    },
+    {
+      "epoch": 1.316,
+      "grad_norm": 32.42262268066406,
+      "learning_rate": 7.370200000000001e-06,
+      "loss": 0.9451,
+      "step": 13160
+    },
+    {
+      "epoch": 1.317,
+      "grad_norm": 11.495752334594727,
+      "learning_rate": 7.368200000000001e-06,
+      "loss": 0.8786,
+      "step": 13170
+    },
+    {
+      "epoch": 1.318,
+      "grad_norm": 22.15064811706543,
+      "learning_rate": 7.366200000000001e-06,
+      "loss": 0.5448,
+      "step": 13180
+    },
+    {
+      "epoch": 1.319,
+      "grad_norm": 28.169740676879883,
+      "learning_rate": 7.3642000000000006e-06,
+      "loss": 0.8506,
+      "step": 13190
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.906099319458008,
+      "learning_rate": 7.3622000000000004e-06,
+      "loss": 0.8319,
+      "step": 13200
+    },
+    {
+      "epoch": 1.321,
+      "grad_norm": 30.913488388061523,
+      "learning_rate": 7.3602e-06,
+      "loss": 0.7243,
+      "step": 13210
+    },
+    {
+      "epoch": 1.322,
+      "grad_norm": 41.180477142333984,
+      "learning_rate": 7.358200000000001e-06,
+      "loss": 0.8648,
+      "step": 13220
+    },
+    {
+      "epoch": 1.323,
+      "grad_norm": 29.278974533081055,
+      "learning_rate": 7.356200000000001e-06,
+      "loss": 0.6343,
+      "step": 13230
+    },
+    {
+      "epoch": 1.324,
+      "grad_norm": 8.021432876586914,
+      "learning_rate": 7.3542e-06,
+      "loss": 0.5605,
+      "step": 13240
+    },
+    {
+      "epoch": 1.325,
+      "grad_norm": 32.5282096862793,
+      "learning_rate": 7.3522000000000005e-06,
+      "loss": 0.8554,
+      "step": 13250
+    },
+    {
+      "epoch": 1.326,
+      "grad_norm": 26.73030662536621,
+      "learning_rate": 7.3502e-06,
+      "loss": 0.6036,
+      "step": 13260
+    },
+    {
+      "epoch": 1.327,
+      "grad_norm": 35.48610305786133,
+      "learning_rate": 7.348200000000001e-06,
+      "loss": 0.786,
+      "step": 13270
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 19.208602905273438,
+      "learning_rate": 7.346200000000001e-06,
+      "loss": 0.6422,
+      "step": 13280
+    },
+    {
+      "epoch": 1.329,
+      "grad_norm": 13.82779598236084,
+      "learning_rate": 7.3442e-06,
+      "loss": 0.6407,
+      "step": 13290
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 33.58254623413086,
+      "learning_rate": 7.342200000000001e-06,
+      "loss": 0.7012,
+      "step": 13300
+    },
+    {
+      "epoch": 1.331,
+      "grad_norm": 16.39531707763672,
+      "learning_rate": 7.3402000000000005e-06,
+      "loss": 0.6333,
+      "step": 13310
+    },
+    {
+      "epoch": 1.332,
+      "grad_norm": 35.258243560791016,
+      "learning_rate": 7.338200000000001e-06,
+      "loss": 0.5802,
+      "step": 13320
+    },
+    {
+      "epoch": 1.333,
+      "grad_norm": 23.17426872253418,
+      "learning_rate": 7.336200000000001e-06,
+      "loss": 0.6976,
+      "step": 13330
+    },
+    {
+      "epoch": 1.334,
+      "grad_norm": 47.89413070678711,
+      "learning_rate": 7.3342e-06,
+      "loss": 0.6137,
+      "step": 13340
+    },
+    {
+      "epoch": 1.335,
+      "grad_norm": 26.017189025878906,
+      "learning_rate": 7.332200000000001e-06,
+      "loss": 0.7701,
+      "step": 13350
+    },
+    {
+      "epoch": 1.336,
+      "grad_norm": 6.929594039916992,
+      "learning_rate": 7.3302000000000006e-06,
+      "loss": 0.6573,
+      "step": 13360
+    },
+    {
+      "epoch": 1.337,
+      "grad_norm": 9.463047981262207,
+      "learning_rate": 7.328200000000001e-06,
+      "loss": 0.6186,
+      "step": 13370
+    },
+    {
+      "epoch": 1.338,
+      "grad_norm": 29.972517013549805,
+      "learning_rate": 7.326200000000001e-06,
+      "loss": 0.5866,
+      "step": 13380
+    },
+    {
+      "epoch": 1.339,
+      "grad_norm": 22.74357032775879,
+      "learning_rate": 7.3242e-06,
+      "loss": 0.8169,
+      "step": 13390
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 28.037187576293945,
+      "learning_rate": 7.322200000000001e-06,
+      "loss": 0.566,
+      "step": 13400
+    },
+    {
+      "epoch": 1.341,
+      "grad_norm": 19.294530868530273,
+      "learning_rate": 7.320200000000001e-06,
+      "loss": 0.5276,
+      "step": 13410
+    },
+    {
+      "epoch": 1.342,
+      "grad_norm": 34.14975357055664,
+      "learning_rate": 7.3182000000000005e-06,
+      "loss": 0.96,
+      "step": 13420
+    },
+    {
+      "epoch": 1.343,
+      "grad_norm": 28.102813720703125,
+      "learning_rate": 7.316200000000001e-06,
+      "loss": 0.7187,
+      "step": 13430
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 24.984050750732422,
+      "learning_rate": 7.3142e-06,
+      "loss": 0.7285,
+      "step": 13440
+    },
+    {
+      "epoch": 1.345,
+      "grad_norm": 8.587176322937012,
+      "learning_rate": 7.3122e-06,
+      "loss": 0.6251,
+      "step": 13450
+    },
+    {
+      "epoch": 1.346,
+      "grad_norm": 48.03047561645508,
+      "learning_rate": 7.310200000000001e-06,
+      "loss": 0.6534,
+      "step": 13460
+    },
+    {
+      "epoch": 1.347,
+      "grad_norm": 41.015052795410156,
+      "learning_rate": 7.308200000000001e-06,
+      "loss": 0.638,
+      "step": 13470
+    },
+    {
+      "epoch": 1.3479999999999999,
+      "grad_norm": 53.580875396728516,
+      "learning_rate": 7.306200000000001e-06,
+      "loss": 0.5571,
+      "step": 13480
+    },
+    {
+      "epoch": 1.349,
+      "grad_norm": 41.472984313964844,
+      "learning_rate": 7.3042e-06,
+      "loss": 0.7513,
+      "step": 13490
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 23.2569580078125,
+      "learning_rate": 7.3022e-06,
+      "loss": 0.5893,
+      "step": 13500
+    },
+    {
+      "epoch": 1.351,
+      "grad_norm": 39.171356201171875,
+      "learning_rate": 7.300200000000001e-06,
+      "loss": 0.7546,
+      "step": 13510
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 23.469484329223633,
+      "learning_rate": 7.298200000000001e-06,
+      "loss": 0.5874,
+      "step": 13520
+    },
+    {
+      "epoch": 1.353,
+      "grad_norm": 22.72773551940918,
+      "learning_rate": 7.296200000000001e-06,
+      "loss": 0.5501,
+      "step": 13530
+    },
+    {
+      "epoch": 1.354,
+      "grad_norm": 27.632083892822266,
+      "learning_rate": 7.2942e-06,
+      "loss": 0.6415,
+      "step": 13540
+    },
+    {
+      "epoch": 1.355,
+      "grad_norm": 14.760106086730957,
+      "learning_rate": 7.2922e-06,
+      "loss": 0.5344,
+      "step": 13550
+    },
+    {
+      "epoch": 1.3559999999999999,
+      "grad_norm": 48.3216667175293,
+      "learning_rate": 7.290200000000001e-06,
+      "loss": 0.6311,
+      "step": 13560
+    },
+    {
+      "epoch": 1.357,
+      "grad_norm": 34.64678192138672,
+      "learning_rate": 7.288200000000001e-06,
+      "loss": 0.6622,
+      "step": 13570
+    },
+    {
+      "epoch": 1.358,
+      "grad_norm": 16.70772361755371,
+      "learning_rate": 7.286200000000001e-06,
+      "loss": 0.6337,
+      "step": 13580
+    },
+    {
+      "epoch": 1.359,
+      "grad_norm": 10.013886451721191,
+      "learning_rate": 7.2842000000000005e-06,
+      "loss": 0.5636,
+      "step": 13590
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 57.176029205322266,
+      "learning_rate": 7.2822e-06,
+      "loss": 0.5612,
+      "step": 13600
+    },
+    {
+      "epoch": 1.361,
+      "grad_norm": 56.01203536987305,
+      "learning_rate": 7.2802e-06,
+      "loss": 0.6763,
+      "step": 13610
+    },
+    {
+      "epoch": 1.362,
+      "grad_norm": 20.71746253967285,
+      "learning_rate": 7.278200000000001e-06,
+      "loss": 0.8182,
+      "step": 13620
+    },
+    {
+      "epoch": 1.363,
+      "grad_norm": 23.728199005126953,
+      "learning_rate": 7.276200000000001e-06,
+      "loss": 0.6961,
+      "step": 13630
+    },
+    {
+      "epoch": 1.3639999999999999,
+      "grad_norm": 39.34861755371094,
+      "learning_rate": 7.2742e-06,
+      "loss": 0.4997,
+      "step": 13640
+    },
+    {
+      "epoch": 1.365,
+      "grad_norm": 35.87208557128906,
+      "learning_rate": 7.2722000000000004e-06,
+      "loss": 0.8257,
+      "step": 13650
+    },
+    {
+      "epoch": 1.366,
+      "grad_norm": 13.711315155029297,
+      "learning_rate": 7.2702e-06,
+      "loss": 0.6184,
+      "step": 13660
+    },
+    {
+      "epoch": 1.367,
+      "grad_norm": 22.51475715637207,
+      "learning_rate": 7.268200000000001e-06,
+      "loss": 0.8108,
+      "step": 13670
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 33.50455093383789,
+      "learning_rate": 7.266200000000001e-06,
+      "loss": 0.8212,
+      "step": 13680
+    },
+    {
+      "epoch": 1.369,
+      "grad_norm": 19.336406707763672,
+      "learning_rate": 7.2642e-06,
+      "loss": 0.4771,
+      "step": 13690
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 24.16560935974121,
+      "learning_rate": 7.2622000000000005e-06,
+      "loss": 0.4932,
+      "step": 13700
+    },
+    {
+      "epoch": 1.371,
+      "grad_norm": 16.50635528564453,
+      "learning_rate": 7.2602e-06,
+      "loss": 0.7304,
+      "step": 13710
+    },
+    {
+      "epoch": 1.3719999999999999,
+      "grad_norm": 18.5286865234375,
+      "learning_rate": 7.258200000000001e-06,
+      "loss": 0.5905,
+      "step": 13720
+    },
+    {
+      "epoch": 1.373,
+      "grad_norm": 23.307600021362305,
+      "learning_rate": 7.256200000000001e-06,
+      "loss": 0.8059,
+      "step": 13730
+    },
+    {
+      "epoch": 1.374,
+      "grad_norm": 43.4228515625,
+      "learning_rate": 7.2542e-06,
+      "loss": 0.6437,
+      "step": 13740
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 25.997020721435547,
+      "learning_rate": 7.252200000000001e-06,
+      "loss": 0.7291,
+      "step": 13750
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 9.636469841003418,
+      "learning_rate": 7.2502000000000005e-06,
+      "loss": 0.6815,
+      "step": 13760
+    },
+    {
+      "epoch": 1.377,
+      "grad_norm": 19.01958465576172,
+      "learning_rate": 7.248200000000001e-06,
+      "loss": 0.6909,
+      "step": 13770
+    },
+    {
+      "epoch": 1.3780000000000001,
+      "grad_norm": 12.587032318115234,
+      "learning_rate": 7.246200000000001e-06,
+      "loss": 0.4865,
+      "step": 13780
+    },
+    {
+      "epoch": 1.379,
+      "grad_norm": 14.733444213867188,
+      "learning_rate": 7.2442e-06,
+      "loss": 0.5581,
+      "step": 13790
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.286008834838867,
+      "learning_rate": 7.242200000000001e-06,
+      "loss": 0.7601,
+      "step": 13800
+    },
+    {
+      "epoch": 1.381,
+      "grad_norm": 55.66062927246094,
+      "learning_rate": 7.240200000000001e-06,
+      "loss": 0.7174,
+      "step": 13810
+    },
+    {
+      "epoch": 1.3820000000000001,
+      "grad_norm": 12.686264991760254,
+      "learning_rate": 7.2382000000000004e-06,
+      "loss": 0.7169,
+      "step": 13820
+    },
+    {
+      "epoch": 1.383,
+      "grad_norm": 33.9859619140625,
+      "learning_rate": 7.236200000000001e-06,
+      "loss": 0.6755,
+      "step": 13830
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 24.982358932495117,
+      "learning_rate": 7.2342e-06,
+      "loss": 0.631,
+      "step": 13840
+    },
+    {
+      "epoch": 1.385,
+      "grad_norm": 29.842527389526367,
+      "learning_rate": 7.2322e-06,
+      "loss": 0.7629,
+      "step": 13850
+    },
+    {
+      "epoch": 1.3860000000000001,
+      "grad_norm": 36.99466323852539,
+      "learning_rate": 7.230200000000001e-06,
+      "loss": 0.978,
+      "step": 13860
+    },
+    {
+      "epoch": 1.387,
+      "grad_norm": 30.951824188232422,
+      "learning_rate": 7.2282000000000005e-06,
+      "loss": 0.5872,
+      "step": 13870
+    },
+    {
+      "epoch": 1.388,
+      "grad_norm": 2.009528398513794,
+      "learning_rate": 7.226200000000001e-06,
+      "loss": 0.5371,
+      "step": 13880
+    },
+    {
+      "epoch": 1.389,
+      "grad_norm": 21.285736083984375,
+      "learning_rate": 7.2242e-06,
+      "loss": 0.6345,
+      "step": 13890
+    },
+    {
+      "epoch": 1.3900000000000001,
+      "grad_norm": 29.036205291748047,
+      "learning_rate": 7.2222e-06,
+      "loss": 0.6506,
+      "step": 13900
+    },
+    {
+      "epoch": 1.391,
+      "grad_norm": 32.15199661254883,
+      "learning_rate": 7.220200000000001e-06,
+      "loss": 0.7098,
+      "step": 13910
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 56.528621673583984,
+      "learning_rate": 7.218200000000001e-06,
+      "loss": 0.7811,
+      "step": 13920
+    },
+    {
+      "epoch": 1.393,
+      "grad_norm": 32.981815338134766,
+      "learning_rate": 7.216200000000001e-06,
+      "loss": 0.6597,
+      "step": 13930
+    },
+    {
+      "epoch": 1.3940000000000001,
+      "grad_norm": 26.504575729370117,
+      "learning_rate": 7.2142e-06,
+      "loss": 0.5865,
+      "step": 13940
+    },
+    {
+      "epoch": 1.395,
+      "grad_norm": 43.711788177490234,
+      "learning_rate": 7.2122e-06,
+      "loss": 0.734,
+      "step": 13950
+    },
+    {
+      "epoch": 1.396,
+      "grad_norm": 18.780986785888672,
+      "learning_rate": 7.210200000000001e-06,
+      "loss": 0.4507,
+      "step": 13960
+    },
+    {
+      "epoch": 1.397,
+      "grad_norm": 31.044410705566406,
+      "learning_rate": 7.208200000000001e-06,
+      "loss": 0.3073,
+      "step": 13970
+    },
+    {
+      "epoch": 1.3980000000000001,
+      "grad_norm": 23.003955841064453,
+      "learning_rate": 7.2062000000000006e-06,
+      "loss": 0.3738,
+      "step": 13980
+    },
+    {
+      "epoch": 1.399,
+      "grad_norm": 39.76530075073242,
+      "learning_rate": 7.2042e-06,
+      "loss": 0.6692,
+      "step": 13990
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.235315322875977,
+      "learning_rate": 7.2022e-06,
+      "loss": 0.6609,
+      "step": 14000
+    },
+    {
+      "epoch": 1.401,
+      "grad_norm": 38.41657257080078,
+      "learning_rate": 7.2002e-06,
+      "loss": 0.8216,
+      "step": 14010
+    },
+    {
+      "epoch": 1.4020000000000001,
+      "grad_norm": 3.7093570232391357,
+      "learning_rate": 7.198200000000001e-06,
+      "loss": 0.4728,
+      "step": 14020
+    },
+    {
+      "epoch": 1.403,
+      "grad_norm": 7.972429275512695,
+      "learning_rate": 7.196200000000001e-06,
+      "loss": 0.788,
+      "step": 14030
+    },
+    {
+      "epoch": 1.404,
+      "grad_norm": 30.618633270263672,
+      "learning_rate": 7.1942e-06,
+      "loss": 0.7229,
+      "step": 14040
+    },
+    {
+      "epoch": 1.405,
+      "grad_norm": 20.31015396118164,
+      "learning_rate": 7.1922e-06,
+      "loss": 0.679,
+      "step": 14050
+    },
+    {
+      "epoch": 1.4060000000000001,
+      "grad_norm": 33.285003662109375,
+      "learning_rate": 7.1902e-06,
+      "loss": 0.6957,
+      "step": 14060
+    },
+    {
+      "epoch": 1.407,
+      "grad_norm": 17.131729125976562,
+      "learning_rate": 7.188200000000001e-06,
+      "loss": 0.4287,
+      "step": 14070
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 37.19723892211914,
+      "learning_rate": 7.186200000000001e-06,
+      "loss": 0.7653,
+      "step": 14080
+    },
+    {
+      "epoch": 1.409,
+      "grad_norm": 24.82241439819336,
+      "learning_rate": 7.1842e-06,
+      "loss": 0.7721,
+      "step": 14090
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 41.999305725097656,
+      "learning_rate": 7.1822000000000005e-06,
+      "loss": 0.3842,
+      "step": 14100
+    },
+    {
+      "epoch": 1.411,
+      "grad_norm": 26.453805923461914,
+      "learning_rate": 7.1802e-06,
+      "loss": 0.8213,
+      "step": 14110
+    },
+    {
+      "epoch": 1.412,
+      "grad_norm": 29.831308364868164,
+      "learning_rate": 7.178200000000001e-06,
+      "loss": 0.965,
+      "step": 14120
+    },
+    {
+      "epoch": 1.413,
+      "grad_norm": 35.91419219970703,
+      "learning_rate": 7.176200000000001e-06,
+      "loss": 0.863,
+      "step": 14130
+    },
+    {
+      "epoch": 1.414,
+      "grad_norm": 17.73390769958496,
+      "learning_rate": 7.1742e-06,
+      "loss": 0.5709,
+      "step": 14140
+    },
+    {
+      "epoch": 1.415,
+      "grad_norm": 7.556557655334473,
+      "learning_rate": 7.1722000000000006e-06,
+      "loss": 0.4825,
+      "step": 14150
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 31.779680252075195,
+      "learning_rate": 7.1702e-06,
+      "loss": 0.8153,
+      "step": 14160
+    },
+    {
+      "epoch": 1.417,
+      "grad_norm": 10.960797309875488,
+      "learning_rate": 7.168200000000001e-06,
+      "loss": 0.595,
+      "step": 14170
+    },
+    {
+      "epoch": 1.418,
+      "grad_norm": 8.104966163635254,
+      "learning_rate": 7.166400000000001e-06,
+      "loss": 0.6693,
+      "step": 14180
+    },
+    {
+      "epoch": 1.419,
+      "grad_norm": 26.086509704589844,
+      "learning_rate": 7.1644e-06,
+      "loss": 0.6027,
+      "step": 14190
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 25.449806213378906,
+      "learning_rate": 7.1624e-06,
+      "loss": 0.557,
+      "step": 14200
+    },
+    {
+      "epoch": 1.421,
+      "grad_norm": 32.112892150878906,
+      "learning_rate": 7.160400000000001e-06,
+      "loss": 0.7366,
+      "step": 14210
+    },
+    {
+      "epoch": 1.422,
+      "grad_norm": 28.12934684753418,
+      "learning_rate": 7.158400000000001e-06,
+      "loss": 0.8341,
+      "step": 14220
+    },
+    {
+      "epoch": 1.423,
+      "grad_norm": 30.7032413482666,
+      "learning_rate": 7.1564000000000015e-06,
+      "loss": 0.6367,
+      "step": 14230
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 16.595033645629883,
+      "learning_rate": 7.1544000000000005e-06,
+      "loss": 0.6789,
+      "step": 14240
+    },
+    {
+      "epoch": 1.425,
+      "grad_norm": 31.310976028442383,
+      "learning_rate": 7.1524e-06,
+      "loss": 0.6617,
+      "step": 14250
+    },
+    {
+      "epoch": 1.426,
+      "grad_norm": 19.14084815979004,
+      "learning_rate": 7.150400000000001e-06,
+      "loss": 0.5745,
+      "step": 14260
+    },
+    {
+      "epoch": 1.427,
+      "grad_norm": 49.51873779296875,
+      "learning_rate": 7.148400000000001e-06,
+      "loss": 0.8873,
+      "step": 14270
+    },
+    {
+      "epoch": 1.428,
+      "grad_norm": 34.733726501464844,
+      "learning_rate": 7.146400000000001e-06,
+      "loss": 0.6153,
+      "step": 14280
+    },
+    {
+      "epoch": 1.429,
+      "grad_norm": 38.35405731201172,
+      "learning_rate": 7.144400000000001e-06,
+      "loss": 0.3432,
+      "step": 14290
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 28.54090118408203,
+      "learning_rate": 7.1424000000000004e-06,
+      "loss": 0.6794,
+      "step": 14300
+    },
+    {
+      "epoch": 1.431,
+      "grad_norm": 156.63278198242188,
+      "learning_rate": 7.1404e-06,
+      "loss": 0.6149,
+      "step": 14310
+    },
+    {
+      "epoch": 1.432,
+      "grad_norm": 44.73384094238281,
+      "learning_rate": 7.138400000000001e-06,
+      "loss": 0.7077,
+      "step": 14320
+    },
+    {
+      "epoch": 1.433,
+      "grad_norm": 32.4652099609375,
+      "learning_rate": 7.136400000000001e-06,
+      "loss": 0.5943,
+      "step": 14330
+    },
+    {
+      "epoch": 1.434,
+      "grad_norm": 15.920811653137207,
+      "learning_rate": 7.1344e-06,
+      "loss": 0.5748,
+      "step": 14340
+    },
+    {
+      "epoch": 1.435,
+      "grad_norm": 14.366812705993652,
+      "learning_rate": 7.1324000000000005e-06,
+      "loss": 0.7162,
+      "step": 14350
+    },
+    {
+      "epoch": 1.436,
+      "grad_norm": 20.413108825683594,
+      "learning_rate": 7.1304e-06,
+      "loss": 0.7665,
+      "step": 14360
+    },
+    {
+      "epoch": 1.437,
+      "grad_norm": 35.08794021606445,
+      "learning_rate": 7.128400000000001e-06,
+      "loss": 0.7175,
+      "step": 14370
+    },
+    {
+      "epoch": 1.438,
+      "grad_norm": 58.03242492675781,
+      "learning_rate": 7.126400000000001e-06,
+      "loss": 0.952,
+      "step": 14380
+    },
+    {
+      "epoch": 1.439,
+      "grad_norm": 24.53120231628418,
+      "learning_rate": 7.1244e-06,
+      "loss": 0.7022,
+      "step": 14390
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 33.467071533203125,
+      "learning_rate": 7.122400000000001e-06,
+      "loss": 0.6428,
+      "step": 14400
+    },
+    {
+      "epoch": 1.441,
+      "grad_norm": 37.66570281982422,
+      "learning_rate": 7.1204000000000005e-06,
+      "loss": 0.8671,
+      "step": 14410
+    },
+    {
+      "epoch": 1.442,
+      "grad_norm": 31.30205535888672,
+      "learning_rate": 7.118400000000001e-06,
+      "loss": 0.5898,
+      "step": 14420
+    },
+    {
+      "epoch": 1.443,
+      "grad_norm": 31.95399284362793,
+      "learning_rate": 7.116400000000001e-06,
+      "loss": 0.6795,
+      "step": 14430
+    },
+    {
+      "epoch": 1.444,
+      "grad_norm": 32.41059494018555,
+      "learning_rate": 7.1144e-06,
+      "loss": 0.4456,
+      "step": 14440
+    },
+    {
+      "epoch": 1.445,
+      "grad_norm": 25.59587287902832,
+      "learning_rate": 7.112400000000001e-06,
+      "loss": 0.7705,
+      "step": 14450
+    },
+    {
+      "epoch": 1.446,
+      "grad_norm": 37.86859893798828,
+      "learning_rate": 7.1104000000000006e-06,
+      "loss": 0.6855,
+      "step": 14460
+    },
+    {
+      "epoch": 1.447,
+      "grad_norm": 7.653895378112793,
+      "learning_rate": 7.108400000000001e-06,
+      "loss": 0.7096,
+      "step": 14470
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 34.88923263549805,
+      "learning_rate": 7.106400000000001e-06,
+      "loss": 0.6658,
+      "step": 14480
+    },
+    {
+      "epoch": 1.449,
+      "grad_norm": 27.103458404541016,
+      "learning_rate": 7.1044e-06,
+      "loss": 0.7682,
+      "step": 14490
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.40449905395508,
+      "learning_rate": 7.102400000000001e-06,
+      "loss": 0.8501,
+      "step": 14500
+    },
+    {
+      "epoch": 1.451,
+      "grad_norm": 52.43901824951172,
+      "learning_rate": 7.100400000000001e-06,
+      "loss": 0.5298,
+      "step": 14510
+    },
+    {
+      "epoch": 1.452,
+      "grad_norm": 36.47237014770508,
+      "learning_rate": 7.0984000000000005e-06,
+      "loss": 0.9262,
+      "step": 14520
+    },
+    {
+      "epoch": 1.453,
+      "grad_norm": 33.24127197265625,
+      "learning_rate": 7.096400000000001e-06,
+      "loss": 0.7582,
+      "step": 14530
+    },
+    {
+      "epoch": 1.454,
+      "grad_norm": 27.120197296142578,
+      "learning_rate": 7.0944e-06,
+      "loss": 0.7854,
+      "step": 14540
+    },
+    {
+      "epoch": 1.455,
+      "grad_norm": 9.86540699005127,
+      "learning_rate": 7.0924e-06,
+      "loss": 0.8409,
+      "step": 14550
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 17.383777618408203,
+      "learning_rate": 7.090400000000001e-06,
+      "loss": 0.4384,
+      "step": 14560
+    },
+    {
+      "epoch": 1.457,
+      "grad_norm": 28.551664352416992,
+      "learning_rate": 7.088400000000001e-06,
+      "loss": 0.8198,
+      "step": 14570
+    },
+    {
+      "epoch": 1.458,
+      "grad_norm": 27.755903244018555,
+      "learning_rate": 7.086400000000001e-06,
+      "loss": 0.6956,
+      "step": 14580
+    },
+    {
+      "epoch": 1.459,
+      "grad_norm": 16.9943904876709,
+      "learning_rate": 7.0844e-06,
+      "loss": 0.6524,
+      "step": 14590
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 19.76603889465332,
+      "learning_rate": 7.0824e-06,
+      "loss": 0.5919,
+      "step": 14600
+    },
+    {
+      "epoch": 1.461,
+      "grad_norm": 4.806751728057861,
+      "learning_rate": 7.080400000000001e-06,
+      "loss": 0.6876,
+      "step": 14610
+    },
+    {
+      "epoch": 1.462,
+      "grad_norm": 18.323942184448242,
+      "learning_rate": 7.078400000000001e-06,
+      "loss": 0.8455,
+      "step": 14620
+    },
+    {
+      "epoch": 1.463,
+      "grad_norm": 23.91664695739746,
+      "learning_rate": 7.076400000000001e-06,
+      "loss": 0.7989,
+      "step": 14630
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 12.201299667358398,
+      "learning_rate": 7.0744e-06,
+      "loss": 0.7662,
+      "step": 14640
+    },
+    {
+      "epoch": 1.465,
+      "grad_norm": 20.28821563720703,
+      "learning_rate": 7.0724e-06,
+      "loss": 0.6595,
+      "step": 14650
+    },
+    {
+      "epoch": 1.466,
+      "grad_norm": 48.507930755615234,
+      "learning_rate": 7.070400000000001e-06,
+      "loss": 0.4343,
+      "step": 14660
+    },
+    {
+      "epoch": 1.467,
+      "grad_norm": 54.569313049316406,
+      "learning_rate": 7.068400000000001e-06,
+      "loss": 0.615,
+      "step": 14670
+    },
+    {
+      "epoch": 1.468,
+      "grad_norm": 22.862001419067383,
+      "learning_rate": 7.066400000000001e-06,
+      "loss": 0.7374,
+      "step": 14680
+    },
+    {
+      "epoch": 1.4689999999999999,
+      "grad_norm": 23.220779418945312,
+      "learning_rate": 7.0644000000000005e-06,
+      "loss": 0.7556,
+      "step": 14690
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 16.430978775024414,
+      "learning_rate": 7.0624e-06,
+      "loss": 0.6466,
+      "step": 14700
+    },
+    {
+      "epoch": 1.471,
+      "grad_norm": 40.130470275878906,
+      "learning_rate": 7.0604e-06,
+      "loss": 0.798,
+      "step": 14710
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 31.032583236694336,
+      "learning_rate": 7.058400000000001e-06,
+      "loss": 0.7023,
+      "step": 14720
+    },
+    {
+      "epoch": 1.4729999999999999,
+      "grad_norm": 33.947750091552734,
+      "learning_rate": 7.056400000000001e-06,
+      "loss": 0.7844,
+      "step": 14730
+    },
+    {
+      "epoch": 1.474,
+      "grad_norm": 20.618391036987305,
+      "learning_rate": 7.0544e-06,
+      "loss": 0.8526,
+      "step": 14740
+    },
+    {
+      "epoch": 1.475,
+      "grad_norm": 30.07219886779785,
+      "learning_rate": 7.0524000000000004e-06,
+      "loss": 0.6605,
+      "step": 14750
+    },
+    {
+      "epoch": 1.476,
+      "grad_norm": 37.258087158203125,
+      "learning_rate": 7.0504e-06,
+      "loss": 0.6999,
+      "step": 14760
+    },
+    {
+      "epoch": 1.4769999999999999,
+      "grad_norm": 36.73057556152344,
+      "learning_rate": 7.048400000000001e-06,
+      "loss": 0.5431,
+      "step": 14770
+    },
+    {
+      "epoch": 1.478,
+      "grad_norm": 45.83023452758789,
+      "learning_rate": 7.046400000000001e-06,
+      "loss": 1.0078,
+      "step": 14780
+    },
+    {
+      "epoch": 1.479,
+      "grad_norm": 29.41172981262207,
+      "learning_rate": 7.0444e-06,
+      "loss": 0.5701,
+      "step": 14790
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.402508497238159,
+      "learning_rate": 7.0424000000000005e-06,
+      "loss": 0.4707,
+      "step": 14800
+    },
+    {
+      "epoch": 1.4809999999999999,
+      "grad_norm": 24.081668853759766,
+      "learning_rate": 7.0404e-06,
+      "loss": 0.6248,
+      "step": 14810
+    },
+    {
+      "epoch": 1.482,
+      "grad_norm": 43.570091247558594,
+      "learning_rate": 7.038400000000001e-06,
+      "loss": 0.676,
+      "step": 14820
+    },
+    {
+      "epoch": 1.483,
+      "grad_norm": 36.47454833984375,
+      "learning_rate": 7.036400000000001e-06,
+      "loss": 0.6858,
+      "step": 14830
+    },
+    {
+      "epoch": 1.484,
+      "grad_norm": 29.922746658325195,
+      "learning_rate": 7.0344e-06,
+      "loss": 0.5933,
+      "step": 14840
+    },
+    {
+      "epoch": 1.4849999999999999,
+      "grad_norm": 8.109268188476562,
+      "learning_rate": 7.032400000000001e-06,
+      "loss": 0.6108,
+      "step": 14850
+    },
+    {
+      "epoch": 1.486,
+      "grad_norm": 7.4250264167785645,
+      "learning_rate": 7.0304000000000005e-06,
+      "loss": 0.6123,
+      "step": 14860
+    },
+    {
+      "epoch": 1.487,
+      "grad_norm": 9.724357604980469,
+      "learning_rate": 7.028400000000001e-06,
+      "loss": 0.6254,
+      "step": 14870
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 7.028298377990723,
+      "learning_rate": 7.026400000000001e-06,
+      "loss": 0.6052,
+      "step": 14880
+    },
+    {
+      "epoch": 1.4889999999999999,
+      "grad_norm": 31.92440414428711,
+      "learning_rate": 7.0244e-06,
+      "loss": 0.8187,
+      "step": 14890
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.212546348571777,
+      "learning_rate": 7.022400000000001e-06,
+      "loss": 0.7831,
+      "step": 14900
+    },
+    {
+      "epoch": 1.491,
+      "grad_norm": 41.331748962402344,
+      "learning_rate": 7.020400000000001e-06,
+      "loss": 0.6803,
+      "step": 14910
+    },
+    {
+      "epoch": 1.492,
+      "grad_norm": 29.76875877380371,
+      "learning_rate": 7.0184000000000004e-06,
+      "loss": 0.9259,
+      "step": 14920
+    },
+    {
+      "epoch": 1.4929999999999999,
+      "grad_norm": 23.765771865844727,
+      "learning_rate": 7.016400000000001e-06,
+      "loss": 0.7033,
+      "step": 14930
+    },
+    {
+      "epoch": 1.494,
+      "grad_norm": 42.97665786743164,
+      "learning_rate": 7.0144e-06,
+      "loss": 0.8351,
+      "step": 14940
+    },
+    {
+      "epoch": 1.495,
+      "grad_norm": 22.49999237060547,
+      "learning_rate": 7.0124e-06,
+      "loss": 0.9859,
+      "step": 14950
+    },
+    {
+      "epoch": 1.496,
+      "grad_norm": 25.587005615234375,
+      "learning_rate": 7.010400000000001e-06,
+      "loss": 0.73,
+      "step": 14960
+    },
+    {
+      "epoch": 1.4969999999999999,
+      "grad_norm": 24.11408805847168,
+      "learning_rate": 7.0084000000000005e-06,
+      "loss": 0.6629,
+      "step": 14970
+    },
+    {
+      "epoch": 1.498,
+      "grad_norm": 33.97454833984375,
+      "learning_rate": 7.006400000000001e-06,
+      "loss": 0.7687,
+      "step": 14980
+    },
+    {
+      "epoch": 1.499,
+      "grad_norm": 27.10246467590332,
+      "learning_rate": 7.0044e-06,
+      "loss": 0.5656,
+      "step": 14990
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 31.772153854370117,
+      "learning_rate": 7.0024e-06,
+      "loss": 0.5767,
+      "step": 15000
+    },
+    {
+      "epoch": 1.501,
+      "grad_norm": 32.298919677734375,
+      "learning_rate": 7.000400000000001e-06,
+      "loss": 0.6371,
+      "step": 15010
+    },
+    {
+      "epoch": 1.502,
+      "grad_norm": 41.70977020263672,
+      "learning_rate": 6.998400000000001e-06,
+      "loss": 0.8802,
+      "step": 15020
+    },
+    {
+      "epoch": 1.5030000000000001,
+      "grad_norm": 5.668227672576904,
+      "learning_rate": 6.996400000000001e-06,
+      "loss": 0.5482,
+      "step": 15030
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 30.462928771972656,
+      "learning_rate": 6.9944e-06,
+      "loss": 0.5476,
+      "step": 15040
+    },
+    {
+      "epoch": 1.505,
+      "grad_norm": 17.560972213745117,
+      "learning_rate": 6.9924e-06,
+      "loss": 0.473,
+      "step": 15050
+    },
+    {
+      "epoch": 1.506,
+      "grad_norm": 7.06756067276001,
+      "learning_rate": 6.990400000000001e-06,
+      "loss": 0.4892,
+      "step": 15060
+    },
+    {
+      "epoch": 1.5070000000000001,
+      "grad_norm": 14.195466995239258,
+      "learning_rate": 6.988400000000001e-06,
+      "loss": 0.6969,
+      "step": 15070
+    },
+    {
+      "epoch": 1.508,
+      "grad_norm": 32.51763153076172,
+      "learning_rate": 6.9864000000000006e-06,
+      "loss": 0.6478,
+      "step": 15080
+    },
+    {
+      "epoch": 1.509,
+      "grad_norm": 34.5938606262207,
+      "learning_rate": 6.9844e-06,
+      "loss": 0.7144,
+      "step": 15090
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 29.059711456298828,
+      "learning_rate": 6.9824e-06,
+      "loss": 0.3797,
+      "step": 15100
+    },
+    {
+      "epoch": 1.5110000000000001,
+      "grad_norm": 42.22810363769531,
+      "learning_rate": 6.9804e-06,
+      "loss": 0.7841,
+      "step": 15110
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 15.14452075958252,
+      "learning_rate": 6.978400000000001e-06,
+      "loss": 0.9623,
+      "step": 15120
+    },
+    {
+      "epoch": 1.513,
+      "grad_norm": 26.8512020111084,
+      "learning_rate": 6.976400000000001e-06,
+      "loss": 0.5488,
+      "step": 15130
+    },
+    {
+      "epoch": 1.514,
+      "grad_norm": 12.91208267211914,
+      "learning_rate": 6.9744e-06,
+      "loss": 0.6394,
+      "step": 15140
+    },
+    {
+      "epoch": 1.5150000000000001,
+      "grad_norm": 19.62507438659668,
+      "learning_rate": 6.9724e-06,
+      "loss": 0.4828,
+      "step": 15150
+    },
+    {
+      "epoch": 1.516,
+      "grad_norm": 40.00102615356445,
+      "learning_rate": 6.9704e-06,
+      "loss": 0.4805,
+      "step": 15160
+    },
+    {
+      "epoch": 1.517,
+      "grad_norm": 68.10365295410156,
+      "learning_rate": 6.968400000000001e-06,
+      "loss": 0.583,
+      "step": 15170
+    },
+    {
+      "epoch": 1.518,
+      "grad_norm": 22.018186569213867,
+      "learning_rate": 6.966400000000001e-06,
+      "loss": 0.8632,
+      "step": 15180
+    },
+    {
+      "epoch": 1.5190000000000001,
+      "grad_norm": 18.3768253326416,
+      "learning_rate": 6.9644e-06,
+      "loss": 0.4432,
+      "step": 15190
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.584394454956055,
+      "learning_rate": 6.9624000000000005e-06,
+      "loss": 0.4675,
+      "step": 15200
+    },
+    {
+      "epoch": 1.521,
+      "grad_norm": 40.143768310546875,
+      "learning_rate": 6.9604e-06,
+      "loss": 0.9586,
+      "step": 15210
+    },
+    {
+      "epoch": 1.522,
+      "grad_norm": 18.579164505004883,
+      "learning_rate": 6.958400000000001e-06,
+      "loss": 0.3957,
+      "step": 15220
+    },
+    {
+      "epoch": 1.5230000000000001,
+      "grad_norm": 18.15229034423828,
+      "learning_rate": 6.956400000000001e-06,
+      "loss": 0.7275,
+      "step": 15230
+    },
+    {
+      "epoch": 1.524,
+      "grad_norm": 17.880271911621094,
+      "learning_rate": 6.9544e-06,
+      "loss": 0.7367,
+      "step": 15240
+    },
+    {
+      "epoch": 1.525,
+      "grad_norm": 5.8904266357421875,
+      "learning_rate": 6.9524000000000006e-06,
+      "loss": 0.7471,
+      "step": 15250
+    },
+    {
+      "epoch": 1.526,
+      "grad_norm": 28.13979721069336,
+      "learning_rate": 6.9504e-06,
+      "loss": 0.8176,
+      "step": 15260
+    },
+    {
+      "epoch": 1.5270000000000001,
+      "grad_norm": 13.52975082397461,
+      "learning_rate": 6.948400000000001e-06,
+      "loss": 0.587,
+      "step": 15270
+    },
+    {
+      "epoch": 1.528,
+      "grad_norm": 34.50688934326172,
+      "learning_rate": 6.946400000000001e-06,
+      "loss": 0.3854,
+      "step": 15280
+    },
+    {
+      "epoch": 1.529,
+      "grad_norm": 26.554487228393555,
+      "learning_rate": 6.9444e-06,
+      "loss": 0.5158,
+      "step": 15290
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.872733116149902,
+      "learning_rate": 6.942400000000001e-06,
+      "loss": 0.5621,
+      "step": 15300
+    },
+    {
+      "epoch": 1.5310000000000001,
+      "grad_norm": 47.18104934692383,
+      "learning_rate": 6.9404000000000005e-06,
+      "loss": 0.5125,
+      "step": 15310
+    },
+    {
+      "epoch": 1.532,
+      "grad_norm": 26.80068588256836,
+      "learning_rate": 6.9384e-06,
+      "loss": 0.5913,
+      "step": 15320
+    },
+    {
+      "epoch": 1.533,
+      "grad_norm": 39.93467712402344,
+      "learning_rate": 6.936400000000001e-06,
+      "loss": 0.5561,
+      "step": 15330
+    },
+    {
+      "epoch": 1.534,
+      "grad_norm": 3.488877534866333,
+      "learning_rate": 6.9344e-06,
+      "loss": 0.6594,
+      "step": 15340
+    },
+    {
+      "epoch": 1.5350000000000001,
+      "grad_norm": 19.96973419189453,
+      "learning_rate": 6.9324e-06,
+      "loss": 0.7096,
+      "step": 15350
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 32.5767936706543,
+      "learning_rate": 6.930600000000001e-06,
+      "loss": 0.6651,
+      "step": 15360
+    },
+    {
+      "epoch": 1.537,
+      "grad_norm": 22.27090835571289,
+      "learning_rate": 6.928600000000001e-06,
+      "loss": 0.6513,
+      "step": 15370
+    },
+    {
+      "epoch": 1.538,
+      "grad_norm": 57.235572814941406,
+      "learning_rate": 6.926600000000001e-06,
+      "loss": 0.8128,
+      "step": 15380
+    },
+    {
+      "epoch": 1.5390000000000001,
+      "grad_norm": 26.9422664642334,
+      "learning_rate": 6.924600000000001e-06,
+      "loss": 0.4483,
+      "step": 15390
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.947314262390137,
+      "learning_rate": 6.9226000000000004e-06,
+      "loss": 0.6516,
+      "step": 15400
+    },
+    {
+      "epoch": 1.541,
+      "grad_norm": 48.50078201293945,
+      "learning_rate": 6.9206e-06,
+      "loss": 0.8598,
+      "step": 15410
+    },
+    {
+      "epoch": 1.542,
+      "grad_norm": 26.725366592407227,
+      "learning_rate": 6.918600000000001e-06,
+      "loss": 0.7719,
+      "step": 15420
+    },
+    {
+      "epoch": 1.5430000000000001,
+      "grad_norm": 23.61861228942871,
+      "learning_rate": 6.916600000000001e-06,
+      "loss": 0.9362,
+      "step": 15430
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 24.58664894104004,
+      "learning_rate": 6.9146e-06,
+      "loss": 0.7702,
+      "step": 15440
+    },
+    {
+      "epoch": 1.545,
+      "grad_norm": 16.35002899169922,
+      "learning_rate": 6.9126000000000005e-06,
+      "loss": 0.5286,
+      "step": 15450
+    },
+    {
+      "epoch": 1.546,
+      "grad_norm": 7.408541202545166,
+      "learning_rate": 6.9106e-06,
+      "loss": 0.6786,
+      "step": 15460
+    },
+    {
+      "epoch": 1.5470000000000002,
+      "grad_norm": 29.840011596679688,
+      "learning_rate": 6.908600000000001e-06,
+      "loss": 0.7321,
+      "step": 15470
+    },
+    {
+      "epoch": 1.548,
+      "grad_norm": 24.454252243041992,
+      "learning_rate": 6.906600000000001e-06,
+      "loss": 0.7347,
+      "step": 15480
+    },
+    {
+      "epoch": 1.549,
+      "grad_norm": 8.857321739196777,
+      "learning_rate": 6.9046e-06,
+      "loss": 0.7328,
+      "step": 15490
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 21.619884490966797,
+      "learning_rate": 6.902600000000001e-06,
+      "loss": 0.7709,
+      "step": 15500
+    },
+    {
+      "epoch": 1.5510000000000002,
+      "grad_norm": 26.981609344482422,
+      "learning_rate": 6.9006000000000005e-06,
+      "loss": 0.7383,
+      "step": 15510
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 24.25237274169922,
+      "learning_rate": 6.898600000000001e-06,
+      "loss": 0.582,
+      "step": 15520
+    },
+    {
+      "epoch": 1.553,
+      "grad_norm": 20.094484329223633,
+      "learning_rate": 6.896600000000001e-06,
+      "loss": 0.8337,
+      "step": 15530
+    },
+    {
+      "epoch": 1.554,
+      "grad_norm": 33.374324798583984,
+      "learning_rate": 6.8946e-06,
+      "loss": 0.634,
+      "step": 15540
+    },
+    {
+      "epoch": 1.5550000000000002,
+      "grad_norm": 23.03681182861328,
+      "learning_rate": 6.892600000000001e-06,
+      "loss": 0.4505,
+      "step": 15550
+    },
+    {
+      "epoch": 1.556,
+      "grad_norm": 36.90275955200195,
+      "learning_rate": 6.8906000000000006e-06,
+      "loss": 0.8079,
+      "step": 15560
+    },
+    {
+      "epoch": 1.557,
+      "grad_norm": 21.511165618896484,
+      "learning_rate": 6.8886e-06,
+      "loss": 0.6231,
+      "step": 15570
+    },
+    {
+      "epoch": 1.558,
+      "grad_norm": 27.871326446533203,
+      "learning_rate": 6.886600000000001e-06,
+      "loss": 0.5137,
+      "step": 15580
+    },
+    {
+      "epoch": 1.5590000000000002,
+      "grad_norm": 32.82701873779297,
+      "learning_rate": 6.8846e-06,
+      "loss": 0.8568,
+      "step": 15590
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.283423900604248,
+      "learning_rate": 6.8826e-06,
+      "loss": 0.6777,
+      "step": 15600
+    },
+    {
+      "epoch": 1.561,
+      "grad_norm": 37.66361999511719,
+      "learning_rate": 6.880600000000001e-06,
+      "loss": 0.3779,
+      "step": 15610
+    },
+    {
+      "epoch": 1.562,
+      "grad_norm": 29.52036476135254,
+      "learning_rate": 6.8786000000000005e-06,
+      "loss": 0.6852,
+      "step": 15620
+    },
+    {
+      "epoch": 1.563,
+      "grad_norm": 8.317204475402832,
+      "learning_rate": 6.876600000000001e-06,
+      "loss": 0.5511,
+      "step": 15630
+    },
+    {
+      "epoch": 1.564,
+      "grad_norm": 41.191375732421875,
+      "learning_rate": 6.8746e-06,
+      "loss": 0.8747,
+      "step": 15640
+    },
+    {
+      "epoch": 1.565,
+      "grad_norm": 39.559818267822266,
+      "learning_rate": 6.8726e-06,
+      "loss": 0.8863,
+      "step": 15650
+    },
+    {
+      "epoch": 1.5659999999999998,
+      "grad_norm": 30.73655891418457,
+      "learning_rate": 6.870600000000001e-06,
+      "loss": 0.7081,
+      "step": 15660
+    },
+    {
+      "epoch": 1.567,
+      "grad_norm": 30.881059646606445,
+      "learning_rate": 6.868600000000001e-06,
+      "loss": 0.7772,
+      "step": 15670
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 31.458768844604492,
+      "learning_rate": 6.866600000000001e-06,
+      "loss": 0.6283,
+      "step": 15680
+    },
+    {
+      "epoch": 1.569,
+      "grad_norm": 18.553871154785156,
+      "learning_rate": 6.8646e-06,
+      "loss": 0.5786,
+      "step": 15690
+    },
+    {
+      "epoch": 1.5699999999999998,
+      "grad_norm": 4.743809223175049,
+      "learning_rate": 6.8626e-06,
+      "loss": 0.7825,
+      "step": 15700
+    },
+    {
+      "epoch": 1.571,
+      "grad_norm": 46.41438293457031,
+      "learning_rate": 6.860600000000001e-06,
+      "loss": 0.4837,
+      "step": 15710
+    },
+    {
+      "epoch": 1.572,
+      "grad_norm": 17.342342376708984,
+      "learning_rate": 6.858600000000001e-06,
+      "loss": 0.4237,
+      "step": 15720
+    },
+    {
+      "epoch": 1.573,
+      "grad_norm": 37.21664047241211,
+      "learning_rate": 6.856600000000001e-06,
+      "loss": 0.6188,
+      "step": 15730
+    },
+    {
+      "epoch": 1.5739999999999998,
+      "grad_norm": 45.30168914794922,
+      "learning_rate": 6.8546e-06,
+      "loss": 0.9802,
+      "step": 15740
+    },
+    {
+      "epoch": 1.575,
+      "grad_norm": 38.685813903808594,
+      "learning_rate": 6.8526e-06,
+      "loss": 0.7087,
+      "step": 15750
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 23.965330123901367,
+      "learning_rate": 6.850600000000001e-06,
+      "loss": 0.5126,
+      "step": 15760
+    },
+    {
+      "epoch": 1.577,
+      "grad_norm": 43.43236541748047,
+      "learning_rate": 6.848600000000001e-06,
+      "loss": 0.6661,
+      "step": 15770
+    },
+    {
+      "epoch": 1.5779999999999998,
+      "grad_norm": 29.220190048217773,
+      "learning_rate": 6.846600000000001e-06,
+      "loss": 0.7872,
+      "step": 15780
+    },
+    {
+      "epoch": 1.579,
+      "grad_norm": 4.1027021408081055,
+      "learning_rate": 6.8446000000000005e-06,
+      "loss": 0.6425,
+      "step": 15790
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 20.48824119567871,
+      "learning_rate": 6.8426e-06,
+      "loss": 0.8026,
+      "step": 15800
+    },
+    {
+      "epoch": 1.581,
+      "grad_norm": 16.290855407714844,
+      "learning_rate": 6.8406e-06,
+      "loss": 0.8203,
+      "step": 15810
+    },
+    {
+      "epoch": 1.5819999999999999,
+      "grad_norm": 14.963516235351562,
+      "learning_rate": 6.838600000000001e-06,
+      "loss": 0.6832,
+      "step": 15820
+    },
+    {
+      "epoch": 1.583,
+      "grad_norm": 37.97867965698242,
+      "learning_rate": 6.836600000000001e-06,
+      "loss": 0.6167,
+      "step": 15830
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 29.7786865234375,
+      "learning_rate": 6.8346e-06,
+      "loss": 0.8014,
+      "step": 15840
+    },
+    {
+      "epoch": 1.585,
+      "grad_norm": 4.787508487701416,
+      "learning_rate": 6.8326000000000004e-06,
+      "loss": 0.7712,
+      "step": 15850
+    },
+    {
+      "epoch": 1.5859999999999999,
+      "grad_norm": 39.955650329589844,
+      "learning_rate": 6.8306e-06,
+      "loss": 0.5754,
+      "step": 15860
+    },
+    {
+      "epoch": 1.587,
+      "grad_norm": 9.246562957763672,
+      "learning_rate": 6.828600000000001e-06,
+      "loss": 0.7853,
+      "step": 15870
+    },
+    {
+      "epoch": 1.588,
+      "grad_norm": 16.79322052001953,
+      "learning_rate": 6.826600000000001e-06,
+      "loss": 1.0043,
+      "step": 15880
+    },
+    {
+      "epoch": 1.589,
+      "grad_norm": 46.77156448364258,
+      "learning_rate": 6.8246e-06,
+      "loss": 0.5613,
+      "step": 15890
+    },
+    {
+      "epoch": 1.5899999999999999,
+      "grad_norm": 30.43254852294922,
+      "learning_rate": 6.8226000000000005e-06,
+      "loss": 0.5192,
+      "step": 15900
+    },
+    {
+      "epoch": 1.591,
+      "grad_norm": 24.879961013793945,
+      "learning_rate": 6.8206e-06,
+      "loss": 0.6208,
+      "step": 15910
+    },
+    {
+      "epoch": 1.592,
+      "grad_norm": 31.248703002929688,
+      "learning_rate": 6.818600000000001e-06,
+      "loss": 0.5469,
+      "step": 15920
+    },
+    {
+      "epoch": 1.593,
+      "grad_norm": 43.610923767089844,
+      "learning_rate": 6.816600000000001e-06,
+      "loss": 0.8376,
+      "step": 15930
+    },
+    {
+      "epoch": 1.5939999999999999,
+      "grad_norm": 27.251039505004883,
+      "learning_rate": 6.8146e-06,
+      "loss": 0.8027,
+      "step": 15940
+    },
+    {
+      "epoch": 1.595,
+      "grad_norm": 52.440940856933594,
+      "learning_rate": 6.812600000000001e-06,
+      "loss": 0.6246,
+      "step": 15950
+    },
+    {
+      "epoch": 1.596,
+      "grad_norm": 52.91548156738281,
+      "learning_rate": 6.8106000000000005e-06,
+      "loss": 0.6263,
+      "step": 15960
+    },
+    {
+      "epoch": 1.597,
+      "grad_norm": 30.92323112487793,
+      "learning_rate": 6.808600000000001e-06,
+      "loss": 0.6969,
+      "step": 15970
+    },
+    {
+      "epoch": 1.5979999999999999,
+      "grad_norm": 76.39839172363281,
+      "learning_rate": 6.806600000000001e-06,
+      "loss": 1.0139,
+      "step": 15980
+    },
+    {
+      "epoch": 1.599,
+      "grad_norm": 21.792797088623047,
+      "learning_rate": 6.8046e-06,
+      "loss": 0.8526,
+      "step": 15990
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.62412166595459,
+      "learning_rate": 6.802600000000001e-06,
+      "loss": 0.6768,
+      "step": 16000
+    },
+    {
+      "epoch": 1.601,
+      "grad_norm": 26.514087677001953,
+      "learning_rate": 6.800600000000001e-06,
+      "loss": 0.6667,
+      "step": 16010
+    },
+    {
+      "epoch": 1.6019999999999999,
+      "grad_norm": 8.58350658416748,
+      "learning_rate": 6.7986000000000004e-06,
+      "loss": 0.3468,
+      "step": 16020
+    },
+    {
+      "epoch": 1.603,
+      "grad_norm": 6.864509582519531,
+      "learning_rate": 6.796600000000001e-06,
+      "loss": 0.6009,
+      "step": 16030
+    },
+    {
+      "epoch": 1.604,
+      "grad_norm": 38.441192626953125,
+      "learning_rate": 6.7946e-06,
+      "loss": 0.9751,
+      "step": 16040
+    },
+    {
+      "epoch": 1.605,
+      "grad_norm": 14.290736198425293,
+      "learning_rate": 6.7926e-06,
+      "loss": 0.5905,
+      "step": 16050
+    },
+    {
+      "epoch": 1.6059999999999999,
+      "grad_norm": 21.486791610717773,
+      "learning_rate": 6.790600000000001e-06,
+      "loss": 0.7175,
+      "step": 16060
+    },
+    {
+      "epoch": 1.607,
+      "grad_norm": 85.47362518310547,
+      "learning_rate": 6.7886000000000005e-06,
+      "loss": 0.7642,
+      "step": 16070
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 38.84056091308594,
+      "learning_rate": 6.786600000000001e-06,
+      "loss": 0.6609,
+      "step": 16080
+    },
+    {
+      "epoch": 1.609,
+      "grad_norm": 5.32833194732666,
+      "learning_rate": 6.7846e-06,
+      "loss": 0.5081,
+      "step": 16090
+    },
+    {
+      "epoch": 1.6099999999999999,
+      "grad_norm": 23.446685791015625,
+      "learning_rate": 6.7826e-06,
+      "loss": 0.7069,
+      "step": 16100
+    },
+    {
+      "epoch": 1.611,
+      "grad_norm": 65.45818328857422,
+      "learning_rate": 6.780600000000001e-06,
+      "loss": 0.7685,
+      "step": 16110
+    },
+    {
+      "epoch": 1.612,
+      "grad_norm": 20.963775634765625,
+      "learning_rate": 6.778600000000001e-06,
+      "loss": 0.6136,
+      "step": 16120
+    },
+    {
+      "epoch": 1.613,
+      "grad_norm": 52.342918395996094,
+      "learning_rate": 6.776600000000001e-06,
+      "loss": 0.8376,
+      "step": 16130
+    },
+    {
+      "epoch": 1.6139999999999999,
+      "grad_norm": 48.46240234375,
+      "learning_rate": 6.7746e-06,
+      "loss": 0.6365,
+      "step": 16140
+    },
+    {
+      "epoch": 1.615,
+      "grad_norm": 44.245235443115234,
+      "learning_rate": 6.7726e-06,
+      "loss": 0.8167,
+      "step": 16150
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 20.52388572692871,
+      "learning_rate": 6.770600000000001e-06,
+      "loss": 0.5238,
+      "step": 16160
+    },
+    {
+      "epoch": 1.617,
+      "grad_norm": 41.138763427734375,
+      "learning_rate": 6.768600000000001e-06,
+      "loss": 0.7287,
+      "step": 16170
+    },
+    {
+      "epoch": 1.6179999999999999,
+      "grad_norm": 26.4044132232666,
+      "learning_rate": 6.7666000000000006e-06,
+      "loss": 0.638,
+      "step": 16180
+    },
+    {
+      "epoch": 1.619,
+      "grad_norm": 28.66182518005371,
+      "learning_rate": 6.7646e-06,
+      "loss": 0.7564,
+      "step": 16190
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 148.91444396972656,
+      "learning_rate": 6.7626e-06,
+      "loss": 0.5724,
+      "step": 16200
+    },
+    {
+      "epoch": 1.621,
+      "grad_norm": 27.070476531982422,
+      "learning_rate": 6.7606e-06,
+      "loss": 0.5099,
+      "step": 16210
+    },
+    {
+      "epoch": 1.6219999999999999,
+      "grad_norm": 29.128446578979492,
+      "learning_rate": 6.758600000000001e-06,
+      "loss": 0.7376,
+      "step": 16220
+    },
+    {
+      "epoch": 1.623,
+      "grad_norm": 40.75493240356445,
+      "learning_rate": 6.756600000000001e-06,
+      "loss": 0.8102,
+      "step": 16230
+    },
+    {
+      "epoch": 1.624,
+      "grad_norm": 13.366748809814453,
+      "learning_rate": 6.7546e-06,
+      "loss": 0.7062,
+      "step": 16240
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 26.001070022583008,
+      "learning_rate": 6.7526e-06,
+      "loss": 0.7237,
+      "step": 16250
+    },
+    {
+      "epoch": 1.626,
+      "grad_norm": 22.817955017089844,
+      "learning_rate": 6.7506e-06,
+      "loss": 0.9075,
+      "step": 16260
+    },
+    {
+      "epoch": 1.627,
+      "grad_norm": 33.775001525878906,
+      "learning_rate": 6.748600000000001e-06,
+      "loss": 0.4802,
+      "step": 16270
+    },
+    {
+      "epoch": 1.6280000000000001,
+      "grad_norm": 26.016061782836914,
+      "learning_rate": 6.746600000000001e-06,
+      "loss": 0.7084,
+      "step": 16280
+    },
+    {
+      "epoch": 1.629,
+      "grad_norm": 29.101972579956055,
+      "learning_rate": 6.7446e-06,
+      "loss": 0.8943,
+      "step": 16290
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 40.83861541748047,
+      "learning_rate": 6.7426000000000005e-06,
+      "loss": 0.5705,
+      "step": 16300
+    },
+    {
+      "epoch": 1.631,
+      "grad_norm": 46.490352630615234,
+      "learning_rate": 6.7406e-06,
+      "loss": 0.523,
+      "step": 16310
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 2.7267861366271973,
+      "learning_rate": 6.738600000000001e-06,
+      "loss": 0.5519,
+      "step": 16320
+    },
+    {
+      "epoch": 1.633,
+      "grad_norm": 35.80474853515625,
+      "learning_rate": 6.736600000000001e-06,
+      "loss": 0.721,
+      "step": 16330
+    },
+    {
+      "epoch": 1.634,
+      "grad_norm": 27.28360939025879,
+      "learning_rate": 6.7346e-06,
+      "loss": 0.7681,
+      "step": 16340
+    },
+    {
+      "epoch": 1.635,
+      "grad_norm": 27.563488006591797,
+      "learning_rate": 6.7326000000000006e-06,
+      "loss": 0.6124,
+      "step": 16350
+    },
+    {
+      "epoch": 1.6360000000000001,
+      "grad_norm": 42.63063430786133,
+      "learning_rate": 6.7306e-06,
+      "loss": 0.7241,
+      "step": 16360
+    },
+    {
+      "epoch": 1.637,
+      "grad_norm": 21.946151733398438,
+      "learning_rate": 6.728600000000001e-06,
+      "loss": 0.6195,
+      "step": 16370
+    },
+    {
+      "epoch": 1.638,
+      "grad_norm": 49.707210540771484,
+      "learning_rate": 6.726600000000001e-06,
+      "loss": 0.7359,
+      "step": 16380
+    },
+    {
+      "epoch": 1.639,
+      "grad_norm": 66.19535827636719,
+      "learning_rate": 6.7246e-06,
+      "loss": 0.3962,
+      "step": 16390
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 26.122718811035156,
+      "learning_rate": 6.722600000000001e-06,
+      "loss": 0.7547,
+      "step": 16400
+    },
+    {
+      "epoch": 1.641,
+      "grad_norm": 13.014486312866211,
+      "learning_rate": 6.7206000000000005e-06,
+      "loss": 0.5428,
+      "step": 16410
+    },
+    {
+      "epoch": 1.642,
+      "grad_norm": 48.25709915161133,
+      "learning_rate": 6.7186e-06,
+      "loss": 0.551,
+      "step": 16420
+    },
+    {
+      "epoch": 1.643,
+      "grad_norm": 20.289514541625977,
+      "learning_rate": 6.716600000000001e-06,
+      "loss": 0.4724,
+      "step": 16430
+    },
+    {
+      "epoch": 1.6440000000000001,
+      "grad_norm": 33.06206130981445,
+      "learning_rate": 6.7146e-06,
+      "loss": 0.8169,
+      "step": 16440
+    },
+    {
+      "epoch": 1.645,
+      "grad_norm": 29.4952392578125,
+      "learning_rate": 6.7126e-06,
+      "loss": 0.4904,
+      "step": 16450
+    },
+    {
+      "epoch": 1.646,
+      "grad_norm": 19.41499900817871,
+      "learning_rate": 6.710600000000001e-06,
+      "loss": 0.3842,
+      "step": 16460
+    },
+    {
+      "epoch": 1.647,
+      "grad_norm": 51.508609771728516,
+      "learning_rate": 6.7086000000000004e-06,
+      "loss": 0.74,
+      "step": 16470
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 46.98064422607422,
+      "learning_rate": 6.706600000000001e-06,
+      "loss": 0.8286,
+      "step": 16480
+    },
+    {
+      "epoch": 1.649,
+      "grad_norm": 21.764245986938477,
+      "learning_rate": 6.7046e-06,
+      "loss": 0.9053,
+      "step": 16490
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.565622329711914,
+      "learning_rate": 6.7026e-06,
+      "loss": 0.6417,
+      "step": 16500
+    },
+    {
+      "epoch": 1.651,
+      "grad_norm": 42.838314056396484,
+      "learning_rate": 6.700600000000001e-06,
+      "loss": 0.4717,
+      "step": 16510
+    },
+    {
+      "epoch": 1.6520000000000001,
+      "grad_norm": 33.5409049987793,
+      "learning_rate": 6.6986000000000005e-06,
+      "loss": 0.544,
+      "step": 16520
+    },
+    {
+      "epoch": 1.653,
+      "grad_norm": 27.133085250854492,
+      "learning_rate": 6.696600000000001e-06,
+      "loss": 0.6666,
+      "step": 16530
+    },
+    {
+      "epoch": 1.654,
+      "grad_norm": 25.304351806640625,
+      "learning_rate": 6.694600000000001e-06,
+      "loss": 0.7644,
+      "step": 16540
+    },
+    {
+      "epoch": 1.655,
+      "grad_norm": 49.61137008666992,
+      "learning_rate": 6.6926e-06,
+      "loss": 0.5898,
+      "step": 16550
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 18.069475173950195,
+      "learning_rate": 6.690600000000001e-06,
+      "loss": 0.7118,
+      "step": 16560
+    },
+    {
+      "epoch": 1.657,
+      "grad_norm": 56.22249984741211,
+      "learning_rate": 6.688600000000001e-06,
+      "loss": 0.728,
+      "step": 16570
+    },
+    {
+      "epoch": 1.658,
+      "grad_norm": 26.706298828125,
+      "learning_rate": 6.6866000000000005e-06,
+      "loss": 0.9413,
+      "step": 16580
+    },
+    {
+      "epoch": 1.659,
+      "grad_norm": 34.80197525024414,
+      "learning_rate": 6.684600000000001e-06,
+      "loss": 0.6705,
+      "step": 16590
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 34.63695526123047,
+      "learning_rate": 6.6826e-06,
+      "loss": 0.7658,
+      "step": 16600
+    },
+    {
+      "epoch": 1.661,
+      "grad_norm": 39.2309684753418,
+      "learning_rate": 6.6806e-06,
+      "loss": 0.6774,
+      "step": 16610
+    },
+    {
+      "epoch": 1.662,
+      "grad_norm": 59.59306335449219,
+      "learning_rate": 6.678600000000001e-06,
+      "loss": 0.7997,
+      "step": 16620
+    },
+    {
+      "epoch": 1.663,
+      "grad_norm": 13.926924705505371,
+      "learning_rate": 6.676600000000001e-06,
+      "loss": 0.5904,
+      "step": 16630
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 47.741844177246094,
+      "learning_rate": 6.674600000000001e-06,
+      "loss": 0.6835,
+      "step": 16640
+    },
+    {
+      "epoch": 1.665,
+      "grad_norm": 19.617435455322266,
+      "learning_rate": 6.6726e-06,
+      "loss": 0.7402,
+      "step": 16650
+    },
+    {
+      "epoch": 1.666,
+      "grad_norm": 11.972737312316895,
+      "learning_rate": 6.6706e-06,
+      "loss": 0.6702,
+      "step": 16660
+    },
+    {
+      "epoch": 1.667,
+      "grad_norm": 18.987348556518555,
+      "learning_rate": 6.668600000000001e-06,
+      "loss": 0.4913,
+      "step": 16670
+    },
+    {
+      "epoch": 1.6680000000000001,
+      "grad_norm": 31.206674575805664,
+      "learning_rate": 6.666600000000001e-06,
+      "loss": 0.5087,
+      "step": 16680
+    },
+    {
+      "epoch": 1.669,
+      "grad_norm": 35.57680892944336,
+      "learning_rate": 6.664600000000001e-06,
+      "loss": 0.6203,
+      "step": 16690
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 23.097082138061523,
+      "learning_rate": 6.6626e-06,
+      "loss": 0.6263,
+      "step": 16700
+    },
+    {
+      "epoch": 1.671,
+      "grad_norm": 2.384589195251465,
+      "learning_rate": 6.6606e-06,
+      "loss": 0.7431,
+      "step": 16710
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 38.233177185058594,
+      "learning_rate": 6.658600000000001e-06,
+      "loss": 0.6329,
+      "step": 16720
+    },
+    {
+      "epoch": 1.673,
+      "grad_norm": 47.652183532714844,
+      "learning_rate": 6.656600000000001e-06,
+      "loss": 0.715,
+      "step": 16730
+    },
+    {
+      "epoch": 1.674,
+      "grad_norm": 14.567292213439941,
+      "learning_rate": 6.6546000000000015e-06,
+      "loss": 0.3611,
+      "step": 16740
+    },
+    {
+      "epoch": 1.675,
+      "grad_norm": 31.54159164428711,
+      "learning_rate": 6.6526000000000005e-06,
+      "loss": 0.55,
+      "step": 16750
+    },
+    {
+      "epoch": 1.6760000000000002,
+      "grad_norm": 47.42003631591797,
+      "learning_rate": 6.6506e-06,
+      "loss": 0.7384,
+      "step": 16760
+    },
+    {
+      "epoch": 1.677,
+      "grad_norm": 25.71277618408203,
+      "learning_rate": 6.648600000000001e-06,
+      "loss": 0.4905,
+      "step": 16770
+    },
+    {
+      "epoch": 1.678,
+      "grad_norm": 39.28430938720703,
+      "learning_rate": 6.646600000000001e-06,
+      "loss": 0.7597,
+      "step": 16780
+    },
+    {
+      "epoch": 1.679,
+      "grad_norm": 29.059282302856445,
+      "learning_rate": 6.644600000000001e-06,
+      "loss": 0.8572,
+      "step": 16790
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 45.53642654418945,
+      "learning_rate": 6.6426000000000006e-06,
+      "loss": 0.6979,
+      "step": 16800
+    },
+    {
+      "epoch": 1.681,
+      "grad_norm": 34.41741943359375,
+      "learning_rate": 6.6406e-06,
+      "loss": 0.5413,
+      "step": 16810
+    },
+    {
+      "epoch": 1.682,
+      "grad_norm": 24.35959815979004,
+      "learning_rate": 6.6386e-06,
+      "loss": 0.6513,
+      "step": 16820
+    },
+    {
+      "epoch": 1.683,
+      "grad_norm": 37.356407165527344,
+      "learning_rate": 6.636600000000001e-06,
+      "loss": 0.6982,
+      "step": 16830
+    },
+    {
+      "epoch": 1.6840000000000002,
+      "grad_norm": 41.323490142822266,
+      "learning_rate": 6.634600000000001e-06,
+      "loss": 0.8573,
+      "step": 16840
+    },
+    {
+      "epoch": 1.685,
+      "grad_norm": 25.554746627807617,
+      "learning_rate": 6.6326e-06,
+      "loss": 0.43,
+      "step": 16850
+    },
+    {
+      "epoch": 1.686,
+      "grad_norm": 18.734989166259766,
+      "learning_rate": 6.6306000000000005e-06,
+      "loss": 0.6716,
+      "step": 16860
+    },
+    {
+      "epoch": 1.687,
+      "grad_norm": 20.203542709350586,
+      "learning_rate": 6.6286e-06,
+      "loss": 0.5459,
+      "step": 16870
+    },
+    {
+      "epoch": 1.688,
+      "grad_norm": 35.660457611083984,
+      "learning_rate": 6.626600000000001e-06,
+      "loss": 0.7359,
+      "step": 16880
+    },
+    {
+      "epoch": 1.689,
+      "grad_norm": 7.850064277648926,
+      "learning_rate": 6.624600000000001e-06,
+      "loss": 0.5398,
+      "step": 16890
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.7819433212280273,
+      "learning_rate": 6.6226e-06,
+      "loss": 0.4942,
+      "step": 16900
+    },
+    {
+      "epoch": 1.6909999999999998,
+      "grad_norm": 39.97076416015625,
+      "learning_rate": 6.620600000000001e-06,
+      "loss": 1.0417,
+      "step": 16910
+    },
+    {
+      "epoch": 1.692,
+      "grad_norm": 38.29095458984375,
+      "learning_rate": 6.6186000000000005e-06,
+      "loss": 0.8314,
+      "step": 16920
+    },
+    {
+      "epoch": 1.693,
+      "grad_norm": 30.200014114379883,
+      "learning_rate": 6.616600000000001e-06,
+      "loss": 0.8102,
+      "step": 16930
+    },
+    {
+      "epoch": 1.694,
+      "grad_norm": 15.969015121459961,
+      "learning_rate": 6.614600000000001e-06,
+      "loss": 0.2612,
+      "step": 16940
+    },
+    {
+      "epoch": 1.6949999999999998,
+      "grad_norm": 23.59886932373047,
+      "learning_rate": 6.6126e-06,
+      "loss": 0.6885,
+      "step": 16950
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 1.8917691707611084,
+      "learning_rate": 6.610600000000001e-06,
+      "loss": 0.6052,
+      "step": 16960
+    },
+    {
+      "epoch": 1.697,
+      "grad_norm": 62.720245361328125,
+      "learning_rate": 6.6086000000000006e-06,
+      "loss": 0.5954,
+      "step": 16970
+    },
+    {
+      "epoch": 1.698,
+      "grad_norm": 5.497705459594727,
+      "learning_rate": 6.6066e-06,
+      "loss": 0.6399,
+      "step": 16980
+    },
+    {
+      "epoch": 1.6989999999999998,
+      "grad_norm": 16.05719566345215,
+      "learning_rate": 6.604600000000001e-06,
+      "loss": 0.5922,
+      "step": 16990
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.563225507736206,
+      "learning_rate": 6.6026e-06,
+      "loss": 0.4885,
+      "step": 17000
+    },
+    {
+      "epoch": 1.701,
+      "grad_norm": 41.32305145263672,
+      "learning_rate": 6.6006e-06,
+      "loss": 0.8704,
+      "step": 17010
+    },
+    {
+      "epoch": 1.702,
+      "grad_norm": 50.50960159301758,
+      "learning_rate": 6.598600000000001e-06,
+      "loss": 0.6659,
+      "step": 17020
+    },
+    {
+      "epoch": 1.7029999999999998,
+      "grad_norm": 7.299249649047852,
+      "learning_rate": 6.5966000000000005e-06,
+      "loss": 0.6752,
+      "step": 17030
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 26.67925262451172,
+      "learning_rate": 6.594600000000001e-06,
+      "loss": 0.7089,
+      "step": 17040
+    },
+    {
+      "epoch": 1.705,
+      "grad_norm": 43.11434555053711,
+      "learning_rate": 6.5926e-06,
+      "loss": 0.4218,
+      "step": 17050
+    },
+    {
+      "epoch": 1.706,
+      "grad_norm": 30.153173446655273,
+      "learning_rate": 6.5906e-06,
+      "loss": 0.9292,
+      "step": 17060
+    },
+    {
+      "epoch": 1.7069999999999999,
+      "grad_norm": 26.88372230529785,
+      "learning_rate": 6.588600000000001e-06,
+      "loss": 0.9461,
+      "step": 17070
+    },
+    {
+      "epoch": 1.708,
+      "grad_norm": 12.442044258117676,
+      "learning_rate": 6.586600000000001e-06,
+      "loss": 0.4348,
+      "step": 17080
+    },
+    {
+      "epoch": 1.709,
+      "grad_norm": 42.74766540527344,
+      "learning_rate": 6.584600000000001e-06,
+      "loss": 0.6239,
+      "step": 17090
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 30.965904235839844,
+      "learning_rate": 6.5826e-06,
+      "loss": 0.9168,
+      "step": 17100
+    },
+    {
+      "epoch": 1.7109999999999999,
+      "grad_norm": 27.245014190673828,
+      "learning_rate": 6.5806e-06,
+      "loss": 0.6842,
+      "step": 17110
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 17.746694564819336,
+      "learning_rate": 6.578600000000001e-06,
+      "loss": 0.8224,
+      "step": 17120
+    },
+    {
+      "epoch": 1.713,
+      "grad_norm": 11.387202262878418,
+      "learning_rate": 6.576600000000001e-06,
+      "loss": 0.5531,
+      "step": 17130
+    },
+    {
+      "epoch": 1.714,
+      "grad_norm": 30.879100799560547,
+      "learning_rate": 6.574600000000001e-06,
+      "loss": 0.6209,
+      "step": 17140
+    },
+    {
+      "epoch": 1.7149999999999999,
+      "grad_norm": 9.201401710510254,
+      "learning_rate": 6.5726e-06,
+      "loss": 0.6358,
+      "step": 17150
+    },
+    {
+      "epoch": 1.716,
+      "grad_norm": 32.12496566772461,
+      "learning_rate": 6.5706e-06,
+      "loss": 0.6251,
+      "step": 17160
+    },
+    {
+      "epoch": 1.717,
+      "grad_norm": 7.410712242126465,
+      "learning_rate": 6.568600000000001e-06,
+      "loss": 0.6895,
+      "step": 17170
+    },
+    {
+      "epoch": 1.718,
+      "grad_norm": 40.6132926940918,
+      "learning_rate": 6.566600000000001e-06,
+      "loss": 0.681,
+      "step": 17180
+    },
+    {
+      "epoch": 1.7189999999999999,
+      "grad_norm": 18.442171096801758,
+      "learning_rate": 6.564600000000001e-06,
+      "loss": 0.5489,
+      "step": 17190
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.694601058959961,
+      "learning_rate": 6.5626000000000005e-06,
+      "loss": 0.6267,
+      "step": 17200
+    },
+    {
+      "epoch": 1.721,
+      "grad_norm": 17.21006202697754,
+      "learning_rate": 6.5606e-06,
+      "loss": 1.1118,
+      "step": 17210
+    },
+    {
+      "epoch": 1.722,
+      "grad_norm": 14.537155151367188,
+      "learning_rate": 6.5586e-06,
+      "loss": 0.8024,
+      "step": 17220
+    },
+    {
+      "epoch": 1.7229999999999999,
+      "grad_norm": 32.91739273071289,
+      "learning_rate": 6.556600000000001e-06,
+      "loss": 0.7176,
+      "step": 17230
+    },
+    {
+      "epoch": 1.724,
+      "grad_norm": 21.398927688598633,
+      "learning_rate": 6.554600000000001e-06,
+      "loss": 0.6832,
+      "step": 17240
+    },
+    {
+      "epoch": 1.725,
+      "grad_norm": 37.23515701293945,
+      "learning_rate": 6.5526e-06,
+      "loss": 0.6286,
+      "step": 17250
+    },
+    {
+      "epoch": 1.726,
+      "grad_norm": 30.678726196289062,
+      "learning_rate": 6.5506000000000004e-06,
+      "loss": 0.6981,
+      "step": 17260
+    },
+    {
+      "epoch": 1.7269999999999999,
+      "grad_norm": 28.972732543945312,
+      "learning_rate": 6.5486e-06,
+      "loss": 0.6119,
+      "step": 17270
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 37.57304000854492,
+      "learning_rate": 6.546600000000001e-06,
+      "loss": 0.7778,
+      "step": 17280
+    },
+    {
+      "epoch": 1.729,
+      "grad_norm": 16.991241455078125,
+      "learning_rate": 6.544600000000001e-06,
+      "loss": 0.6761,
+      "step": 17290
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 24.713104248046875,
+      "learning_rate": 6.5426e-06,
+      "loss": 0.8166,
+      "step": 17300
+    },
+    {
+      "epoch": 1.7309999999999999,
+      "grad_norm": 18.130779266357422,
+      "learning_rate": 6.5406000000000005e-06,
+      "loss": 0.5623,
+      "step": 17310
+    },
+    {
+      "epoch": 1.732,
+      "grad_norm": 6.897360324859619,
+      "learning_rate": 6.5386e-06,
+      "loss": 0.4677,
+      "step": 17320
+    },
+    {
+      "epoch": 1.733,
+      "grad_norm": 15.79992961883545,
+      "learning_rate": 6.536600000000001e-06,
+      "loss": 0.6267,
+      "step": 17330
+    },
+    {
+      "epoch": 1.734,
+      "grad_norm": 27.68032455444336,
+      "learning_rate": 6.534600000000001e-06,
+      "loss": 0.6454,
+      "step": 17340
+    },
+    {
+      "epoch": 1.7349999999999999,
+      "grad_norm": 28.13858413696289,
+      "learning_rate": 6.5326e-06,
+      "loss": 0.7075,
+      "step": 17350
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 7.756270408630371,
+      "learning_rate": 6.530600000000001e-06,
+      "loss": 0.5348,
+      "step": 17360
+    },
+    {
+      "epoch": 1.737,
+      "grad_norm": 35.861351013183594,
+      "learning_rate": 6.5286000000000005e-06,
+      "loss": 0.8423,
+      "step": 17370
+    },
+    {
+      "epoch": 1.738,
+      "grad_norm": 14.177804946899414,
+      "learning_rate": 6.5266e-06,
+      "loss": 0.5516,
+      "step": 17380
+    },
+    {
+      "epoch": 1.7389999999999999,
+      "grad_norm": 23.096698760986328,
+      "learning_rate": 6.524600000000001e-06,
+      "loss": 0.8236,
+      "step": 17390
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 38.693511962890625,
+      "learning_rate": 6.5226e-06,
+      "loss": 0.7141,
+      "step": 17400
+    },
+    {
+      "epoch": 1.741,
+      "grad_norm": 25.841833114624023,
+      "learning_rate": 6.5206e-06,
+      "loss": 0.7845,
+      "step": 17410
+    },
+    {
+      "epoch": 1.742,
+      "grad_norm": 24.260513305664062,
+      "learning_rate": 6.5186000000000006e-06,
+      "loss": 0.6659,
+      "step": 17420
+    },
+    {
+      "epoch": 1.7429999999999999,
+      "grad_norm": 24.908824920654297,
+      "learning_rate": 6.5166000000000004e-06,
+      "loss": 0.5314,
+      "step": 17430
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 25.634967803955078,
+      "learning_rate": 6.514600000000001e-06,
+      "loss": 0.7284,
+      "step": 17440
+    },
+    {
+      "epoch": 1.745,
+      "grad_norm": 15.970552444458008,
+      "learning_rate": 6.5126e-06,
+      "loss": 0.9723,
+      "step": 17450
+    },
+    {
+      "epoch": 1.746,
+      "grad_norm": 28.712669372558594,
+      "learning_rate": 6.5106e-06,
+      "loss": 0.4265,
+      "step": 17460
+    },
+    {
+      "epoch": 1.7469999999999999,
+      "grad_norm": 27.592082977294922,
+      "learning_rate": 6.508600000000001e-06,
+      "loss": 0.5927,
+      "step": 17470
+    },
+    {
+      "epoch": 1.748,
+      "grad_norm": 42.75637435913086,
+      "learning_rate": 6.5066000000000005e-06,
+      "loss": 0.7495,
+      "step": 17480
+    },
+    {
+      "epoch": 1.749,
+      "grad_norm": 39.701114654541016,
+      "learning_rate": 6.504600000000001e-06,
+      "loss": 0.5361,
+      "step": 17490
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 19.57805633544922,
+      "learning_rate": 6.5026e-06,
+      "loss": 0.9434,
+      "step": 17500
+    },
+    {
+      "epoch": 1.751,
+      "grad_norm": 24.80259132385254,
+      "learning_rate": 6.5006e-06,
+      "loss": 0.533,
+      "step": 17510
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 30.908178329467773,
+      "learning_rate": 6.498600000000001e-06,
+      "loss": 0.9068,
+      "step": 17520
+    },
+    {
+      "epoch": 1.7530000000000001,
+      "grad_norm": 48.78810119628906,
+      "learning_rate": 6.496600000000001e-06,
+      "loss": 0.8086,
+      "step": 17530
+    },
+    {
+      "epoch": 1.754,
+      "grad_norm": 36.677059173583984,
+      "learning_rate": 6.494600000000001e-06,
+      "loss": 0.7572,
+      "step": 17540
+    },
+    {
+      "epoch": 1.755,
+      "grad_norm": 22.354658126831055,
+      "learning_rate": 6.4926e-06,
+      "loss": 0.6053,
+      "step": 17550
+    },
+    {
+      "epoch": 1.756,
+      "grad_norm": 32.97988510131836,
+      "learning_rate": 6.4906e-06,
+      "loss": 0.5435,
+      "step": 17560
+    },
+    {
+      "epoch": 1.7570000000000001,
+      "grad_norm": 18.22273063659668,
+      "learning_rate": 6.488600000000001e-06,
+      "loss": 0.456,
+      "step": 17570
+    },
+    {
+      "epoch": 1.758,
+      "grad_norm": 24.372108459472656,
+      "learning_rate": 6.486600000000001e-06,
+      "loss": 0.5838,
+      "step": 17580
+    },
+    {
+      "epoch": 1.759,
+      "grad_norm": 12.150777816772461,
+      "learning_rate": 6.4846000000000006e-06,
+      "loss": 0.4555,
+      "step": 17590
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.295907497406006,
+      "learning_rate": 6.4826e-06,
+      "loss": 0.4124,
+      "step": 17600
+    },
+    {
+      "epoch": 1.7610000000000001,
+      "grad_norm": 28.266719818115234,
+      "learning_rate": 6.4806e-06,
+      "loss": 0.7032,
+      "step": 17610
+    },
+    {
+      "epoch": 1.762,
+      "grad_norm": 51.440616607666016,
+      "learning_rate": 6.4786e-06,
+      "loss": 0.6094,
+      "step": 17620
+    },
+    {
+      "epoch": 1.763,
+      "grad_norm": 16.108505249023438,
+      "learning_rate": 6.476600000000001e-06,
+      "loss": 0.5252,
+      "step": 17630
+    },
+    {
+      "epoch": 1.764,
+      "grad_norm": 42.28464889526367,
+      "learning_rate": 6.474600000000001e-06,
+      "loss": 0.7524,
+      "step": 17640
+    },
+    {
+      "epoch": 1.7650000000000001,
+      "grad_norm": 10.865596771240234,
+      "learning_rate": 6.4726e-06,
+      "loss": 0.6395,
+      "step": 17650
+    },
+    {
+      "epoch": 1.766,
+      "grad_norm": 22.09259033203125,
+      "learning_rate": 6.4706e-06,
+      "loss": 0.472,
+      "step": 17660
+    },
+    {
+      "epoch": 1.767,
+      "grad_norm": 10.556814193725586,
+      "learning_rate": 6.4686e-06,
+      "loss": 0.3279,
+      "step": 17670
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 69.1151351928711,
+      "learning_rate": 6.466600000000001e-06,
+      "loss": 0.8119,
+      "step": 17680
+    },
+    {
+      "epoch": 1.7690000000000001,
+      "grad_norm": 49.826446533203125,
+      "learning_rate": 6.464600000000001e-06,
+      "loss": 0.7345,
+      "step": 17690
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 37.7425537109375,
+      "learning_rate": 6.4626e-06,
+      "loss": 0.7113,
+      "step": 17700
+    },
+    {
+      "epoch": 1.771,
+      "grad_norm": 25.57686996459961,
+      "learning_rate": 6.4606000000000005e-06,
+      "loss": 0.7251,
+      "step": 17710
+    },
+    {
+      "epoch": 1.772,
+      "grad_norm": 18.124256134033203,
+      "learning_rate": 6.4586e-06,
+      "loss": 0.797,
+      "step": 17720
+    },
+    {
+      "epoch": 1.7730000000000001,
+      "grad_norm": 65.94688415527344,
+      "learning_rate": 6.456600000000001e-06,
+      "loss": 0.6839,
+      "step": 17730
+    },
+    {
+      "epoch": 1.774,
+      "grad_norm": 30.200220108032227,
+      "learning_rate": 6.454600000000001e-06,
+      "loss": 0.4463,
+      "step": 17740
+    },
+    {
+      "epoch": 1.775,
+      "grad_norm": 23.35943031311035,
+      "learning_rate": 6.4526e-06,
+      "loss": 0.5262,
+      "step": 17750
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 41.625030517578125,
+      "learning_rate": 6.4506000000000005e-06,
+      "loss": 0.5537,
+      "step": 17760
+    },
+    {
+      "epoch": 1.7770000000000001,
+      "grad_norm": 31.092355728149414,
+      "learning_rate": 6.4486e-06,
+      "loss": 0.6805,
+      "step": 17770
+    },
+    {
+      "epoch": 1.778,
+      "grad_norm": 6.774554252624512,
+      "learning_rate": 6.4466e-06,
+      "loss": 0.6299,
+      "step": 17780
+    },
+    {
+      "epoch": 1.779,
+      "grad_norm": 22.500181198120117,
+      "learning_rate": 6.444600000000001e-06,
+      "loss": 0.5055,
+      "step": 17790
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 40.5450553894043,
+      "learning_rate": 6.442600000000001e-06,
+      "loss": 0.5327,
+      "step": 17800
+    },
+    {
+      "epoch": 1.7810000000000001,
+      "grad_norm": 16.847360610961914,
+      "learning_rate": 6.4406e-06,
+      "loss": 0.6484,
+      "step": 17810
+    },
+    {
+      "epoch": 1.782,
+      "grad_norm": 25.066572189331055,
+      "learning_rate": 6.4386000000000005e-06,
+      "loss": 0.4881,
+      "step": 17820
+    },
+    {
+      "epoch": 1.783,
+      "grad_norm": 31.86751365661621,
+      "learning_rate": 6.4366e-06,
+      "loss": 0.4612,
+      "step": 17830
+    },
+    {
+      "epoch": 1.784,
+      "grad_norm": 27.578292846679688,
+      "learning_rate": 6.434600000000001e-06,
+      "loss": 0.8895,
+      "step": 17840
+    },
+    {
+      "epoch": 1.7850000000000001,
+      "grad_norm": 29.506324768066406,
+      "learning_rate": 6.432600000000001e-06,
+      "loss": 0.7933,
+      "step": 17850
+    },
+    {
+      "epoch": 1.786,
+      "grad_norm": 4.340175628662109,
+      "learning_rate": 6.4306e-06,
+      "loss": 0.5756,
+      "step": 17860
+    },
+    {
+      "epoch": 1.787,
+      "grad_norm": 20.72159767150879,
+      "learning_rate": 6.428600000000001e-06,
+      "loss": 0.695,
+      "step": 17870
+    },
+    {
+      "epoch": 1.788,
+      "grad_norm": 22.929832458496094,
+      "learning_rate": 6.4266000000000004e-06,
+      "loss": 0.4567,
+      "step": 17880
+    },
+    {
+      "epoch": 1.7890000000000001,
+      "grad_norm": 13.51591968536377,
+      "learning_rate": 6.424600000000001e-06,
+      "loss": 0.6086,
+      "step": 17890
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 48.332122802734375,
+      "learning_rate": 6.422600000000001e-06,
+      "loss": 0.6551,
+      "step": 17900
+    },
+    {
+      "epoch": 1.791,
+      "grad_norm": 22.90260124206543,
+      "learning_rate": 6.4206e-06,
+      "loss": 0.6961,
+      "step": 17910
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 16.833646774291992,
+      "learning_rate": 6.418600000000001e-06,
+      "loss": 0.4337,
+      "step": 17920
+    },
+    {
+      "epoch": 1.7930000000000001,
+      "grad_norm": 35.251670837402344,
+      "learning_rate": 6.4166000000000005e-06,
+      "loss": 0.7702,
+      "step": 17930
+    },
+    {
+      "epoch": 1.794,
+      "grad_norm": 34.65640640258789,
+      "learning_rate": 6.414600000000001e-06,
+      "loss": 0.7296,
+      "step": 17940
+    },
+    {
+      "epoch": 1.795,
+      "grad_norm": 53.794921875,
+      "learning_rate": 6.412600000000001e-06,
+      "loss": 0.7949,
+      "step": 17950
+    },
+    {
+      "epoch": 1.796,
+      "grad_norm": 36.07053756713867,
+      "learning_rate": 6.4106e-06,
+      "loss": 0.7731,
+      "step": 17960
+    },
+    {
+      "epoch": 1.7970000000000002,
+      "grad_norm": 50.78809356689453,
+      "learning_rate": 6.408600000000001e-06,
+      "loss": 0.5625,
+      "step": 17970
+    },
+    {
+      "epoch": 1.798,
+      "grad_norm": 28.448030471801758,
+      "learning_rate": 6.406600000000001e-06,
+      "loss": 0.555,
+      "step": 17980
+    },
+    {
+      "epoch": 1.799,
+      "grad_norm": 22.320802688598633,
+      "learning_rate": 6.4046000000000005e-06,
+      "loss": 0.6709,
+      "step": 17990
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.228553771972656,
+      "learning_rate": 6.402600000000001e-06,
+      "loss": 0.6453,
+      "step": 18000
+    },
+    {
+      "epoch": 1.8010000000000002,
+      "grad_norm": 36.92367172241211,
+      "learning_rate": 6.4006e-06,
+      "loss": 0.9075,
+      "step": 18010
+    },
+    {
+      "epoch": 1.802,
+      "grad_norm": 30.44162368774414,
+      "learning_rate": 6.3986e-06,
+      "loss": 0.6903,
+      "step": 18020
+    },
+    {
+      "epoch": 1.803,
+      "grad_norm": 39.92036056518555,
+      "learning_rate": 6.396600000000001e-06,
+      "loss": 0.8592,
+      "step": 18030
+    },
+    {
+      "epoch": 1.804,
+      "grad_norm": 44.56736373901367,
+      "learning_rate": 6.3946000000000006e-06,
+      "loss": 0.4859,
+      "step": 18040
+    },
+    {
+      "epoch": 1.8050000000000002,
+      "grad_norm": 25.35263442993164,
+      "learning_rate": 6.392600000000001e-06,
+      "loss": 0.6489,
+      "step": 18050
+    },
+    {
+      "epoch": 1.806,
+      "grad_norm": 73.92831420898438,
+      "learning_rate": 6.3906e-06,
+      "loss": 0.7088,
+      "step": 18060
+    },
+    {
+      "epoch": 1.807,
+      "grad_norm": 75.07518768310547,
+      "learning_rate": 6.3886e-06,
+      "loss": 0.7994,
+      "step": 18070
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 35.053138732910156,
+      "learning_rate": 6.386600000000001e-06,
+      "loss": 1.0469,
+      "step": 18080
+    },
+    {
+      "epoch": 1.8090000000000002,
+      "grad_norm": 21.79502296447754,
+      "learning_rate": 6.384600000000001e-06,
+      "loss": 0.7169,
+      "step": 18090
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.465487003326416,
+      "learning_rate": 6.382600000000001e-06,
+      "loss": 0.9225,
+      "step": 18100
+    },
+    {
+      "epoch": 1.811,
+      "grad_norm": 45.01616287231445,
+      "learning_rate": 6.3806e-06,
+      "loss": 0.5866,
+      "step": 18110
+    },
+    {
+      "epoch": 1.812,
+      "grad_norm": 59.63362503051758,
+      "learning_rate": 6.3786e-06,
+      "loss": 0.8389,
+      "step": 18120
+    },
+    {
+      "epoch": 1.813,
+      "grad_norm": 5.388458728790283,
+      "learning_rate": 6.376600000000001e-06,
+      "loss": 0.4796,
+      "step": 18130
+    },
+    {
+      "epoch": 1.814,
+      "grad_norm": 2.917116165161133,
+      "learning_rate": 6.374600000000001e-06,
+      "loss": 0.4393,
+      "step": 18140
+    },
+    {
+      "epoch": 1.815,
+      "grad_norm": 18.18976593017578,
+      "learning_rate": 6.372600000000001e-06,
+      "loss": 0.8059,
+      "step": 18150
+    },
+    {
+      "epoch": 1.8159999999999998,
+      "grad_norm": 30.708412170410156,
+      "learning_rate": 6.3706000000000005e-06,
+      "loss": 0.6153,
+      "step": 18160
+    },
+    {
+      "epoch": 1.817,
+      "grad_norm": 29.721654891967773,
+      "learning_rate": 6.3686e-06,
+      "loss": 0.5348,
+      "step": 18170
+    },
+    {
+      "epoch": 1.818,
+      "grad_norm": 29.426971435546875,
+      "learning_rate": 6.3666e-06,
+      "loss": 0.564,
+      "step": 18180
+    },
+    {
+      "epoch": 1.819,
+      "grad_norm": 10.837160110473633,
+      "learning_rate": 6.364600000000001e-06,
+      "loss": 0.5361,
+      "step": 18190
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 51.735557556152344,
+      "learning_rate": 6.362600000000001e-06,
+      "loss": 0.6383,
+      "step": 18200
+    },
+    {
+      "epoch": 1.821,
+      "grad_norm": 28.849987030029297,
+      "learning_rate": 6.3606e-06,
+      "loss": 0.5701,
+      "step": 18210
+    },
+    {
+      "epoch": 1.822,
+      "grad_norm": 0.3885299265384674,
+      "learning_rate": 6.3586e-06,
+      "loss": 0.2929,
+      "step": 18220
+    },
+    {
+      "epoch": 1.823,
+      "grad_norm": 41.65828323364258,
+      "learning_rate": 6.3566e-06,
+      "loss": 0.8778,
+      "step": 18230
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 46.33951187133789,
+      "learning_rate": 6.354600000000001e-06,
+      "loss": 0.9669,
+      "step": 18240
+    },
+    {
+      "epoch": 1.825,
+      "grad_norm": 116.5798568725586,
+      "learning_rate": 6.352600000000001e-06,
+      "loss": 0.5901,
+      "step": 18250
+    },
+    {
+      "epoch": 1.826,
+      "grad_norm": 11.925887107849121,
+      "learning_rate": 6.3506e-06,
+      "loss": 0.5316,
+      "step": 18260
+    },
+    {
+      "epoch": 1.827,
+      "grad_norm": 17.010337829589844,
+      "learning_rate": 6.3486000000000005e-06,
+      "loss": 0.7052,
+      "step": 18270
+    },
+    {
+      "epoch": 1.8279999999999998,
+      "grad_norm": 15.04338264465332,
+      "learning_rate": 6.3466e-06,
+      "loss": 0.5075,
+      "step": 18280
+    },
+    {
+      "epoch": 1.829,
+      "grad_norm": 53.3408203125,
+      "learning_rate": 6.344600000000001e-06,
+      "loss": 0.5449,
+      "step": 18290
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.94815444946289,
+      "learning_rate": 6.342600000000001e-06,
+      "loss": 0.6648,
+      "step": 18300
+    },
+    {
+      "epoch": 1.831,
+      "grad_norm": 12.17188549041748,
+      "learning_rate": 6.3406e-06,
+      "loss": 0.6536,
+      "step": 18310
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 39.76250457763672,
+      "learning_rate": 6.338600000000001e-06,
+      "loss": 0.7032,
+      "step": 18320
+    },
+    {
+      "epoch": 1.833,
+      "grad_norm": 40.960411071777344,
+      "learning_rate": 6.3366000000000005e-06,
+      "loss": 0.6843,
+      "step": 18330
+    },
+    {
+      "epoch": 1.834,
+      "grad_norm": 57.23612976074219,
+      "learning_rate": 6.334600000000001e-06,
+      "loss": 0.6186,
+      "step": 18340
+    },
+    {
+      "epoch": 1.835,
+      "grad_norm": 32.05949020385742,
+      "learning_rate": 6.332600000000001e-06,
+      "loss": 0.5321,
+      "step": 18350
+    },
+    {
+      "epoch": 1.8359999999999999,
+      "grad_norm": 27.393653869628906,
+      "learning_rate": 6.3306e-06,
+      "loss": 0.8525,
+      "step": 18360
+    },
+    {
+      "epoch": 1.837,
+      "grad_norm": 17.125869750976562,
+      "learning_rate": 6.328600000000001e-06,
+      "loss": 0.4895,
+      "step": 18370
+    },
+    {
+      "epoch": 1.838,
+      "grad_norm": 54.76142883300781,
+      "learning_rate": 6.3266000000000005e-06,
+      "loss": 0.7369,
+      "step": 18380
+    },
+    {
+      "epoch": 1.839,
+      "grad_norm": 41.17182159423828,
+      "learning_rate": 6.3246e-06,
+      "loss": 0.7504,
+      "step": 18390
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 18.484394073486328,
+      "learning_rate": 6.322600000000001e-06,
+      "loss": 0.7947,
+      "step": 18400
+    },
+    {
+      "epoch": 1.841,
+      "grad_norm": 10.778761863708496,
+      "learning_rate": 6.3206e-06,
+      "loss": 0.5847,
+      "step": 18410
+    },
+    {
+      "epoch": 1.842,
+      "grad_norm": 36.59711837768555,
+      "learning_rate": 6.3186e-06,
+      "loss": 0.5724,
+      "step": 18420
+    },
+    {
+      "epoch": 1.843,
+      "grad_norm": 21.72235870361328,
+      "learning_rate": 6.316600000000001e-06,
+      "loss": 0.5684,
+      "step": 18430
+    },
+    {
+      "epoch": 1.8439999999999999,
+      "grad_norm": 28.071380615234375,
+      "learning_rate": 6.3146000000000005e-06,
+      "loss": 0.6841,
+      "step": 18440
+    },
+    {
+      "epoch": 1.845,
+      "grad_norm": 15.306550025939941,
+      "learning_rate": 6.312600000000001e-06,
+      "loss": 0.5565,
+      "step": 18450
+    },
+    {
+      "epoch": 1.846,
+      "grad_norm": 49.12186813354492,
+      "learning_rate": 6.3106e-06,
+      "loss": 0.7218,
+      "step": 18460
+    },
+    {
+      "epoch": 1.847,
+      "grad_norm": 27.92928695678711,
+      "learning_rate": 6.3086e-06,
+      "loss": 0.7262,
+      "step": 18470
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 30.815195083618164,
+      "learning_rate": 6.306600000000001e-06,
+      "loss": 0.6921,
+      "step": 18480
+    },
+    {
+      "epoch": 1.849,
+      "grad_norm": 9.238826751708984,
+      "learning_rate": 6.304600000000001e-06,
+      "loss": 0.6324,
+      "step": 18490
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 33.87525177001953,
+      "learning_rate": 6.302600000000001e-06,
+      "loss": 0.9252,
+      "step": 18500
+    },
+    {
+      "epoch": 1.851,
+      "grad_norm": 23.98631477355957,
+      "learning_rate": 6.3006e-06,
+      "loss": 0.6777,
+      "step": 18510
+    },
+    {
+      "epoch": 1.8519999999999999,
+      "grad_norm": 55.89791488647461,
+      "learning_rate": 6.2986e-06,
+      "loss": 0.9501,
+      "step": 18520
+    },
+    {
+      "epoch": 1.853,
+      "grad_norm": 49.16728591918945,
+      "learning_rate": 6.296600000000001e-06,
+      "loss": 0.6926,
+      "step": 18530
+    },
+    {
+      "epoch": 1.854,
+      "grad_norm": 49.215492248535156,
+      "learning_rate": 6.294600000000001e-06,
+      "loss": 0.7731,
+      "step": 18540
+    },
+    {
+      "epoch": 1.855,
+      "grad_norm": 20.07518196105957,
+      "learning_rate": 6.2926000000000005e-06,
+      "loss": 0.6389,
+      "step": 18550
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 11.670626640319824,
+      "learning_rate": 6.2906e-06,
+      "loss": 0.5868,
+      "step": 18560
+    },
+    {
+      "epoch": 1.857,
+      "grad_norm": 37.362640380859375,
+      "learning_rate": 6.2886e-06,
+      "loss": 0.6444,
+      "step": 18570
+    },
+    {
+      "epoch": 1.858,
+      "grad_norm": 34.82246017456055,
+      "learning_rate": 6.2866e-06,
+      "loss": 0.4443,
+      "step": 18580
+    },
+    {
+      "epoch": 1.859,
+      "grad_norm": 39.475189208984375,
+      "learning_rate": 6.284600000000001e-06,
+      "loss": 0.4061,
+      "step": 18590
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 44.0338249206543,
+      "learning_rate": 6.282600000000001e-06,
+      "loss": 0.7411,
+      "step": 18600
+    },
+    {
+      "epoch": 1.861,
+      "grad_norm": 37.729103088378906,
+      "learning_rate": 6.2806e-06,
+      "loss": 0.799,
+      "step": 18610
+    },
+    {
+      "epoch": 1.862,
+      "grad_norm": 12.825973510742188,
+      "learning_rate": 6.2786e-06,
+      "loss": 0.3664,
+      "step": 18620
+    },
+    {
+      "epoch": 1.863,
+      "grad_norm": 30.061628341674805,
+      "learning_rate": 6.2766e-06,
+      "loss": 0.88,
+      "step": 18630
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 6.273592948913574,
+      "learning_rate": 6.274600000000001e-06,
+      "loss": 0.526,
+      "step": 18640
+    },
+    {
+      "epoch": 1.865,
+      "grad_norm": 28.17704200744629,
+      "learning_rate": 6.272600000000001e-06,
+      "loss": 0.7294,
+      "step": 18650
+    },
+    {
+      "epoch": 1.866,
+      "grad_norm": 45.9304084777832,
+      "learning_rate": 6.2706e-06,
+      "loss": 0.6401,
+      "step": 18660
+    },
+    {
+      "epoch": 1.867,
+      "grad_norm": 5.665647506713867,
+      "learning_rate": 6.2686000000000004e-06,
+      "loss": 0.5935,
+      "step": 18670
+    },
+    {
+      "epoch": 1.8679999999999999,
+      "grad_norm": 11.19158935546875,
+      "learning_rate": 6.2666e-06,
+      "loss": 0.4248,
+      "step": 18680
+    },
+    {
+      "epoch": 1.869,
+      "grad_norm": 22.568464279174805,
+      "learning_rate": 6.264600000000001e-06,
+      "loss": 0.6085,
+      "step": 18690
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 58.15055847167969,
+      "learning_rate": 6.262600000000001e-06,
+      "loss": 0.5467,
+      "step": 18700
+    },
+    {
+      "epoch": 1.871,
+      "grad_norm": 28.386558532714844,
+      "learning_rate": 6.2606e-06,
+      "loss": 0.6753,
+      "step": 18710
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 44.6679573059082,
+      "learning_rate": 6.2586000000000005e-06,
+      "loss": 0.9636,
+      "step": 18720
+    },
+    {
+      "epoch": 1.873,
+      "grad_norm": 27.022314071655273,
+      "learning_rate": 6.2566e-06,
+      "loss": 0.6403,
+      "step": 18730
+    },
+    {
+      "epoch": 1.874,
+      "grad_norm": 36.570796966552734,
+      "learning_rate": 6.254600000000001e-06,
+      "loss": 0.7745,
+      "step": 18740
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 34.11301040649414,
+      "learning_rate": 6.252600000000001e-06,
+      "loss": 0.6801,
+      "step": 18750
+    },
+    {
+      "epoch": 1.876,
+      "grad_norm": 66.19660186767578,
+      "learning_rate": 6.2506e-06,
+      "loss": 0.6163,
+      "step": 18760
+    },
+    {
+      "epoch": 1.877,
+      "grad_norm": 9.71645736694336,
+      "learning_rate": 6.248600000000001e-06,
+      "loss": 0.4776,
+      "step": 18770
+    },
+    {
+      "epoch": 1.8780000000000001,
+      "grad_norm": 23.817113876342773,
+      "learning_rate": 6.2466000000000005e-06,
+      "loss": 0.587,
+      "step": 18780
+    },
+    {
+      "epoch": 1.879,
+      "grad_norm": 64.90528869628906,
+      "learning_rate": 6.2446e-06,
+      "loss": 0.7563,
+      "step": 18790
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 31.6854190826416,
+      "learning_rate": 6.242600000000001e-06,
+      "loss": 0.6346,
+      "step": 18800
+    },
+    {
+      "epoch": 1.881,
+      "grad_norm": 38.89235305786133,
+      "learning_rate": 6.2406e-06,
+      "loss": 0.7503,
+      "step": 18810
+    },
+    {
+      "epoch": 1.8820000000000001,
+      "grad_norm": 24.65252685546875,
+      "learning_rate": 6.2386e-06,
+      "loss": 0.7953,
+      "step": 18820
+    },
+    {
+      "epoch": 1.883,
+      "grad_norm": 46.37358474731445,
+      "learning_rate": 6.2366000000000006e-06,
+      "loss": 0.7451,
+      "step": 18830
+    },
+    {
+      "epoch": 1.884,
+      "grad_norm": 80.3861083984375,
+      "learning_rate": 6.2346e-06,
+      "loss": 0.8484,
+      "step": 18840
+    },
+    {
+      "epoch": 1.885,
+      "grad_norm": 36.111209869384766,
+      "learning_rate": 6.232600000000001e-06,
+      "loss": 0.7848,
+      "step": 18850
+    },
+    {
+      "epoch": 1.8860000000000001,
+      "grad_norm": 31.149621963500977,
+      "learning_rate": 6.2306e-06,
+      "loss": 0.8536,
+      "step": 18860
+    },
+    {
+      "epoch": 1.887,
+      "grad_norm": 31.306196212768555,
+      "learning_rate": 6.2286e-06,
+      "loss": 0.6689,
+      "step": 18870
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 46.65980911254883,
+      "learning_rate": 6.226600000000001e-06,
+      "loss": 0.656,
+      "step": 18880
+    },
+    {
+      "epoch": 1.889,
+      "grad_norm": 37.753135681152344,
+      "learning_rate": 6.2246000000000005e-06,
+      "loss": 0.7444,
+      "step": 18890
+    },
+    {
+      "epoch": 1.8900000000000001,
+      "grad_norm": 36.70481491088867,
+      "learning_rate": 6.222600000000001e-06,
+      "loss": 0.5093,
+      "step": 18900
+    },
+    {
+      "epoch": 1.891,
+      "grad_norm": 25.8837833404541,
+      "learning_rate": 6.2206e-06,
+      "loss": 0.7465,
+      "step": 18910
+    },
+    {
+      "epoch": 1.892,
+      "grad_norm": 16.100866317749023,
+      "learning_rate": 6.2186e-06,
+      "loss": 0.5148,
+      "step": 18920
+    },
+    {
+      "epoch": 1.893,
+      "grad_norm": 37.50735092163086,
+      "learning_rate": 6.216600000000001e-06,
+      "loss": 0.5519,
+      "step": 18930
+    },
+    {
+      "epoch": 1.8940000000000001,
+      "grad_norm": 37.880340576171875,
+      "learning_rate": 6.214600000000001e-06,
+      "loss": 0.6402,
+      "step": 18940
+    },
+    {
+      "epoch": 1.895,
+      "grad_norm": 19.375473022460938,
+      "learning_rate": 6.2126000000000005e-06,
+      "loss": 0.4894,
+      "step": 18950
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 24.145061492919922,
+      "learning_rate": 6.2106e-06,
+      "loss": 0.766,
+      "step": 18960
+    },
+    {
+      "epoch": 1.897,
+      "grad_norm": 9.419266700744629,
+      "learning_rate": 6.2086e-06,
+      "loss": 0.8279,
+      "step": 18970
+    },
+    {
+      "epoch": 1.8980000000000001,
+      "grad_norm": 19.154048919677734,
+      "learning_rate": 6.2066e-06,
+      "loss": 0.6761,
+      "step": 18980
+    },
+    {
+      "epoch": 1.899,
+      "grad_norm": 33.68185043334961,
+      "learning_rate": 6.204600000000001e-06,
+      "loss": 0.5217,
+      "step": 18990
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 23.5676212310791,
+      "learning_rate": 6.2026000000000006e-06,
+      "loss": 0.619,
+      "step": 19000
+    },
+    {
+      "epoch": 1.901,
+      "grad_norm": 28.72123146057129,
+      "learning_rate": 6.2005999999999996e-06,
+      "loss": 0.8742,
+      "step": 19010
+    },
+    {
+      "epoch": 1.9020000000000001,
+      "grad_norm": 9.565010070800781,
+      "learning_rate": 6.1986e-06,
+      "loss": 0.5942,
+      "step": 19020
+    },
+    {
+      "epoch": 1.903,
+      "grad_norm": 20.975059509277344,
+      "learning_rate": 6.1966e-06,
+      "loss": 0.5888,
+      "step": 19030
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 15.559560775756836,
+      "learning_rate": 6.194600000000001e-06,
+      "loss": 0.5595,
+      "step": 19040
+    },
+    {
+      "epoch": 1.905,
+      "grad_norm": 31.841747283935547,
+      "learning_rate": 6.192600000000001e-06,
+      "loss": 1.067,
+      "step": 19050
+    },
+    {
+      "epoch": 1.9060000000000001,
+      "grad_norm": 7.151304244995117,
+      "learning_rate": 6.190600000000001e-06,
+      "loss": 0.6589,
+      "step": 19060
+    },
+    {
+      "epoch": 1.907,
+      "grad_norm": 42.31881332397461,
+      "learning_rate": 6.1886e-06,
+      "loss": 0.7246,
+      "step": 19070
+    },
+    {
+      "epoch": 1.908,
+      "grad_norm": 10.477072715759277,
+      "learning_rate": 6.1866e-06,
+      "loss": 0.5011,
+      "step": 19080
+    },
+    {
+      "epoch": 1.909,
+      "grad_norm": 24.97276496887207,
+      "learning_rate": 6.184600000000001e-06,
+      "loss": 0.712,
+      "step": 19090
+    },
+    {
+      "epoch": 1.9100000000000001,
+      "grad_norm": 56.96651840209961,
+      "learning_rate": 6.182600000000001e-06,
+      "loss": 0.7178,
+      "step": 19100
+    },
+    {
+      "epoch": 1.911,
+      "grad_norm": 39.4375,
+      "learning_rate": 6.1806000000000014e-06,
+      "loss": 0.666,
+      "step": 19110
+    },
+    {
+      "epoch": 1.912,
+      "grad_norm": 66.98754119873047,
+      "learning_rate": 6.1786000000000004e-06,
+      "loss": 1.0206,
+      "step": 19120
+    },
+    {
+      "epoch": 1.913,
+      "grad_norm": 29.384794235229492,
+      "learning_rate": 6.1766e-06,
+      "loss": 0.8413,
+      "step": 19130
+    },
+    {
+      "epoch": 1.9140000000000001,
+      "grad_norm": 22.646846771240234,
+      "learning_rate": 6.174600000000001e-06,
+      "loss": 0.7687,
+      "step": 19140
+    },
+    {
+      "epoch": 1.915,
+      "grad_norm": 32.5865592956543,
+      "learning_rate": 6.172600000000001e-06,
+      "loss": 0.7559,
+      "step": 19150
+    },
+    {
+      "epoch": 1.916,
+      "grad_norm": 14.892390251159668,
+      "learning_rate": 6.170600000000001e-06,
+      "loss": 0.7594,
+      "step": 19160
+    },
+    {
+      "epoch": 1.917,
+      "grad_norm": 32.36692810058594,
+      "learning_rate": 6.1686000000000005e-06,
+      "loss": 0.6105,
+      "step": 19170
+    },
+    {
+      "epoch": 1.9180000000000001,
+      "grad_norm": 22.858366012573242,
+      "learning_rate": 6.1666e-06,
+      "loss": 0.6479,
+      "step": 19180
+    },
+    {
+      "epoch": 1.919,
+      "grad_norm": 34.3131217956543,
+      "learning_rate": 6.1646e-06,
+      "loss": 0.6657,
+      "step": 19190
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 14.601320266723633,
+      "learning_rate": 6.162600000000001e-06,
+      "loss": 0.6711,
+      "step": 19200
+    },
+    {
+      "epoch": 1.921,
+      "grad_norm": 27.481473922729492,
+      "learning_rate": 6.160600000000001e-06,
+      "loss": 0.6777,
+      "step": 19210
+    },
+    {
+      "epoch": 1.9220000000000002,
+      "grad_norm": 19.15306282043457,
+      "learning_rate": 6.1586e-06,
+      "loss": 0.8164,
+      "step": 19220
+    },
+    {
+      "epoch": 1.923,
+      "grad_norm": 18.720388412475586,
+      "learning_rate": 6.1566000000000005e-06,
+      "loss": 0.5016,
+      "step": 19230
+    },
+    {
+      "epoch": 1.924,
+      "grad_norm": 21.875688552856445,
+      "learning_rate": 6.1546e-06,
+      "loss": 0.6871,
+      "step": 19240
+    },
+    {
+      "epoch": 1.925,
+      "grad_norm": 19.90581512451172,
+      "learning_rate": 6.152600000000001e-06,
+      "loss": 0.8132,
+      "step": 19250
+    },
+    {
+      "epoch": 1.9260000000000002,
+      "grad_norm": 38.14246368408203,
+      "learning_rate": 6.150600000000001e-06,
+      "loss": 0.6122,
+      "step": 19260
+    },
+    {
+      "epoch": 1.927,
+      "grad_norm": 16.996580123901367,
+      "learning_rate": 6.1486e-06,
+      "loss": 0.561,
+      "step": 19270
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 15.28274154663086,
+      "learning_rate": 6.146600000000001e-06,
+      "loss": 0.4495,
+      "step": 19280
+    },
+    {
+      "epoch": 1.929,
+      "grad_norm": 32.38314437866211,
+      "learning_rate": 6.1446000000000004e-06,
+      "loss": 0.6292,
+      "step": 19290
+    },
+    {
+      "epoch": 1.9300000000000002,
+      "grad_norm": 19.988807678222656,
+      "learning_rate": 6.142600000000001e-06,
+      "loss": 0.7398,
+      "step": 19300
+    },
+    {
+      "epoch": 1.931,
+      "grad_norm": 40.12685012817383,
+      "learning_rate": 6.140600000000001e-06,
+      "loss": 0.8673,
+      "step": 19310
+    },
+    {
+      "epoch": 1.932,
+      "grad_norm": 23.791723251342773,
+      "learning_rate": 6.1386e-06,
+      "loss": 0.7318,
+      "step": 19320
+    },
+    {
+      "epoch": 1.933,
+      "grad_norm": 5.818343162536621,
+      "learning_rate": 6.136600000000001e-06,
+      "loss": 0.4892,
+      "step": 19330
+    },
+    {
+      "epoch": 1.9340000000000002,
+      "grad_norm": 32.458740234375,
+      "learning_rate": 6.1346000000000005e-06,
+      "loss": 0.4111,
+      "step": 19340
+    },
+    {
+      "epoch": 1.935,
+      "grad_norm": 23.6914119720459,
+      "learning_rate": 6.1326e-06,
+      "loss": 0.417,
+      "step": 19350
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 39.17317199707031,
+      "learning_rate": 6.130600000000001e-06,
+      "loss": 0.8256,
+      "step": 19360
+    },
+    {
+      "epoch": 1.937,
+      "grad_norm": 4.7436957359313965,
+      "learning_rate": 6.1286e-06,
+      "loss": 0.5982,
+      "step": 19370
+    },
+    {
+      "epoch": 1.938,
+      "grad_norm": 16.462995529174805,
+      "learning_rate": 6.1266e-06,
+      "loss": 0.7216,
+      "step": 19380
+    },
+    {
+      "epoch": 1.939,
+      "grad_norm": 12.548359870910645,
+      "learning_rate": 6.124600000000001e-06,
+      "loss": 0.7207,
+      "step": 19390
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.830649375915527,
+      "learning_rate": 6.1226000000000005e-06,
+      "loss": 0.4089,
+      "step": 19400
+    },
+    {
+      "epoch": 1.9409999999999998,
+      "grad_norm": 12.117684364318848,
+      "learning_rate": 6.120600000000001e-06,
+      "loss": 0.5172,
+      "step": 19410
+    },
+    {
+      "epoch": 1.942,
+      "grad_norm": 25.5642147064209,
+      "learning_rate": 6.1186e-06,
+      "loss": 0.4171,
+      "step": 19420
+    },
+    {
+      "epoch": 1.943,
+      "grad_norm": 42.071712493896484,
+      "learning_rate": 6.1166e-06,
+      "loss": 0.7356,
+      "step": 19430
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 2.329131603240967,
+      "learning_rate": 6.114600000000001e-06,
+      "loss": 0.4786,
+      "step": 19440
+    },
+    {
+      "epoch": 1.9449999999999998,
+      "grad_norm": 41.95789337158203,
+      "learning_rate": 6.1126000000000006e-06,
+      "loss": 0.5854,
+      "step": 19450
+    },
+    {
+      "epoch": 1.946,
+      "grad_norm": 14.594151496887207,
+      "learning_rate": 6.110600000000001e-06,
+      "loss": 0.687,
+      "step": 19460
+    },
+    {
+      "epoch": 1.947,
+      "grad_norm": 74.92787170410156,
+      "learning_rate": 6.1086e-06,
+      "loss": 0.8889,
+      "step": 19470
+    },
+    {
+      "epoch": 1.948,
+      "grad_norm": 42.688167572021484,
+      "learning_rate": 6.1066e-06,
+      "loss": 0.507,
+      "step": 19480
+    },
+    {
+      "epoch": 1.9489999999999998,
+      "grad_norm": 37.9570198059082,
+      "learning_rate": 6.104600000000001e-06,
+      "loss": 0.7797,
+      "step": 19490
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 29.89501953125,
+      "learning_rate": 6.102600000000001e-06,
+      "loss": 0.7377,
+      "step": 19500
+    },
+    {
+      "epoch": 1.951,
+      "grad_norm": 8.201680183410645,
+      "learning_rate": 6.100600000000001e-06,
+      "loss": 0.9134,
+      "step": 19510
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 28.5087890625,
+      "learning_rate": 6.0986e-06,
+      "loss": 0.6407,
+      "step": 19520
+    },
+    {
+      "epoch": 1.9529999999999998,
+      "grad_norm": 13.32079029083252,
+      "learning_rate": 6.0966e-06,
+      "loss": 0.6662,
+      "step": 19530
+    },
+    {
+      "epoch": 1.954,
+      "grad_norm": 16.217845916748047,
+      "learning_rate": 6.094600000000001e-06,
+      "loss": 0.5368,
+      "step": 19540
+    },
+    {
+      "epoch": 1.955,
+      "grad_norm": 25.147905349731445,
+      "learning_rate": 6.092600000000001e-06,
+      "loss": 0.7631,
+      "step": 19550
+    },
+    {
+      "epoch": 1.956,
+      "grad_norm": 36.97098922729492,
+      "learning_rate": 6.090600000000001e-06,
+      "loss": 0.7657,
+      "step": 19560
+    },
+    {
+      "epoch": 1.9569999999999999,
+      "grad_norm": 32.233978271484375,
+      "learning_rate": 6.0886000000000005e-06,
+      "loss": 0.7187,
+      "step": 19570
+    },
+    {
+      "epoch": 1.958,
+      "grad_norm": 22.951942443847656,
+      "learning_rate": 6.0866e-06,
+      "loss": 0.8062,
+      "step": 19580
+    },
+    {
+      "epoch": 1.959,
+      "grad_norm": 23.635204315185547,
+      "learning_rate": 6.0846e-06,
+      "loss": 0.6772,
+      "step": 19590
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 33.38811492919922,
+      "learning_rate": 6.082600000000001e-06,
+      "loss": 0.6484,
+      "step": 19600
+    },
+    {
+      "epoch": 1.9609999999999999,
+      "grad_norm": 22.453102111816406,
+      "learning_rate": 6.080600000000001e-06,
+      "loss": 0.7624,
+      "step": 19610
+    },
+    {
+      "epoch": 1.962,
+      "grad_norm": 50.9228401184082,
+      "learning_rate": 6.0786e-06,
+      "loss": 0.7692,
+      "step": 19620
+    },
+    {
+      "epoch": 1.963,
+      "grad_norm": 12.734217643737793,
+      "learning_rate": 6.0766e-06,
+      "loss": 0.8659,
+      "step": 19630
+    },
+    {
+      "epoch": 1.964,
+      "grad_norm": 22.461109161376953,
+      "learning_rate": 6.0746e-06,
+      "loss": 0.8955,
+      "step": 19640
+    },
+    {
+      "epoch": 1.9649999999999999,
+      "grad_norm": 31.198320388793945,
+      "learning_rate": 6.072600000000001e-06,
+      "loss": 0.6954,
+      "step": 19650
+    },
+    {
+      "epoch": 1.966,
+      "grad_norm": 17.338960647583008,
+      "learning_rate": 6.070600000000001e-06,
+      "loss": 0.6051,
+      "step": 19660
+    },
+    {
+      "epoch": 1.967,
+      "grad_norm": 19.843788146972656,
+      "learning_rate": 6.0686e-06,
+      "loss": 0.6954,
+      "step": 19670
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 28.969507217407227,
+      "learning_rate": 6.0666000000000005e-06,
+      "loss": 0.5193,
+      "step": 19680
+    },
+    {
+      "epoch": 1.9689999999999999,
+      "grad_norm": 13.265837669372559,
+      "learning_rate": 6.0646e-06,
+      "loss": 0.6375,
+      "step": 19690
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.410111427307129,
+      "learning_rate": 6.062600000000001e-06,
+      "loss": 0.7512,
+      "step": 19700
+    },
+    {
+      "epoch": 1.971,
+      "grad_norm": 34.03276062011719,
+      "learning_rate": 6.060600000000001e-06,
+      "loss": 0.6919,
+      "step": 19710
+    },
+    {
+      "epoch": 1.972,
+      "grad_norm": 9.901426315307617,
+      "learning_rate": 6.0586e-06,
+      "loss": 0.4584,
+      "step": 19720
+    },
+    {
+      "epoch": 1.9729999999999999,
+      "grad_norm": 26.24072265625,
+      "learning_rate": 6.056600000000001e-06,
+      "loss": 0.5403,
+      "step": 19730
+    },
+    {
+      "epoch": 1.974,
+      "grad_norm": 38.819496154785156,
+      "learning_rate": 6.0546000000000004e-06,
+      "loss": 0.7342,
+      "step": 19740
+    },
+    {
+      "epoch": 1.975,
+      "grad_norm": 30.7410888671875,
+      "learning_rate": 6.0526e-06,
+      "loss": 0.5699,
+      "step": 19750
+    },
+    {
+      "epoch": 1.976,
+      "grad_norm": 19.44249153137207,
+      "learning_rate": 6.050600000000001e-06,
+      "loss": 0.5331,
+      "step": 19760
+    },
+    {
+      "epoch": 1.9769999999999999,
+      "grad_norm": 5.10331392288208,
+      "learning_rate": 6.0486e-06,
+      "loss": 0.5346,
+      "step": 19770
+    },
+    {
+      "epoch": 1.978,
+      "grad_norm": 21.655014038085938,
+      "learning_rate": 6.0466e-06,
+      "loss": 0.7944,
+      "step": 19780
+    },
+    {
+      "epoch": 1.979,
+      "grad_norm": 29.06139373779297,
+      "learning_rate": 6.0446000000000005e-06,
+      "loss": 0.8133,
+      "step": 19790
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 41.67628479003906,
+      "learning_rate": 6.0426e-06,
+      "loss": 0.6099,
+      "step": 19800
+    },
+    {
+      "epoch": 1.9809999999999999,
+      "grad_norm": 58.43052673339844,
+      "learning_rate": 6.040600000000001e-06,
+      "loss": 0.7782,
+      "step": 19810
+    },
+    {
+      "epoch": 1.982,
+      "grad_norm": 24.17974853515625,
+      "learning_rate": 6.0386e-06,
+      "loss": 0.5837,
+      "step": 19820
+    },
+    {
+      "epoch": 1.983,
+      "grad_norm": 20.998645782470703,
+      "learning_rate": 6.0366e-06,
+      "loss": 0.8658,
+      "step": 19830
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 9.17137336730957,
+      "learning_rate": 6.034600000000001e-06,
+      "loss": 0.5392,
+      "step": 19840
+    },
+    {
+      "epoch": 1.9849999999999999,
+      "grad_norm": 21.920196533203125,
+      "learning_rate": 6.0326000000000005e-06,
+      "loss": 0.6472,
+      "step": 19850
+    },
+    {
+      "epoch": 1.986,
+      "grad_norm": 11.771697044372559,
+      "learning_rate": 6.030600000000001e-06,
+      "loss": 0.6313,
+      "step": 19860
+    },
+    {
+      "epoch": 1.987,
+      "grad_norm": 24.2471981048584,
+      "learning_rate": 6.0286e-06,
+      "loss": 0.6562,
+      "step": 19870
+    },
+    {
+      "epoch": 1.988,
+      "grad_norm": 22.034807205200195,
+      "learning_rate": 6.0266e-06,
+      "loss": 0.6229,
+      "step": 19880
+    },
+    {
+      "epoch": 1.9889999999999999,
+      "grad_norm": 36.45222854614258,
+      "learning_rate": 6.024600000000001e-06,
+      "loss": 0.7467,
+      "step": 19890
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 33.47389221191406,
+      "learning_rate": 6.022600000000001e-06,
+      "loss": 0.6745,
+      "step": 19900
+    },
+    {
+      "epoch": 1.991,
+      "grad_norm": 68.49720764160156,
+      "learning_rate": 6.020600000000001e-06,
+      "loss": 0.7931,
+      "step": 19910
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 44.77395248413086,
+      "learning_rate": 6.0186e-06,
+      "loss": 0.7077,
+      "step": 19920
+    },
+    {
+      "epoch": 1.9929999999999999,
+      "grad_norm": 26.6000919342041,
+      "learning_rate": 6.0166e-06,
+      "loss": 0.7528,
+      "step": 19930
+    },
+    {
+      "epoch": 1.994,
+      "grad_norm": 26.399738311767578,
+      "learning_rate": 6.014600000000001e-06,
+      "loss": 0.3916,
+      "step": 19940
+    },
+    {
+      "epoch": 1.995,
+      "grad_norm": 12.024267196655273,
+      "learning_rate": 6.012600000000001e-06,
+      "loss": 0.838,
+      "step": 19950
+    },
+    {
+      "epoch": 1.996,
+      "grad_norm": 0.7568420171737671,
+      "learning_rate": 6.0106000000000005e-06,
+      "loss": 0.5821,
+      "step": 19960
+    },
+    {
+      "epoch": 1.9969999999999999,
+      "grad_norm": 19.045326232910156,
+      "learning_rate": 6.0086e-06,
+      "loss": 0.4877,
+      "step": 19970
+    },
+    {
+      "epoch": 1.998,
+      "grad_norm": 30.190324783325195,
+      "learning_rate": 6.0066e-06,
+      "loss": 0.8412,
+      "step": 19980
+    },
+    {
+      "epoch": 1.999,
+      "grad_norm": 20.587814331054688,
+      "learning_rate": 6.0046e-06,
+      "loss": 0.7776,
+      "step": 19990
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 47.415870666503906,
+      "learning_rate": 6.002600000000001e-06,
+      "loss": 0.8361,
+      "step": 20000
+    },
+    {
+      "epoch": 2.001,
+      "grad_norm": 34.35898971557617,
+      "learning_rate": 6.000600000000001e-06,
+      "loss": 0.6689,
+      "step": 20010
+    },
+    {
+      "epoch": 2.002,
+      "grad_norm": 22.506715774536133,
+      "learning_rate": 5.9986e-06,
+      "loss": 0.4416,
+      "step": 20020
+    },
+    {
+      "epoch": 2.003,
+      "grad_norm": 33.02839279174805,
+      "learning_rate": 5.9966e-06,
+      "loss": 0.6362,
+      "step": 20030
+    },
+    {
+      "epoch": 2.004,
+      "grad_norm": 52.246097564697266,
+      "learning_rate": 5.9946e-06,
+      "loss": 0.6402,
+      "step": 20040
+    },
+    {
+      "epoch": 2.005,
+      "grad_norm": 17.559724807739258,
+      "learning_rate": 5.992600000000001e-06,
+      "loss": 0.6283,
+      "step": 20050
+    },
+    {
+      "epoch": 2.006,
+      "grad_norm": 41.119911193847656,
+      "learning_rate": 5.990600000000001e-06,
+      "loss": 0.856,
+      "step": 20060
+    },
+    {
+      "epoch": 2.007,
+      "grad_norm": 39.30539321899414,
+      "learning_rate": 5.9886e-06,
+      "loss": 0.6278,
+      "step": 20070
+    },
+    {
+      "epoch": 2.008,
+      "grad_norm": 30.116628646850586,
+      "learning_rate": 5.9866e-06,
+      "loss": 0.5567,
+      "step": 20080
+    },
+    {
+      "epoch": 2.009,
+      "grad_norm": 42.305233001708984,
+      "learning_rate": 5.9846e-06,
+      "loss": 0.5619,
+      "step": 20090
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 23.330278396606445,
+      "learning_rate": 5.982600000000001e-06,
+      "loss": 0.5062,
+      "step": 20100
+    },
+    {
+      "epoch": 2.011,
+      "grad_norm": 0.366443008184433,
+      "learning_rate": 5.980600000000001e-06,
+      "loss": 0.5095,
+      "step": 20110
+    },
+    {
+      "epoch": 2.012,
+      "grad_norm": 22.624025344848633,
+      "learning_rate": 5.9786e-06,
+      "loss": 0.5983,
+      "step": 20120
+    },
+    {
+      "epoch": 2.013,
+      "grad_norm": 55.463741302490234,
+      "learning_rate": 5.9766000000000005e-06,
+      "loss": 0.7777,
+      "step": 20130
+    },
+    {
+      "epoch": 2.014,
+      "grad_norm": 26.516307830810547,
+      "learning_rate": 5.9746e-06,
+      "loss": 0.5525,
+      "step": 20140
+    },
+    {
+      "epoch": 2.015,
+      "grad_norm": 24.06287384033203,
+      "learning_rate": 5.9726e-06,
+      "loss": 0.7574,
+      "step": 20150
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 50.38890075683594,
+      "learning_rate": 5.970600000000001e-06,
+      "loss": 0.5241,
+      "step": 20160
+    },
+    {
+      "epoch": 2.017,
+      "grad_norm": 33.88356018066406,
+      "learning_rate": 5.9686e-06,
+      "loss": 0.4269,
+      "step": 20170
+    },
+    {
+      "epoch": 2.018,
+      "grad_norm": 26.697917938232422,
+      "learning_rate": 5.9666e-06,
+      "loss": 0.6051,
+      "step": 20180
+    },
+    {
+      "epoch": 2.019,
+      "grad_norm": 15.631553649902344,
+      "learning_rate": 5.9646000000000005e-06,
+      "loss": 0.6533,
+      "step": 20190
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.76919937133789,
+      "learning_rate": 5.9626e-06,
+      "loss": 0.7292,
+      "step": 20200
+    },
+    {
+      "epoch": 2.021,
+      "grad_norm": 29.44046974182129,
+      "learning_rate": 5.960600000000001e-06,
+      "loss": 0.7371,
+      "step": 20210
+    },
+    {
+      "epoch": 2.022,
+      "grad_norm": 26.83331298828125,
+      "learning_rate": 5.9586e-06,
+      "loss": 0.8509,
+      "step": 20220
+    },
+    {
+      "epoch": 2.023,
+      "grad_norm": 18.271591186523438,
+      "learning_rate": 5.9566e-06,
+      "loss": 0.5269,
+      "step": 20230
+    },
+    {
+      "epoch": 2.024,
+      "grad_norm": 81.81536102294922,
+      "learning_rate": 5.9546000000000006e-06,
+      "loss": 0.7498,
+      "step": 20240
+    },
+    {
+      "epoch": 2.025,
+      "grad_norm": 42.30961608886719,
+      "learning_rate": 5.9526e-06,
+      "loss": 0.5365,
+      "step": 20250
+    },
+    {
+      "epoch": 2.026,
+      "grad_norm": 22.208972930908203,
+      "learning_rate": 5.950600000000001e-06,
+      "loss": 0.5537,
+      "step": 20260
+    },
+    {
+      "epoch": 2.027,
+      "grad_norm": 1.5519413948059082,
+      "learning_rate": 5.948600000000001e-06,
+      "loss": 0.7741,
+      "step": 20270
+    },
+    {
+      "epoch": 2.028,
+      "grad_norm": 24.482988357543945,
+      "learning_rate": 5.9466e-06,
+      "loss": 0.4574,
+      "step": 20280
+    },
+    {
+      "epoch": 2.029,
+      "grad_norm": 7.925387382507324,
+      "learning_rate": 5.944600000000001e-06,
+      "loss": 0.6198,
+      "step": 20290
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 35.66553497314453,
+      "learning_rate": 5.9426000000000005e-06,
+      "loss": 0.8091,
+      "step": 20300
+    },
+    {
+      "epoch": 2.031,
+      "grad_norm": 36.801700592041016,
+      "learning_rate": 5.940600000000001e-06,
+      "loss": 0.5905,
+      "step": 20310
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 16.00217628479004,
+      "learning_rate": 5.938600000000001e-06,
+      "loss": 0.6229,
+      "step": 20320
+    },
+    {
+      "epoch": 2.033,
+      "grad_norm": 41.883907318115234,
+      "learning_rate": 5.9366e-06,
+      "loss": 0.617,
+      "step": 20330
+    },
+    {
+      "epoch": 2.034,
+      "grad_norm": 42.81610870361328,
+      "learning_rate": 5.934600000000001e-06,
+      "loss": 0.5014,
+      "step": 20340
+    },
+    {
+      "epoch": 2.035,
+      "grad_norm": 40.49055862426758,
+      "learning_rate": 5.932600000000001e-06,
+      "loss": 0.4363,
+      "step": 20350
+    },
+    {
+      "epoch": 2.036,
+      "grad_norm": 21.71941566467285,
+      "learning_rate": 5.9306000000000004e-06,
+      "loss": 0.853,
+      "step": 20360
+    },
+    {
+      "epoch": 2.037,
+      "grad_norm": 63.479583740234375,
+      "learning_rate": 5.928600000000001e-06,
+      "loss": 0.8359,
+      "step": 20370
+    },
+    {
+      "epoch": 2.038,
+      "grad_norm": 11.940189361572266,
+      "learning_rate": 5.9266e-06,
+      "loss": 0.4842,
+      "step": 20380
+    },
+    {
+      "epoch": 2.039,
+      "grad_norm": 13.797748565673828,
+      "learning_rate": 5.9246e-06,
+      "loss": 0.3111,
+      "step": 20390
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 26.523082733154297,
+      "learning_rate": 5.922600000000001e-06,
+      "loss": 0.8263,
+      "step": 20400
+    },
+    {
+      "epoch": 2.041,
+      "grad_norm": 15.565614700317383,
+      "learning_rate": 5.9206000000000005e-06,
+      "loss": 0.2726,
+      "step": 20410
+    },
+    {
+      "epoch": 2.042,
+      "grad_norm": 23.820993423461914,
+      "learning_rate": 5.918600000000001e-06,
+      "loss": 0.8273,
+      "step": 20420
+    },
+    {
+      "epoch": 2.043,
+      "grad_norm": 22.535480499267578,
+      "learning_rate": 5.9166e-06,
+      "loss": 0.614,
+      "step": 20430
+    },
+    {
+      "epoch": 2.044,
+      "grad_norm": 60.1475830078125,
+      "learning_rate": 5.9146e-06,
+      "loss": 0.7601,
+      "step": 20440
+    },
+    {
+      "epoch": 2.045,
+      "grad_norm": 27.413978576660156,
+      "learning_rate": 5.912600000000001e-06,
+      "loss": 0.7644,
+      "step": 20450
+    },
+    {
+      "epoch": 2.046,
+      "grad_norm": 7.223193645477295,
+      "learning_rate": 5.910600000000001e-06,
+      "loss": 0.5678,
+      "step": 20460
+    },
+    {
+      "epoch": 2.047,
+      "grad_norm": 15.275238990783691,
+      "learning_rate": 5.908600000000001e-06,
+      "loss": 0.7498,
+      "step": 20470
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 68.40691375732422,
+      "learning_rate": 5.9066e-06,
+      "loss": 0.9087,
+      "step": 20480
+    },
+    {
+      "epoch": 2.049,
+      "grad_norm": 18.788619995117188,
+      "learning_rate": 5.9046e-06,
+      "loss": 0.8295,
+      "step": 20490
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 57.324222564697266,
+      "learning_rate": 5.902600000000001e-06,
+      "loss": 0.7998,
+      "step": 20500
+    },
+    {
+      "epoch": 2.051,
+      "grad_norm": 6.82062292098999,
+      "learning_rate": 5.900600000000001e-06,
+      "loss": 0.4767,
+      "step": 20510
+    },
+    {
+      "epoch": 2.052,
+      "grad_norm": 35.272586822509766,
+      "learning_rate": 5.898600000000001e-06,
+      "loss": 0.5846,
+      "step": 20520
+    },
+    {
+      "epoch": 2.053,
+      "grad_norm": 29.768474578857422,
+      "learning_rate": 5.8966000000000004e-06,
+      "loss": 0.6642,
+      "step": 20530
+    },
+    {
+      "epoch": 2.054,
+      "grad_norm": 29.56476593017578,
+      "learning_rate": 5.8946e-06,
+      "loss": 0.7453,
+      "step": 20540
+    },
+    {
+      "epoch": 2.055,
+      "grad_norm": 55.21421813964844,
+      "learning_rate": 5.8926e-06,
+      "loss": 0.6613,
+      "step": 20550
+    },
+    {
+      "epoch": 2.056,
+      "grad_norm": 38.4114875793457,
+      "learning_rate": 5.890600000000001e-06,
+      "loss": 0.6084,
+      "step": 20560
+    },
+    {
+      "epoch": 2.057,
+      "grad_norm": 6.092259883880615,
+      "learning_rate": 5.888600000000001e-06,
+      "loss": 0.3543,
+      "step": 20570
+    },
+    {
+      "epoch": 2.058,
+      "grad_norm": 30.972808837890625,
+      "learning_rate": 5.8866e-06,
+      "loss": 0.6939,
+      "step": 20580
+    },
+    {
+      "epoch": 2.059,
+      "grad_norm": 48.587074279785156,
+      "learning_rate": 5.8846e-06,
+      "loss": 0.7431,
+      "step": 20590
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.721481323242188,
+      "learning_rate": 5.8826e-06,
+      "loss": 0.6013,
+      "step": 20600
+    },
+    {
+      "epoch": 2.061,
+      "grad_norm": 55.05281066894531,
+      "learning_rate": 5.880600000000001e-06,
+      "loss": 0.8299,
+      "step": 20610
+    },
+    {
+      "epoch": 2.062,
+      "grad_norm": 13.547633171081543,
+      "learning_rate": 5.878600000000001e-06,
+      "loss": 0.6287,
+      "step": 20620
+    },
+    {
+      "epoch": 2.063,
+      "grad_norm": 43.30170440673828,
+      "learning_rate": 5.8766e-06,
+      "loss": 0.9062,
+      "step": 20630
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 69.8653335571289,
+      "learning_rate": 5.8746000000000005e-06,
+      "loss": 0.6167,
+      "step": 20640
+    },
+    {
+      "epoch": 2.065,
+      "grad_norm": 13.36156940460205,
+      "learning_rate": 5.8726e-06,
+      "loss": 0.5346,
+      "step": 20650
+    },
+    {
+      "epoch": 2.066,
+      "grad_norm": 55.982967376708984,
+      "learning_rate": 5.870600000000001e-06,
+      "loss": 0.6719,
+      "step": 20660
+    },
+    {
+      "epoch": 2.067,
+      "grad_norm": 19.473783493041992,
+      "learning_rate": 5.868600000000001e-06,
+      "loss": 0.6472,
+      "step": 20670
+    },
+    {
+      "epoch": 2.068,
+      "grad_norm": 47.84219741821289,
+      "learning_rate": 5.8666e-06,
+      "loss": 0.6803,
+      "step": 20680
+    },
+    {
+      "epoch": 2.069,
+      "grad_norm": 22.7913761138916,
+      "learning_rate": 5.8646000000000006e-06,
+      "loss": 0.9931,
+      "step": 20690
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 30.682554244995117,
+      "learning_rate": 5.8626e-06,
+      "loss": 0.5268,
+      "step": 20700
+    },
+    {
+      "epoch": 2.071,
+      "grad_norm": 30.542743682861328,
+      "learning_rate": 5.860600000000001e-06,
+      "loss": 0.5187,
+      "step": 20710
+    },
+    {
+      "epoch": 2.072,
+      "grad_norm": 33.08385467529297,
+      "learning_rate": 5.858600000000001e-06,
+      "loss": 0.7148,
+      "step": 20720
+    },
+    {
+      "epoch": 2.073,
+      "grad_norm": 43.1998291015625,
+      "learning_rate": 5.8566e-06,
+      "loss": 0.5574,
+      "step": 20730
+    },
+    {
+      "epoch": 2.074,
+      "grad_norm": 35.44502258300781,
+      "learning_rate": 5.854600000000001e-06,
+      "loss": 0.5502,
+      "step": 20740
+    },
+    {
+      "epoch": 2.075,
+      "grad_norm": 23.778690338134766,
+      "learning_rate": 5.8526000000000005e-06,
+      "loss": 0.7837,
+      "step": 20750
+    },
+    {
+      "epoch": 2.076,
+      "grad_norm": 15.768932342529297,
+      "learning_rate": 5.8506e-06,
+      "loss": 0.5188,
+      "step": 20760
+    },
+    {
+      "epoch": 2.077,
+      "grad_norm": 57.871437072753906,
+      "learning_rate": 5.848600000000001e-06,
+      "loss": 0.523,
+      "step": 20770
+    },
+    {
+      "epoch": 2.078,
+      "grad_norm": 45.32200622558594,
+      "learning_rate": 5.8466e-06,
+      "loss": 0.8176,
+      "step": 20780
+    },
+    {
+      "epoch": 2.079,
+      "grad_norm": 32.9245491027832,
+      "learning_rate": 5.8446e-06,
+      "loss": 0.5147,
+      "step": 20790
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 44.795677185058594,
+      "learning_rate": 5.842600000000001e-06,
+      "loss": 0.745,
+      "step": 20800
+    },
+    {
+      "epoch": 2.081,
+      "grad_norm": 49.823299407958984,
+      "learning_rate": 5.8406000000000005e-06,
+      "loss": 0.7354,
+      "step": 20810
+    },
+    {
+      "epoch": 2.082,
+      "grad_norm": 6.510171413421631,
+      "learning_rate": 5.838600000000001e-06,
+      "loss": 0.4389,
+      "step": 20820
+    },
+    {
+      "epoch": 2.083,
+      "grad_norm": 24.822555541992188,
+      "learning_rate": 5.8366e-06,
+      "loss": 0.7396,
+      "step": 20830
+    },
+    {
+      "epoch": 2.084,
+      "grad_norm": 35.97170639038086,
+      "learning_rate": 5.8346e-06,
+      "loss": 0.5066,
+      "step": 20840
+    },
+    {
+      "epoch": 2.085,
+      "grad_norm": 62.91329574584961,
+      "learning_rate": 5.832600000000001e-06,
+      "loss": 1.0351,
+      "step": 20850
+    },
+    {
+      "epoch": 2.086,
+      "grad_norm": 40.740325927734375,
+      "learning_rate": 5.8306000000000006e-06,
+      "loss": 0.4666,
+      "step": 20860
+    },
+    {
+      "epoch": 2.087,
+      "grad_norm": 35.693050384521484,
+      "learning_rate": 5.828600000000001e-06,
+      "loss": 0.6951,
+      "step": 20870
+    },
+    {
+      "epoch": 2.088,
+      "grad_norm": 7.824371814727783,
+      "learning_rate": 5.8266e-06,
+      "loss": 0.6165,
+      "step": 20880
+    },
+    {
+      "epoch": 2.089,
+      "grad_norm": 18.26198959350586,
+      "learning_rate": 5.8246e-06,
+      "loss": 0.6349,
+      "step": 20890
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 33.34101486206055,
+      "learning_rate": 5.822600000000001e-06,
+      "loss": 0.5883,
+      "step": 20900
+    },
+    {
+      "epoch": 2.091,
+      "grad_norm": 9.032116889953613,
+      "learning_rate": 5.820600000000001e-06,
+      "loss": 0.8677,
+      "step": 20910
+    },
+    {
+      "epoch": 2.092,
+      "grad_norm": 61.170379638671875,
+      "learning_rate": 5.8186000000000005e-06,
+      "loss": 0.6158,
+      "step": 20920
+    },
+    {
+      "epoch": 2.093,
+      "grad_norm": 21.601675033569336,
+      "learning_rate": 5.8166e-06,
+      "loss": 0.7612,
+      "step": 20930
+    },
+    {
+      "epoch": 2.094,
+      "grad_norm": 22.31391716003418,
+      "learning_rate": 5.8146e-06,
+      "loss": 0.6095,
+      "step": 20940
+    },
+    {
+      "epoch": 2.095,
+      "grad_norm": 33.5605583190918,
+      "learning_rate": 5.8126e-06,
+      "loss": 0.503,
+      "step": 20950
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 15.684220314025879,
+      "learning_rate": 5.810600000000001e-06,
+      "loss": 0.5778,
+      "step": 20960
+    },
+    {
+      "epoch": 2.097,
+      "grad_norm": 45.926124572753906,
+      "learning_rate": 5.808600000000001e-06,
+      "loss": 0.7229,
+      "step": 20970
+    },
+    {
+      "epoch": 2.098,
+      "grad_norm": 27.19575309753418,
+      "learning_rate": 5.8066e-06,
+      "loss": 0.5593,
+      "step": 20980
+    },
+    {
+      "epoch": 2.099,
+      "grad_norm": 28.116910934448242,
+      "learning_rate": 5.8046e-06,
+      "loss": 0.8757,
+      "step": 20990
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.367252349853516,
+      "learning_rate": 5.8026e-06,
+      "loss": 0.4212,
+      "step": 21000
+    },
+    {
+      "epoch": 2.101,
+      "grad_norm": 39.175296783447266,
+      "learning_rate": 5.800600000000001e-06,
+      "loss": 0.4437,
+      "step": 21010
+    },
+    {
+      "epoch": 2.102,
+      "grad_norm": 49.59600830078125,
+      "learning_rate": 5.798600000000001e-06,
+      "loss": 0.7073,
+      "step": 21020
+    },
+    {
+      "epoch": 2.103,
+      "grad_norm": 10.100749969482422,
+      "learning_rate": 5.7966e-06,
+      "loss": 0.8202,
+      "step": 21030
+    },
+    {
+      "epoch": 2.104,
+      "grad_norm": 29.555816650390625,
+      "learning_rate": 5.7946e-06,
+      "loss": 0.5118,
+      "step": 21040
+    },
+    {
+      "epoch": 2.105,
+      "grad_norm": 22.045875549316406,
+      "learning_rate": 5.7926e-06,
+      "loss": 0.4967,
+      "step": 21050
+    },
+    {
+      "epoch": 2.106,
+      "grad_norm": 43.3510856628418,
+      "learning_rate": 5.790600000000001e-06,
+      "loss": 0.9821,
+      "step": 21060
+    },
+    {
+      "epoch": 2.107,
+      "grad_norm": 43.989376068115234,
+      "learning_rate": 5.788600000000001e-06,
+      "loss": 0.5552,
+      "step": 21070
+    },
+    {
+      "epoch": 2.108,
+      "grad_norm": 30.01483917236328,
+      "learning_rate": 5.7866e-06,
+      "loss": 0.5964,
+      "step": 21080
+    },
+    {
+      "epoch": 2.109,
+      "grad_norm": 33.281497955322266,
+      "learning_rate": 5.7846000000000005e-06,
+      "loss": 0.7582,
+      "step": 21090
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 69.60244750976562,
+      "learning_rate": 5.7826e-06,
+      "loss": 0.5143,
+      "step": 21100
+    },
+    {
+      "epoch": 2.111,
+      "grad_norm": 2.7783589363098145,
+      "learning_rate": 5.780600000000001e-06,
+      "loss": 0.6714,
+      "step": 21110
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 60.577178955078125,
+      "learning_rate": 5.778600000000001e-06,
+      "loss": 0.5207,
+      "step": 21120
+    },
+    {
+      "epoch": 2.113,
+      "grad_norm": 29.13699722290039,
+      "learning_rate": 5.7766e-06,
+      "loss": 0.5172,
+      "step": 21130
+    },
+    {
+      "epoch": 2.114,
+      "grad_norm": 48.14480209350586,
+      "learning_rate": 5.774600000000001e-06,
+      "loss": 0.6832,
+      "step": 21140
+    },
+    {
+      "epoch": 2.115,
+      "grad_norm": 14.728019714355469,
+      "learning_rate": 5.7726000000000004e-06,
+      "loss": 0.8305,
+      "step": 21150
+    },
+    {
+      "epoch": 2.116,
+      "grad_norm": 5.194901466369629,
+      "learning_rate": 5.7706e-06,
+      "loss": 0.6312,
+      "step": 21160
+    },
+    {
+      "epoch": 2.117,
+      "grad_norm": 48.74283218383789,
+      "learning_rate": 5.768600000000001e-06,
+      "loss": 0.6754,
+      "step": 21170
+    },
+    {
+      "epoch": 2.118,
+      "grad_norm": 22.42253303527832,
+      "learning_rate": 5.7666e-06,
+      "loss": 0.8935,
+      "step": 21180
+    },
+    {
+      "epoch": 2.1189999999999998,
+      "grad_norm": 24.238758087158203,
+      "learning_rate": 5.7646e-06,
+      "loss": 0.6695,
+      "step": 21190
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 39.905906677246094,
+      "learning_rate": 5.7626000000000005e-06,
+      "loss": 0.6241,
+      "step": 21200
+    },
+    {
+      "epoch": 2.121,
+      "grad_norm": 35.234169006347656,
+      "learning_rate": 5.7606e-06,
+      "loss": 0.6531,
+      "step": 21210
+    },
+    {
+      "epoch": 2.122,
+      "grad_norm": 12.21348762512207,
+      "learning_rate": 5.758600000000001e-06,
+      "loss": 0.7264,
+      "step": 21220
+    },
+    {
+      "epoch": 2.123,
+      "grad_norm": 32.99053192138672,
+      "learning_rate": 5.7566e-06,
+      "loss": 0.4499,
+      "step": 21230
+    },
+    {
+      "epoch": 2.124,
+      "grad_norm": 53.96305465698242,
+      "learning_rate": 5.7546e-06,
+      "loss": 0.7433,
+      "step": 21240
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 9.550210952758789,
+      "learning_rate": 5.752600000000001e-06,
+      "loss": 0.7014,
+      "step": 21250
+    },
+    {
+      "epoch": 2.126,
+      "grad_norm": 1.4421324729919434,
+      "learning_rate": 5.7506000000000005e-06,
+      "loss": 0.4575,
+      "step": 21260
+    },
+    {
+      "epoch": 2.127,
+      "grad_norm": 13.814619064331055,
+      "learning_rate": 5.748600000000001e-06,
+      "loss": 0.5578,
+      "step": 21270
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 1.734191656112671,
+      "learning_rate": 5.7466e-06,
+      "loss": 0.8801,
+      "step": 21280
+    },
+    {
+      "epoch": 2.129,
+      "grad_norm": 11.361309051513672,
+      "learning_rate": 5.7446e-06,
+      "loss": 0.4905,
+      "step": 21290
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 50.14918899536133,
+      "learning_rate": 5.742600000000001e-06,
+      "loss": 0.7295,
+      "step": 21300
+    },
+    {
+      "epoch": 2.1310000000000002,
+      "grad_norm": 34.93165588378906,
+      "learning_rate": 5.7406000000000006e-06,
+      "loss": 0.7161,
+      "step": 21310
+    },
+    {
+      "epoch": 2.132,
+      "grad_norm": 22.720502853393555,
+      "learning_rate": 5.7386e-06,
+      "loss": 0.5065,
+      "step": 21320
+    },
+    {
+      "epoch": 2.133,
+      "grad_norm": 39.20326614379883,
+      "learning_rate": 5.7366e-06,
+      "loss": 0.7608,
+      "step": 21330
+    },
+    {
+      "epoch": 2.134,
+      "grad_norm": 39.12361145019531,
+      "learning_rate": 5.7346e-06,
+      "loss": 0.6045,
+      "step": 21340
+    },
+    {
+      "epoch": 2.135,
+      "grad_norm": 19.994277954101562,
+      "learning_rate": 5.7326e-06,
+      "loss": 0.3712,
+      "step": 21350
+    },
+    {
+      "epoch": 2.136,
+      "grad_norm": 66.90753173828125,
+      "learning_rate": 5.730600000000001e-06,
+      "loss": 0.6658,
+      "step": 21360
+    },
+    {
+      "epoch": 2.137,
+      "grad_norm": 36.23524856567383,
+      "learning_rate": 5.7286000000000005e-06,
+      "loss": 0.6202,
+      "step": 21370
+    },
+    {
+      "epoch": 2.138,
+      "grad_norm": 4.688386917114258,
+      "learning_rate": 5.7265999999999995e-06,
+      "loss": 0.6627,
+      "step": 21380
+    },
+    {
+      "epoch": 2.1390000000000002,
+      "grad_norm": 24.567197799682617,
+      "learning_rate": 5.7246e-06,
+      "loss": 0.6649,
+      "step": 21390
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.481902599334717,
+      "learning_rate": 5.7226e-06,
+      "loss": 0.8944,
+      "step": 21400
+    },
+    {
+      "epoch": 2.141,
+      "grad_norm": 18.482616424560547,
+      "learning_rate": 5.720600000000001e-06,
+      "loss": 0.4641,
+      "step": 21410
+    },
+    {
+      "epoch": 2.142,
+      "grad_norm": 15.100506782531738,
+      "learning_rate": 5.718600000000001e-06,
+      "loss": 0.502,
+      "step": 21420
+    },
+    {
+      "epoch": 2.143,
+      "grad_norm": 16.71766471862793,
+      "learning_rate": 5.7166e-06,
+      "loss": 0.7693,
+      "step": 21430
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 37.1085205078125,
+      "learning_rate": 5.7146e-06,
+      "loss": 0.8897,
+      "step": 21440
+    },
+    {
+      "epoch": 2.145,
+      "grad_norm": 23.989057540893555,
+      "learning_rate": 5.7126e-06,
+      "loss": 0.614,
+      "step": 21450
+    },
+    {
+      "epoch": 2.146,
+      "grad_norm": 9.404507637023926,
+      "learning_rate": 5.710600000000001e-06,
+      "loss": 0.6387,
+      "step": 21460
+    },
+    {
+      "epoch": 2.147,
+      "grad_norm": 21.427249908447266,
+      "learning_rate": 5.708600000000001e-06,
+      "loss": 0.695,
+      "step": 21470
+    },
+    {
+      "epoch": 2.148,
+      "grad_norm": 28.601951599121094,
+      "learning_rate": 5.7066e-06,
+      "loss": 0.6689,
+      "step": 21480
+    },
+    {
+      "epoch": 2.149,
+      "grad_norm": 37.28999328613281,
+      "learning_rate": 5.7046e-06,
+      "loss": 0.6496,
+      "step": 21490
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 12.694632530212402,
+      "learning_rate": 5.7026e-06,
+      "loss": 0.5615,
+      "step": 21500
+    },
+    {
+      "epoch": 2.151,
+      "grad_norm": 19.564849853515625,
+      "learning_rate": 5.700600000000001e-06,
+      "loss": 0.5441,
+      "step": 21510
+    },
+    {
+      "epoch": 2.152,
+      "grad_norm": 31.632789611816406,
+      "learning_rate": 5.698600000000001e-06,
+      "loss": 0.5756,
+      "step": 21520
+    },
+    {
+      "epoch": 2.153,
+      "grad_norm": 14.230611801147461,
+      "learning_rate": 5.696600000000001e-06,
+      "loss": 0.5026,
+      "step": 21530
+    },
+    {
+      "epoch": 2.154,
+      "grad_norm": 44.073448181152344,
+      "learning_rate": 5.6946000000000005e-06,
+      "loss": 0.6345,
+      "step": 21540
+    },
+    {
+      "epoch": 2.155,
+      "grad_norm": 58.922603607177734,
+      "learning_rate": 5.6926e-06,
+      "loss": 0.6142,
+      "step": 21550
+    },
+    {
+      "epoch": 2.156,
+      "grad_norm": 14.919763565063477,
+      "learning_rate": 5.6906e-06,
+      "loss": 0.4576,
+      "step": 21560
+    },
+    {
+      "epoch": 2.157,
+      "grad_norm": 25.46698760986328,
+      "learning_rate": 5.688600000000001e-06,
+      "loss": 0.8048,
+      "step": 21570
+    },
+    {
+      "epoch": 2.158,
+      "grad_norm": 54.56586837768555,
+      "learning_rate": 5.686600000000001e-06,
+      "loss": 0.5692,
+      "step": 21580
+    },
+    {
+      "epoch": 2.159,
+      "grad_norm": 35.88685607910156,
+      "learning_rate": 5.6846e-06,
+      "loss": 0.4866,
+      "step": 21590
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 23.260847091674805,
+      "learning_rate": 5.6826000000000004e-06,
+      "loss": 0.488,
+      "step": 21600
+    },
+    {
+      "epoch": 2.161,
+      "grad_norm": 51.36183547973633,
+      "learning_rate": 5.6806e-06,
+      "loss": 0.7119,
+      "step": 21610
+    },
+    {
+      "epoch": 2.162,
+      "grad_norm": 26.24199867248535,
+      "learning_rate": 5.678600000000001e-06,
+      "loss": 0.784,
+      "step": 21620
+    },
+    {
+      "epoch": 2.163,
+      "grad_norm": 28.365392684936523,
+      "learning_rate": 5.676600000000001e-06,
+      "loss": 0.7343,
+      "step": 21630
+    },
+    {
+      "epoch": 2.164,
+      "grad_norm": 50.15872573852539,
+      "learning_rate": 5.6746e-06,
+      "loss": 0.5722,
+      "step": 21640
+    },
+    {
+      "epoch": 2.165,
+      "grad_norm": 45.75239562988281,
+      "learning_rate": 5.6726000000000005e-06,
+      "loss": 0.5819,
+      "step": 21650
+    },
+    {
+      "epoch": 2.166,
+      "grad_norm": 26.17607307434082,
+      "learning_rate": 5.6706e-06,
+      "loss": 0.4819,
+      "step": 21660
+    },
+    {
+      "epoch": 2.167,
+      "grad_norm": 6.468771457672119,
+      "learning_rate": 5.668600000000001e-06,
+      "loss": 0.6419,
+      "step": 21670
+    },
+    {
+      "epoch": 2.168,
+      "grad_norm": 0.5758807063102722,
+      "learning_rate": 5.666600000000001e-06,
+      "loss": 0.6369,
+      "step": 21680
+    },
+    {
+      "epoch": 2.169,
+      "grad_norm": 41.119407653808594,
+      "learning_rate": 5.6646e-06,
+      "loss": 0.5396,
+      "step": 21690
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 20.98927116394043,
+      "learning_rate": 5.662600000000001e-06,
+      "loss": 0.5572,
+      "step": 21700
+    },
+    {
+      "epoch": 2.171,
+      "grad_norm": 2.9193451404571533,
+      "learning_rate": 5.6606000000000005e-06,
+      "loss": 0.6213,
+      "step": 21710
+    },
+    {
+      "epoch": 2.172,
+      "grad_norm": 41.307823181152344,
+      "learning_rate": 5.6586e-06,
+      "loss": 0.7824,
+      "step": 21720
+    },
+    {
+      "epoch": 2.173,
+      "grad_norm": 49.53284454345703,
+      "learning_rate": 5.656600000000001e-06,
+      "loss": 0.7281,
+      "step": 21730
+    },
+    {
+      "epoch": 2.174,
+      "grad_norm": 80.82706451416016,
+      "learning_rate": 5.6546e-06,
+      "loss": 0.525,
+      "step": 21740
+    },
+    {
+      "epoch": 2.175,
+      "grad_norm": 10.112666130065918,
+      "learning_rate": 5.6526e-06,
+      "loss": 0.8426,
+      "step": 21750
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 30.3128719329834,
+      "learning_rate": 5.650600000000001e-06,
+      "loss": 0.5265,
+      "step": 21760
+    },
+    {
+      "epoch": 2.177,
+      "grad_norm": 11.140154838562012,
+      "learning_rate": 5.6486000000000004e-06,
+      "loss": 0.7528,
+      "step": 21770
+    },
+    {
+      "epoch": 2.178,
+      "grad_norm": 32.421485900878906,
+      "learning_rate": 5.646600000000001e-06,
+      "loss": 0.5861,
+      "step": 21780
+    },
+    {
+      "epoch": 2.179,
+      "grad_norm": 21.485401153564453,
+      "learning_rate": 5.6446e-06,
+      "loss": 0.4652,
+      "step": 21790
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.479448318481445,
+      "learning_rate": 5.6426e-06,
+      "loss": 0.8183,
+      "step": 21800
+    },
+    {
+      "epoch": 2.181,
+      "grad_norm": 26.815753936767578,
+      "learning_rate": 5.640600000000001e-06,
+      "loss": 0.5524,
+      "step": 21810
+    },
+    {
+      "epoch": 2.182,
+      "grad_norm": 6.771590232849121,
+      "learning_rate": 5.6386000000000005e-06,
+      "loss": 0.6103,
+      "step": 21820
+    },
+    {
+      "epoch": 2.183,
+      "grad_norm": 7.171220302581787,
+      "learning_rate": 5.636600000000001e-06,
+      "loss": 0.5156,
+      "step": 21830
+    },
+    {
+      "epoch": 2.184,
+      "grad_norm": 49.12594985961914,
+      "learning_rate": 5.6346e-06,
+      "loss": 0.7338,
+      "step": 21840
+    },
+    {
+      "epoch": 2.185,
+      "grad_norm": 26.85982322692871,
+      "learning_rate": 5.6326e-06,
+      "loss": 0.9972,
+      "step": 21850
+    },
+    {
+      "epoch": 2.186,
+      "grad_norm": 22.835695266723633,
+      "learning_rate": 5.630600000000001e-06,
+      "loss": 0.7974,
+      "step": 21860
+    },
+    {
+      "epoch": 2.187,
+      "grad_norm": 8.738889694213867,
+      "learning_rate": 5.628600000000001e-06,
+      "loss": 0.6405,
+      "step": 21870
+    },
+    {
+      "epoch": 2.188,
+      "grad_norm": 38.73388671875,
+      "learning_rate": 5.626600000000001e-06,
+      "loss": 0.9766,
+      "step": 21880
+    },
+    {
+      "epoch": 2.189,
+      "grad_norm": 42.84649658203125,
+      "learning_rate": 5.6246e-06,
+      "loss": 0.7369,
+      "step": 21890
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 22.246782302856445,
+      "learning_rate": 5.6226e-06,
+      "loss": 0.6748,
+      "step": 21900
+    },
+    {
+      "epoch": 2.191,
+      "grad_norm": 41.804264068603516,
+      "learning_rate": 5.620600000000001e-06,
+      "loss": 0.5704,
+      "step": 21910
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 55.34584045410156,
+      "learning_rate": 5.618600000000001e-06,
+      "loss": 0.7983,
+      "step": 21920
+    },
+    {
+      "epoch": 2.193,
+      "grad_norm": 24.581790924072266,
+      "learning_rate": 5.6166000000000006e-06,
+      "loss": 0.6806,
+      "step": 21930
+    },
+    {
+      "epoch": 2.194,
+      "grad_norm": 55.127342224121094,
+      "learning_rate": 5.6146e-06,
+      "loss": 0.6151,
+      "step": 21940
+    },
+    {
+      "epoch": 2.195,
+      "grad_norm": 29.649892807006836,
+      "learning_rate": 5.6126e-06,
+      "loss": 0.5617,
+      "step": 21950
+    },
+    {
+      "epoch": 2.196,
+      "grad_norm": 27.13589096069336,
+      "learning_rate": 5.6108000000000006e-06,
+      "loss": 0.8283,
+      "step": 21960
+    },
+    {
+      "epoch": 2.197,
+      "grad_norm": 12.190668106079102,
+      "learning_rate": 5.608800000000001e-06,
+      "loss": 0.6444,
+      "step": 21970
+    },
+    {
+      "epoch": 2.198,
+      "grad_norm": 30.363269805908203,
+      "learning_rate": 5.6068e-06,
+      "loss": 0.4189,
+      "step": 21980
+    },
+    {
+      "epoch": 2.199,
+      "grad_norm": 40.20530700683594,
+      "learning_rate": 5.6048e-06,
+      "loss": 0.5882,
+      "step": 21990
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 13.80125617980957,
+      "learning_rate": 5.602800000000001e-06,
+      "loss": 0.3865,
+      "step": 22000
+    },
+    {
+      "epoch": 2.201,
+      "grad_norm": 37.33003616333008,
+      "learning_rate": 5.600800000000001e-06,
+      "loss": 0.6091,
+      "step": 22010
+    },
+    {
+      "epoch": 2.202,
+      "grad_norm": 98.78890228271484,
+      "learning_rate": 5.5988000000000005e-06,
+      "loss": 0.62,
+      "step": 22020
+    },
+    {
+      "epoch": 2.203,
+      "grad_norm": 41.13166427612305,
+      "learning_rate": 5.5968e-06,
+      "loss": 0.5836,
+      "step": 22030
+    },
+    {
+      "epoch": 2.204,
+      "grad_norm": 42.02532958984375,
+      "learning_rate": 5.5948e-06,
+      "loss": 0.5915,
+      "step": 22040
+    },
+    {
+      "epoch": 2.205,
+      "grad_norm": 42.591224670410156,
+      "learning_rate": 5.5928e-06,
+      "loss": 0.4783,
+      "step": 22050
+    },
+    {
+      "epoch": 2.206,
+      "grad_norm": 30.582590103149414,
+      "learning_rate": 5.590800000000001e-06,
+      "loss": 0.7235,
+      "step": 22060
+    },
+    {
+      "epoch": 2.207,
+      "grad_norm": 32.897727966308594,
+      "learning_rate": 5.588800000000001e-06,
+      "loss": 0.7734,
+      "step": 22070
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 18.399417877197266,
+      "learning_rate": 5.5868e-06,
+      "loss": 0.7114,
+      "step": 22080
+    },
+    {
+      "epoch": 2.209,
+      "grad_norm": 8.900433540344238,
+      "learning_rate": 5.5848e-06,
+      "loss": 0.5118,
+      "step": 22090
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 42.656253814697266,
+      "learning_rate": 5.5828e-06,
+      "loss": 0.6707,
+      "step": 22100
+    },
+    {
+      "epoch": 2.211,
+      "grad_norm": 15.351293563842773,
+      "learning_rate": 5.580800000000001e-06,
+      "loss": 0.6547,
+      "step": 22110
+    },
+    {
+      "epoch": 2.212,
+      "grad_norm": 35.78485107421875,
+      "learning_rate": 5.578800000000001e-06,
+      "loss": 0.8694,
+      "step": 22120
+    },
+    {
+      "epoch": 2.213,
+      "grad_norm": 90.62995147705078,
+      "learning_rate": 5.5768e-06,
+      "loss": 0.8345,
+      "step": 22130
+    },
+    {
+      "epoch": 2.214,
+      "grad_norm": 24.1240177154541,
+      "learning_rate": 5.5748e-06,
+      "loss": 0.6986,
+      "step": 22140
+    },
+    {
+      "epoch": 2.215,
+      "grad_norm": 37.8781623840332,
+      "learning_rate": 5.5728e-06,
+      "loss": 0.8083,
+      "step": 22150
+    },
+    {
+      "epoch": 2.216,
+      "grad_norm": 4.367500305175781,
+      "learning_rate": 5.570800000000001e-06,
+      "loss": 0.2719,
+      "step": 22160
+    },
+    {
+      "epoch": 2.217,
+      "grad_norm": 40.212860107421875,
+      "learning_rate": 5.568800000000001e-06,
+      "loss": 0.5333,
+      "step": 22170
+    },
+    {
+      "epoch": 2.218,
+      "grad_norm": 41.37190628051758,
+      "learning_rate": 5.5668e-06,
+      "loss": 0.483,
+      "step": 22180
+    },
+    {
+      "epoch": 2.219,
+      "grad_norm": 55.49161148071289,
+      "learning_rate": 5.5648000000000005e-06,
+      "loss": 0.7027,
+      "step": 22190
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 47.93917465209961,
+      "learning_rate": 5.5628e-06,
+      "loss": 0.6786,
+      "step": 22200
+    },
+    {
+      "epoch": 2.221,
+      "grad_norm": 45.09898376464844,
+      "learning_rate": 5.560800000000001e-06,
+      "loss": 0.8251,
+      "step": 22210
+    },
+    {
+      "epoch": 2.222,
+      "grad_norm": 43.21870422363281,
+      "learning_rate": 5.558800000000001e-06,
+      "loss": 0.9235,
+      "step": 22220
+    },
+    {
+      "epoch": 2.223,
+      "grad_norm": 42.25673294067383,
+      "learning_rate": 5.5568e-06,
+      "loss": 0.85,
+      "step": 22230
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 22.41086196899414,
+      "learning_rate": 5.554800000000001e-06,
+      "loss": 0.5968,
+      "step": 22240
+    },
+    {
+      "epoch": 2.225,
+      "grad_norm": 21.879209518432617,
+      "learning_rate": 5.5528000000000004e-06,
+      "loss": 0.6044,
+      "step": 22250
+    },
+    {
+      "epoch": 2.226,
+      "grad_norm": 33.574073791503906,
+      "learning_rate": 5.5508e-06,
+      "loss": 0.52,
+      "step": 22260
+    },
+    {
+      "epoch": 2.227,
+      "grad_norm": 0.5494199395179749,
+      "learning_rate": 5.548800000000001e-06,
+      "loss": 0.7395,
+      "step": 22270
+    },
+    {
+      "epoch": 2.228,
+      "grad_norm": 26.909727096557617,
+      "learning_rate": 5.5468e-06,
+      "loss": 0.4841,
+      "step": 22280
+    },
+    {
+      "epoch": 2.229,
+      "grad_norm": 35.930931091308594,
+      "learning_rate": 5.5448e-06,
+      "loss": 0.8255,
+      "step": 22290
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 30.434680938720703,
+      "learning_rate": 5.5428000000000005e-06,
+      "loss": 0.7121,
+      "step": 22300
+    },
+    {
+      "epoch": 2.231,
+      "grad_norm": 24.429750442504883,
+      "learning_rate": 5.5408e-06,
+      "loss": 0.8339,
+      "step": 22310
+    },
+    {
+      "epoch": 2.232,
+      "grad_norm": 31.06974983215332,
+      "learning_rate": 5.538800000000001e-06,
+      "loss": 0.8602,
+      "step": 22320
+    },
+    {
+      "epoch": 2.233,
+      "grad_norm": 54.97211837768555,
+      "learning_rate": 5.5368e-06,
+      "loss": 0.694,
+      "step": 22330
+    },
+    {
+      "epoch": 2.234,
+      "grad_norm": 25.800743103027344,
+      "learning_rate": 5.5348e-06,
+      "loss": 0.5709,
+      "step": 22340
+    },
+    {
+      "epoch": 2.235,
+      "grad_norm": 15.872865676879883,
+      "learning_rate": 5.532800000000001e-06,
+      "loss": 0.5848,
+      "step": 22350
+    },
+    {
+      "epoch": 2.2359999999999998,
+      "grad_norm": 8.24520206451416,
+      "learning_rate": 5.5308000000000005e-06,
+      "loss": 0.5325,
+      "step": 22360
+    },
+    {
+      "epoch": 2.237,
+      "grad_norm": 6.170074462890625,
+      "learning_rate": 5.528800000000001e-06,
+      "loss": 0.6424,
+      "step": 22370
+    },
+    {
+      "epoch": 2.238,
+      "grad_norm": 21.991802215576172,
+      "learning_rate": 5.5268e-06,
+      "loss": 0.5698,
+      "step": 22380
+    },
+    {
+      "epoch": 2.239,
+      "grad_norm": 19.199386596679688,
+      "learning_rate": 5.5248e-06,
+      "loss": 0.4919,
+      "step": 22390
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 49.97930908203125,
+      "learning_rate": 5.522800000000001e-06,
+      "loss": 0.7104,
+      "step": 22400
+    },
+    {
+      "epoch": 2.241,
+      "grad_norm": 36.936805725097656,
+      "learning_rate": 5.5208000000000006e-06,
+      "loss": 0.986,
+      "step": 22410
+    },
+    {
+      "epoch": 2.242,
+      "grad_norm": 26.627878189086914,
+      "learning_rate": 5.5188e-06,
+      "loss": 0.8035,
+      "step": 22420
+    },
+    {
+      "epoch": 2.243,
+      "grad_norm": 11.643138885498047,
+      "learning_rate": 5.5168e-06,
+      "loss": 0.5016,
+      "step": 22430
+    },
+    {
+      "epoch": 2.2439999999999998,
+      "grad_norm": 37.188358306884766,
+      "learning_rate": 5.5148e-06,
+      "loss": 0.8121,
+      "step": 22440
+    },
+    {
+      "epoch": 2.245,
+      "grad_norm": 10.066282272338867,
+      "learning_rate": 5.5128e-06,
+      "loss": 0.6035,
+      "step": 22450
+    },
+    {
+      "epoch": 2.246,
+      "grad_norm": 13.461889266967773,
+      "learning_rate": 5.510800000000001e-06,
+      "loss": 0.5461,
+      "step": 22460
+    },
+    {
+      "epoch": 2.247,
+      "grad_norm": 49.27286911010742,
+      "learning_rate": 5.5088000000000005e-06,
+      "loss": 0.7651,
+      "step": 22470
+    },
+    {
+      "epoch": 2.248,
+      "grad_norm": 30.79641342163086,
+      "learning_rate": 5.5067999999999995e-06,
+      "loss": 0.5639,
+      "step": 22480
+    },
+    {
+      "epoch": 2.249,
+      "grad_norm": 22.9559268951416,
+      "learning_rate": 5.5048e-06,
+      "loss": 0.7904,
+      "step": 22490
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.752798557281494,
+      "learning_rate": 5.5028e-06,
+      "loss": 0.7704,
+      "step": 22500
+    },
+    {
+      "epoch": 2.251,
+      "grad_norm": 9.406001091003418,
+      "learning_rate": 5.500800000000001e-06,
+      "loss": 0.867,
+      "step": 22510
+    },
+    {
+      "epoch": 2.252,
+      "grad_norm": 13.422652244567871,
+      "learning_rate": 5.498800000000001e-06,
+      "loss": 0.4994,
+      "step": 22520
+    },
+    {
+      "epoch": 2.253,
+      "grad_norm": 24.475555419921875,
+      "learning_rate": 5.4968e-06,
+      "loss": 0.74,
+      "step": 22530
+    },
+    {
+      "epoch": 2.254,
+      "grad_norm": 3.736154317855835,
+      "learning_rate": 5.4948e-06,
+      "loss": 0.5785,
+      "step": 22540
+    },
+    {
+      "epoch": 2.255,
+      "grad_norm": 1.4272233247756958,
+      "learning_rate": 5.4928e-06,
+      "loss": 1.0276,
+      "step": 22550
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 25.722667694091797,
+      "learning_rate": 5.490800000000001e-06,
+      "loss": 1.0477,
+      "step": 22560
+    },
+    {
+      "epoch": 2.257,
+      "grad_norm": 25.320680618286133,
+      "learning_rate": 5.488800000000001e-06,
+      "loss": 0.5366,
+      "step": 22570
+    },
+    {
+      "epoch": 2.258,
+      "grad_norm": 13.106619834899902,
+      "learning_rate": 5.4868e-06,
+      "loss": 0.4721,
+      "step": 22580
+    },
+    {
+      "epoch": 2.259,
+      "grad_norm": 1.7035945653915405,
+      "learning_rate": 5.4848e-06,
+      "loss": 0.6387,
+      "step": 22590
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.009597778320312,
+      "learning_rate": 5.4828e-06,
+      "loss": 0.6352,
+      "step": 22600
+    },
+    {
+      "epoch": 2.261,
+      "grad_norm": 12.074342727661133,
+      "learning_rate": 5.480800000000001e-06,
+      "loss": 0.7541,
+      "step": 22610
+    },
+    {
+      "epoch": 2.262,
+      "grad_norm": 38.93974304199219,
+      "learning_rate": 5.478800000000001e-06,
+      "loss": 0.8047,
+      "step": 22620
+    },
+    {
+      "epoch": 2.263,
+      "grad_norm": 15.713391304016113,
+      "learning_rate": 5.4768e-06,
+      "loss": 0.5585,
+      "step": 22630
+    },
+    {
+      "epoch": 2.2640000000000002,
+      "grad_norm": 20.313886642456055,
+      "learning_rate": 5.4748000000000005e-06,
+      "loss": 0.3573,
+      "step": 22640
+    },
+    {
+      "epoch": 2.265,
+      "grad_norm": 9.712434768676758,
+      "learning_rate": 5.4728e-06,
+      "loss": 0.5595,
+      "step": 22650
+    },
+    {
+      "epoch": 2.266,
+      "grad_norm": 32.309200286865234,
+      "learning_rate": 5.4708e-06,
+      "loss": 0.407,
+      "step": 22660
+    },
+    {
+      "epoch": 2.267,
+      "grad_norm": 46.730037689208984,
+      "learning_rate": 5.468800000000001e-06,
+      "loss": 0.8093,
+      "step": 22670
+    },
+    {
+      "epoch": 2.268,
+      "grad_norm": 11.148622512817383,
+      "learning_rate": 5.466800000000001e-06,
+      "loss": 0.6787,
+      "step": 22680
+    },
+    {
+      "epoch": 2.269,
+      "grad_norm": 47.33662033081055,
+      "learning_rate": 5.4648e-06,
+      "loss": 0.6553,
+      "step": 22690
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.559915542602539,
+      "learning_rate": 5.4628000000000005e-06,
+      "loss": 0.7154,
+      "step": 22700
+    },
+    {
+      "epoch": 2.271,
+      "grad_norm": 30.307464599609375,
+      "learning_rate": 5.4608e-06,
+      "loss": 0.5123,
+      "step": 22710
+    },
+    {
+      "epoch": 2.2720000000000002,
+      "grad_norm": 20.74936294555664,
+      "learning_rate": 5.458800000000001e-06,
+      "loss": 0.6701,
+      "step": 22720
+    },
+    {
+      "epoch": 2.273,
+      "grad_norm": 8.080650329589844,
+      "learning_rate": 5.456800000000001e-06,
+      "loss": 0.6648,
+      "step": 22730
+    },
+    {
+      "epoch": 2.274,
+      "grad_norm": 13.168947219848633,
+      "learning_rate": 5.4548e-06,
+      "loss": 0.4532,
+      "step": 22740
+    },
+    {
+      "epoch": 2.275,
+      "grad_norm": 42.89329147338867,
+      "learning_rate": 5.4528000000000005e-06,
+      "loss": 0.5806,
+      "step": 22750
+    },
+    {
+      "epoch": 2.276,
+      "grad_norm": 24.9879093170166,
+      "learning_rate": 5.4508e-06,
+      "loss": 0.6691,
+      "step": 22760
+    },
+    {
+      "epoch": 2.277,
+      "grad_norm": 28.52879524230957,
+      "learning_rate": 5.448800000000001e-06,
+      "loss": 0.4086,
+      "step": 22770
+    },
+    {
+      "epoch": 2.278,
+      "grad_norm": 20.092578887939453,
+      "learning_rate": 5.446800000000001e-06,
+      "loss": 0.7222,
+      "step": 22780
+    },
+    {
+      "epoch": 2.279,
+      "grad_norm": 34.03709411621094,
+      "learning_rate": 5.4448e-06,
+      "loss": 0.8185,
+      "step": 22790
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 25.46604347229004,
+      "learning_rate": 5.442800000000001e-06,
+      "loss": 0.643,
+      "step": 22800
+    },
+    {
+      "epoch": 2.281,
+      "grad_norm": 25.507854461669922,
+      "learning_rate": 5.4408000000000005e-06,
+      "loss": 0.5392,
+      "step": 22810
+    },
+    {
+      "epoch": 2.282,
+      "grad_norm": 20.695667266845703,
+      "learning_rate": 5.4388e-06,
+      "loss": 0.8377,
+      "step": 22820
+    },
+    {
+      "epoch": 2.283,
+      "grad_norm": 34.970245361328125,
+      "learning_rate": 5.436800000000001e-06,
+      "loss": 0.6591,
+      "step": 22830
+    },
+    {
+      "epoch": 2.284,
+      "grad_norm": 59.59330749511719,
+      "learning_rate": 5.4348e-06,
+      "loss": 0.9524,
+      "step": 22840
+    },
+    {
+      "epoch": 2.285,
+      "grad_norm": 46.58605194091797,
+      "learning_rate": 5.4328e-06,
+      "loss": 0.6674,
+      "step": 22850
+    },
+    {
+      "epoch": 2.286,
+      "grad_norm": 13.24176025390625,
+      "learning_rate": 5.430800000000001e-06,
+      "loss": 0.6838,
+      "step": 22860
+    },
+    {
+      "epoch": 2.287,
+      "grad_norm": 48.0534782409668,
+      "learning_rate": 5.4288000000000004e-06,
+      "loss": 0.6875,
+      "step": 22870
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 47.62519073486328,
+      "learning_rate": 5.426800000000001e-06,
+      "loss": 0.7306,
+      "step": 22880
+    },
+    {
+      "epoch": 2.289,
+      "grad_norm": 27.308670043945312,
+      "learning_rate": 5.4248e-06,
+      "loss": 0.5339,
+      "step": 22890
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 26.38911247253418,
+      "learning_rate": 5.4228e-06,
+      "loss": 0.7351,
+      "step": 22900
+    },
+    {
+      "epoch": 2.291,
+      "grad_norm": 41.1791877746582,
+      "learning_rate": 5.420800000000001e-06,
+      "loss": 0.5436,
+      "step": 22910
+    },
+    {
+      "epoch": 2.292,
+      "grad_norm": 26.171401977539062,
+      "learning_rate": 5.4188000000000005e-06,
+      "loss": 0.4981,
+      "step": 22920
+    },
+    {
+      "epoch": 2.293,
+      "grad_norm": 19.563926696777344,
+      "learning_rate": 5.416800000000001e-06,
+      "loss": 0.748,
+      "step": 22930
+    },
+    {
+      "epoch": 2.294,
+      "grad_norm": 47.672515869140625,
+      "learning_rate": 5.4148e-06,
+      "loss": 0.7037,
+      "step": 22940
+    },
+    {
+      "epoch": 2.295,
+      "grad_norm": 62.735050201416016,
+      "learning_rate": 5.4128e-06,
+      "loss": 0.6723,
+      "step": 22950
+    },
+    {
+      "epoch": 2.296,
+      "grad_norm": 34.751407623291016,
+      "learning_rate": 5.410800000000001e-06,
+      "loss": 0.5159,
+      "step": 22960
+    },
+    {
+      "epoch": 2.297,
+      "grad_norm": 7.955129623413086,
+      "learning_rate": 5.408800000000001e-06,
+      "loss": 0.5092,
+      "step": 22970
+    },
+    {
+      "epoch": 2.298,
+      "grad_norm": 26.632680892944336,
+      "learning_rate": 5.406800000000001e-06,
+      "loss": 0.5358,
+      "step": 22980
+    },
+    {
+      "epoch": 2.299,
+      "grad_norm": 8.946920394897461,
+      "learning_rate": 5.4048e-06,
+      "loss": 0.6629,
+      "step": 22990
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 24.032062530517578,
+      "learning_rate": 5.4028e-06,
+      "loss": 0.7799,
+      "step": 23000
+    },
+    {
+      "epoch": 2.301,
+      "grad_norm": 31.619062423706055,
+      "learning_rate": 5.400800000000001e-06,
+      "loss": 0.693,
+      "step": 23010
+    },
+    {
+      "epoch": 2.302,
+      "grad_norm": 57.403263092041016,
+      "learning_rate": 5.398800000000001e-06,
+      "loss": 0.5805,
+      "step": 23020
+    },
+    {
+      "epoch": 2.303,
+      "grad_norm": 40.945125579833984,
+      "learning_rate": 5.3968000000000006e-06,
+      "loss": 0.6838,
+      "step": 23030
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 26.93997573852539,
+      "learning_rate": 5.3948000000000004e-06,
+      "loss": 0.4315,
+      "step": 23040
+    },
+    {
+      "epoch": 2.305,
+      "grad_norm": 24.222333908081055,
+      "learning_rate": 5.3928e-06,
+      "loss": 0.5856,
+      "step": 23050
+    },
+    {
+      "epoch": 2.306,
+      "grad_norm": 29.424251556396484,
+      "learning_rate": 5.3908e-06,
+      "loss": 0.4816,
+      "step": 23060
+    },
+    {
+      "epoch": 2.307,
+      "grad_norm": 51.99330139160156,
+      "learning_rate": 5.388800000000001e-06,
+      "loss": 0.7163,
+      "step": 23070
+    },
+    {
+      "epoch": 2.308,
+      "grad_norm": 12.303630828857422,
+      "learning_rate": 5.386800000000001e-06,
+      "loss": 0.3429,
+      "step": 23080
+    },
+    {
+      "epoch": 2.309,
+      "grad_norm": 57.72380828857422,
+      "learning_rate": 5.3848e-06,
+      "loss": 0.4764,
+      "step": 23090
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 29.551307678222656,
+      "learning_rate": 5.3828e-06,
+      "loss": 0.8045,
+      "step": 23100
+    },
+    {
+      "epoch": 2.311,
+      "grad_norm": 39.42962646484375,
+      "learning_rate": 5.3808e-06,
+      "loss": 0.682,
+      "step": 23110
+    },
+    {
+      "epoch": 2.312,
+      "grad_norm": 8.102534294128418,
+      "learning_rate": 5.378800000000001e-06,
+      "loss": 0.6236,
+      "step": 23120
+    },
+    {
+      "epoch": 2.313,
+      "grad_norm": 34.7684211730957,
+      "learning_rate": 5.376800000000001e-06,
+      "loss": 0.4118,
+      "step": 23130
+    },
+    {
+      "epoch": 2.314,
+      "grad_norm": 18.475805282592773,
+      "learning_rate": 5.3748e-06,
+      "loss": 0.5975,
+      "step": 23140
+    },
+    {
+      "epoch": 2.315,
+      "grad_norm": 34.82725524902344,
+      "learning_rate": 5.3728000000000005e-06,
+      "loss": 0.6382,
+      "step": 23150
+    },
+    {
+      "epoch": 2.316,
+      "grad_norm": 33.135921478271484,
+      "learning_rate": 5.3708e-06,
+      "loss": 0.6132,
+      "step": 23160
+    },
+    {
+      "epoch": 2.317,
+      "grad_norm": 31.031635284423828,
+      "learning_rate": 5.368800000000001e-06,
+      "loss": 0.3783,
+      "step": 23170
+    },
+    {
+      "epoch": 2.318,
+      "grad_norm": 44.96309280395508,
+      "learning_rate": 5.366800000000001e-06,
+      "loss": 0.8038,
+      "step": 23180
+    },
+    {
+      "epoch": 2.319,
+      "grad_norm": 50.742454528808594,
+      "learning_rate": 5.3648e-06,
+      "loss": 0.6723,
+      "step": 23190
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 19.635595321655273,
+      "learning_rate": 5.3628000000000006e-06,
+      "loss": 0.4662,
+      "step": 23200
+    },
+    {
+      "epoch": 2.321,
+      "grad_norm": 36.878971099853516,
+      "learning_rate": 5.3608e-06,
+      "loss": 0.6226,
+      "step": 23210
+    },
+    {
+      "epoch": 2.322,
+      "grad_norm": 6.729045867919922,
+      "learning_rate": 5.3588e-06,
+      "loss": 0.4275,
+      "step": 23220
+    },
+    {
+      "epoch": 2.323,
+      "grad_norm": 52.66419982910156,
+      "learning_rate": 5.356800000000001e-06,
+      "loss": 0.5837,
+      "step": 23230
+    },
+    {
+      "epoch": 2.324,
+      "grad_norm": 44.77753448486328,
+      "learning_rate": 5.3548e-06,
+      "loss": 0.7493,
+      "step": 23240
+    },
+    {
+      "epoch": 2.325,
+      "grad_norm": 14.312146186828613,
+      "learning_rate": 5.3528e-06,
+      "loss": 0.5975,
+      "step": 23250
+    },
+    {
+      "epoch": 2.326,
+      "grad_norm": 47.50712203979492,
+      "learning_rate": 5.3508000000000005e-06,
+      "loss": 0.6969,
+      "step": 23260
+    },
+    {
+      "epoch": 2.327,
+      "grad_norm": 55.93048095703125,
+      "learning_rate": 5.3488e-06,
+      "loss": 0.9151,
+      "step": 23270
+    },
+    {
+      "epoch": 2.328,
+      "grad_norm": 33.65269470214844,
+      "learning_rate": 5.346800000000001e-06,
+      "loss": 0.611,
+      "step": 23280
+    },
+    {
+      "epoch": 2.329,
+      "grad_norm": 17.26365852355957,
+      "learning_rate": 5.3448e-06,
+      "loss": 0.675,
+      "step": 23290
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 25.64186668395996,
+      "learning_rate": 5.3428e-06,
+      "loss": 0.6604,
+      "step": 23300
+    },
+    {
+      "epoch": 2.331,
+      "grad_norm": 48.147682189941406,
+      "learning_rate": 5.340800000000001e-06,
+      "loss": 0.8392,
+      "step": 23310
+    },
+    {
+      "epoch": 2.332,
+      "grad_norm": 25.545421600341797,
+      "learning_rate": 5.3388000000000005e-06,
+      "loss": 0.7228,
+      "step": 23320
+    },
+    {
+      "epoch": 2.333,
+      "grad_norm": 28.70657730102539,
+      "learning_rate": 5.336800000000001e-06,
+      "loss": 0.6491,
+      "step": 23330
+    },
+    {
+      "epoch": 2.334,
+      "grad_norm": 34.852413177490234,
+      "learning_rate": 5.3348e-06,
+      "loss": 0.5739,
+      "step": 23340
+    },
+    {
+      "epoch": 2.335,
+      "grad_norm": 9.7340087890625,
+      "learning_rate": 5.3328e-06,
+      "loss": 0.8116,
+      "step": 23350
+    },
+    {
+      "epoch": 2.336,
+      "grad_norm": 26.43950843811035,
+      "learning_rate": 5.330800000000001e-06,
+      "loss": 0.5883,
+      "step": 23360
+    },
+    {
+      "epoch": 2.337,
+      "grad_norm": 23.425874710083008,
+      "learning_rate": 5.3288000000000005e-06,
+      "loss": 0.7699,
+      "step": 23370
+    },
+    {
+      "epoch": 2.338,
+      "grad_norm": 44.046539306640625,
+      "learning_rate": 5.326800000000001e-06,
+      "loss": 0.926,
+      "step": 23380
+    },
+    {
+      "epoch": 2.339,
+      "grad_norm": 39.17654800415039,
+      "learning_rate": 5.3248e-06,
+      "loss": 0.5595,
+      "step": 23390
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 36.046775817871094,
+      "learning_rate": 5.3228e-06,
+      "loss": 0.57,
+      "step": 23400
+    },
+    {
+      "epoch": 2.341,
+      "grad_norm": 17.362285614013672,
+      "learning_rate": 5.320800000000001e-06,
+      "loss": 0.6189,
+      "step": 23410
+    },
+    {
+      "epoch": 2.342,
+      "grad_norm": 17.513729095458984,
+      "learning_rate": 5.318800000000001e-06,
+      "loss": 0.54,
+      "step": 23420
+    },
+    {
+      "epoch": 2.343,
+      "grad_norm": 45.768333435058594,
+      "learning_rate": 5.3168000000000005e-06,
+      "loss": 0.752,
+      "step": 23430
+    },
+    {
+      "epoch": 2.344,
+      "grad_norm": 36.07986831665039,
+      "learning_rate": 5.3148e-06,
+      "loss": 0.649,
+      "step": 23440
+    },
+    {
+      "epoch": 2.3449999999999998,
+      "grad_norm": 50.08828353881836,
+      "learning_rate": 5.3128e-06,
+      "loss": 0.6256,
+      "step": 23450
+    },
+    {
+      "epoch": 2.346,
+      "grad_norm": 21.56818389892578,
+      "learning_rate": 5.3108e-06,
+      "loss": 0.5439,
+      "step": 23460
+    },
+    {
+      "epoch": 2.347,
+      "grad_norm": 12.028852462768555,
+      "learning_rate": 5.308800000000001e-06,
+      "loss": 0.2928,
+      "step": 23470
+    },
+    {
+      "epoch": 2.348,
+      "grad_norm": 25.907634735107422,
+      "learning_rate": 5.306800000000001e-06,
+      "loss": 0.4197,
+      "step": 23480
+    },
+    {
+      "epoch": 2.349,
+      "grad_norm": 25.22545623779297,
+      "learning_rate": 5.3048e-06,
+      "loss": 0.4871,
+      "step": 23490
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 27.278261184692383,
+      "learning_rate": 5.3028e-06,
+      "loss": 0.5702,
+      "step": 23500
+    },
+    {
+      "epoch": 2.351,
+      "grad_norm": 6.1130194664001465,
+      "learning_rate": 5.3008e-06,
+      "loss": 0.5626,
+      "step": 23510
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 28.173141479492188,
+      "learning_rate": 5.298800000000001e-06,
+      "loss": 0.6642,
+      "step": 23520
+    },
+    {
+      "epoch": 2.3529999999999998,
+      "grad_norm": 45.549293518066406,
+      "learning_rate": 5.296800000000001e-06,
+      "loss": 0.7794,
+      "step": 23530
+    },
+    {
+      "epoch": 2.354,
+      "grad_norm": 59.046600341796875,
+      "learning_rate": 5.2948e-06,
+      "loss": 0.8257,
+      "step": 23540
+    },
+    {
+      "epoch": 2.355,
+      "grad_norm": 30.231481552124023,
+      "learning_rate": 5.2928e-06,
+      "loss": 0.5337,
+      "step": 23550
+    },
+    {
+      "epoch": 2.356,
+      "grad_norm": 28.655134201049805,
+      "learning_rate": 5.2908e-06,
+      "loss": 0.6839,
+      "step": 23560
+    },
+    {
+      "epoch": 2.357,
+      "grad_norm": 57.767459869384766,
+      "learning_rate": 5.288800000000001e-06,
+      "loss": 0.9569,
+      "step": 23570
+    },
+    {
+      "epoch": 2.358,
+      "grad_norm": 55.075984954833984,
+      "learning_rate": 5.286800000000001e-06,
+      "loss": 0.6479,
+      "step": 23580
+    },
+    {
+      "epoch": 2.359,
+      "grad_norm": 14.50174617767334,
+      "learning_rate": 5.2848e-06,
+      "loss": 0.7389,
+      "step": 23590
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 36.16502380371094,
+      "learning_rate": 5.2828000000000005e-06,
+      "loss": 0.7146,
+      "step": 23600
+    },
+    {
+      "epoch": 2.3609999999999998,
+      "grad_norm": 20.909074783325195,
+      "learning_rate": 5.2808e-06,
+      "loss": 0.4444,
+      "step": 23610
+    },
+    {
+      "epoch": 2.362,
+      "grad_norm": 34.6703987121582,
+      "learning_rate": 5.2788e-06,
+      "loss": 0.7386,
+      "step": 23620
+    },
+    {
+      "epoch": 2.363,
+      "grad_norm": 10.016300201416016,
+      "learning_rate": 5.276800000000001e-06,
+      "loss": 0.5363,
+      "step": 23630
+    },
+    {
+      "epoch": 2.364,
+      "grad_norm": 15.961995124816895,
+      "learning_rate": 5.2748e-06,
+      "loss": 0.434,
+      "step": 23640
+    },
+    {
+      "epoch": 2.365,
+      "grad_norm": 29.54912757873535,
+      "learning_rate": 5.2728e-06,
+      "loss": 0.5253,
+      "step": 23650
+    },
+    {
+      "epoch": 2.366,
+      "grad_norm": 46.137672424316406,
+      "learning_rate": 5.2708000000000004e-06,
+      "loss": 0.6896,
+      "step": 23660
+    },
+    {
+      "epoch": 2.367,
+      "grad_norm": 26.812349319458008,
+      "learning_rate": 5.2688e-06,
+      "loss": 0.5173,
+      "step": 23670
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 25.67329216003418,
+      "learning_rate": 5.266800000000001e-06,
+      "loss": 0.3728,
+      "step": 23680
+    },
+    {
+      "epoch": 2.3689999999999998,
+      "grad_norm": 56.716339111328125,
+      "learning_rate": 5.2648e-06,
+      "loss": 0.6709,
+      "step": 23690
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 81.84050750732422,
+      "learning_rate": 5.2628e-06,
+      "loss": 0.8044,
+      "step": 23700
+    },
+    {
+      "epoch": 2.371,
+      "grad_norm": 45.409515380859375,
+      "learning_rate": 5.2608000000000005e-06,
+      "loss": 0.491,
+      "step": 23710
+    },
+    {
+      "epoch": 2.372,
+      "grad_norm": 48.80219650268555,
+      "learning_rate": 5.2588e-06,
+      "loss": 0.4556,
+      "step": 23720
+    },
+    {
+      "epoch": 2.373,
+      "grad_norm": 23.4346981048584,
+      "learning_rate": 5.256800000000001e-06,
+      "loss": 0.8579,
+      "step": 23730
+    },
+    {
+      "epoch": 2.374,
+      "grad_norm": 33.4806022644043,
+      "learning_rate": 5.2548e-06,
+      "loss": 0.8229,
+      "step": 23740
+    },
+    {
+      "epoch": 2.375,
+      "grad_norm": 49.55470657348633,
+      "learning_rate": 5.2528e-06,
+      "loss": 0.6262,
+      "step": 23750
+    },
+    {
+      "epoch": 2.376,
+      "grad_norm": 27.586681365966797,
+      "learning_rate": 5.250800000000001e-06,
+      "loss": 0.6945,
+      "step": 23760
+    },
+    {
+      "epoch": 2.377,
+      "grad_norm": 56.0458869934082,
+      "learning_rate": 5.2488000000000005e-06,
+      "loss": 0.9446,
+      "step": 23770
+    },
+    {
+      "epoch": 2.378,
+      "grad_norm": 33.53122329711914,
+      "learning_rate": 5.246800000000001e-06,
+      "loss": 0.7489,
+      "step": 23780
+    },
+    {
+      "epoch": 2.379,
+      "grad_norm": 24.26375961303711,
+      "learning_rate": 5.2448e-06,
+      "loss": 0.5203,
+      "step": 23790
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 16.765478134155273,
+      "learning_rate": 5.2428e-06,
+      "loss": 0.8502,
+      "step": 23800
+    },
+    {
+      "epoch": 2.3810000000000002,
+      "grad_norm": 52.03241729736328,
+      "learning_rate": 5.240800000000001e-06,
+      "loss": 0.4692,
+      "step": 23810
+    },
+    {
+      "epoch": 2.382,
+      "grad_norm": 26.394987106323242,
+      "learning_rate": 5.2388000000000006e-06,
+      "loss": 0.5537,
+      "step": 23820
+    },
+    {
+      "epoch": 2.383,
+      "grad_norm": 47.837158203125,
+      "learning_rate": 5.2368e-06,
+      "loss": 0.6629,
+      "step": 23830
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 4.176435947418213,
+      "learning_rate": 5.2348e-06,
+      "loss": 0.56,
+      "step": 23840
+    },
+    {
+      "epoch": 2.385,
+      "grad_norm": 29.736774444580078,
+      "learning_rate": 5.2328e-06,
+      "loss": 0.6556,
+      "step": 23850
+    },
+    {
+      "epoch": 2.386,
+      "grad_norm": 28.325538635253906,
+      "learning_rate": 5.2308e-06,
+      "loss": 0.7128,
+      "step": 23860
+    },
+    {
+      "epoch": 2.387,
+      "grad_norm": 33.87845230102539,
+      "learning_rate": 5.228800000000001e-06,
+      "loss": 0.7559,
+      "step": 23870
+    },
+    {
+      "epoch": 2.388,
+      "grad_norm": 27.20892906188965,
+      "learning_rate": 5.2268000000000005e-06,
+      "loss": 0.5474,
+      "step": 23880
+    },
+    {
+      "epoch": 2.3890000000000002,
+      "grad_norm": 4.944998741149902,
+      "learning_rate": 5.224800000000001e-06,
+      "loss": 0.6466,
+      "step": 23890
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 62.19276428222656,
+      "learning_rate": 5.2228e-06,
+      "loss": 0.6979,
+      "step": 23900
+    },
+    {
+      "epoch": 2.391,
+      "grad_norm": 47.98637390136719,
+      "learning_rate": 5.2208e-06,
+      "loss": 0.6794,
+      "step": 23910
+    },
+    {
+      "epoch": 2.392,
+      "grad_norm": 40.98355484008789,
+      "learning_rate": 5.218800000000001e-06,
+      "loss": 0.7471,
+      "step": 23920
+    },
+    {
+      "epoch": 2.393,
+      "grad_norm": 44.224998474121094,
+      "learning_rate": 5.216800000000001e-06,
+      "loss": 0.5519,
+      "step": 23930
+    },
+    {
+      "epoch": 2.394,
+      "grad_norm": 33.322776794433594,
+      "learning_rate": 5.214800000000001e-06,
+      "loss": 0.4908,
+      "step": 23940
+    },
+    {
+      "epoch": 2.395,
+      "grad_norm": 33.787879943847656,
+      "learning_rate": 5.2128e-06,
+      "loss": 0.7247,
+      "step": 23950
+    },
+    {
+      "epoch": 2.396,
+      "grad_norm": 28.912538528442383,
+      "learning_rate": 5.2108e-06,
+      "loss": 0.5081,
+      "step": 23960
+    },
+    {
+      "epoch": 2.3970000000000002,
+      "grad_norm": 28.065685272216797,
+      "learning_rate": 5.208800000000001e-06,
+      "loss": 0.8082,
+      "step": 23970
+    },
+    {
+      "epoch": 2.398,
+      "grad_norm": 2.713496446609497,
+      "learning_rate": 5.206800000000001e-06,
+      "loss": 0.5147,
+      "step": 23980
+    },
+    {
+      "epoch": 2.399,
+      "grad_norm": 18.171648025512695,
+      "learning_rate": 5.2048000000000005e-06,
+      "loss": 0.817,
+      "step": 23990
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 23.39317512512207,
+      "learning_rate": 5.2028e-06,
+      "loss": 0.6705,
+      "step": 24000
+    },
+    {
+      "epoch": 2.401,
+      "grad_norm": 9.348867416381836,
+      "learning_rate": 5.2008e-06,
+      "loss": 0.5186,
+      "step": 24010
+    },
+    {
+      "epoch": 2.402,
+      "grad_norm": 52.66146469116211,
+      "learning_rate": 5.1988e-06,
+      "loss": 0.6322,
+      "step": 24020
+    },
+    {
+      "epoch": 2.403,
+      "grad_norm": 25.316165924072266,
+      "learning_rate": 5.196800000000001e-06,
+      "loss": 0.5371,
+      "step": 24030
+    },
+    {
+      "epoch": 2.404,
+      "grad_norm": 14.026839256286621,
+      "learning_rate": 5.194800000000001e-06,
+      "loss": 0.4817,
+      "step": 24040
+    },
+    {
+      "epoch": 2.4050000000000002,
+      "grad_norm": 37.170867919921875,
+      "learning_rate": 5.1928e-06,
+      "loss": 1.0024,
+      "step": 24050
+    },
+    {
+      "epoch": 2.406,
+      "grad_norm": 12.929289817810059,
+      "learning_rate": 5.1908e-06,
+      "loss": 0.5727,
+      "step": 24060
+    },
+    {
+      "epoch": 2.407,
+      "grad_norm": 16.34523582458496,
+      "learning_rate": 5.1888e-06,
+      "loss": 0.5111,
+      "step": 24070
+    },
+    {
+      "epoch": 2.408,
+      "grad_norm": 68.75968933105469,
+      "learning_rate": 5.186800000000001e-06,
+      "loss": 0.7928,
+      "step": 24080
+    },
+    {
+      "epoch": 2.409,
+      "grad_norm": 23.91437339782715,
+      "learning_rate": 5.184800000000001e-06,
+      "loss": 0.7094,
+      "step": 24090
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 52.49360656738281,
+      "learning_rate": 5.1828e-06,
+      "loss": 0.5987,
+      "step": 24100
+    },
+    {
+      "epoch": 2.411,
+      "grad_norm": 35.196495056152344,
+      "learning_rate": 5.1808000000000004e-06,
+      "loss": 0.6937,
+      "step": 24110
+    },
+    {
+      "epoch": 2.412,
+      "grad_norm": 53.06798553466797,
+      "learning_rate": 5.1788e-06,
+      "loss": 0.4998,
+      "step": 24120
+    },
+    {
+      "epoch": 2.413,
+      "grad_norm": 33.701969146728516,
+      "learning_rate": 5.176800000000001e-06,
+      "loss": 0.814,
+      "step": 24130
+    },
+    {
+      "epoch": 2.414,
+      "grad_norm": 18.027755737304688,
+      "learning_rate": 5.174800000000001e-06,
+      "loss": 0.7363,
+      "step": 24140
+    },
+    {
+      "epoch": 2.415,
+      "grad_norm": 47.720741271972656,
+      "learning_rate": 5.1728e-06,
+      "loss": 0.6642,
+      "step": 24150
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 62.25669860839844,
+      "learning_rate": 5.1708000000000005e-06,
+      "loss": 0.8371,
+      "step": 24160
+    },
+    {
+      "epoch": 2.417,
+      "grad_norm": 33.065185546875,
+      "learning_rate": 5.1688e-06,
+      "loss": 0.7936,
+      "step": 24170
+    },
+    {
+      "epoch": 2.418,
+      "grad_norm": 13.660017013549805,
+      "learning_rate": 5.166800000000001e-06,
+      "loss": 0.7092,
+      "step": 24180
+    },
+    {
+      "epoch": 2.419,
+      "grad_norm": 28.82988739013672,
+      "learning_rate": 5.164800000000001e-06,
+      "loss": 0.6911,
+      "step": 24190
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.214262008666992,
+      "learning_rate": 5.1628e-06,
+      "loss": 0.7027,
+      "step": 24200
+    },
+    {
+      "epoch": 2.421,
+      "grad_norm": 60.720298767089844,
+      "learning_rate": 5.160800000000001e-06,
+      "loss": 0.7199,
+      "step": 24210
+    },
+    {
+      "epoch": 2.422,
+      "grad_norm": 31.063989639282227,
+      "learning_rate": 5.1588000000000005e-06,
+      "loss": 0.5332,
+      "step": 24220
+    },
+    {
+      "epoch": 2.423,
+      "grad_norm": 383.9568176269531,
+      "learning_rate": 5.1568e-06,
+      "loss": 0.6992,
+      "step": 24230
+    },
+    {
+      "epoch": 2.424,
+      "grad_norm": 46.2714729309082,
+      "learning_rate": 5.154800000000001e-06,
+      "loss": 0.6366,
+      "step": 24240
+    },
+    {
+      "epoch": 2.425,
+      "grad_norm": 35.338539123535156,
+      "learning_rate": 5.1528e-06,
+      "loss": 0.8147,
+      "step": 24250
+    },
+    {
+      "epoch": 2.426,
+      "grad_norm": 34.018646240234375,
+      "learning_rate": 5.1508e-06,
+      "loss": 0.7693,
+      "step": 24260
+    },
+    {
+      "epoch": 2.427,
+      "grad_norm": 46.39628982543945,
+      "learning_rate": 5.148800000000001e-06,
+      "loss": 0.6542,
+      "step": 24270
+    },
+    {
+      "epoch": 2.428,
+      "grad_norm": 8.005655288696289,
+      "learning_rate": 5.1468000000000004e-06,
+      "loss": 0.4196,
+      "step": 24280
+    },
+    {
+      "epoch": 2.429,
+      "grad_norm": 31.6107234954834,
+      "learning_rate": 5.144800000000001e-06,
+      "loss": 0.6921,
+      "step": 24290
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.957141876220703,
+      "learning_rate": 5.1428e-06,
+      "loss": 0.5301,
+      "step": 24300
+    },
+    {
+      "epoch": 2.431,
+      "grad_norm": 39.83584213256836,
+      "learning_rate": 5.1408e-06,
+      "loss": 0.6387,
+      "step": 24310
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 24.404521942138672,
+      "learning_rate": 5.138800000000001e-06,
+      "loss": 0.4989,
+      "step": 24320
+    },
+    {
+      "epoch": 2.433,
+      "grad_norm": 20.31212043762207,
+      "learning_rate": 5.1368000000000005e-06,
+      "loss": 0.5232,
+      "step": 24330
+    },
+    {
+      "epoch": 2.434,
+      "grad_norm": 44.953060150146484,
+      "learning_rate": 5.134800000000001e-06,
+      "loss": 0.6488,
+      "step": 24340
+    },
+    {
+      "epoch": 2.435,
+      "grad_norm": 42.72674560546875,
+      "learning_rate": 5.1328e-06,
+      "loss": 0.6257,
+      "step": 24350
+    },
+    {
+      "epoch": 2.436,
+      "grad_norm": 42.03239440917969,
+      "learning_rate": 5.1308e-06,
+      "loss": 0.7274,
+      "step": 24360
+    },
+    {
+      "epoch": 2.437,
+      "grad_norm": 40.29648971557617,
+      "learning_rate": 5.128800000000001e-06,
+      "loss": 0.4868,
+      "step": 24370
+    },
+    {
+      "epoch": 2.438,
+      "grad_norm": 9.804975509643555,
+      "learning_rate": 5.126800000000001e-06,
+      "loss": 0.6022,
+      "step": 24380
+    },
+    {
+      "epoch": 2.439,
+      "grad_norm": 18.629207611083984,
+      "learning_rate": 5.1248000000000005e-06,
+      "loss": 0.6451,
+      "step": 24390
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 52.884403228759766,
+      "learning_rate": 5.1228e-06,
+      "loss": 0.5821,
+      "step": 24400
+    },
+    {
+      "epoch": 2.441,
+      "grad_norm": 41.7515754699707,
+      "learning_rate": 5.1208e-06,
+      "loss": 0.7286,
+      "step": 24410
+    },
+    {
+      "epoch": 2.442,
+      "grad_norm": 44.79233169555664,
+      "learning_rate": 5.1188e-06,
+      "loss": 0.5648,
+      "step": 24420
+    },
+    {
+      "epoch": 2.443,
+      "grad_norm": 17.482986450195312,
+      "learning_rate": 5.116800000000001e-06,
+      "loss": 0.6169,
+      "step": 24430
+    },
+    {
+      "epoch": 2.444,
+      "grad_norm": 49.76847457885742,
+      "learning_rate": 5.1148000000000006e-06,
+      "loss": 0.5268,
+      "step": 24440
+    },
+    {
+      "epoch": 2.445,
+      "grad_norm": 88.84921264648438,
+      "learning_rate": 5.1127999999999996e-06,
+      "loss": 0.9343,
+      "step": 24450
+    },
+    {
+      "epoch": 2.446,
+      "grad_norm": 52.439720153808594,
+      "learning_rate": 5.1108e-06,
+      "loss": 0.6185,
+      "step": 24460
+    },
+    {
+      "epoch": 2.447,
+      "grad_norm": 3.1405117511749268,
+      "learning_rate": 5.1090000000000006e-06,
+      "loss": 0.3354,
+      "step": 24470
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 37.97247314453125,
+      "learning_rate": 5.107000000000001e-06,
+      "loss": 0.5834,
+      "step": 24480
+    },
+    {
+      "epoch": 2.449,
+      "grad_norm": 3.9048521518707275,
+      "learning_rate": 5.105e-06,
+      "loss": 0.5524,
+      "step": 24490
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 45.24668502807617,
+      "learning_rate": 5.103e-06,
+      "loss": 0.4898,
+      "step": 24500
+    },
+    {
+      "epoch": 2.451,
+      "grad_norm": 46.744415283203125,
+      "learning_rate": 5.101000000000001e-06,
+      "loss": 0.7093,
+      "step": 24510
+    },
+    {
+      "epoch": 2.452,
+      "grad_norm": 36.82590866088867,
+      "learning_rate": 5.099000000000001e-06,
+      "loss": 0.5571,
+      "step": 24520
+    },
+    {
+      "epoch": 2.453,
+      "grad_norm": 14.219284057617188,
+      "learning_rate": 5.0970000000000005e-06,
+      "loss": 0.3158,
+      "step": 24530
+    },
+    {
+      "epoch": 2.454,
+      "grad_norm": 4.428883075714111,
+      "learning_rate": 5.095e-06,
+      "loss": 0.6263,
+      "step": 24540
+    },
+    {
+      "epoch": 2.455,
+      "grad_norm": 40.982872009277344,
+      "learning_rate": 5.093e-06,
+      "loss": 0.7157,
+      "step": 24550
+    },
+    {
+      "epoch": 2.456,
+      "grad_norm": 36.43859100341797,
+      "learning_rate": 5.091e-06,
+      "loss": 0.7041,
+      "step": 24560
+    },
+    {
+      "epoch": 2.457,
+      "grad_norm": 21.157794952392578,
+      "learning_rate": 5.089000000000001e-06,
+      "loss": 0.4459,
+      "step": 24570
+    },
+    {
+      "epoch": 2.458,
+      "grad_norm": 15.586324691772461,
+      "learning_rate": 5.087000000000001e-06,
+      "loss": 0.7045,
+      "step": 24580
+    },
+    {
+      "epoch": 2.459,
+      "grad_norm": 23.27480697631836,
+      "learning_rate": 5.085e-06,
+      "loss": 0.4485,
+      "step": 24590
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 30.3853702545166,
+      "learning_rate": 5.083e-06,
+      "loss": 0.7559,
+      "step": 24600
+    },
+    {
+      "epoch": 2.461,
+      "grad_norm": 62.69881820678711,
+      "learning_rate": 5.081e-06,
+      "loss": 0.7755,
+      "step": 24610
+    },
+    {
+      "epoch": 2.462,
+      "grad_norm": 35.382137298583984,
+      "learning_rate": 5.079000000000001e-06,
+      "loss": 0.7302,
+      "step": 24620
+    },
+    {
+      "epoch": 2.463,
+      "grad_norm": 5.012226104736328,
+      "learning_rate": 5.077000000000001e-06,
+      "loss": 0.4293,
+      "step": 24630
+    },
+    {
+      "epoch": 2.464,
+      "grad_norm": 30.396087646484375,
+      "learning_rate": 5.075e-06,
+      "loss": 0.5469,
+      "step": 24640
+    },
+    {
+      "epoch": 2.465,
+      "grad_norm": 23.97662925720215,
+      "learning_rate": 5.073e-06,
+      "loss": 0.4722,
+      "step": 24650
+    },
+    {
+      "epoch": 2.466,
+      "grad_norm": 61.57948303222656,
+      "learning_rate": 5.071e-06,
+      "loss": 0.5139,
+      "step": 24660
+    },
+    {
+      "epoch": 2.467,
+      "grad_norm": 64.28836822509766,
+      "learning_rate": 5.069000000000001e-06,
+      "loss": 0.5281,
+      "step": 24670
+    },
+    {
+      "epoch": 2.468,
+      "grad_norm": 17.3576602935791,
+      "learning_rate": 5.067000000000001e-06,
+      "loss": 0.8598,
+      "step": 24680
+    },
+    {
+      "epoch": 2.469,
+      "grad_norm": 48.70146942138672,
+      "learning_rate": 5.065e-06,
+      "loss": 0.6562,
+      "step": 24690
+    },
+    {
+      "epoch": 2.4699999999999998,
+      "grad_norm": 55.60419845581055,
+      "learning_rate": 5.0630000000000005e-06,
+      "loss": 0.6233,
+      "step": 24700
+    },
+    {
+      "epoch": 2.471,
+      "grad_norm": 12.617278099060059,
+      "learning_rate": 5.061e-06,
+      "loss": 0.5946,
+      "step": 24710
+    },
+    {
+      "epoch": 2.472,
+      "grad_norm": 10.002517700195312,
+      "learning_rate": 5.059e-06,
+      "loss": 0.6976,
+      "step": 24720
+    },
+    {
+      "epoch": 2.473,
+      "grad_norm": 44.83880615234375,
+      "learning_rate": 5.057000000000001e-06,
+      "loss": 0.6083,
+      "step": 24730
+    },
+    {
+      "epoch": 2.474,
+      "grad_norm": 40.62992858886719,
+      "learning_rate": 5.055e-06,
+      "loss": 0.5168,
+      "step": 24740
+    },
+    {
+      "epoch": 2.475,
+      "grad_norm": 39.006439208984375,
+      "learning_rate": 5.053e-06,
+      "loss": 0.6952,
+      "step": 24750
+    },
+    {
+      "epoch": 2.476,
+      "grad_norm": 22.533584594726562,
+      "learning_rate": 5.0510000000000004e-06,
+      "loss": 0.5019,
+      "step": 24760
+    },
+    {
+      "epoch": 2.477,
+      "grad_norm": 37.3040771484375,
+      "learning_rate": 5.049e-06,
+      "loss": 0.5401,
+      "step": 24770
+    },
+    {
+      "epoch": 2.4779999999999998,
+      "grad_norm": 53.25998306274414,
+      "learning_rate": 5.047000000000001e-06,
+      "loss": 0.6991,
+      "step": 24780
+    },
+    {
+      "epoch": 2.479,
+      "grad_norm": 11.337082862854004,
+      "learning_rate": 5.045e-06,
+      "loss": 0.4653,
+      "step": 24790
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.860915660858154,
+      "learning_rate": 5.043e-06,
+      "loss": 0.7585,
+      "step": 24800
+    },
+    {
+      "epoch": 2.481,
+      "grad_norm": 41.77631759643555,
+      "learning_rate": 5.0410000000000005e-06,
+      "loss": 0.8433,
+      "step": 24810
+    },
+    {
+      "epoch": 2.482,
+      "grad_norm": 13.934993743896484,
+      "learning_rate": 5.039e-06,
+      "loss": 0.5707,
+      "step": 24820
+    },
+    {
+      "epoch": 2.483,
+      "grad_norm": 69.27035522460938,
+      "learning_rate": 5.037000000000001e-06,
+      "loss": 0.6715,
+      "step": 24830
+    },
+    {
+      "epoch": 2.484,
+      "grad_norm": 19.001020431518555,
+      "learning_rate": 5.035e-06,
+      "loss": 0.3987,
+      "step": 24840
+    },
+    {
+      "epoch": 2.485,
+      "grad_norm": 20.378995895385742,
+      "learning_rate": 5.033e-06,
+      "loss": 0.8881,
+      "step": 24850
+    },
+    {
+      "epoch": 2.4859999999999998,
+      "grad_norm": 36.096031188964844,
+      "learning_rate": 5.031000000000001e-06,
+      "loss": 0.6329,
+      "step": 24860
+    },
+    {
+      "epoch": 2.487,
+      "grad_norm": 53.7542610168457,
+      "learning_rate": 5.0290000000000005e-06,
+      "loss": 0.6449,
+      "step": 24870
+    },
+    {
+      "epoch": 2.488,
+      "grad_norm": 24.860681533813477,
+      "learning_rate": 5.027000000000001e-06,
+      "loss": 0.3337,
+      "step": 24880
+    },
+    {
+      "epoch": 2.489,
+      "grad_norm": 51.24372482299805,
+      "learning_rate": 5.025e-06,
+      "loss": 1.04,
+      "step": 24890
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 30.71172523498535,
+      "learning_rate": 5.023e-06,
+      "loss": 0.6922,
+      "step": 24900
+    },
+    {
+      "epoch": 2.491,
+      "grad_norm": 29.371429443359375,
+      "learning_rate": 5.021000000000001e-06,
+      "loss": 0.6405,
+      "step": 24910
+    },
+    {
+      "epoch": 2.492,
+      "grad_norm": 26.474454879760742,
+      "learning_rate": 5.0190000000000006e-06,
+      "loss": 0.2826,
+      "step": 24920
+    },
+    {
+      "epoch": 2.493,
+      "grad_norm": 43.77047348022461,
+      "learning_rate": 5.017e-06,
+      "loss": 0.9044,
+      "step": 24930
+    },
+    {
+      "epoch": 2.4939999999999998,
+      "grad_norm": 36.24064254760742,
+      "learning_rate": 5.015e-06,
+      "loss": 0.5805,
+      "step": 24940
+    },
+    {
+      "epoch": 2.495,
+      "grad_norm": 3.8243775367736816,
+      "learning_rate": 5.013e-06,
+      "loss": 0.7752,
+      "step": 24950
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 29.713932037353516,
+      "learning_rate": 5.011e-06,
+      "loss": 0.5483,
+      "step": 24960
+    },
+    {
+      "epoch": 2.497,
+      "grad_norm": 31.567214965820312,
+      "learning_rate": 5.009000000000001e-06,
+      "loss": 0.5147,
+      "step": 24970
+    },
+    {
+      "epoch": 2.498,
+      "grad_norm": 28.092866897583008,
+      "learning_rate": 5.0070000000000005e-06,
+      "loss": 0.8079,
+      "step": 24980
+    },
+    {
+      "epoch": 2.499,
+      "grad_norm": 12.45438003540039,
+      "learning_rate": 5.0049999999999995e-06,
+      "loss": 0.5923,
+      "step": 24990
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 41.411476135253906,
+      "learning_rate": 5.003e-06,
+      "loss": 0.5853,
+      "step": 25000
+    },
+    {
+      "epoch": 2.501,
+      "grad_norm": 26.281492233276367,
+      "learning_rate": 5.001e-06,
+      "loss": 0.7854,
+      "step": 25010
+    },
+    {
+      "epoch": 2.502,
+      "grad_norm": 31.148704528808594,
+      "learning_rate": 4.999000000000001e-06,
+      "loss": 0.5544,
+      "step": 25020
+    },
+    {
+      "epoch": 2.503,
+      "grad_norm": 19.910507202148438,
+      "learning_rate": 4.997000000000001e-06,
+      "loss": 0.6938,
+      "step": 25030
+    },
+    {
+      "epoch": 2.504,
+      "grad_norm": 13.199621200561523,
+      "learning_rate": 4.9950000000000005e-06,
+      "loss": 0.6362,
+      "step": 25040
+    },
+    {
+      "epoch": 2.505,
+      "grad_norm": 43.331302642822266,
+      "learning_rate": 4.993e-06,
+      "loss": 0.6495,
+      "step": 25050
+    },
+    {
+      "epoch": 2.5060000000000002,
+      "grad_norm": 4.908505439758301,
+      "learning_rate": 4.991e-06,
+      "loss": 0.7268,
+      "step": 25060
+    },
+    {
+      "epoch": 2.507,
+      "grad_norm": 11.888862609863281,
+      "learning_rate": 4.989000000000001e-06,
+      "loss": 0.6008,
+      "step": 25070
+    },
+    {
+      "epoch": 2.508,
+      "grad_norm": 8.154284477233887,
+      "learning_rate": 4.987e-06,
+      "loss": 0.5632,
+      "step": 25080
+    },
+    {
+      "epoch": 2.509,
+      "grad_norm": 38.40000534057617,
+      "learning_rate": 4.9850000000000006e-06,
+      "loss": 0.9932,
+      "step": 25090
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.843054294586182,
+      "learning_rate": 4.983e-06,
+      "loss": 0.5249,
+      "step": 25100
+    },
+    {
+      "epoch": 2.511,
+      "grad_norm": 24.743989944458008,
+      "learning_rate": 4.981e-06,
+      "loss": 0.6258,
+      "step": 25110
+    },
+    {
+      "epoch": 2.512,
+      "grad_norm": 41.41858673095703,
+      "learning_rate": 4.979e-06,
+      "loss": 0.7094,
+      "step": 25120
+    },
+    {
+      "epoch": 2.513,
+      "grad_norm": 37.239830017089844,
+      "learning_rate": 4.977e-06,
+      "loss": 0.7222,
+      "step": 25130
+    },
+    {
+      "epoch": 2.5140000000000002,
+      "grad_norm": 16.851455688476562,
+      "learning_rate": 4.975000000000001e-06,
+      "loss": 0.6925,
+      "step": 25140
+    },
+    {
+      "epoch": 2.515,
+      "grad_norm": 21.424264907836914,
+      "learning_rate": 4.9730000000000005e-06,
+      "loss": 0.6786,
+      "step": 25150
+    },
+    {
+      "epoch": 2.516,
+      "grad_norm": 37.35005569458008,
+      "learning_rate": 4.971e-06,
+      "loss": 0.5343,
+      "step": 25160
+    },
+    {
+      "epoch": 2.517,
+      "grad_norm": 12.154038429260254,
+      "learning_rate": 4.969e-06,
+      "loss": 0.649,
+      "step": 25170
+    },
+    {
+      "epoch": 2.518,
+      "grad_norm": 21.416759490966797,
+      "learning_rate": 4.967e-06,
+      "loss": 0.5783,
+      "step": 25180
+    },
+    {
+      "epoch": 2.519,
+      "grad_norm": 15.972697257995605,
+      "learning_rate": 4.965000000000001e-06,
+      "loss": 0.8398,
+      "step": 25190
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 35.27529525756836,
+      "learning_rate": 4.963000000000001e-06,
+      "loss": 0.6247,
+      "step": 25200
+    },
+    {
+      "epoch": 2.521,
+      "grad_norm": 59.43956756591797,
+      "learning_rate": 4.9610000000000004e-06,
+      "loss": 0.797,
+      "step": 25210
+    },
+    {
+      "epoch": 2.5220000000000002,
+      "grad_norm": 8.644207954406738,
+      "learning_rate": 4.959e-06,
+      "loss": 0.4781,
+      "step": 25220
+    },
+    {
+      "epoch": 2.523,
+      "grad_norm": 32.61981964111328,
+      "learning_rate": 4.957e-06,
+      "loss": 0.9126,
+      "step": 25230
+    },
+    {
+      "epoch": 2.524,
+      "grad_norm": 52.15250778198242,
+      "learning_rate": 4.955e-06,
+      "loss": 0.532,
+      "step": 25240
+    },
+    {
+      "epoch": 2.525,
+      "grad_norm": 23.468826293945312,
+      "learning_rate": 4.953000000000001e-06,
+      "loss": 0.7001,
+      "step": 25250
+    },
+    {
+      "epoch": 2.526,
+      "grad_norm": 12.018453598022461,
+      "learning_rate": 4.9510000000000005e-06,
+      "loss": 0.5967,
+      "step": 25260
+    },
+    {
+      "epoch": 2.527,
+      "grad_norm": 35.626129150390625,
+      "learning_rate": 4.949e-06,
+      "loss": 0.5483,
+      "step": 25270
+    },
+    {
+      "epoch": 2.528,
+      "grad_norm": 21.014564514160156,
+      "learning_rate": 4.947e-06,
+      "loss": 0.7496,
+      "step": 25280
+    },
+    {
+      "epoch": 2.529,
+      "grad_norm": 35.65088653564453,
+      "learning_rate": 4.945e-06,
+      "loss": 0.6784,
+      "step": 25290
+    },
+    {
+      "epoch": 2.5300000000000002,
+      "grad_norm": 21.193424224853516,
+      "learning_rate": 4.943000000000001e-06,
+      "loss": 0.9482,
+      "step": 25300
+    },
+    {
+      "epoch": 2.531,
+      "grad_norm": 94.68802642822266,
+      "learning_rate": 4.941000000000001e-06,
+      "loss": 0.9929,
+      "step": 25310
+    },
+    {
+      "epoch": 2.532,
+      "grad_norm": 15.581934928894043,
+      "learning_rate": 4.9390000000000005e-06,
+      "loss": 0.7156,
+      "step": 25320
+    },
+    {
+      "epoch": 2.533,
+      "grad_norm": 29.984956741333008,
+      "learning_rate": 4.937e-06,
+      "loss": 0.5357,
+      "step": 25330
+    },
+    {
+      "epoch": 2.534,
+      "grad_norm": 54.73834991455078,
+      "learning_rate": 4.935e-06,
+      "loss": 0.8771,
+      "step": 25340
+    },
+    {
+      "epoch": 2.535,
+      "grad_norm": 14.674304962158203,
+      "learning_rate": 4.933000000000001e-06,
+      "loss": 0.6677,
+      "step": 25350
+    },
+    {
+      "epoch": 2.536,
+      "grad_norm": 17.07393455505371,
+      "learning_rate": 4.931e-06,
+      "loss": 0.6853,
+      "step": 25360
+    },
+    {
+      "epoch": 2.537,
+      "grad_norm": 7.97967529296875,
+      "learning_rate": 4.929000000000001e-06,
+      "loss": 0.4164,
+      "step": 25370
+    },
+    {
+      "epoch": 2.5380000000000003,
+      "grad_norm": 31.6351261138916,
+      "learning_rate": 4.9270000000000004e-06,
+      "loss": 0.6275,
+      "step": 25380
+    },
+    {
+      "epoch": 2.539,
+      "grad_norm": 49.94380187988281,
+      "learning_rate": 4.925e-06,
+      "loss": 0.7284,
+      "step": 25390
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 32.98598861694336,
+      "learning_rate": 4.923000000000001e-06,
+      "loss": 0.7223,
+      "step": 25400
+    },
+    {
+      "epoch": 2.541,
+      "grad_norm": 81.89440155029297,
+      "learning_rate": 4.921e-06,
+      "loss": 0.7486,
+      "step": 25410
+    },
+    {
+      "epoch": 2.542,
+      "grad_norm": 34.58148193359375,
+      "learning_rate": 4.919000000000001e-06,
+      "loss": 0.6063,
+      "step": 25420
+    },
+    {
+      "epoch": 2.543,
+      "grad_norm": 23.96933364868164,
+      "learning_rate": 4.9170000000000005e-06,
+      "loss": 0.6273,
+      "step": 25430
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 2.428050994873047,
+      "learning_rate": 4.915e-06,
+      "loss": 0.4913,
+      "step": 25440
+    },
+    {
+      "epoch": 2.545,
+      "grad_norm": 26.93573570251465,
+      "learning_rate": 4.913e-06,
+      "loss": 0.889,
+      "step": 25450
+    },
+    {
+      "epoch": 2.5460000000000003,
+      "grad_norm": 7.789152145385742,
+      "learning_rate": 4.911e-06,
+      "loss": 0.3632,
+      "step": 25460
+    },
+    {
+      "epoch": 2.547,
+      "grad_norm": 18.7591495513916,
+      "learning_rate": 4.909000000000001e-06,
+      "loss": 0.6128,
+      "step": 25470
+    },
+    {
+      "epoch": 2.548,
+      "grad_norm": 21.817312240600586,
+      "learning_rate": 4.907000000000001e-06,
+      "loss": 0.7535,
+      "step": 25480
+    },
+    {
+      "epoch": 2.549,
+      "grad_norm": 22.82774543762207,
+      "learning_rate": 4.9050000000000005e-06,
+      "loss": 0.9183,
+      "step": 25490
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 27.486326217651367,
+      "learning_rate": 4.903e-06,
+      "loss": 0.6347,
+      "step": 25500
+    },
+    {
+      "epoch": 2.551,
+      "grad_norm": 18.89615821838379,
+      "learning_rate": 4.901e-06,
+      "loss": 0.6479,
+      "step": 25510
+    },
+    {
+      "epoch": 2.552,
+      "grad_norm": 39.137046813964844,
+      "learning_rate": 4.899e-06,
+      "loss": 0.5722,
+      "step": 25520
+    },
+    {
+      "epoch": 2.553,
+      "grad_norm": 17.233186721801758,
+      "learning_rate": 4.897000000000001e-06,
+      "loss": 0.7472,
+      "step": 25530
+    },
+    {
+      "epoch": 2.5540000000000003,
+      "grad_norm": 15.910088539123535,
+      "learning_rate": 4.8950000000000006e-06,
+      "loss": 0.5386,
+      "step": 25540
+    },
+    {
+      "epoch": 2.555,
+      "grad_norm": 36.60475158691406,
+      "learning_rate": 4.893e-06,
+      "loss": 0.7298,
+      "step": 25550
+    },
+    {
+      "epoch": 2.556,
+      "grad_norm": 37.67909240722656,
+      "learning_rate": 4.891e-06,
+      "loss": 0.6772,
+      "step": 25560
+    },
+    {
+      "epoch": 2.557,
+      "grad_norm": 6.002618789672852,
+      "learning_rate": 4.889e-06,
+      "loss": 0.4854,
+      "step": 25570
+    },
+    {
+      "epoch": 2.558,
+      "grad_norm": 22.503517150878906,
+      "learning_rate": 4.887000000000001e-06,
+      "loss": 0.7424,
+      "step": 25580
+    },
+    {
+      "epoch": 2.559,
+      "grad_norm": 38.0008430480957,
+      "learning_rate": 4.885000000000001e-06,
+      "loss": 0.686,
+      "step": 25590
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 31.778282165527344,
+      "learning_rate": 4.8830000000000005e-06,
+      "loss": 0.6194,
+      "step": 25600
+    },
+    {
+      "epoch": 2.561,
+      "grad_norm": 39.77953338623047,
+      "learning_rate": 4.881e-06,
+      "loss": 0.6395,
+      "step": 25610
+    },
+    {
+      "epoch": 2.5620000000000003,
+      "grad_norm": 5.725687503814697,
+      "learning_rate": 4.879e-06,
+      "loss": 0.41,
+      "step": 25620
+    },
+    {
+      "epoch": 2.5629999999999997,
+      "grad_norm": 28.99380874633789,
+      "learning_rate": 4.877000000000001e-06,
+      "loss": 0.5985,
+      "step": 25630
+    },
+    {
+      "epoch": 2.564,
+      "grad_norm": 18.07556915283203,
+      "learning_rate": 4.875e-06,
+      "loss": 0.7484,
+      "step": 25640
+    },
+    {
+      "epoch": 2.565,
+      "grad_norm": 17.21390724182129,
+      "learning_rate": 4.873000000000001e-06,
+      "loss": 0.5817,
+      "step": 25650
+    },
+    {
+      "epoch": 2.566,
+      "grad_norm": 11.937161445617676,
+      "learning_rate": 4.8710000000000005e-06,
+      "loss": 0.4783,
+      "step": 25660
+    },
+    {
+      "epoch": 2.567,
+      "grad_norm": 18.838775634765625,
+      "learning_rate": 4.869e-06,
+      "loss": 0.7863,
+      "step": 25670
+    },
+    {
+      "epoch": 2.568,
+      "grad_norm": 35.052398681640625,
+      "learning_rate": 4.867000000000001e-06,
+      "loss": 0.5906,
+      "step": 25680
+    },
+    {
+      "epoch": 2.569,
+      "grad_norm": 18.534929275512695,
+      "learning_rate": 4.865e-06,
+      "loss": 0.6528,
+      "step": 25690
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 15.753058433532715,
+      "learning_rate": 4.863000000000001e-06,
+      "loss": 0.7879,
+      "step": 25700
+    },
+    {
+      "epoch": 2.5709999999999997,
+      "grad_norm": 25.699085235595703,
+      "learning_rate": 4.8610000000000006e-06,
+      "loss": 0.5737,
+      "step": 25710
+    },
+    {
+      "epoch": 2.572,
+      "grad_norm": 28.720382690429688,
+      "learning_rate": 4.859e-06,
+      "loss": 0.6099,
+      "step": 25720
+    },
+    {
+      "epoch": 2.573,
+      "grad_norm": 7.138073921203613,
+      "learning_rate": 4.857e-06,
+      "loss": 0.6004,
+      "step": 25730
+    },
+    {
+      "epoch": 2.574,
+      "grad_norm": 42.17637252807617,
+      "learning_rate": 4.855e-06,
+      "loss": 0.7334,
+      "step": 25740
+    },
+    {
+      "epoch": 2.575,
+      "grad_norm": 19.05852508544922,
+      "learning_rate": 4.853000000000001e-06,
+      "loss": 0.6474,
+      "step": 25750
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 24.66946029663086,
+      "learning_rate": 4.851e-06,
+      "loss": 0.7538,
+      "step": 25760
+    },
+    {
+      "epoch": 2.577,
+      "grad_norm": 25.733699798583984,
+      "learning_rate": 4.8490000000000005e-06,
+      "loss": 0.6507,
+      "step": 25770
+    },
+    {
+      "epoch": 2.578,
+      "grad_norm": 6.0958709716796875,
+      "learning_rate": 4.847e-06,
+      "loss": 0.3898,
+      "step": 25780
+    },
+    {
+      "epoch": 2.5789999999999997,
+      "grad_norm": 9.826627731323242,
+      "learning_rate": 4.845e-06,
+      "loss": 0.8563,
+      "step": 25790
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 20.660486221313477,
+      "learning_rate": 4.843000000000001e-06,
+      "loss": 0.6671,
+      "step": 25800
+    },
+    {
+      "epoch": 2.581,
+      "grad_norm": 35.543087005615234,
+      "learning_rate": 4.841e-06,
+      "loss": 0.6269,
+      "step": 25810
+    },
+    {
+      "epoch": 2.582,
+      "grad_norm": 9.25235366821289,
+      "learning_rate": 4.839000000000001e-06,
+      "loss": 0.6317,
+      "step": 25820
+    },
+    {
+      "epoch": 2.583,
+      "grad_norm": 11.168290138244629,
+      "learning_rate": 4.8370000000000004e-06,
+      "loss": 0.6362,
+      "step": 25830
+    },
+    {
+      "epoch": 2.584,
+      "grad_norm": 36.224430084228516,
+      "learning_rate": 4.835e-06,
+      "loss": 0.5701,
+      "step": 25840
+    },
+    {
+      "epoch": 2.585,
+      "grad_norm": 35.46279525756836,
+      "learning_rate": 4.833e-06,
+      "loss": 0.5058,
+      "step": 25850
+    },
+    {
+      "epoch": 2.586,
+      "grad_norm": 11.398242950439453,
+      "learning_rate": 4.831e-06,
+      "loss": 0.5375,
+      "step": 25860
+    },
+    {
+      "epoch": 2.5869999999999997,
+      "grad_norm": 35.309444427490234,
+      "learning_rate": 4.829000000000001e-06,
+      "loss": 0.8348,
+      "step": 25870
+    },
+    {
+      "epoch": 2.588,
+      "grad_norm": 9.43244743347168,
+      "learning_rate": 4.8270000000000005e-06,
+      "loss": 0.5513,
+      "step": 25880
+    },
+    {
+      "epoch": 2.589,
+      "grad_norm": 37.240135192871094,
+      "learning_rate": 4.825e-06,
+      "loss": 0.6013,
+      "step": 25890
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 47.575035095214844,
+      "learning_rate": 4.823e-06,
+      "loss": 0.7614,
+      "step": 25900
+    },
+    {
+      "epoch": 2.591,
+      "grad_norm": 8.541173934936523,
+      "learning_rate": 4.821e-06,
+      "loss": 0.4809,
+      "step": 25910
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 22.348037719726562,
+      "learning_rate": 4.819e-06,
+      "loss": 0.582,
+      "step": 25920
+    },
+    {
+      "epoch": 2.593,
+      "grad_norm": 2.8846850395202637,
+      "learning_rate": 4.817000000000001e-06,
+      "loss": 0.4284,
+      "step": 25930
+    },
+    {
+      "epoch": 2.594,
+      "grad_norm": 15.262609481811523,
+      "learning_rate": 4.8150000000000005e-06,
+      "loss": 0.6563,
+      "step": 25940
+    },
+    {
+      "epoch": 2.5949999999999998,
+      "grad_norm": 42.77225112915039,
+      "learning_rate": 4.813e-06,
+      "loss": 0.7155,
+      "step": 25950
+    },
+    {
+      "epoch": 2.596,
+      "grad_norm": 38.20702362060547,
+      "learning_rate": 4.811000000000001e-06,
+      "loss": 0.6124,
+      "step": 25960
+    },
+    {
+      "epoch": 2.597,
+      "grad_norm": 16.295507431030273,
+      "learning_rate": 4.809e-06,
+      "loss": 0.5062,
+      "step": 25970
+    },
+    {
+      "epoch": 2.598,
+      "grad_norm": 45.26594543457031,
+      "learning_rate": 4.807000000000001e-06,
+      "loss": 0.7953,
+      "step": 25980
+    },
+    {
+      "epoch": 2.599,
+      "grad_norm": 47.709781646728516,
+      "learning_rate": 4.805000000000001e-06,
+      "loss": 0.566,
+      "step": 25990
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.982378005981445,
+      "learning_rate": 4.8030000000000004e-06,
+      "loss": 0.5725,
+      "step": 26000
+    },
+    {
+      "epoch": 2.601,
+      "grad_norm": 44.133052825927734,
+      "learning_rate": 4.801e-06,
+      "loss": 0.5826,
+      "step": 26010
+    },
+    {
+      "epoch": 2.602,
+      "grad_norm": 15.568506240844727,
+      "learning_rate": 4.799e-06,
+      "loss": 0.5797,
+      "step": 26020
+    },
+    {
+      "epoch": 2.6029999999999998,
+      "grad_norm": 33.09275436401367,
+      "learning_rate": 4.797000000000001e-06,
+      "loss": 0.6618,
+      "step": 26030
+    },
+    {
+      "epoch": 2.604,
+      "grad_norm": 41.33121871948242,
+      "learning_rate": 4.795e-06,
+      "loss": 1.0158,
+      "step": 26040
+    },
+    {
+      "epoch": 2.605,
+      "grad_norm": 51.52129364013672,
+      "learning_rate": 4.7930000000000005e-06,
+      "loss": 0.6314,
+      "step": 26050
+    },
+    {
+      "epoch": 2.606,
+      "grad_norm": 6.395187854766846,
+      "learning_rate": 4.791e-06,
+      "loss": 0.5432,
+      "step": 26060
+    },
+    {
+      "epoch": 2.607,
+      "grad_norm": 39.273780822753906,
+      "learning_rate": 4.789e-06,
+      "loss": 0.5501,
+      "step": 26070
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 35.81938934326172,
+      "learning_rate": 4.787000000000001e-06,
+      "loss": 0.3602,
+      "step": 26080
+    },
+    {
+      "epoch": 2.609,
+      "grad_norm": 35.3363151550293,
+      "learning_rate": 4.785e-06,
+      "loss": 0.6854,
+      "step": 26090
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 45.12687301635742,
+      "learning_rate": 4.783000000000001e-06,
+      "loss": 0.6236,
+      "step": 26100
+    },
+    {
+      "epoch": 2.6109999999999998,
+      "grad_norm": 37.093116760253906,
+      "learning_rate": 4.7810000000000005e-06,
+      "loss": 0.9301,
+      "step": 26110
+    },
+    {
+      "epoch": 2.612,
+      "grad_norm": 13.176533699035645,
+      "learning_rate": 4.779e-06,
+      "loss": 0.492,
+      "step": 26120
+    },
+    {
+      "epoch": 2.613,
+      "grad_norm": 48.988155364990234,
+      "learning_rate": 4.777e-06,
+      "loss": 0.8739,
+      "step": 26130
+    },
+    {
+      "epoch": 2.614,
+      "grad_norm": 17.3516788482666,
+      "learning_rate": 4.775e-06,
+      "loss": 0.5989,
+      "step": 26140
+    },
+    {
+      "epoch": 2.615,
+      "grad_norm": 3.241021156311035,
+      "learning_rate": 4.773000000000001e-06,
+      "loss": 0.6567,
+      "step": 26150
+    },
+    {
+      "epoch": 2.616,
+      "grad_norm": 27.495573043823242,
+      "learning_rate": 4.7710000000000006e-06,
+      "loss": 0.4817,
+      "step": 26160
+    },
+    {
+      "epoch": 2.617,
+      "grad_norm": 45.31413650512695,
+      "learning_rate": 4.769e-06,
+      "loss": 0.5875,
+      "step": 26170
+    },
+    {
+      "epoch": 2.618,
+      "grad_norm": 43.09683609008789,
+      "learning_rate": 4.767e-06,
+      "loss": 0.752,
+      "step": 26180
+    },
+    {
+      "epoch": 2.6189999999999998,
+      "grad_norm": 17.2843074798584,
+      "learning_rate": 4.765e-06,
+      "loss": 0.6858,
+      "step": 26190
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 47.570281982421875,
+      "learning_rate": 4.763000000000001e-06,
+      "loss": 0.6619,
+      "step": 26200
+    },
+    {
+      "epoch": 2.621,
+      "grad_norm": 2.8026413917541504,
+      "learning_rate": 4.761000000000001e-06,
+      "loss": 0.6908,
+      "step": 26210
+    },
+    {
+      "epoch": 2.622,
+      "grad_norm": 51.79669952392578,
+      "learning_rate": 4.7590000000000005e-06,
+      "loss": 0.5145,
+      "step": 26220
+    },
+    {
+      "epoch": 2.623,
+      "grad_norm": 19.143081665039062,
+      "learning_rate": 4.757e-06,
+      "loss": 0.7878,
+      "step": 26230
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 39.0579948425293,
+      "learning_rate": 4.755e-06,
+      "loss": 0.5953,
+      "step": 26240
+    },
+    {
+      "epoch": 2.625,
+      "grad_norm": 52.51986312866211,
+      "learning_rate": 4.753e-06,
+      "loss": 0.5562,
+      "step": 26250
+    },
+    {
+      "epoch": 2.626,
+      "grad_norm": 53.69147872924805,
+      "learning_rate": 4.751000000000001e-06,
+      "loss": 0.6106,
+      "step": 26260
+    },
+    {
+      "epoch": 2.627,
+      "grad_norm": 38.23792266845703,
+      "learning_rate": 4.749000000000001e-06,
+      "loss": 0.5109,
+      "step": 26270
+    },
+    {
+      "epoch": 2.628,
+      "grad_norm": 46.235511779785156,
+      "learning_rate": 4.7470000000000005e-06,
+      "loss": 0.5813,
+      "step": 26280
+    },
+    {
+      "epoch": 2.629,
+      "grad_norm": 19.168806076049805,
+      "learning_rate": 4.745e-06,
+      "loss": 0.3629,
+      "step": 26290
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.286655426025391,
+      "learning_rate": 4.743e-06,
+      "loss": 0.3528,
+      "step": 26300
+    },
+    {
+      "epoch": 2.6310000000000002,
+      "grad_norm": 52.93207550048828,
+      "learning_rate": 4.741000000000001e-06,
+      "loss": 0.6646,
+      "step": 26310
+    },
+    {
+      "epoch": 2.632,
+      "grad_norm": 6.3632588386535645,
+      "learning_rate": 4.739e-06,
+      "loss": 0.2838,
+      "step": 26320
+    },
+    {
+      "epoch": 2.633,
+      "grad_norm": 42.63784408569336,
+      "learning_rate": 4.7370000000000006e-06,
+      "loss": 0.4457,
+      "step": 26330
+    },
+    {
+      "epoch": 2.634,
+      "grad_norm": 37.047508239746094,
+      "learning_rate": 4.735e-06,
+      "loss": 0.8789,
+      "step": 26340
+    },
+    {
+      "epoch": 2.635,
+      "grad_norm": 4.293346881866455,
+      "learning_rate": 4.733e-06,
+      "loss": 0.41,
+      "step": 26350
+    },
+    {
+      "epoch": 2.636,
+      "grad_norm": 36.892662048339844,
+      "learning_rate": 4.731000000000001e-06,
+      "loss": 0.7638,
+      "step": 26360
+    },
+    {
+      "epoch": 2.637,
+      "grad_norm": 36.081783294677734,
+      "learning_rate": 4.729e-06,
+      "loss": 0.9168,
+      "step": 26370
+    },
+    {
+      "epoch": 2.638,
+      "grad_norm": 3.6100282669067383,
+      "learning_rate": 4.727000000000001e-06,
+      "loss": 0.6065,
+      "step": 26380
+    },
+    {
+      "epoch": 2.6390000000000002,
+      "grad_norm": 51.46051025390625,
+      "learning_rate": 4.7250000000000005e-06,
+      "loss": 0.6926,
+      "step": 26390
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.446451187133789,
+      "learning_rate": 4.723e-06,
+      "loss": 0.5111,
+      "step": 26400
+    },
+    {
+      "epoch": 2.641,
+      "grad_norm": 59.62565612792969,
+      "learning_rate": 4.721e-06,
+      "loss": 0.7106,
+      "step": 26410
+    },
+    {
+      "epoch": 2.642,
+      "grad_norm": 19.761140823364258,
+      "learning_rate": 4.719e-06,
+      "loss": 0.6368,
+      "step": 26420
+    },
+    {
+      "epoch": 2.643,
+      "grad_norm": 47.500484466552734,
+      "learning_rate": 4.717000000000001e-06,
+      "loss": 0.5708,
+      "step": 26430
+    },
+    {
+      "epoch": 2.644,
+      "grad_norm": 27.15022850036621,
+      "learning_rate": 4.715e-06,
+      "loss": 0.6257,
+      "step": 26440
+    },
+    {
+      "epoch": 2.645,
+      "grad_norm": 45.335968017578125,
+      "learning_rate": 4.7130000000000004e-06,
+      "loss": 0.5052,
+      "step": 26450
+    },
+    {
+      "epoch": 2.646,
+      "grad_norm": 38.73992919921875,
+      "learning_rate": 4.711e-06,
+      "loss": 0.6991,
+      "step": 26460
+    },
+    {
+      "epoch": 2.6470000000000002,
+      "grad_norm": 37.85356140136719,
+      "learning_rate": 4.709e-06,
+      "loss": 0.6061,
+      "step": 26470
+    },
+    {
+      "epoch": 2.648,
+      "grad_norm": 24.0386962890625,
+      "learning_rate": 4.707000000000001e-06,
+      "loss": 0.7666,
+      "step": 26480
+    },
+    {
+      "epoch": 2.649,
+      "grad_norm": 28.294811248779297,
+      "learning_rate": 4.705e-06,
+      "loss": 0.5678,
+      "step": 26490
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 41.239707946777344,
+      "learning_rate": 4.7030000000000005e-06,
+      "loss": 0.8175,
+      "step": 26500
+    },
+    {
+      "epoch": 2.651,
+      "grad_norm": 62.72034454345703,
+      "learning_rate": 4.701e-06,
+      "loss": 0.5212,
+      "step": 26510
+    },
+    {
+      "epoch": 2.652,
+      "grad_norm": 38.574623107910156,
+      "learning_rate": 4.699e-06,
+      "loss": 0.7919,
+      "step": 26520
+    },
+    {
+      "epoch": 2.653,
+      "grad_norm": 11.876582145690918,
+      "learning_rate": 4.697e-06,
+      "loss": 0.7385,
+      "step": 26530
+    },
+    {
+      "epoch": 2.654,
+      "grad_norm": 14.082344055175781,
+      "learning_rate": 4.695e-06,
+      "loss": 0.3954,
+      "step": 26540
+    },
+    {
+      "epoch": 2.6550000000000002,
+      "grad_norm": 6.629438877105713,
+      "learning_rate": 4.693000000000001e-06,
+      "loss": 0.5655,
+      "step": 26550
+    },
+    {
+      "epoch": 2.656,
+      "grad_norm": 31.33914566040039,
+      "learning_rate": 4.6910000000000005e-06,
+      "loss": 0.7101,
+      "step": 26560
+    },
+    {
+      "epoch": 2.657,
+      "grad_norm": 32.01548767089844,
+      "learning_rate": 4.689e-06,
+      "loss": 0.7707,
+      "step": 26570
+    },
+    {
+      "epoch": 2.658,
+      "grad_norm": 20.46829605102539,
+      "learning_rate": 4.687e-06,
+      "loss": 0.585,
+      "step": 26580
+    },
+    {
+      "epoch": 2.659,
+      "grad_norm": 21.68409538269043,
+      "learning_rate": 4.685000000000001e-06,
+      "loss": 0.6834,
+      "step": 26590
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 53.09114456176758,
+      "learning_rate": 4.683000000000001e-06,
+      "loss": 0.6068,
+      "step": 26600
+    },
+    {
+      "epoch": 2.661,
+      "grad_norm": 7.67567253112793,
+      "learning_rate": 4.681000000000001e-06,
+      "loss": 0.4884,
+      "step": 26610
+    },
+    {
+      "epoch": 2.662,
+      "grad_norm": 19.89399528503418,
+      "learning_rate": 4.6790000000000004e-06,
+      "loss": 0.7026,
+      "step": 26620
+    },
+    {
+      "epoch": 2.6630000000000003,
+      "grad_norm": 22.56330108642578,
+      "learning_rate": 4.677e-06,
+      "loss": 0.6527,
+      "step": 26630
+    },
+    {
+      "epoch": 2.664,
+      "grad_norm": 7.4913105964660645,
+      "learning_rate": 4.675000000000001e-06,
+      "loss": 0.5722,
+      "step": 26640
+    },
+    {
+      "epoch": 2.665,
+      "grad_norm": 21.277095794677734,
+      "learning_rate": 4.673e-06,
+      "loss": 0.6265,
+      "step": 26650
+    },
+    {
+      "epoch": 2.666,
+      "grad_norm": 10.212080955505371,
+      "learning_rate": 4.671000000000001e-06,
+      "loss": 0.6366,
+      "step": 26660
+    },
+    {
+      "epoch": 2.667,
+      "grad_norm": 73.72460174560547,
+      "learning_rate": 4.6690000000000005e-06,
+      "loss": 0.5917,
+      "step": 26670
+    },
+    {
+      "epoch": 2.668,
+      "grad_norm": 36.61571502685547,
+      "learning_rate": 4.667e-06,
+      "loss": 0.6464,
+      "step": 26680
+    },
+    {
+      "epoch": 2.669,
+      "grad_norm": 42.48518371582031,
+      "learning_rate": 4.665e-06,
+      "loss": 0.4937,
+      "step": 26690
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 61.28166580200195,
+      "learning_rate": 4.663e-06,
+      "loss": 0.7738,
+      "step": 26700
+    },
+    {
+      "epoch": 2.6710000000000003,
+      "grad_norm": 6.222008228302002,
+      "learning_rate": 4.661000000000001e-06,
+      "loss": 0.2444,
+      "step": 26710
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 38.14292907714844,
+      "learning_rate": 4.659e-06,
+      "loss": 0.5493,
+      "step": 26720
+    },
+    {
+      "epoch": 2.673,
+      "grad_norm": 34.81486129760742,
+      "learning_rate": 4.6570000000000005e-06,
+      "loss": 0.6373,
+      "step": 26730
+    },
+    {
+      "epoch": 2.674,
+      "grad_norm": 22.477020263671875,
+      "learning_rate": 4.655e-06,
+      "loss": 0.7369,
+      "step": 26740
+    },
+    {
+      "epoch": 2.675,
+      "grad_norm": 41.28529357910156,
+      "learning_rate": 4.653e-06,
+      "loss": 0.4623,
+      "step": 26750
+    },
+    {
+      "epoch": 2.676,
+      "grad_norm": 34.939598083496094,
+      "learning_rate": 4.651000000000001e-06,
+      "loss": 0.5825,
+      "step": 26760
+    },
+    {
+      "epoch": 2.677,
+      "grad_norm": 17.477685928344727,
+      "learning_rate": 4.649e-06,
+      "loss": 0.4918,
+      "step": 26770
+    },
+    {
+      "epoch": 2.678,
+      "grad_norm": 19.78916358947754,
+      "learning_rate": 4.6470000000000006e-06,
+      "loss": 0.7126,
+      "step": 26780
+    },
+    {
+      "epoch": 2.6790000000000003,
+      "grad_norm": 33.3621826171875,
+      "learning_rate": 4.645e-06,
+      "loss": 0.7059,
+      "step": 26790
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 32.21095657348633,
+      "learning_rate": 4.643e-06,
+      "loss": 0.5107,
+      "step": 26800
+    },
+    {
+      "epoch": 2.681,
+      "grad_norm": 21.85660171508789,
+      "learning_rate": 4.641e-06,
+      "loss": 0.595,
+      "step": 26810
+    },
+    {
+      "epoch": 2.682,
+      "grad_norm": 29.33839988708496,
+      "learning_rate": 4.639e-06,
+      "loss": 0.5435,
+      "step": 26820
+    },
+    {
+      "epoch": 2.683,
+      "grad_norm": 19.039310455322266,
+      "learning_rate": 4.637000000000001e-06,
+      "loss": 0.7826,
+      "step": 26830
+    },
+    {
+      "epoch": 2.684,
+      "grad_norm": 31.831010818481445,
+      "learning_rate": 4.6350000000000005e-06,
+      "loss": 0.9578,
+      "step": 26840
+    },
+    {
+      "epoch": 2.685,
+      "grad_norm": 12.974231719970703,
+      "learning_rate": 4.633e-06,
+      "loss": 0.6917,
+      "step": 26850
+    },
+    {
+      "epoch": 2.686,
+      "grad_norm": 42.49304962158203,
+      "learning_rate": 4.631e-06,
+      "loss": 0.6856,
+      "step": 26860
+    },
+    {
+      "epoch": 2.6870000000000003,
+      "grad_norm": 5.649697780609131,
+      "learning_rate": 4.629e-06,
+      "loss": 0.4211,
+      "step": 26870
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 28.178577423095703,
+      "learning_rate": 4.627000000000001e-06,
+      "loss": 0.514,
+      "step": 26880
+    },
+    {
+      "epoch": 2.689,
+      "grad_norm": 14.685444831848145,
+      "learning_rate": 4.625000000000001e-06,
+      "loss": 0.6962,
+      "step": 26890
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 32.58005142211914,
+      "learning_rate": 4.6230000000000005e-06,
+      "loss": 0.6088,
+      "step": 26900
+    },
+    {
+      "epoch": 2.691,
+      "grad_norm": 58.40959548950195,
+      "learning_rate": 4.621e-06,
+      "loss": 0.5932,
+      "step": 26910
+    },
+    {
+      "epoch": 2.692,
+      "grad_norm": 2.781147003173828,
+      "learning_rate": 4.619e-06,
+      "loss": 0.5929,
+      "step": 26920
+    },
+    {
+      "epoch": 2.693,
+      "grad_norm": 43.64352798461914,
+      "learning_rate": 4.617e-06,
+      "loss": 0.678,
+      "step": 26930
+    },
+    {
+      "epoch": 2.694,
+      "grad_norm": 39.059879302978516,
+      "learning_rate": 4.615000000000001e-06,
+      "loss": 0.7061,
+      "step": 26940
+    },
+    {
+      "epoch": 2.695,
+      "grad_norm": 36.203399658203125,
+      "learning_rate": 4.6130000000000006e-06,
+      "loss": 0.4879,
+      "step": 26950
+    },
+    {
+      "epoch": 2.6959999999999997,
+      "grad_norm": 53.63970184326172,
+      "learning_rate": 4.611e-06,
+      "loss": 0.6703,
+      "step": 26960
+    },
+    {
+      "epoch": 2.697,
+      "grad_norm": 4.569289207458496,
+      "learning_rate": 4.609e-06,
+      "loss": 0.5606,
+      "step": 26970
+    },
+    {
+      "epoch": 2.698,
+      "grad_norm": 25.914350509643555,
+      "learning_rate": 4.607e-06,
+      "loss": 0.5693,
+      "step": 26980
+    },
+    {
+      "epoch": 2.699,
+      "grad_norm": 15.915177345275879,
+      "learning_rate": 4.605000000000001e-06,
+      "loss": 0.4112,
+      "step": 26990
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 57.12289810180664,
+      "learning_rate": 4.603000000000001e-06,
+      "loss": 0.6674,
+      "step": 27000
+    },
+    {
+      "epoch": 2.701,
+      "grad_norm": 5.237196445465088,
+      "learning_rate": 4.6010000000000005e-06,
+      "loss": 0.6634,
+      "step": 27010
+    },
+    {
+      "epoch": 2.702,
+      "grad_norm": 29.42719078063965,
+      "learning_rate": 4.599e-06,
+      "loss": 0.5788,
+      "step": 27020
+    },
+    {
+      "epoch": 2.703,
+      "grad_norm": 43.75193405151367,
+      "learning_rate": 4.597e-06,
+      "loss": 0.7219,
+      "step": 27030
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 21.26077651977539,
+      "learning_rate": 4.595000000000001e-06,
+      "loss": 0.6192,
+      "step": 27040
+    },
+    {
+      "epoch": 2.705,
+      "grad_norm": 39.62782287597656,
+      "learning_rate": 4.593e-06,
+      "loss": 0.7536,
+      "step": 27050
+    },
+    {
+      "epoch": 2.706,
+      "grad_norm": 19.186992645263672,
+      "learning_rate": 4.591000000000001e-06,
+      "loss": 0.6155,
+      "step": 27060
+    },
+    {
+      "epoch": 2.707,
+      "grad_norm": 17.996379852294922,
+      "learning_rate": 4.5890000000000004e-06,
+      "loss": 0.5156,
+      "step": 27070
+    },
+    {
+      "epoch": 2.708,
+      "grad_norm": 59.57747268676758,
+      "learning_rate": 4.587e-06,
+      "loss": 0.5751,
+      "step": 27080
+    },
+    {
+      "epoch": 2.709,
+      "grad_norm": 28.020484924316406,
+      "learning_rate": 4.585e-06,
+      "loss": 0.6283,
+      "step": 27090
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 16.961652755737305,
+      "learning_rate": 4.583e-06,
+      "loss": 0.5277,
+      "step": 27100
+    },
+    {
+      "epoch": 2.711,
+      "grad_norm": 63.39067077636719,
+      "learning_rate": 4.581000000000001e-06,
+      "loss": 0.5592,
+      "step": 27110
+    },
+    {
+      "epoch": 2.7119999999999997,
+      "grad_norm": 8.494537353515625,
+      "learning_rate": 4.579e-06,
+      "loss": 0.3698,
+      "step": 27120
+    },
+    {
+      "epoch": 2.713,
+      "grad_norm": 41.40669250488281,
+      "learning_rate": 4.577e-06,
+      "loss": 0.6026,
+      "step": 27130
+    },
+    {
+      "epoch": 2.714,
+      "grad_norm": 20.117542266845703,
+      "learning_rate": 4.575e-06,
+      "loss": 0.4716,
+      "step": 27140
+    },
+    {
+      "epoch": 2.715,
+      "grad_norm": 42.48599624633789,
+      "learning_rate": 4.573e-06,
+      "loss": 0.7185,
+      "step": 27150
+    },
+    {
+      "epoch": 2.716,
+      "grad_norm": 75.84381866455078,
+      "learning_rate": 4.571000000000001e-06,
+      "loss": 0.8573,
+      "step": 27160
+    },
+    {
+      "epoch": 2.717,
+      "grad_norm": 40.775611877441406,
+      "learning_rate": 4.569e-06,
+      "loss": 0.6571,
+      "step": 27170
+    },
+    {
+      "epoch": 2.718,
+      "grad_norm": 16.275890350341797,
+      "learning_rate": 4.5670000000000005e-06,
+      "loss": 0.3679,
+      "step": 27180
+    },
+    {
+      "epoch": 2.719,
+      "grad_norm": 23.729806900024414,
+      "learning_rate": 4.565e-06,
+      "loss": 0.5482,
+      "step": 27190
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 8.010773658752441,
+      "learning_rate": 4.563e-06,
+      "loss": 0.5101,
+      "step": 27200
+    },
+    {
+      "epoch": 2.721,
+      "grad_norm": 19.037765502929688,
+      "learning_rate": 4.561e-06,
+      "loss": 0.6863,
+      "step": 27210
+    },
+    {
+      "epoch": 2.722,
+      "grad_norm": 26.438081741333008,
+      "learning_rate": 4.559000000000001e-06,
+      "loss": 0.866,
+      "step": 27220
+    },
+    {
+      "epoch": 2.723,
+      "grad_norm": 13.23498821258545,
+      "learning_rate": 4.557000000000001e-06,
+      "loss": 0.6461,
+      "step": 27230
+    },
+    {
+      "epoch": 2.724,
+      "grad_norm": 13.73941421508789,
+      "learning_rate": 4.5550000000000004e-06,
+      "loss": 0.7792,
+      "step": 27240
+    },
+    {
+      "epoch": 2.725,
+      "grad_norm": 30.74558448791504,
+      "learning_rate": 4.553e-06,
+      "loss": 0.5246,
+      "step": 27250
+    },
+    {
+      "epoch": 2.726,
+      "grad_norm": 34.99910354614258,
+      "learning_rate": 4.551e-06,
+      "loss": 0.6406,
+      "step": 27260
+    },
+    {
+      "epoch": 2.727,
+      "grad_norm": 31.215740203857422,
+      "learning_rate": 4.549000000000001e-06,
+      "loss": 0.778,
+      "step": 27270
+    },
+    {
+      "epoch": 2.7279999999999998,
+      "grad_norm": 10.44588851928711,
+      "learning_rate": 4.547000000000001e-06,
+      "loss": 0.6892,
+      "step": 27280
+    },
+    {
+      "epoch": 2.729,
+      "grad_norm": 13.254731178283691,
+      "learning_rate": 4.5450000000000005e-06,
+      "loss": 0.7861,
+      "step": 27290
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 29.335412979125977,
+      "learning_rate": 4.543e-06,
+      "loss": 0.6654,
+      "step": 27300
+    },
+    {
+      "epoch": 2.731,
+      "grad_norm": 26.643966674804688,
+      "learning_rate": 4.541e-06,
+      "loss": 0.7136,
+      "step": 27310
+    },
+    {
+      "epoch": 2.732,
+      "grad_norm": 5.667937755584717,
+      "learning_rate": 4.539000000000001e-06,
+      "loss": 0.4284,
+      "step": 27320
+    },
+    {
+      "epoch": 2.733,
+      "grad_norm": 40.35080337524414,
+      "learning_rate": 4.537e-06,
+      "loss": 0.7426,
+      "step": 27330
+    },
+    {
+      "epoch": 2.734,
+      "grad_norm": 39.766109466552734,
+      "learning_rate": 4.535000000000001e-06,
+      "loss": 0.8392,
+      "step": 27340
+    },
+    {
+      "epoch": 2.735,
+      "grad_norm": 17.853336334228516,
+      "learning_rate": 4.5330000000000005e-06,
+      "loss": 0.5092,
+      "step": 27350
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 34.687530517578125,
+      "learning_rate": 4.531e-06,
+      "loss": 0.4737,
+      "step": 27360
+    },
+    {
+      "epoch": 2.737,
+      "grad_norm": 33.80595016479492,
+      "learning_rate": 4.529000000000001e-06,
+      "loss": 0.7412,
+      "step": 27370
+    },
+    {
+      "epoch": 2.738,
+      "grad_norm": 45.17950439453125,
+      "learning_rate": 4.527e-06,
+      "loss": 0.5256,
+      "step": 27380
+    },
+    {
+      "epoch": 2.739,
+      "grad_norm": 38.902565002441406,
+      "learning_rate": 4.525000000000001e-06,
+      "loss": 0.7947,
+      "step": 27390
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 40.43433380126953,
+      "learning_rate": 4.5230000000000006e-06,
+      "loss": 0.7601,
+      "step": 27400
+    },
+    {
+      "epoch": 2.741,
+      "grad_norm": 48.44124984741211,
+      "learning_rate": 4.521e-06,
+      "loss": 0.688,
+      "step": 27410
+    },
+    {
+      "epoch": 2.742,
+      "grad_norm": 37.88941955566406,
+      "learning_rate": 4.519e-06,
+      "loss": 0.7972,
+      "step": 27420
+    },
+    {
+      "epoch": 2.743,
+      "grad_norm": 34.65846252441406,
+      "learning_rate": 4.517e-06,
+      "loss": 0.7524,
+      "step": 27430
+    },
+    {
+      "epoch": 2.7439999999999998,
+      "grad_norm": 36.840789794921875,
+      "learning_rate": 4.515000000000001e-06,
+      "loss": 0.4491,
+      "step": 27440
+    },
+    {
+      "epoch": 2.745,
+      "grad_norm": 14.328316688537598,
+      "learning_rate": 4.513e-06,
+      "loss": 0.5018,
+      "step": 27450
+    },
+    {
+      "epoch": 2.746,
+      "grad_norm": 18.487655639648438,
+      "learning_rate": 4.5110000000000005e-06,
+      "loss": 0.7067,
+      "step": 27460
+    },
+    {
+      "epoch": 2.747,
+      "grad_norm": 5.489202499389648,
+      "learning_rate": 4.509e-06,
+      "loss": 0.5449,
+      "step": 27470
+    },
+    {
+      "epoch": 2.748,
+      "grad_norm": 29.96932601928711,
+      "learning_rate": 4.507e-06,
+      "loss": 0.5671,
+      "step": 27480
+    },
+    {
+      "epoch": 2.749,
+      "grad_norm": 23.164758682250977,
+      "learning_rate": 4.505e-06,
+      "loss": 0.7109,
+      "step": 27490
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 37.39947509765625,
+      "learning_rate": 4.503e-06,
+      "loss": 0.4134,
+      "step": 27500
+    },
+    {
+      "epoch": 2.751,
+      "grad_norm": 9.628229141235352,
+      "learning_rate": 4.501000000000001e-06,
+      "loss": 0.5925,
+      "step": 27510
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 23.049732208251953,
+      "learning_rate": 4.4990000000000005e-06,
+      "loss": 0.4716,
+      "step": 27520
+    },
+    {
+      "epoch": 2.753,
+      "grad_norm": 27.242902755737305,
+      "learning_rate": 4.497e-06,
+      "loss": 0.5761,
+      "step": 27530
+    },
+    {
+      "epoch": 2.754,
+      "grad_norm": 22.390472412109375,
+      "learning_rate": 4.495e-06,
+      "loss": 0.6094,
+      "step": 27540
+    },
+    {
+      "epoch": 2.755,
+      "grad_norm": 33.42397689819336,
+      "learning_rate": 4.493e-06,
+      "loss": 0.5837,
+      "step": 27550
+    },
+    {
+      "epoch": 2.7560000000000002,
+      "grad_norm": 33.163333892822266,
+      "learning_rate": 4.491000000000001e-06,
+      "loss": 0.5067,
+      "step": 27560
+    },
+    {
+      "epoch": 2.757,
+      "grad_norm": 14.397940635681152,
+      "learning_rate": 4.4890000000000006e-06,
+      "loss": 0.6467,
+      "step": 27570
+    },
+    {
+      "epoch": 2.758,
+      "grad_norm": 34.04242706298828,
+      "learning_rate": 4.487e-06,
+      "loss": 0.6445,
+      "step": 27580
+    },
+    {
+      "epoch": 2.759,
+      "grad_norm": 15.349383354187012,
+      "learning_rate": 4.485e-06,
+      "loss": 0.5121,
+      "step": 27590
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.300539970397949,
+      "learning_rate": 4.483e-06,
+      "loss": 0.4665,
+      "step": 27600
+    },
+    {
+      "epoch": 2.761,
+      "grad_norm": 39.58720016479492,
+      "learning_rate": 4.481e-06,
+      "loss": 0.66,
+      "step": 27610
+    },
+    {
+      "epoch": 2.762,
+      "grad_norm": 26.696575164794922,
+      "learning_rate": 4.479000000000001e-06,
+      "loss": 0.8141,
+      "step": 27620
+    },
+    {
+      "epoch": 2.763,
+      "grad_norm": 43.283931732177734,
+      "learning_rate": 4.4770000000000005e-06,
+      "loss": 0.6543,
+      "step": 27630
+    },
+    {
+      "epoch": 2.7640000000000002,
+      "grad_norm": 45.54264831542969,
+      "learning_rate": 4.475e-06,
+      "loss": 0.8184,
+      "step": 27640
+    },
+    {
+      "epoch": 2.765,
+      "grad_norm": 33.091976165771484,
+      "learning_rate": 4.473e-06,
+      "loss": 0.5188,
+      "step": 27650
+    },
+    {
+      "epoch": 2.766,
+      "grad_norm": 21.730209350585938,
+      "learning_rate": 4.471e-06,
+      "loss": 0.4846,
+      "step": 27660
+    },
+    {
+      "epoch": 2.767,
+      "grad_norm": 4.529099464416504,
+      "learning_rate": 4.469000000000001e-06,
+      "loss": 1.1101,
+      "step": 27670
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 27.221635818481445,
+      "learning_rate": 4.467000000000001e-06,
+      "loss": 0.7868,
+      "step": 27680
+    },
+    {
+      "epoch": 2.769,
+      "grad_norm": 30.304941177368164,
+      "learning_rate": 4.4650000000000004e-06,
+      "loss": 0.5997,
+      "step": 27690
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 35.67657470703125,
+      "learning_rate": 4.463e-06,
+      "loss": 0.6457,
+      "step": 27700
+    },
+    {
+      "epoch": 2.771,
+      "grad_norm": 50.44115447998047,
+      "learning_rate": 4.461e-06,
+      "loss": 0.6635,
+      "step": 27710
+    },
+    {
+      "epoch": 2.7720000000000002,
+      "grad_norm": 19.699462890625,
+      "learning_rate": 4.459000000000001e-06,
+      "loss": 0.4893,
+      "step": 27720
+    },
+    {
+      "epoch": 2.773,
+      "grad_norm": 23.55537986755371,
+      "learning_rate": 4.457e-06,
+      "loss": 0.6317,
+      "step": 27730
+    },
+    {
+      "epoch": 2.774,
+      "grad_norm": 81.08770751953125,
+      "learning_rate": 4.4550000000000005e-06,
+      "loss": 0.7006,
+      "step": 27740
+    },
+    {
+      "epoch": 2.775,
+      "grad_norm": 20.223026275634766,
+      "learning_rate": 4.453e-06,
+      "loss": 0.5312,
+      "step": 27750
+    },
+    {
+      "epoch": 2.776,
+      "grad_norm": 46.003047943115234,
+      "learning_rate": 4.451e-06,
+      "loss": 0.7338,
+      "step": 27760
+    },
+    {
+      "epoch": 2.777,
+      "grad_norm": 7.982237339019775,
+      "learning_rate": 4.449000000000001e-06,
+      "loss": 0.4504,
+      "step": 27770
+    },
+    {
+      "epoch": 2.778,
+      "grad_norm": 44.6984977722168,
+      "learning_rate": 4.447e-06,
+      "loss": 0.8292,
+      "step": 27780
+    },
+    {
+      "epoch": 2.779,
+      "grad_norm": 17.92458724975586,
+      "learning_rate": 4.445000000000001e-06,
+      "loss": 0.5918,
+      "step": 27790
+    },
+    {
+      "epoch": 2.7800000000000002,
+      "grad_norm": 5.827061176300049,
+      "learning_rate": 4.4430000000000005e-06,
+      "loss": 0.549,
+      "step": 27800
+    },
+    {
+      "epoch": 2.781,
+      "grad_norm": 35.08589553833008,
+      "learning_rate": 4.441e-06,
+      "loss": 0.4181,
+      "step": 27810
+    },
+    {
+      "epoch": 2.782,
+      "grad_norm": 38.79449462890625,
+      "learning_rate": 4.439e-06,
+      "loss": 0.5933,
+      "step": 27820
+    },
+    {
+      "epoch": 2.783,
+      "grad_norm": 3.3199753761291504,
+      "learning_rate": 4.437e-06,
+      "loss": 0.4137,
+      "step": 27830
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 30.860759735107422,
+      "learning_rate": 4.435000000000001e-06,
+      "loss": 0.5869,
+      "step": 27840
+    },
+    {
+      "epoch": 2.785,
+      "grad_norm": 37.540809631347656,
+      "learning_rate": 4.433000000000001e-06,
+      "loss": 0.9034,
+      "step": 27850
+    },
+    {
+      "epoch": 2.786,
+      "grad_norm": 7.0977396965026855,
+      "learning_rate": 4.4310000000000004e-06,
+      "loss": 0.7018,
+      "step": 27860
+    },
+    {
+      "epoch": 2.787,
+      "grad_norm": 25.47239875793457,
+      "learning_rate": 4.429e-06,
+      "loss": 0.7053,
+      "step": 27870
+    },
+    {
+      "epoch": 2.7880000000000003,
+      "grad_norm": 32.19316864013672,
+      "learning_rate": 4.427e-06,
+      "loss": 0.6039,
+      "step": 27880
+    },
+    {
+      "epoch": 2.789,
+      "grad_norm": 46.36806869506836,
+      "learning_rate": 4.425e-06,
+      "loss": 0.6423,
+      "step": 27890
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 14.628954887390137,
+      "learning_rate": 4.423000000000001e-06,
+      "loss": 0.5092,
+      "step": 27900
+    },
+    {
+      "epoch": 2.791,
+      "grad_norm": 39.353546142578125,
+      "learning_rate": 4.4210000000000005e-06,
+      "loss": 0.5355,
+      "step": 27910
+    },
+    {
+      "epoch": 2.792,
+      "grad_norm": 28.54851531982422,
+      "learning_rate": 4.419e-06,
+      "loss": 0.5379,
+      "step": 27920
+    },
+    {
+      "epoch": 2.793,
+      "grad_norm": 24.745616912841797,
+      "learning_rate": 4.417e-06,
+      "loss": 0.695,
+      "step": 27930
+    },
+    {
+      "epoch": 2.794,
+      "grad_norm": 15.356302261352539,
+      "learning_rate": 4.415e-06,
+      "loss": 0.5868,
+      "step": 27940
+    },
+    {
+      "epoch": 2.795,
+      "grad_norm": 56.78498458862305,
+      "learning_rate": 4.413000000000001e-06,
+      "loss": 0.5065,
+      "step": 27950
+    },
+    {
+      "epoch": 2.7960000000000003,
+      "grad_norm": 27.250751495361328,
+      "learning_rate": 4.411000000000001e-06,
+      "loss": 0.6082,
+      "step": 27960
+    },
+    {
+      "epoch": 2.797,
+      "grad_norm": 14.399949073791504,
+      "learning_rate": 4.4090000000000005e-06,
+      "loss": 0.4962,
+      "step": 27970
+    },
+    {
+      "epoch": 2.798,
+      "grad_norm": 15.190604209899902,
+      "learning_rate": 4.407e-06,
+      "loss": 0.5516,
+      "step": 27980
+    },
+    {
+      "epoch": 2.799,
+      "grad_norm": 40.881858825683594,
+      "learning_rate": 4.405e-06,
+      "loss": 0.914,
+      "step": 27990
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 11.181133270263672,
+      "learning_rate": 4.403000000000001e-06,
+      "loss": 1.1243,
+      "step": 28000
+    },
+    {
+      "epoch": 2.801,
+      "grad_norm": 25.786142349243164,
+      "learning_rate": 4.401e-06,
+      "loss": 0.4991,
+      "step": 28010
+    },
+    {
+      "epoch": 2.802,
+      "grad_norm": 42.99445724487305,
+      "learning_rate": 4.3990000000000006e-06,
+      "loss": 0.5364,
+      "step": 28020
+    },
+    {
+      "epoch": 2.803,
+      "grad_norm": 49.733707427978516,
+      "learning_rate": 4.397e-06,
+      "loss": 0.773,
+      "step": 28030
+    },
+    {
+      "epoch": 2.8040000000000003,
+      "grad_norm": 44.46137237548828,
+      "learning_rate": 4.395e-06,
+      "loss": 0.5945,
+      "step": 28040
+    },
+    {
+      "epoch": 2.805,
+      "grad_norm": 59.004329681396484,
+      "learning_rate": 4.393000000000001e-06,
+      "loss": 0.6318,
+      "step": 28050
+    },
+    {
+      "epoch": 2.806,
+      "grad_norm": 32.86210632324219,
+      "learning_rate": 4.391e-06,
+      "loss": 0.5343,
+      "step": 28060
+    },
+    {
+      "epoch": 2.807,
+      "grad_norm": 15.308101654052734,
+      "learning_rate": 4.389000000000001e-06,
+      "loss": 0.6305,
+      "step": 28070
+    },
+    {
+      "epoch": 2.808,
+      "grad_norm": 37.713905334472656,
+      "learning_rate": 4.3870000000000005e-06,
+      "loss": 0.4288,
+      "step": 28080
+    },
+    {
+      "epoch": 2.809,
+      "grad_norm": 22.348554611206055,
+      "learning_rate": 4.385e-06,
+      "loss": 0.4645,
+      "step": 28090
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 36.5361328125,
+      "learning_rate": 4.383e-06,
+      "loss": 1.1116,
+      "step": 28100
+    },
+    {
+      "epoch": 2.811,
+      "grad_norm": 39.90386962890625,
+      "learning_rate": 4.381e-06,
+      "loss": 0.6124,
+      "step": 28110
+    },
+    {
+      "epoch": 2.8120000000000003,
+      "grad_norm": 49.790897369384766,
+      "learning_rate": 4.379000000000001e-06,
+      "loss": 0.7894,
+      "step": 28120
+    },
+    {
+      "epoch": 2.8129999999999997,
+      "grad_norm": 64.48717498779297,
+      "learning_rate": 4.377e-06,
+      "loss": 0.6046,
+      "step": 28130
+    },
+    {
+      "epoch": 2.814,
+      "grad_norm": 12.735812187194824,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.8096,
+      "step": 28140
+    },
+    {
+      "epoch": 2.815,
+      "grad_norm": 41.18169403076172,
+      "learning_rate": 4.373e-06,
+      "loss": 0.455,
+      "step": 28150
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 16.28331184387207,
+      "learning_rate": 4.371e-06,
+      "loss": 0.5784,
+      "step": 28160
+    },
+    {
+      "epoch": 2.817,
+      "grad_norm": 39.03124237060547,
+      "learning_rate": 4.369000000000001e-06,
+      "loss": 0.6922,
+      "step": 28170
+    },
+    {
+      "epoch": 2.818,
+      "grad_norm": 18.438169479370117,
+      "learning_rate": 4.367e-06,
+      "loss": 0.5654,
+      "step": 28180
+    },
+    {
+      "epoch": 2.819,
+      "grad_norm": 18.553783416748047,
+      "learning_rate": 4.3650000000000006e-06,
+      "loss": 0.7399,
+      "step": 28190
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 28.252033233642578,
+      "learning_rate": 4.363e-06,
+      "loss": 0.663,
+      "step": 28200
+    },
+    {
+      "epoch": 2.8209999999999997,
+      "grad_norm": 14.199079513549805,
+      "learning_rate": 4.361e-06,
+      "loss": 0.4705,
+      "step": 28210
+    },
+    {
+      "epoch": 2.822,
+      "grad_norm": 25.68846321105957,
+      "learning_rate": 4.359e-06,
+      "loss": 0.5264,
+      "step": 28220
+    },
+    {
+      "epoch": 2.823,
+      "grad_norm": 55.090919494628906,
+      "learning_rate": 4.357e-06,
+      "loss": 0.5085,
+      "step": 28230
+    },
+    {
+      "epoch": 2.824,
+      "grad_norm": 45.92644500732422,
+      "learning_rate": 4.355000000000001e-06,
+      "loss": 1.0086,
+      "step": 28240
+    },
+    {
+      "epoch": 2.825,
+      "grad_norm": 42.4690055847168,
+      "learning_rate": 4.3530000000000005e-06,
+      "loss": 0.9663,
+      "step": 28250
+    },
+    {
+      "epoch": 2.826,
+      "grad_norm": 43.58584976196289,
+      "learning_rate": 4.351e-06,
+      "loss": 0.8671,
+      "step": 28260
+    },
+    {
+      "epoch": 2.827,
+      "grad_norm": 38.93071365356445,
+      "learning_rate": 4.349e-06,
+      "loss": 0.5779,
+      "step": 28270
+    },
+    {
+      "epoch": 2.828,
+      "grad_norm": 31.59266471862793,
+      "learning_rate": 4.347e-06,
+      "loss": 0.6636,
+      "step": 28280
+    },
+    {
+      "epoch": 2.8289999999999997,
+      "grad_norm": 41.18907928466797,
+      "learning_rate": 4.345000000000001e-06,
+      "loss": 0.7013,
+      "step": 28290
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 10.073530197143555,
+      "learning_rate": 4.343000000000001e-06,
+      "loss": 0.6073,
+      "step": 28300
+    },
+    {
+      "epoch": 2.831,
+      "grad_norm": 20.655839920043945,
+      "learning_rate": 4.3410000000000005e-06,
+      "loss": 0.3729,
+      "step": 28310
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 18.784809112548828,
+      "learning_rate": 4.339e-06,
+      "loss": 0.7303,
+      "step": 28320
+    },
+    {
+      "epoch": 2.833,
+      "grad_norm": 57.55241775512695,
+      "learning_rate": 4.337e-06,
+      "loss": 0.9426,
+      "step": 28330
+    },
+    {
+      "epoch": 2.834,
+      "grad_norm": 33.25703811645508,
+      "learning_rate": 4.335e-06,
+      "loss": 0.8751,
+      "step": 28340
+    },
+    {
+      "epoch": 2.835,
+      "grad_norm": 8.003162384033203,
+      "learning_rate": 4.333000000000001e-06,
+      "loss": 0.57,
+      "step": 28350
+    },
+    {
+      "epoch": 2.836,
+      "grad_norm": 77.80805969238281,
+      "learning_rate": 4.3310000000000005e-06,
+      "loss": 0.5509,
+      "step": 28360
+    },
+    {
+      "epoch": 2.8369999999999997,
+      "grad_norm": 10.62796401977539,
+      "learning_rate": 4.329e-06,
+      "loss": 0.495,
+      "step": 28370
+    },
+    {
+      "epoch": 2.838,
+      "grad_norm": 32.467105865478516,
+      "learning_rate": 4.327e-06,
+      "loss": 0.4321,
+      "step": 28380
+    },
+    {
+      "epoch": 2.839,
+      "grad_norm": 22.827064514160156,
+      "learning_rate": 4.325e-06,
+      "loss": 0.5089,
+      "step": 28390
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 31.53234100341797,
+      "learning_rate": 4.323000000000001e-06,
+      "loss": 0.6055,
+      "step": 28400
+    },
+    {
+      "epoch": 2.841,
+      "grad_norm": 7.039924144744873,
+      "learning_rate": 4.321e-06,
+      "loss": 0.6939,
+      "step": 28410
+    },
+    {
+      "epoch": 2.842,
+      "grad_norm": 14.15356731414795,
+      "learning_rate": 4.3190000000000005e-06,
+      "loss": 0.7508,
+      "step": 28420
+    },
+    {
+      "epoch": 2.843,
+      "grad_norm": 59.17403030395508,
+      "learning_rate": 4.317e-06,
+      "loss": 0.5693,
+      "step": 28430
+    },
+    {
+      "epoch": 2.844,
+      "grad_norm": 8.260176658630371,
+      "learning_rate": 4.315e-06,
+      "loss": 0.4824,
+      "step": 28440
+    },
+    {
+      "epoch": 2.8449999999999998,
+      "grad_norm": 24.74396324157715,
+      "learning_rate": 4.313000000000001e-06,
+      "loss": 0.5549,
+      "step": 28450
+    },
+    {
+      "epoch": 2.846,
+      "grad_norm": 30.768203735351562,
+      "learning_rate": 4.311e-06,
+      "loss": 0.6408,
+      "step": 28460
+    },
+    {
+      "epoch": 2.847,
+      "grad_norm": 37.09162139892578,
+      "learning_rate": 4.309000000000001e-06,
+      "loss": 0.6981,
+      "step": 28470
+    },
+    {
+      "epoch": 2.848,
+      "grad_norm": 21.815000534057617,
+      "learning_rate": 4.3070000000000004e-06,
+      "loss": 0.607,
+      "step": 28480
+    },
+    {
+      "epoch": 2.849,
+      "grad_norm": 43.1075553894043,
+      "learning_rate": 4.305e-06,
+      "loss": 0.6963,
+      "step": 28490
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 61.59611129760742,
+      "learning_rate": 4.303e-06,
+      "loss": 0.661,
+      "step": 28500
+    },
+    {
+      "epoch": 2.851,
+      "grad_norm": 18.589096069335938,
+      "learning_rate": 4.301e-06,
+      "loss": 0.7131,
+      "step": 28510
+    },
+    {
+      "epoch": 2.852,
+      "grad_norm": 5.096174716949463,
+      "learning_rate": 4.299000000000001e-06,
+      "loss": 0.4882,
+      "step": 28520
+    },
+    {
+      "epoch": 2.8529999999999998,
+      "grad_norm": 20.526193618774414,
+      "learning_rate": 4.2970000000000005e-06,
+      "loss": 0.681,
+      "step": 28530
+    },
+    {
+      "epoch": 2.854,
+      "grad_norm": 26.11942481994629,
+      "learning_rate": 4.295e-06,
+      "loss": 0.7666,
+      "step": 28540
+    },
+    {
+      "epoch": 2.855,
+      "grad_norm": 26.176834106445312,
+      "learning_rate": 4.293e-06,
+      "loss": 0.5955,
+      "step": 28550
+    },
+    {
+      "epoch": 2.856,
+      "grad_norm": 27.885841369628906,
+      "learning_rate": 4.291e-06,
+      "loss": 0.6494,
+      "step": 28560
+    },
+    {
+      "epoch": 2.857,
+      "grad_norm": 33.89963150024414,
+      "learning_rate": 4.289000000000001e-06,
+      "loss": 0.2812,
+      "step": 28570
+    },
+    {
+      "epoch": 2.858,
+      "grad_norm": 41.28437805175781,
+      "learning_rate": 4.287000000000001e-06,
+      "loss": 0.7791,
+      "step": 28580
+    },
+    {
+      "epoch": 2.859,
+      "grad_norm": 31.66999053955078,
+      "learning_rate": 4.2850000000000005e-06,
+      "loss": 0.6356,
+      "step": 28590
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 38.8473014831543,
+      "learning_rate": 4.283e-06,
+      "loss": 0.6462,
+      "step": 28600
+    },
+    {
+      "epoch": 2.8609999999999998,
+      "grad_norm": 38.425048828125,
+      "learning_rate": 4.281e-06,
+      "loss": 0.5853,
+      "step": 28610
+    },
+    {
+      "epoch": 2.862,
+      "grad_norm": 38.103275299072266,
+      "learning_rate": 4.279e-06,
+      "loss": 0.4751,
+      "step": 28620
+    },
+    {
+      "epoch": 2.863,
+      "grad_norm": 44.24375915527344,
+      "learning_rate": 4.277000000000001e-06,
+      "loss": 0.7975,
+      "step": 28630
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 27.316816329956055,
+      "learning_rate": 4.2750000000000006e-06,
+      "loss": 0.7389,
+      "step": 28640
+    },
+    {
+      "epoch": 2.865,
+      "grad_norm": 76.10896301269531,
+      "learning_rate": 4.2730000000000004e-06,
+      "loss": 0.3966,
+      "step": 28650
+    },
+    {
+      "epoch": 2.866,
+      "grad_norm": 47.929649353027344,
+      "learning_rate": 4.271e-06,
+      "loss": 0.62,
+      "step": 28660
+    },
+    {
+      "epoch": 2.867,
+      "grad_norm": 52.1750602722168,
+      "learning_rate": 4.269e-06,
+      "loss": 0.7978,
+      "step": 28670
+    },
+    {
+      "epoch": 2.868,
+      "grad_norm": 52.91067886352539,
+      "learning_rate": 4.267000000000001e-06,
+      "loss": 0.6759,
+      "step": 28680
+    },
+    {
+      "epoch": 2.8689999999999998,
+      "grad_norm": 30.69292449951172,
+      "learning_rate": 4.265000000000001e-06,
+      "loss": 0.5969,
+      "step": 28690
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 45.15058517456055,
+      "learning_rate": 4.2630000000000005e-06,
+      "loss": 0.422,
+      "step": 28700
+    },
+    {
+      "epoch": 2.871,
+      "grad_norm": 9.530531883239746,
+      "learning_rate": 4.261e-06,
+      "loss": 0.5713,
+      "step": 28710
+    },
+    {
+      "epoch": 2.872,
+      "grad_norm": 15.413444519042969,
+      "learning_rate": 4.259e-06,
+      "loss": 0.3489,
+      "step": 28720
+    },
+    {
+      "epoch": 2.873,
+      "grad_norm": 33.19585418701172,
+      "learning_rate": 4.257000000000001e-06,
+      "loss": 0.6475,
+      "step": 28730
+    },
+    {
+      "epoch": 2.874,
+      "grad_norm": 31.359365463256836,
+      "learning_rate": 4.255e-06,
+      "loss": 0.5289,
+      "step": 28740
+    },
+    {
+      "epoch": 2.875,
+      "grad_norm": 37.34211349487305,
+      "learning_rate": 4.253000000000001e-06,
+      "loss": 0.6632,
+      "step": 28750
+    },
+    {
+      "epoch": 2.876,
+      "grad_norm": 46.908180236816406,
+      "learning_rate": 4.2510000000000005e-06,
+      "loss": 0.6369,
+      "step": 28760
+    },
+    {
+      "epoch": 2.877,
+      "grad_norm": 33.455013275146484,
+      "learning_rate": 4.249e-06,
+      "loss": 0.4806,
+      "step": 28770
+    },
+    {
+      "epoch": 2.878,
+      "grad_norm": 38.421104431152344,
+      "learning_rate": 4.247e-06,
+      "loss": 0.5913,
+      "step": 28780
+    },
+    {
+      "epoch": 2.879,
+      "grad_norm": 27.120136260986328,
+      "learning_rate": 4.245e-06,
+      "loss": 0.5004,
+      "step": 28790
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 54.891231536865234,
+      "learning_rate": 4.243000000000001e-06,
+      "loss": 0.6929,
+      "step": 28800
+    },
+    {
+      "epoch": 2.8810000000000002,
+      "grad_norm": 27.580013275146484,
+      "learning_rate": 4.241e-06,
+      "loss": 0.7942,
+      "step": 28810
+    },
+    {
+      "epoch": 2.882,
+      "grad_norm": 34.77034378051758,
+      "learning_rate": 4.239e-06,
+      "loss": 0.7068,
+      "step": 28820
+    },
+    {
+      "epoch": 2.883,
+      "grad_norm": 17.92949104309082,
+      "learning_rate": 4.237e-06,
+      "loss": 0.5559,
+      "step": 28830
+    },
+    {
+      "epoch": 2.884,
+      "grad_norm": 30.56000328063965,
+      "learning_rate": 4.235e-06,
+      "loss": 0.5818,
+      "step": 28840
+    },
+    {
+      "epoch": 2.885,
+      "grad_norm": 57.818504333496094,
+      "learning_rate": 4.233000000000001e-06,
+      "loss": 0.7016,
+      "step": 28850
+    },
+    {
+      "epoch": 2.886,
+      "grad_norm": 31.524658203125,
+      "learning_rate": 4.231e-06,
+      "loss": 0.5322,
+      "step": 28860
+    },
+    {
+      "epoch": 2.887,
+      "grad_norm": 27.571067810058594,
+      "learning_rate": 4.2290000000000005e-06,
+      "loss": 0.6619,
+      "step": 28870
+    },
+    {
+      "epoch": 2.888,
+      "grad_norm": 6.408551216125488,
+      "learning_rate": 4.227e-06,
+      "loss": 0.6884,
+      "step": 28880
+    },
+    {
+      "epoch": 2.8890000000000002,
+      "grad_norm": 51.78606033325195,
+      "learning_rate": 4.225e-06,
+      "loss": 0.4897,
+      "step": 28890
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 12.073798179626465,
+      "learning_rate": 4.223e-06,
+      "loss": 0.6119,
+      "step": 28900
+    },
+    {
+      "epoch": 2.891,
+      "grad_norm": 127.30245208740234,
+      "learning_rate": 4.221e-06,
+      "loss": 0.5858,
+      "step": 28910
+    },
+    {
+      "epoch": 2.892,
+      "grad_norm": 5.639313220977783,
+      "learning_rate": 4.219000000000001e-06,
+      "loss": 0.4539,
+      "step": 28920
+    },
+    {
+      "epoch": 2.893,
+      "grad_norm": 3.212557077407837,
+      "learning_rate": 4.2170000000000005e-06,
+      "loss": 0.5321,
+      "step": 28930
+    },
+    {
+      "epoch": 2.894,
+      "grad_norm": 34.34654235839844,
+      "learning_rate": 4.215e-06,
+      "loss": 0.5977,
+      "step": 28940
+    },
+    {
+      "epoch": 2.895,
+      "grad_norm": 25.9399471282959,
+      "learning_rate": 4.213e-06,
+      "loss": 0.75,
+      "step": 28950
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 70.75745391845703,
+      "learning_rate": 4.211e-06,
+      "loss": 0.5391,
+      "step": 28960
+    },
+    {
+      "epoch": 2.8970000000000002,
+      "grad_norm": 35.240882873535156,
+      "learning_rate": 4.209000000000001e-06,
+      "loss": 0.619,
+      "step": 28970
+    },
+    {
+      "epoch": 2.898,
+      "grad_norm": 43.350284576416016,
+      "learning_rate": 4.2070000000000005e-06,
+      "loss": 0.6964,
+      "step": 28980
+    },
+    {
+      "epoch": 2.899,
+      "grad_norm": 27.253419876098633,
+      "learning_rate": 4.205e-06,
+      "loss": 0.6312,
+      "step": 28990
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 22.365201950073242,
+      "learning_rate": 4.203e-06,
+      "loss": 0.7312,
+      "step": 29000
+    },
+    {
+      "epoch": 2.901,
+      "grad_norm": 33.48457336425781,
+      "learning_rate": 4.201e-06,
+      "loss": 0.7824,
+      "step": 29010
+    },
+    {
+      "epoch": 2.902,
+      "grad_norm": 20.902122497558594,
+      "learning_rate": 4.199e-06,
+      "loss": 0.568,
+      "step": 29020
+    },
+    {
+      "epoch": 2.903,
+      "grad_norm": 34.6548957824707,
+      "learning_rate": 4.197000000000001e-06,
+      "loss": 0.3214,
+      "step": 29030
+    },
+    {
+      "epoch": 2.904,
+      "grad_norm": 13.812738418579102,
+      "learning_rate": 4.1950000000000005e-06,
+      "loss": 0.5746,
+      "step": 29040
+    },
+    {
+      "epoch": 2.9050000000000002,
+      "grad_norm": 15.632746696472168,
+      "learning_rate": 4.193e-06,
+      "loss": 0.5344,
+      "step": 29050
+    },
+    {
+      "epoch": 2.906,
+      "grad_norm": 5.410891532897949,
+      "learning_rate": 4.191e-06,
+      "loss": 0.5319,
+      "step": 29060
+    },
+    {
+      "epoch": 2.907,
+      "grad_norm": 18.57172966003418,
+      "learning_rate": 4.189e-06,
+      "loss": 0.6795,
+      "step": 29070
+    },
+    {
+      "epoch": 2.908,
+      "grad_norm": 38.506141662597656,
+      "learning_rate": 4.187000000000001e-06,
+      "loss": 0.8386,
+      "step": 29080
+    },
+    {
+      "epoch": 2.909,
+      "grad_norm": 62.42726135253906,
+      "learning_rate": 4.185000000000001e-06,
+      "loss": 0.4449,
+      "step": 29090
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": Infinity,
+      "learning_rate": 4.183200000000001e-06,
+      "loss": 0.7186,
+      "step": 29100
+    },
+    {
+      "epoch": 2.911,
+      "grad_norm": 42.34231948852539,
+      "learning_rate": 4.1812e-06,
+      "loss": 0.5659,
+      "step": 29110
+    },
+    {
+      "epoch": 2.912,
+      "grad_norm": 40.03559875488281,
+      "learning_rate": 4.179200000000001e-06,
+      "loss": 0.5508,
+      "step": 29120
+    },
+    {
+      "epoch": 2.9130000000000003,
+      "grad_norm": 49.353240966796875,
+      "learning_rate": 4.1772000000000004e-06,
+      "loss": 0.68,
+      "step": 29130
+    },
+    {
+      "epoch": 2.914,
+      "grad_norm": 38.63571548461914,
+      "learning_rate": 4.1752e-06,
+      "loss": 0.7239,
+      "step": 29140
+    },
+    {
+      "epoch": 2.915,
+      "grad_norm": 7.519901275634766,
+      "learning_rate": 4.173200000000001e-06,
+      "loss": 0.4593,
+      "step": 29150
+    },
+    {
+      "epoch": 2.916,
+      "grad_norm": 32.73497772216797,
+      "learning_rate": 4.1712e-06,
+      "loss": 0.4446,
+      "step": 29160
+    },
+    {
+      "epoch": 2.917,
+      "grad_norm": 35.581703186035156,
+      "learning_rate": 4.169200000000001e-06,
+      "loss": 0.2908,
+      "step": 29170
+    },
+    {
+      "epoch": 2.918,
+      "grad_norm": 20.578540802001953,
+      "learning_rate": 4.1672000000000005e-06,
+      "loss": 0.7993,
+      "step": 29180
+    },
+    {
+      "epoch": 2.919,
+      "grad_norm": 22.452545166015625,
+      "learning_rate": 4.1652e-06,
+      "loss": 0.5683,
+      "step": 29190
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 19.33148765563965,
+      "learning_rate": 4.1632e-06,
+      "loss": 0.4966,
+      "step": 29200
+    },
+    {
+      "epoch": 2.9210000000000003,
+      "grad_norm": 20.982643127441406,
+      "learning_rate": 4.1612e-06,
+      "loss": 0.6517,
+      "step": 29210
+    },
+    {
+      "epoch": 2.922,
+      "grad_norm": 45.04250717163086,
+      "learning_rate": 4.159200000000001e-06,
+      "loss": 0.6315,
+      "step": 29220
+    },
+    {
+      "epoch": 2.923,
+      "grad_norm": 40.241294860839844,
+      "learning_rate": 4.1572e-06,
+      "loss": 0.7002,
+      "step": 29230
+    },
+    {
+      "epoch": 2.924,
+      "grad_norm": 31.673635482788086,
+      "learning_rate": 4.1552000000000005e-06,
+      "loss": 0.7658,
+      "step": 29240
+    },
+    {
+      "epoch": 2.925,
+      "grad_norm": 29.901514053344727,
+      "learning_rate": 4.1532e-06,
+      "loss": 0.8841,
+      "step": 29250
+    },
+    {
+      "epoch": 2.926,
+      "grad_norm": 12.362701416015625,
+      "learning_rate": 4.1512e-06,
+      "loss": 0.5445,
+      "step": 29260
+    },
+    {
+      "epoch": 2.927,
+      "grad_norm": 36.973838806152344,
+      "learning_rate": 4.149200000000001e-06,
+      "loss": 0.5255,
+      "step": 29270
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 1.602242112159729,
+      "learning_rate": 4.1472e-06,
+      "loss": 0.7783,
+      "step": 29280
+    },
+    {
+      "epoch": 2.9290000000000003,
+      "grad_norm": 33.756126403808594,
+      "learning_rate": 4.1452000000000006e-06,
+      "loss": 0.5491,
+      "step": 29290
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 68.7015380859375,
+      "learning_rate": 4.1432e-06,
+      "loss": 0.7136,
+      "step": 29300
+    },
+    {
+      "epoch": 2.931,
+      "grad_norm": 27.283906936645508,
+      "learning_rate": 4.1412e-06,
+      "loss": 0.8519,
+      "step": 29310
+    },
+    {
+      "epoch": 2.932,
+      "grad_norm": 29.23966407775879,
+      "learning_rate": 4.1392e-06,
+      "loss": 0.736,
+      "step": 29320
+    },
+    {
+      "epoch": 2.933,
+      "grad_norm": 30.30097198486328,
+      "learning_rate": 4.1372e-06,
+      "loss": 0.4569,
+      "step": 29330
+    },
+    {
+      "epoch": 2.934,
+      "grad_norm": 60.6238899230957,
+      "learning_rate": 4.135200000000001e-06,
+      "loss": 0.6256,
+      "step": 29340
+    },
+    {
+      "epoch": 2.935,
+      "grad_norm": 48.743343353271484,
+      "learning_rate": 4.1332000000000005e-06,
+      "loss": 0.7367,
+      "step": 29350
+    },
+    {
+      "epoch": 2.936,
+      "grad_norm": 30.004465103149414,
+      "learning_rate": 4.1312e-06,
+      "loss": 1.0274,
+      "step": 29360
+    },
+    {
+      "epoch": 2.9370000000000003,
+      "grad_norm": 26.581308364868164,
+      "learning_rate": 4.1292e-06,
+      "loss": 0.5193,
+      "step": 29370
+    },
+    {
+      "epoch": 2.9379999999999997,
+      "grad_norm": 5.948087215423584,
+      "learning_rate": 4.1272e-06,
+      "loss": 0.44,
+      "step": 29380
+    },
+    {
+      "epoch": 2.939,
+      "grad_norm": 47.7027473449707,
+      "learning_rate": 4.1252e-06,
+      "loss": 0.7333,
+      "step": 29390
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 39.193851470947266,
+      "learning_rate": 4.123200000000001e-06,
+      "loss": 0.6778,
+      "step": 29400
+    },
+    {
+      "epoch": 2.941,
+      "grad_norm": 27.26168441772461,
+      "learning_rate": 4.1212000000000005e-06,
+      "loss": 0.6049,
+      "step": 29410
+    },
+    {
+      "epoch": 2.942,
+      "grad_norm": 22.667373657226562,
+      "learning_rate": 4.1192e-06,
+      "loss": 0.5667,
+      "step": 29420
+    },
+    {
+      "epoch": 2.943,
+      "grad_norm": 33.34720993041992,
+      "learning_rate": 4.1172e-06,
+      "loss": 0.5668,
+      "step": 29430
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 37.20878601074219,
+      "learning_rate": 4.1152e-06,
+      "loss": 0.916,
+      "step": 29440
+    },
+    {
+      "epoch": 2.945,
+      "grad_norm": 10.755010604858398,
+      "learning_rate": 4.113200000000001e-06,
+      "loss": 0.6121,
+      "step": 29450
+    },
+    {
+      "epoch": 2.9459999999999997,
+      "grad_norm": 13.625876426696777,
+      "learning_rate": 4.1112000000000006e-06,
+      "loss": 0.5307,
+      "step": 29460
+    },
+    {
+      "epoch": 2.947,
+      "grad_norm": 57.56064224243164,
+      "learning_rate": 4.1092e-06,
+      "loss": 0.4517,
+      "step": 29470
+    },
+    {
+      "epoch": 2.948,
+      "grad_norm": 27.347122192382812,
+      "learning_rate": 4.1072e-06,
+      "loss": 0.6075,
+      "step": 29480
+    },
+    {
+      "epoch": 2.949,
+      "grad_norm": 64.42729949951172,
+      "learning_rate": 4.1052e-06,
+      "loss": 0.7204,
+      "step": 29490
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 42.93135452270508,
+      "learning_rate": 4.103200000000001e-06,
+      "loss": 0.586,
+      "step": 29500
+    },
+    {
+      "epoch": 2.951,
+      "grad_norm": 36.54902648925781,
+      "learning_rate": 4.1012e-06,
+      "loss": 0.7982,
+      "step": 29510
+    },
+    {
+      "epoch": 2.952,
+      "grad_norm": 25.70467758178711,
+      "learning_rate": 4.0992000000000005e-06,
+      "loss": 0.5134,
+      "step": 29520
+    },
+    {
+      "epoch": 2.953,
+      "grad_norm": 22.512723922729492,
+      "learning_rate": 4.0972e-06,
+      "loss": 0.8235,
+      "step": 29530
+    },
+    {
+      "epoch": 2.9539999999999997,
+      "grad_norm": 50.21841049194336,
+      "learning_rate": 4.0952e-06,
+      "loss": 0.6306,
+      "step": 29540
+    },
+    {
+      "epoch": 2.955,
+      "grad_norm": 17.58340835571289,
+      "learning_rate": 4.093200000000001e-06,
+      "loss": 0.5652,
+      "step": 29550
+    },
+    {
+      "epoch": 2.956,
+      "grad_norm": 19.082971572875977,
+      "learning_rate": 4.0912e-06,
+      "loss": 0.5947,
+      "step": 29560
+    },
+    {
+      "epoch": 2.957,
+      "grad_norm": 47.442440032958984,
+      "learning_rate": 4.089200000000001e-06,
+      "loss": 0.7297,
+      "step": 29570
+    },
+    {
+      "epoch": 2.958,
+      "grad_norm": 17.213855743408203,
+      "learning_rate": 4.0872000000000004e-06,
+      "loss": 0.4932,
+      "step": 29580
+    },
+    {
+      "epoch": 2.959,
+      "grad_norm": 40.99843978881836,
+      "learning_rate": 4.0852e-06,
+      "loss": 0.6037,
+      "step": 29590
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 29.501218795776367,
+      "learning_rate": 4.0832e-06,
+      "loss": 0.5786,
+      "step": 29600
+    },
+    {
+      "epoch": 2.961,
+      "grad_norm": 28.239242553710938,
+      "learning_rate": 4.0812e-06,
+      "loss": 0.788,
+      "step": 29610
+    },
+    {
+      "epoch": 2.9619999999999997,
+      "grad_norm": 53.01205062866211,
+      "learning_rate": 4.079200000000001e-06,
+      "loss": 0.6026,
+      "step": 29620
+    },
+    {
+      "epoch": 2.963,
+      "grad_norm": 22.547651290893555,
+      "learning_rate": 4.0772000000000005e-06,
+      "loss": 0.6091,
+      "step": 29630
+    },
+    {
+      "epoch": 2.964,
+      "grad_norm": 45.12074279785156,
+      "learning_rate": 4.0752e-06,
+      "loss": 0.5451,
+      "step": 29640
+    },
+    {
+      "epoch": 2.965,
+      "grad_norm": 15.948826789855957,
+      "learning_rate": 4.0732e-06,
+      "loss": 0.3129,
+      "step": 29650
+    },
+    {
+      "epoch": 2.966,
+      "grad_norm": 8.485644340515137,
+      "learning_rate": 4.0712e-06,
+      "loss": 0.6199,
+      "step": 29660
+    },
+    {
+      "epoch": 2.967,
+      "grad_norm": 9.05318546295166,
+      "learning_rate": 4.069200000000001e-06,
+      "loss": 0.7764,
+      "step": 29670
+    },
+    {
+      "epoch": 2.968,
+      "grad_norm": 14.031304359436035,
+      "learning_rate": 4.067200000000001e-06,
+      "loss": 0.4388,
+      "step": 29680
+    },
+    {
+      "epoch": 2.969,
+      "grad_norm": 34.62077713012695,
+      "learning_rate": 4.0652000000000005e-06,
+      "loss": 0.6206,
+      "step": 29690
+    },
+    {
+      "epoch": 2.9699999999999998,
+      "grad_norm": 29.150148391723633,
+      "learning_rate": 4.0632e-06,
+      "loss": 0.4939,
+      "step": 29700
+    },
+    {
+      "epoch": 2.971,
+      "grad_norm": 29.395282745361328,
+      "learning_rate": 4.0612e-06,
+      "loss": 0.4836,
+      "step": 29710
+    },
+    {
+      "epoch": 2.972,
+      "grad_norm": 50.002437591552734,
+      "learning_rate": 4.0592e-06,
+      "loss": 0.5221,
+      "step": 29720
+    },
+    {
+      "epoch": 2.973,
+      "grad_norm": 21.285947799682617,
+      "learning_rate": 4.057200000000001e-06,
+      "loss": 0.6336,
+      "step": 29730
+    },
+    {
+      "epoch": 2.974,
+      "grad_norm": 18.822994232177734,
+      "learning_rate": 4.055200000000001e-06,
+      "loss": 0.6895,
+      "step": 29740
+    },
+    {
+      "epoch": 2.975,
+      "grad_norm": 59.74067306518555,
+      "learning_rate": 4.0532000000000004e-06,
+      "loss": 0.6264,
+      "step": 29750
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 46.628448486328125,
+      "learning_rate": 4.0512e-06,
+      "loss": 0.5139,
+      "step": 29760
+    },
+    {
+      "epoch": 2.977,
+      "grad_norm": 4.769842147827148,
+      "learning_rate": 4.0492e-06,
+      "loss": 0.4523,
+      "step": 29770
+    },
+    {
+      "epoch": 2.9779999999999998,
+      "grad_norm": 42.81772994995117,
+      "learning_rate": 4.047200000000001e-06,
+      "loss": 0.5483,
+      "step": 29780
+    },
+    {
+      "epoch": 2.979,
+      "grad_norm": 37.86781692504883,
+      "learning_rate": 4.0452e-06,
+      "loss": 0.4806,
+      "step": 29790
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 28.3906192779541,
+      "learning_rate": 4.0432000000000005e-06,
+      "loss": 0.7016,
+      "step": 29800
+    },
+    {
+      "epoch": 2.981,
+      "grad_norm": 9.775362014770508,
+      "learning_rate": 4.0412e-06,
+      "loss": 0.6332,
+      "step": 29810
+    },
+    {
+      "epoch": 2.982,
+      "grad_norm": 5.316655158996582,
+      "learning_rate": 4.0392e-06,
+      "loss": 0.6472,
+      "step": 29820
+    },
+    {
+      "epoch": 2.983,
+      "grad_norm": 48.42181396484375,
+      "learning_rate": 4.037200000000001e-06,
+      "loss": 0.6845,
+      "step": 29830
+    },
+    {
+      "epoch": 2.984,
+      "grad_norm": 33.597869873046875,
+      "learning_rate": 4.0352e-06,
+      "loss": 0.6101,
+      "step": 29840
+    },
+    {
+      "epoch": 2.985,
+      "grad_norm": 38.29161071777344,
+      "learning_rate": 4.033200000000001e-06,
+      "loss": 0.7222,
+      "step": 29850
+    },
+    {
+      "epoch": 2.9859999999999998,
+      "grad_norm": 33.69076919555664,
+      "learning_rate": 4.0312000000000005e-06,
+      "loss": 0.6477,
+      "step": 29860
+    },
+    {
+      "epoch": 2.987,
+      "grad_norm": 19.68893051147461,
+      "learning_rate": 4.0292e-06,
+      "loss": 0.4098,
+      "step": 29870
+    },
+    {
+      "epoch": 2.988,
+      "grad_norm": 46.982784271240234,
+      "learning_rate": 4.0272e-06,
+      "loss": 0.5056,
+      "step": 29880
+    },
+    {
+      "epoch": 2.989,
+      "grad_norm": 15.345257759094238,
+      "learning_rate": 4.0252e-06,
+      "loss": 0.8451,
+      "step": 29890
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 34.063167572021484,
+      "learning_rate": 4.023200000000001e-06,
+      "loss": 1.02,
+      "step": 29900
+    },
+    {
+      "epoch": 2.991,
+      "grad_norm": 23.77176284790039,
+      "learning_rate": 4.0212e-06,
+      "loss": 0.8993,
+      "step": 29910
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 7.723876476287842,
+      "learning_rate": 4.0192e-06,
+      "loss": 0.4682,
+      "step": 29920
+    },
+    {
+      "epoch": 2.993,
+      "grad_norm": 9.764384269714355,
+      "learning_rate": 4.0172e-06,
+      "loss": 0.4996,
+      "step": 29930
+    },
+    {
+      "epoch": 2.9939999999999998,
+      "grad_norm": 33.22534942626953,
+      "learning_rate": 4.0152e-06,
+      "loss": 0.6259,
+      "step": 29940
+    },
+    {
+      "epoch": 2.995,
+      "grad_norm": 17.53177833557129,
+      "learning_rate": 4.013200000000001e-06,
+      "loss": 0.5254,
+      "step": 29950
+    },
+    {
+      "epoch": 2.996,
+      "grad_norm": 53.74732971191406,
+      "learning_rate": 4.0112e-06,
+      "loss": 0.9515,
+      "step": 29960
+    },
+    {
+      "epoch": 2.997,
+      "grad_norm": 12.981660842895508,
+      "learning_rate": 4.0092000000000005e-06,
+      "loss": 0.5202,
+      "step": 29970
+    },
+    {
+      "epoch": 2.998,
+      "grad_norm": 46.99820327758789,
+      "learning_rate": 4.0072e-06,
+      "loss": 0.4536,
+      "step": 29980
+    },
+    {
+      "epoch": 2.999,
+      "grad_norm": 40.24980926513672,
+      "learning_rate": 4.0052e-06,
+      "loss": 0.7089,
+      "step": 29990
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.793490409851074,
+      "learning_rate": 4.0032e-06,
+      "loss": 0.7395,
+      "step": 30000
+    },
+    {
+      "epoch": 3.001,
+      "grad_norm": 31.98954963684082,
+      "learning_rate": 4.0012e-06,
+      "loss": 0.4896,
+      "step": 30010
+    },
+    {
+      "epoch": 3.002,
+      "grad_norm": 52.853424072265625,
+      "learning_rate": 3.999200000000001e-06,
+      "loss": 0.3648,
+      "step": 30020
+    },
+    {
+      "epoch": 3.003,
+      "grad_norm": 49.65998077392578,
+      "learning_rate": 3.9972000000000005e-06,
+      "loss": 0.5501,
+      "step": 30030
+    },
+    {
+      "epoch": 3.004,
+      "grad_norm": 50.96385955810547,
+      "learning_rate": 3.9952e-06,
+      "loss": 0.6445,
+      "step": 30040
+    },
+    {
+      "epoch": 3.005,
+      "grad_norm": 36.44496154785156,
+      "learning_rate": 3.9932e-06,
+      "loss": 0.6699,
+      "step": 30050
+    },
+    {
+      "epoch": 3.006,
+      "grad_norm": 17.591075897216797,
+      "learning_rate": 3.9912e-06,
+      "loss": 0.7079,
+      "step": 30060
+    },
+    {
+      "epoch": 3.007,
+      "grad_norm": 78.94903564453125,
+      "learning_rate": 3.989200000000001e-06,
+      "loss": 0.6877,
+      "step": 30070
+    },
+    {
+      "epoch": 3.008,
+      "grad_norm": 51.7623291015625,
+      "learning_rate": 3.9872000000000006e-06,
+      "loss": 0.7178,
+      "step": 30080
+    },
+    {
+      "epoch": 3.009,
+      "grad_norm": 37.82719421386719,
+      "learning_rate": 3.9852e-06,
+      "loss": 0.7465,
+      "step": 30090
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 36.30388641357422,
+      "learning_rate": 3.9832e-06,
+      "loss": 0.5996,
+      "step": 30100
+    },
+    {
+      "epoch": 3.011,
+      "grad_norm": 4.858558654785156,
+      "learning_rate": 3.9812e-06,
+      "loss": 0.5496,
+      "step": 30110
+    },
+    {
+      "epoch": 3.012,
+      "grad_norm": 34.84873962402344,
+      "learning_rate": 3.9792e-06,
+      "loss": 0.6838,
+      "step": 30120
+    },
+    {
+      "epoch": 3.013,
+      "grad_norm": 33.88477325439453,
+      "learning_rate": 3.977200000000001e-06,
+      "loss": 0.4307,
+      "step": 30130
+    },
+    {
+      "epoch": 3.014,
+      "grad_norm": 39.399967193603516,
+      "learning_rate": 3.9752000000000005e-06,
+      "loss": 0.6973,
+      "step": 30140
+    },
+    {
+      "epoch": 3.015,
+      "grad_norm": 20.011003494262695,
+      "learning_rate": 3.9732e-06,
+      "loss": 0.695,
+      "step": 30150
+    },
+    {
+      "epoch": 3.016,
+      "grad_norm": 28.336265563964844,
+      "learning_rate": 3.9712e-06,
+      "loss": 0.5254,
+      "step": 30160
+    },
+    {
+      "epoch": 3.017,
+      "grad_norm": 40.91427230834961,
+      "learning_rate": 3.9692e-06,
+      "loss": 0.4747,
+      "step": 30170
+    },
+    {
+      "epoch": 3.018,
+      "grad_norm": 17.782920837402344,
+      "learning_rate": 3.967200000000001e-06,
+      "loss": 0.5615,
+      "step": 30180
+    },
+    {
+      "epoch": 3.019,
+      "grad_norm": 58.19817352294922,
+      "learning_rate": 3.9652e-06,
+      "loss": 0.8336,
+      "step": 30190
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 32.32502746582031,
+      "learning_rate": 3.9632000000000004e-06,
+      "loss": 0.7315,
+      "step": 30200
+    },
+    {
+      "epoch": 3.021,
+      "grad_norm": 5.095047950744629,
+      "learning_rate": 3.9612e-06,
+      "loss": 0.456,
+      "step": 30210
+    },
+    {
+      "epoch": 3.022,
+      "grad_norm": 41.41087341308594,
+      "learning_rate": 3.9592e-06,
+      "loss": 0.8262,
+      "step": 30220
+    },
+    {
+      "epoch": 3.023,
+      "grad_norm": 1.5623220205307007,
+      "learning_rate": 3.957200000000001e-06,
+      "loss": 0.5899,
+      "step": 30230
+    },
+    {
+      "epoch": 3.024,
+      "grad_norm": 56.326873779296875,
+      "learning_rate": 3.9552e-06,
+      "loss": 0.7977,
+      "step": 30240
+    },
+    {
+      "epoch": 3.025,
+      "grad_norm": 13.570528030395508,
+      "learning_rate": 3.9532000000000005e-06,
+      "loss": 0.5588,
+      "step": 30250
+    },
+    {
+      "epoch": 3.026,
+      "grad_norm": 19.437191009521484,
+      "learning_rate": 3.9512e-06,
+      "loss": 0.391,
+      "step": 30260
+    },
+    {
+      "epoch": 3.027,
+      "grad_norm": 99.22734069824219,
+      "learning_rate": 3.9492e-06,
+      "loss": 0.6078,
+      "step": 30270
+    },
+    {
+      "epoch": 3.028,
+      "grad_norm": 36.386932373046875,
+      "learning_rate": 3.9472e-06,
+      "loss": 0.5783,
+      "step": 30280
+    },
+    {
+      "epoch": 3.029,
+      "grad_norm": 6.841565132141113,
+      "learning_rate": 3.9452e-06,
+      "loss": 0.5239,
+      "step": 30290
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 53.24142074584961,
+      "learning_rate": 3.943200000000001e-06,
+      "loss": 0.486,
+      "step": 30300
+    },
+    {
+      "epoch": 3.031,
+      "grad_norm": 42.58986282348633,
+      "learning_rate": 3.9412000000000005e-06,
+      "loss": 0.6734,
+      "step": 30310
+    },
+    {
+      "epoch": 3.032,
+      "grad_norm": 13.315511703491211,
+      "learning_rate": 3.9392e-06,
+      "loss": 0.3784,
+      "step": 30320
+    },
+    {
+      "epoch": 3.033,
+      "grad_norm": 30.482662200927734,
+      "learning_rate": 3.9372e-06,
+      "loss": 0.7804,
+      "step": 30330
+    },
+    {
+      "epoch": 3.034,
+      "grad_norm": 58.506553649902344,
+      "learning_rate": 3.9352e-06,
+      "loss": 0.6049,
+      "step": 30340
+    },
+    {
+      "epoch": 3.035,
+      "grad_norm": 55.03916549682617,
+      "learning_rate": 3.933200000000001e-06,
+      "loss": 0.6667,
+      "step": 30350
+    },
+    {
+      "epoch": 3.036,
+      "grad_norm": 20.70362091064453,
+      "learning_rate": 3.931200000000001e-06,
+      "loss": 0.4529,
+      "step": 30360
+    },
+    {
+      "epoch": 3.037,
+      "grad_norm": 4.082492351531982,
+      "learning_rate": 3.9292000000000004e-06,
+      "loss": 0.6196,
+      "step": 30370
+    },
+    {
+      "epoch": 3.038,
+      "grad_norm": 40.38673400878906,
+      "learning_rate": 3.9272e-06,
+      "loss": 0.619,
+      "step": 30380
+    },
+    {
+      "epoch": 3.039,
+      "grad_norm": 35.989933013916016,
+      "learning_rate": 3.9252e-06,
+      "loss": 0.6164,
+      "step": 30390
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 12.340978622436523,
+      "learning_rate": 3.9232e-06,
+      "loss": 0.5541,
+      "step": 30400
+    },
+    {
+      "epoch": 3.041,
+      "grad_norm": 67.83113098144531,
+      "learning_rate": 3.921200000000001e-06,
+      "loss": 0.7538,
+      "step": 30410
+    },
+    {
+      "epoch": 3.042,
+      "grad_norm": 40.05186462402344,
+      "learning_rate": 3.9192000000000005e-06,
+      "loss": 0.4375,
+      "step": 30420
+    },
+    {
+      "epoch": 3.043,
+      "grad_norm": 27.47614097595215,
+      "learning_rate": 3.9172e-06,
+      "loss": 0.5773,
+      "step": 30430
+    },
+    {
+      "epoch": 3.044,
+      "grad_norm": 26.46270179748535,
+      "learning_rate": 3.9152e-06,
+      "loss": 0.6322,
+      "step": 30440
+    },
+    {
+      "epoch": 3.045,
+      "grad_norm": 8.434654235839844,
+      "learning_rate": 3.9132e-06,
+      "loss": 0.7884,
+      "step": 30450
+    },
+    {
+      "epoch": 3.046,
+      "grad_norm": 43.2637825012207,
+      "learning_rate": 3.911200000000001e-06,
+      "loss": 0.4878,
+      "step": 30460
+    },
+    {
+      "epoch": 3.047,
+      "grad_norm": 24.111162185668945,
+      "learning_rate": 3.909200000000001e-06,
+      "loss": 0.5232,
+      "step": 30470
+    },
+    {
+      "epoch": 3.048,
+      "grad_norm": 59.05705261230469,
+      "learning_rate": 3.9072000000000005e-06,
+      "loss": 0.7774,
+      "step": 30480
+    },
+    {
+      "epoch": 3.049,
+      "grad_norm": 54.003746032714844,
+      "learning_rate": 3.9052e-06,
+      "loss": 0.801,
+      "step": 30490
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 25.38113021850586,
+      "learning_rate": 3.9032e-06,
+      "loss": 0.8444,
+      "step": 30500
+    },
+    {
+      "epoch": 3.051,
+      "grad_norm": 2.1326496601104736,
+      "learning_rate": 3.901200000000001e-06,
+      "loss": 0.551,
+      "step": 30510
+    },
+    {
+      "epoch": 3.052,
+      "grad_norm": 52.77695083618164,
+      "learning_rate": 3.8992e-06,
+      "loss": 0.617,
+      "step": 30520
+    },
+    {
+      "epoch": 3.053,
+      "grad_norm": 55.81608963012695,
+      "learning_rate": 3.8972000000000006e-06,
+      "loss": 0.4779,
+      "step": 30530
+    },
+    {
+      "epoch": 3.054,
+      "grad_norm": 33.85139083862305,
+      "learning_rate": 3.8952e-06,
+      "loss": 0.6629,
+      "step": 30540
+    },
+    {
+      "epoch": 3.055,
+      "grad_norm": 40.588417053222656,
+      "learning_rate": 3.8932e-06,
+      "loss": 0.6148,
+      "step": 30550
+    },
+    {
+      "epoch": 3.056,
+      "grad_norm": 32.17471694946289,
+      "learning_rate": 3.8912e-06,
+      "loss": 0.772,
+      "step": 30560
+    },
+    {
+      "epoch": 3.057,
+      "grad_norm": 26.596405029296875,
+      "learning_rate": 3.8892e-06,
+      "loss": 0.6355,
+      "step": 30570
+    },
+    {
+      "epoch": 3.058,
+      "grad_norm": 28.121776580810547,
+      "learning_rate": 3.887200000000001e-06,
+      "loss": 0.7167,
+      "step": 30580
+    },
+    {
+      "epoch": 3.059,
+      "grad_norm": 28.322776794433594,
+      "learning_rate": 3.8852e-06,
+      "loss": 0.6618,
+      "step": 30590
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 22.134326934814453,
+      "learning_rate": 3.8832e-06,
+      "loss": 0.6055,
+      "step": 30600
+    },
+    {
+      "epoch": 3.061,
+      "grad_norm": 63.12453842163086,
+      "learning_rate": 3.8812e-06,
+      "loss": 0.6147,
+      "step": 30610
+    },
+    {
+      "epoch": 3.062,
+      "grad_norm": 31.899450302124023,
+      "learning_rate": 3.8792e-06,
+      "loss": 0.4047,
+      "step": 30620
+    },
+    {
+      "epoch": 3.063,
+      "grad_norm": 46.35337448120117,
+      "learning_rate": 3.877200000000001e-06,
+      "loss": 0.7266,
+      "step": 30630
+    },
+    {
+      "epoch": 3.064,
+      "grad_norm": 46.468223571777344,
+      "learning_rate": 3.8752e-06,
+      "loss": 0.5428,
+      "step": 30640
+    },
+    {
+      "epoch": 3.065,
+      "grad_norm": 6.490557670593262,
+      "learning_rate": 3.8732000000000005e-06,
+      "loss": 0.3811,
+      "step": 30650
+    },
+    {
+      "epoch": 3.066,
+      "grad_norm": 44.99673843383789,
+      "learning_rate": 3.8712e-06,
+      "loss": 0.5279,
+      "step": 30660
+    },
+    {
+      "epoch": 3.067,
+      "grad_norm": 34.744102478027344,
+      "learning_rate": 3.8692e-06,
+      "loss": 0.5125,
+      "step": 30670
+    },
+    {
+      "epoch": 3.068,
+      "grad_norm": 27.284143447875977,
+      "learning_rate": 3.8672e-06,
+      "loss": 0.4262,
+      "step": 30680
+    },
+    {
+      "epoch": 3.069,
+      "grad_norm": 70.38201904296875,
+      "learning_rate": 3.8652e-06,
+      "loss": 0.8814,
+      "step": 30690
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 48.69472122192383,
+      "learning_rate": 3.8632000000000006e-06,
+      "loss": 0.6056,
+      "step": 30700
+    },
+    {
+      "epoch": 3.071,
+      "grad_norm": 42.86852264404297,
+      "learning_rate": 3.8612e-06,
+      "loss": 0.7168,
+      "step": 30710
+    },
+    {
+      "epoch": 3.072,
+      "grad_norm": 5.89202880859375,
+      "learning_rate": 3.8592e-06,
+      "loss": 0.5317,
+      "step": 30720
+    },
+    {
+      "epoch": 3.073,
+      "grad_norm": 62.24650573730469,
+      "learning_rate": 3.8572e-06,
+      "loss": 0.8052,
+      "step": 30730
+    },
+    {
+      "epoch": 3.074,
+      "grad_norm": 109.9715347290039,
+      "learning_rate": 3.8552e-06,
+      "loss": 0.7197,
+      "step": 30740
+    },
+    {
+      "epoch": 3.075,
+      "grad_norm": 38.530025482177734,
+      "learning_rate": 3.853200000000001e-06,
+      "loss": 0.7395,
+      "step": 30750
+    },
+    {
+      "epoch": 3.076,
+      "grad_norm": 42.872074127197266,
+      "learning_rate": 3.8512000000000005e-06,
+      "loss": 0.7683,
+      "step": 30760
+    },
+    {
+      "epoch": 3.077,
+      "grad_norm": 23.42620086669922,
+      "learning_rate": 3.8492e-06,
+      "loss": 0.5626,
+      "step": 30770
+    },
+    {
+      "epoch": 3.078,
+      "grad_norm": 63.241119384765625,
+      "learning_rate": 3.8472e-06,
+      "loss": 0.5732,
+      "step": 30780
+    },
+    {
+      "epoch": 3.079,
+      "grad_norm": 55.224063873291016,
+      "learning_rate": 3.8452e-06,
+      "loss": 0.6857,
+      "step": 30790
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 33.85127639770508,
+      "learning_rate": 3.8432e-06,
+      "loss": 0.5499,
+      "step": 30800
+    },
+    {
+      "epoch": 3.081,
+      "grad_norm": 39.279022216796875,
+      "learning_rate": 3.841200000000001e-06,
+      "loss": 0.9333,
+      "step": 30810
+    },
+    {
+      "epoch": 3.082,
+      "grad_norm": 12.963658332824707,
+      "learning_rate": 3.8392000000000004e-06,
+      "loss": 0.2521,
+      "step": 30820
+    },
+    {
+      "epoch": 3.083,
+      "grad_norm": 8.555622100830078,
+      "learning_rate": 3.8372e-06,
+      "loss": 0.7496,
+      "step": 30830
+    },
+    {
+      "epoch": 3.084,
+      "grad_norm": 26.538875579833984,
+      "learning_rate": 3.835200000000001e-06,
+      "loss": 0.5641,
+      "step": 30840
+    },
+    {
+      "epoch": 3.085,
+      "grad_norm": 26.513824462890625,
+      "learning_rate": 3.8332e-06,
+      "loss": 0.7684,
+      "step": 30850
+    },
+    {
+      "epoch": 3.086,
+      "grad_norm": 50.69865798950195,
+      "learning_rate": 3.831200000000001e-06,
+      "loss": 0.3673,
+      "step": 30860
+    },
+    {
+      "epoch": 3.087,
+      "grad_norm": 31.16800880432129,
+      "learning_rate": 3.8292000000000005e-06,
+      "loss": 0.3716,
+      "step": 30870
+    },
+    {
+      "epoch": 3.088,
+      "grad_norm": 7.409222602844238,
+      "learning_rate": 3.8272e-06,
+      "loss": 0.6453,
+      "step": 30880
+    },
+    {
+      "epoch": 3.089,
+      "grad_norm": 21.64834213256836,
+      "learning_rate": 3.8252e-06,
+      "loss": 0.514,
+      "step": 30890
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 21.66585350036621,
+      "learning_rate": 3.8232e-06,
+      "loss": 0.5815,
+      "step": 30900
+    },
+    {
+      "epoch": 3.091,
+      "grad_norm": 48.6446418762207,
+      "learning_rate": 3.8214e-06,
+      "loss": 0.646,
+      "step": 30910
+    },
+    {
+      "epoch": 3.092,
+      "grad_norm": 47.077335357666016,
+      "learning_rate": 3.8194e-06,
+      "loss": 0.4413,
+      "step": 30920
+    },
+    {
+      "epoch": 3.093,
+      "grad_norm": 43.14842224121094,
+      "learning_rate": 3.817400000000001e-06,
+      "loss": 0.52,
+      "step": 30930
+    },
+    {
+      "epoch": 3.094,
+      "grad_norm": 42.64856719970703,
+      "learning_rate": 3.8154e-06,
+      "loss": 0.4608,
+      "step": 30940
+    },
+    {
+      "epoch": 3.095,
+      "grad_norm": 26.505043029785156,
+      "learning_rate": 3.8134000000000006e-06,
+      "loss": 0.8754,
+      "step": 30950
+    },
+    {
+      "epoch": 3.096,
+      "grad_norm": 25.05315589904785,
+      "learning_rate": 3.8114e-06,
+      "loss": 0.5399,
+      "step": 30960
+    },
+    {
+      "epoch": 3.097,
+      "grad_norm": 3.38507080078125,
+      "learning_rate": 3.8094000000000003e-06,
+      "loss": 0.5687,
+      "step": 30970
+    },
+    {
+      "epoch": 3.098,
+      "grad_norm": 25.73077392578125,
+      "learning_rate": 3.8074000000000006e-06,
+      "loss": 0.667,
+      "step": 30980
+    },
+    {
+      "epoch": 3.099,
+      "grad_norm": 51.63774871826172,
+      "learning_rate": 3.8054e-06,
+      "loss": 0.5895,
+      "step": 30990
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 36.685298919677734,
+      "learning_rate": 3.8034000000000003e-06,
+      "loss": 0.8263,
+      "step": 31000
+    },
+    {
+      "epoch": 3.101,
+      "grad_norm": 20.661434173583984,
+      "learning_rate": 3.8014e-06,
+      "loss": 0.5372,
+      "step": 31010
+    },
+    {
+      "epoch": 3.102,
+      "grad_norm": 1.888520359992981,
+      "learning_rate": 3.7994000000000004e-06,
+      "loss": 0.4881,
+      "step": 31020
+    },
+    {
+      "epoch": 3.103,
+      "grad_norm": 61.00127029418945,
+      "learning_rate": 3.7974000000000007e-06,
+      "loss": 0.4988,
+      "step": 31030
+    },
+    {
+      "epoch": 3.104,
+      "grad_norm": 52.48198699951172,
+      "learning_rate": 3.7954e-06,
+      "loss": 0.6888,
+      "step": 31040
+    },
+    {
+      "epoch": 3.105,
+      "grad_norm": 20.20423698425293,
+      "learning_rate": 3.7934000000000004e-06,
+      "loss": 0.6551,
+      "step": 31050
+    },
+    {
+      "epoch": 3.106,
+      "grad_norm": 8.83470344543457,
+      "learning_rate": 3.7914000000000002e-06,
+      "loss": 0.8729,
+      "step": 31060
+    },
+    {
+      "epoch": 3.107,
+      "grad_norm": 17.743305206298828,
+      "learning_rate": 3.7894e-06,
+      "loss": 0.3913,
+      "step": 31070
+    },
+    {
+      "epoch": 3.108,
+      "grad_norm": 33.55481719970703,
+      "learning_rate": 3.7874000000000004e-06,
+      "loss": 0.6961,
+      "step": 31080
+    },
+    {
+      "epoch": 3.109,
+      "grad_norm": 45.66580581665039,
+      "learning_rate": 3.7854000000000002e-06,
+      "loss": 0.6243,
+      "step": 31090
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 25.366012573242188,
+      "learning_rate": 3.7834000000000005e-06,
+      "loss": 0.5794,
+      "step": 31100
+    },
+    {
+      "epoch": 3.111,
+      "grad_norm": 40.58540725708008,
+      "learning_rate": 3.7814e-06,
+      "loss": 0.6811,
+      "step": 31110
+    },
+    {
+      "epoch": 3.112,
+      "grad_norm": 19.951820373535156,
+      "learning_rate": 3.7794e-06,
+      "loss": 0.4759,
+      "step": 31120
+    },
+    {
+      "epoch": 3.113,
+      "grad_norm": 35.00603103637695,
+      "learning_rate": 3.7774000000000005e-06,
+      "loss": 0.5049,
+      "step": 31130
+    },
+    {
+      "epoch": 3.114,
+      "grad_norm": 32.95356369018555,
+      "learning_rate": 3.7754000000000003e-06,
+      "loss": 0.4614,
+      "step": 31140
+    },
+    {
+      "epoch": 3.115,
+      "grad_norm": 25.622833251953125,
+      "learning_rate": 3.7734000000000006e-06,
+      "loss": 0.7006,
+      "step": 31150
+    },
+    {
+      "epoch": 3.116,
+      "grad_norm": 55.83294677734375,
+      "learning_rate": 3.7714e-06,
+      "loss": 0.6517,
+      "step": 31160
+    },
+    {
+      "epoch": 3.117,
+      "grad_norm": 18.034406661987305,
+      "learning_rate": 3.7694000000000003e-06,
+      "loss": 0.7242,
+      "step": 31170
+    },
+    {
+      "epoch": 3.118,
+      "grad_norm": 14.074854850769043,
+      "learning_rate": 3.7674000000000006e-06,
+      "loss": 0.5512,
+      "step": 31180
+    },
+    {
+      "epoch": 3.1189999999999998,
+      "grad_norm": 30.482887268066406,
+      "learning_rate": 3.7654e-06,
+      "loss": 0.5969,
+      "step": 31190
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 52.35322952270508,
+      "learning_rate": 3.7634000000000003e-06,
+      "loss": 0.6838,
+      "step": 31200
+    },
+    {
+      "epoch": 3.121,
+      "grad_norm": 67.32181549072266,
+      "learning_rate": 3.7614e-06,
+      "loss": 0.5823,
+      "step": 31210
+    },
+    {
+      "epoch": 3.122,
+      "grad_norm": 24.50666046142578,
+      "learning_rate": 3.7594000000000004e-06,
+      "loss": 0.4469,
+      "step": 31220
+    },
+    {
+      "epoch": 3.123,
+      "grad_norm": 35.51641845703125,
+      "learning_rate": 3.7574000000000007e-06,
+      "loss": 0.7058,
+      "step": 31230
+    },
+    {
+      "epoch": 3.124,
+      "grad_norm": 23.89848518371582,
+      "learning_rate": 3.7554e-06,
+      "loss": 0.7185,
+      "step": 31240
+    },
+    {
+      "epoch": 3.125,
+      "grad_norm": 32.228782653808594,
+      "learning_rate": 3.7534000000000004e-06,
+      "loss": 0.4426,
+      "step": 31250
+    },
+    {
+      "epoch": 3.126,
+      "grad_norm": 16.477025985717773,
+      "learning_rate": 3.7514e-06,
+      "loss": 0.5915,
+      "step": 31260
+    },
+    {
+      "epoch": 3.127,
+      "grad_norm": 4.443339824676514,
+      "learning_rate": 3.7494000000000005e-06,
+      "loss": 0.4272,
+      "step": 31270
+    },
+    {
+      "epoch": 3.128,
+      "grad_norm": 32.48780822753906,
+      "learning_rate": 3.7474000000000003e-06,
+      "loss": 0.5259,
+      "step": 31280
+    },
+    {
+      "epoch": 3.129,
+      "grad_norm": 7.203727722167969,
+      "learning_rate": 3.7454e-06,
+      "loss": 0.624,
+      "step": 31290
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 71.5589828491211,
+      "learning_rate": 3.7434000000000004e-06,
+      "loss": 0.9032,
+      "step": 31300
+    },
+    {
+      "epoch": 3.1310000000000002,
+      "grad_norm": 40.37437057495117,
+      "learning_rate": 3.7414e-06,
+      "loss": 0.607,
+      "step": 31310
+    },
+    {
+      "epoch": 3.132,
+      "grad_norm": 8.436080932617188,
+      "learning_rate": 3.7394e-06,
+      "loss": 0.5178,
+      "step": 31320
+    },
+    {
+      "epoch": 3.133,
+      "grad_norm": 11.011701583862305,
+      "learning_rate": 3.7374000000000004e-06,
+      "loss": 0.5794,
+      "step": 31330
+    },
+    {
+      "epoch": 3.134,
+      "grad_norm": 27.312786102294922,
+      "learning_rate": 3.7354000000000003e-06,
+      "loss": 0.5639,
+      "step": 31340
+    },
+    {
+      "epoch": 3.135,
+      "grad_norm": 34.613746643066406,
+      "learning_rate": 3.7334000000000005e-06,
+      "loss": 0.5785,
+      "step": 31350
+    },
+    {
+      "epoch": 3.136,
+      "grad_norm": 29.656906127929688,
+      "learning_rate": 3.7314000000000004e-06,
+      "loss": 0.6471,
+      "step": 31360
+    },
+    {
+      "epoch": 3.137,
+      "grad_norm": 39.9609489440918,
+      "learning_rate": 3.7294000000000002e-06,
+      "loss": 0.9502,
+      "step": 31370
+    },
+    {
+      "epoch": 3.138,
+      "grad_norm": 28.447853088378906,
+      "learning_rate": 3.7274000000000005e-06,
+      "loss": 0.3478,
+      "step": 31380
+    },
+    {
+      "epoch": 3.1390000000000002,
+      "grad_norm": 8.006779670715332,
+      "learning_rate": 3.7254e-06,
+      "loss": 0.363,
+      "step": 31390
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 53.307796478271484,
+      "learning_rate": 3.7234000000000002e-06,
+      "loss": 0.7523,
+      "step": 31400
+    },
+    {
+      "epoch": 3.141,
+      "grad_norm": 40.8818359375,
+      "learning_rate": 3.7214000000000005e-06,
+      "loss": 0.8793,
+      "step": 31410
+    },
+    {
+      "epoch": 3.142,
+      "grad_norm": 23.048547744750977,
+      "learning_rate": 3.7194000000000003e-06,
+      "loss": 0.4126,
+      "step": 31420
+    },
+    {
+      "epoch": 3.143,
+      "grad_norm": 55.69511795043945,
+      "learning_rate": 3.7174000000000006e-06,
+      "loss": 0.7602,
+      "step": 31430
+    },
+    {
+      "epoch": 3.144,
+      "grad_norm": 49.32610321044922,
+      "learning_rate": 3.7154e-06,
+      "loss": 0.6945,
+      "step": 31440
+    },
+    {
+      "epoch": 3.145,
+      "grad_norm": 78.19044494628906,
+      "learning_rate": 3.7134000000000003e-06,
+      "loss": 0.8383,
+      "step": 31450
+    },
+    {
+      "epoch": 3.146,
+      "grad_norm": 14.405440330505371,
+      "learning_rate": 3.7114000000000006e-06,
+      "loss": 0.5729,
+      "step": 31460
+    },
+    {
+      "epoch": 3.147,
+      "grad_norm": 46.739463806152344,
+      "learning_rate": 3.7094000000000004e-06,
+      "loss": 0.5877,
+      "step": 31470
+    },
+    {
+      "epoch": 3.148,
+      "grad_norm": 14.353507995605469,
+      "learning_rate": 3.7074000000000003e-06,
+      "loss": 0.4201,
+      "step": 31480
+    },
+    {
+      "epoch": 3.149,
+      "grad_norm": 55.31502151489258,
+      "learning_rate": 3.7054e-06,
+      "loss": 0.5387,
+      "step": 31490
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 40.76209259033203,
+      "learning_rate": 3.7034000000000004e-06,
+      "loss": 0.2956,
+      "step": 31500
+    },
+    {
+      "epoch": 3.151,
+      "grad_norm": 35.83966827392578,
+      "learning_rate": 3.7014000000000007e-06,
+      "loss": 0.8846,
+      "step": 31510
+    },
+    {
+      "epoch": 3.152,
+      "grad_norm": 42.301090240478516,
+      "learning_rate": 3.6994e-06,
+      "loss": 0.6257,
+      "step": 31520
+    },
+    {
+      "epoch": 3.153,
+      "grad_norm": 31.590164184570312,
+      "learning_rate": 3.6974000000000004e-06,
+      "loss": 0.2422,
+      "step": 31530
+    },
+    {
+      "epoch": 3.154,
+      "grad_norm": 12.676555633544922,
+      "learning_rate": 3.6954000000000002e-06,
+      "loss": 0.7386,
+      "step": 31540
+    },
+    {
+      "epoch": 3.155,
+      "grad_norm": 54.29704284667969,
+      "learning_rate": 3.6934000000000005e-06,
+      "loss": 0.4906,
+      "step": 31550
+    },
+    {
+      "epoch": 3.156,
+      "grad_norm": 32.339134216308594,
+      "learning_rate": 3.6914000000000004e-06,
+      "loss": 0.6436,
+      "step": 31560
+    },
+    {
+      "epoch": 3.157,
+      "grad_norm": 53.582733154296875,
+      "learning_rate": 3.6894e-06,
+      "loss": 0.5199,
+      "step": 31570
+    },
+    {
+      "epoch": 3.158,
+      "grad_norm": 20.8098087310791,
+      "learning_rate": 3.6874000000000005e-06,
+      "loss": 0.5713,
+      "step": 31580
+    },
+    {
+      "epoch": 3.159,
+      "grad_norm": 12.076930046081543,
+      "learning_rate": 3.6854e-06,
+      "loss": 0.5682,
+      "step": 31590
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 39.23326110839844,
+      "learning_rate": 3.6834e-06,
+      "loss": 0.7824,
+      "step": 31600
+    },
+    {
+      "epoch": 3.161,
+      "grad_norm": 1.6527082920074463,
+      "learning_rate": 3.6814000000000004e-06,
+      "loss": 0.8088,
+      "step": 31610
+    },
+    {
+      "epoch": 3.162,
+      "grad_norm": 46.88147735595703,
+      "learning_rate": 3.6794000000000003e-06,
+      "loss": 0.7926,
+      "step": 31620
+    },
+    {
+      "epoch": 3.163,
+      "grad_norm": 43.62668991088867,
+      "learning_rate": 3.6774000000000006e-06,
+      "loss": 0.603,
+      "step": 31630
+    },
+    {
+      "epoch": 3.164,
+      "grad_norm": 39.14570236206055,
+      "learning_rate": 3.6754e-06,
+      "loss": 0.4348,
+      "step": 31640
+    },
+    {
+      "epoch": 3.165,
+      "grad_norm": 33.04401779174805,
+      "learning_rate": 3.6734000000000003e-06,
+      "loss": 0.4615,
+      "step": 31650
+    },
+    {
+      "epoch": 3.166,
+      "grad_norm": 30.718061447143555,
+      "learning_rate": 3.6714000000000005e-06,
+      "loss": 0.6487,
+      "step": 31660
+    },
+    {
+      "epoch": 3.167,
+      "grad_norm": 40.128971099853516,
+      "learning_rate": 3.6694000000000004e-06,
+      "loss": 0.6791,
+      "step": 31670
+    },
+    {
+      "epoch": 3.168,
+      "grad_norm": 34.56775665283203,
+      "learning_rate": 3.6674000000000002e-06,
+      "loss": 0.4019,
+      "step": 31680
+    },
+    {
+      "epoch": 3.169,
+      "grad_norm": 47.323760986328125,
+      "learning_rate": 3.6654e-06,
+      "loss": 0.79,
+      "step": 31690
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 42.98324203491211,
+      "learning_rate": 3.6634000000000004e-06,
+      "loss": 0.7067,
+      "step": 31700
+    },
+    {
+      "epoch": 3.171,
+      "grad_norm": 16.420366287231445,
+      "learning_rate": 3.6614000000000006e-06,
+      "loss": 0.7854,
+      "step": 31710
+    },
+    {
+      "epoch": 3.172,
+      "grad_norm": 45.52265930175781,
+      "learning_rate": 3.6594e-06,
+      "loss": 0.6981,
+      "step": 31720
+    },
+    {
+      "epoch": 3.173,
+      "grad_norm": 30.123014450073242,
+      "learning_rate": 3.6574000000000003e-06,
+      "loss": 0.6198,
+      "step": 31730
+    },
+    {
+      "epoch": 3.174,
+      "grad_norm": 24.232975006103516,
+      "learning_rate": 3.6554e-06,
+      "loss": 0.6919,
+      "step": 31740
+    },
+    {
+      "epoch": 3.175,
+      "grad_norm": 20.821025848388672,
+      "learning_rate": 3.6534000000000005e-06,
+      "loss": 0.6695,
+      "step": 31750
+    },
+    {
+      "epoch": 3.176,
+      "grad_norm": 51.86601638793945,
+      "learning_rate": 3.6514000000000003e-06,
+      "loss": 0.6286,
+      "step": 31760
+    },
+    {
+      "epoch": 3.177,
+      "grad_norm": 11.700900077819824,
+      "learning_rate": 3.6494e-06,
+      "loss": 0.5536,
+      "step": 31770
+    },
+    {
+      "epoch": 3.178,
+      "grad_norm": 35.8469352722168,
+      "learning_rate": 3.6474000000000004e-06,
+      "loss": 0.4811,
+      "step": 31780
+    },
+    {
+      "epoch": 3.179,
+      "grad_norm": 31.125680923461914,
+      "learning_rate": 3.6454e-06,
+      "loss": 0.4386,
+      "step": 31790
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 40.396820068359375,
+      "learning_rate": 3.6434e-06,
+      "loss": 0.934,
+      "step": 31800
+    },
+    {
+      "epoch": 3.181,
+      "grad_norm": 2.2530946731567383,
+      "learning_rate": 3.6414000000000004e-06,
+      "loss": 0.5532,
+      "step": 31810
+    },
+    {
+      "epoch": 3.182,
+      "grad_norm": 29.33017921447754,
+      "learning_rate": 3.6394000000000003e-06,
+      "loss": 0.6204,
+      "step": 31820
+    },
+    {
+      "epoch": 3.183,
+      "grad_norm": 55.61162185668945,
+      "learning_rate": 3.6374000000000005e-06,
+      "loss": 0.4463,
+      "step": 31830
+    },
+    {
+      "epoch": 3.184,
+      "grad_norm": 4.762482166290283,
+      "learning_rate": 3.6354e-06,
+      "loss": 0.5826,
+      "step": 31840
+    },
+    {
+      "epoch": 3.185,
+      "grad_norm": 33.20708465576172,
+      "learning_rate": 3.6334000000000002e-06,
+      "loss": 0.6001,
+      "step": 31850
+    },
+    {
+      "epoch": 3.186,
+      "grad_norm": 16.60824966430664,
+      "learning_rate": 3.6314000000000005e-06,
+      "loss": 0.7306,
+      "step": 31860
+    },
+    {
+      "epoch": 3.187,
+      "grad_norm": 48.7596435546875,
+      "learning_rate": 3.6294000000000004e-06,
+      "loss": 0.5614,
+      "step": 31870
+    },
+    {
+      "epoch": 3.188,
+      "grad_norm": 11.902722358703613,
+      "learning_rate": 3.6274e-06,
+      "loss": 0.692,
+      "step": 31880
+    },
+    {
+      "epoch": 3.189,
+      "grad_norm": 47.085758209228516,
+      "learning_rate": 3.6254e-06,
+      "loss": 0.566,
+      "step": 31890
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 42.062076568603516,
+      "learning_rate": 3.6234000000000003e-06,
+      "loss": 0.5515,
+      "step": 31900
+    },
+    {
+      "epoch": 3.191,
+      "grad_norm": 3.4789304733276367,
+      "learning_rate": 3.6214000000000006e-06,
+      "loss": 0.586,
+      "step": 31910
+    },
+    {
+      "epoch": 3.192,
+      "grad_norm": 63.67318344116211,
+      "learning_rate": 3.6194e-06,
+      "loss": 0.8436,
+      "step": 31920
+    },
+    {
+      "epoch": 3.193,
+      "grad_norm": 27.702524185180664,
+      "learning_rate": 3.6174000000000003e-06,
+      "loss": 0.3121,
+      "step": 31930
+    },
+    {
+      "epoch": 3.194,
+      "grad_norm": 46.160911560058594,
+      "learning_rate": 3.6154e-06,
+      "loss": 0.5511,
+      "step": 31940
+    },
+    {
+      "epoch": 3.195,
+      "grad_norm": 24.993696212768555,
+      "learning_rate": 3.6134000000000004e-06,
+      "loss": 0.7147,
+      "step": 31950
+    },
+    {
+      "epoch": 3.196,
+      "grad_norm": 28.18203353881836,
+      "learning_rate": 3.6114000000000003e-06,
+      "loss": 0.6309,
+      "step": 31960
+    },
+    {
+      "epoch": 3.197,
+      "grad_norm": 18.851085662841797,
+      "learning_rate": 3.6094e-06,
+      "loss": 0.4695,
+      "step": 31970
+    },
+    {
+      "epoch": 3.198,
+      "grad_norm": 85.22747039794922,
+      "learning_rate": 3.6074000000000004e-06,
+      "loss": 0.7747,
+      "step": 31980
+    },
+    {
+      "epoch": 3.199,
+      "grad_norm": 48.64833068847656,
+      "learning_rate": 3.6054000000000007e-06,
+      "loss": 0.5871,
+      "step": 31990
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 51.850101470947266,
+      "learning_rate": 3.6034e-06,
+      "loss": 0.7215,
+      "step": 32000
+    },
+    {
+      "epoch": 3.201,
+      "grad_norm": 24.21227264404297,
+      "learning_rate": 3.6014000000000004e-06,
+      "loss": 0.8802,
+      "step": 32010
+    },
+    {
+      "epoch": 3.202,
+      "grad_norm": 17.991352081298828,
+      "learning_rate": 3.5994000000000002e-06,
+      "loss": 0.5883,
+      "step": 32020
+    },
+    {
+      "epoch": 3.203,
+      "grad_norm": 69.04742431640625,
+      "learning_rate": 3.5974000000000005e-06,
+      "loss": 0.577,
+      "step": 32030
+    },
+    {
+      "epoch": 3.204,
+      "grad_norm": 52.01885986328125,
+      "learning_rate": 3.5954000000000008e-06,
+      "loss": 0.6458,
+      "step": 32040
+    },
+    {
+      "epoch": 3.205,
+      "grad_norm": 28.483524322509766,
+      "learning_rate": 3.5934e-06,
+      "loss": 0.4858,
+      "step": 32050
+    },
+    {
+      "epoch": 3.206,
+      "grad_norm": 37.120445251464844,
+      "learning_rate": 3.5914000000000005e-06,
+      "loss": 0.2048,
+      "step": 32060
+    },
+    {
+      "epoch": 3.207,
+      "grad_norm": 14.233193397521973,
+      "learning_rate": 3.5894000000000003e-06,
+      "loss": 0.417,
+      "step": 32070
+    },
+    {
+      "epoch": 3.208,
+      "grad_norm": 17.76689910888672,
+      "learning_rate": 3.5874e-06,
+      "loss": 0.5522,
+      "step": 32080
+    },
+    {
+      "epoch": 3.209,
+      "grad_norm": 50.82288360595703,
+      "learning_rate": 3.5854000000000004e-06,
+      "loss": 0.3946,
+      "step": 32090
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 3.2764945030212402,
+      "learning_rate": 3.5834000000000003e-06,
+      "loss": 0.4198,
+      "step": 32100
+    },
+    {
+      "epoch": 3.211,
+      "grad_norm": 60.76835632324219,
+      "learning_rate": 3.5814000000000006e-06,
+      "loss": 0.6575,
+      "step": 32110
+    },
+    {
+      "epoch": 3.212,
+      "grad_norm": 68.07122039794922,
+      "learning_rate": 3.5794e-06,
+      "loss": 0.7174,
+      "step": 32120
+    },
+    {
+      "epoch": 3.213,
+      "grad_norm": 6.457009315490723,
+      "learning_rate": 3.5774000000000003e-06,
+      "loss": 0.7412,
+      "step": 32130
+    },
+    {
+      "epoch": 3.214,
+      "grad_norm": 32.04698181152344,
+      "learning_rate": 3.5754000000000005e-06,
+      "loss": 1.0493,
+      "step": 32140
+    },
+    {
+      "epoch": 3.215,
+      "grad_norm": 11.740108489990234,
+      "learning_rate": 3.5734000000000004e-06,
+      "loss": 0.7469,
+      "step": 32150
+    },
+    {
+      "epoch": 3.216,
+      "grad_norm": 46.86805725097656,
+      "learning_rate": 3.5714000000000002e-06,
+      "loss": 0.7554,
+      "step": 32160
+    },
+    {
+      "epoch": 3.217,
+      "grad_norm": 34.02867126464844,
+      "learning_rate": 3.5694e-06,
+      "loss": 0.6535,
+      "step": 32170
+    },
+    {
+      "epoch": 3.218,
+      "grad_norm": 29.15270233154297,
+      "learning_rate": 3.5674000000000004e-06,
+      "loss": 0.7064,
+      "step": 32180
+    },
+    {
+      "epoch": 3.219,
+      "grad_norm": 23.48900032043457,
+      "learning_rate": 3.5654000000000006e-06,
+      "loss": 0.3669,
+      "step": 32190
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 37.315250396728516,
+      "learning_rate": 3.5634e-06,
+      "loss": 0.856,
+      "step": 32200
+    },
+    {
+      "epoch": 3.221,
+      "grad_norm": 23.128664016723633,
+      "learning_rate": 3.5614000000000003e-06,
+      "loss": 0.528,
+      "step": 32210
+    },
+    {
+      "epoch": 3.222,
+      "grad_norm": 21.8521671295166,
+      "learning_rate": 3.5594e-06,
+      "loss": 0.431,
+      "step": 32220
+    },
+    {
+      "epoch": 3.223,
+      "grad_norm": 13.725652694702148,
+      "learning_rate": 3.5574000000000004e-06,
+      "loss": 0.7434,
+      "step": 32230
+    },
+    {
+      "epoch": 3.224,
+      "grad_norm": 32.80021286010742,
+      "learning_rate": 3.5554000000000007e-06,
+      "loss": 0.5275,
+      "step": 32240
+    },
+    {
+      "epoch": 3.225,
+      "grad_norm": 3.1503541469573975,
+      "learning_rate": 3.5534e-06,
+      "loss": 0.4221,
+      "step": 32250
+    },
+    {
+      "epoch": 3.226,
+      "grad_norm": 71.9884262084961,
+      "learning_rate": 3.5514000000000004e-06,
+      "loss": 0.4528,
+      "step": 32260
+    },
+    {
+      "epoch": 3.227,
+      "grad_norm": 3.422281503677368,
+      "learning_rate": 3.5494000000000003e-06,
+      "loss": 0.6253,
+      "step": 32270
+    },
+    {
+      "epoch": 3.228,
+      "grad_norm": 6.156475067138672,
+      "learning_rate": 3.5474e-06,
+      "loss": 0.5486,
+      "step": 32280
+    },
+    {
+      "epoch": 3.229,
+      "grad_norm": 26.414377212524414,
+      "learning_rate": 3.5454000000000004e-06,
+      "loss": 0.6165,
+      "step": 32290
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 46.42730712890625,
+      "learning_rate": 3.5434000000000002e-06,
+      "loss": 1.0238,
+      "step": 32300
+    },
+    {
+      "epoch": 3.231,
+      "grad_norm": 11.5466890335083,
+      "learning_rate": 3.5414000000000005e-06,
+      "loss": 0.3546,
+      "step": 32310
+    },
+    {
+      "epoch": 3.232,
+      "grad_norm": 29.019729614257812,
+      "learning_rate": 3.5394e-06,
+      "loss": 0.4857,
+      "step": 32320
+    },
+    {
+      "epoch": 3.233,
+      "grad_norm": 27.857799530029297,
+      "learning_rate": 3.5374000000000002e-06,
+      "loss": 1.0558,
+      "step": 32330
+    },
+    {
+      "epoch": 3.234,
+      "grad_norm": 15.012434005737305,
+      "learning_rate": 3.5354000000000005e-06,
+      "loss": 0.7907,
+      "step": 32340
+    },
+    {
+      "epoch": 3.235,
+      "grad_norm": 21.44322395324707,
+      "learning_rate": 3.5334000000000003e-06,
+      "loss": 0.7334,
+      "step": 32350
+    },
+    {
+      "epoch": 3.2359999999999998,
+      "grad_norm": 32.07785415649414,
+      "learning_rate": 3.5314e-06,
+      "loss": 0.6319,
+      "step": 32360
+    },
+    {
+      "epoch": 3.237,
+      "grad_norm": 34.226993560791016,
+      "learning_rate": 3.5294e-06,
+      "loss": 0.5482,
+      "step": 32370
+    },
+    {
+      "epoch": 3.238,
+      "grad_norm": 30.83748435974121,
+      "learning_rate": 3.5274000000000003e-06,
+      "loss": 0.3989,
+      "step": 32380
+    },
+    {
+      "epoch": 3.239,
+      "grad_norm": 34.788299560546875,
+      "learning_rate": 3.5254000000000006e-06,
+      "loss": 0.8568,
+      "step": 32390
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 10.510351181030273,
+      "learning_rate": 3.5234e-06,
+      "loss": 0.6174,
+      "step": 32400
+    },
+    {
+      "epoch": 3.241,
+      "grad_norm": 7.551955223083496,
+      "learning_rate": 3.5214000000000003e-06,
+      "loss": 0.4735,
+      "step": 32410
+    },
+    {
+      "epoch": 3.242,
+      "grad_norm": 53.604469299316406,
+      "learning_rate": 3.5194e-06,
+      "loss": 0.5829,
+      "step": 32420
+    },
+    {
+      "epoch": 3.243,
+      "grad_norm": 35.08150100708008,
+      "learning_rate": 3.5174000000000004e-06,
+      "loss": 0.4748,
+      "step": 32430
+    },
+    {
+      "epoch": 3.2439999999999998,
+      "grad_norm": 36.20558166503906,
+      "learning_rate": 3.5154000000000007e-06,
+      "loss": 0.7476,
+      "step": 32440
+    },
+    {
+      "epoch": 3.245,
+      "grad_norm": 29.312915802001953,
+      "learning_rate": 3.5134e-06,
+      "loss": 0.9648,
+      "step": 32450
+    },
+    {
+      "epoch": 3.246,
+      "grad_norm": 28.11040687561035,
+      "learning_rate": 3.5114000000000004e-06,
+      "loss": 0.6069,
+      "step": 32460
+    },
+    {
+      "epoch": 3.247,
+      "grad_norm": 39.97213363647461,
+      "learning_rate": 3.5094000000000002e-06,
+      "loss": 0.6488,
+      "step": 32470
+    },
+    {
+      "epoch": 3.248,
+      "grad_norm": 33.253658294677734,
+      "learning_rate": 3.5074e-06,
+      "loss": 0.5124,
+      "step": 32480
+    },
+    {
+      "epoch": 3.249,
+      "grad_norm": 35.87740707397461,
+      "learning_rate": 3.5054000000000004e-06,
+      "loss": 0.5967,
+      "step": 32490
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 29.033767700195312,
+      "learning_rate": 3.5034e-06,
+      "loss": 0.7158,
+      "step": 32500
+    },
+    {
+      "epoch": 3.251,
+      "grad_norm": 32.10422897338867,
+      "learning_rate": 3.5014000000000005e-06,
+      "loss": 0.5006,
+      "step": 32510
+    },
+    {
+      "epoch": 3.252,
+      "grad_norm": 47.43272399902344,
+      "learning_rate": 3.4994e-06,
+      "loss": 0.6308,
+      "step": 32520
+    },
+    {
+      "epoch": 3.253,
+      "grad_norm": 37.08671188354492,
+      "learning_rate": 3.4974e-06,
+      "loss": 0.498,
+      "step": 32530
+    },
+    {
+      "epoch": 3.254,
+      "grad_norm": 20.07958984375,
+      "learning_rate": 3.4954000000000004e-06,
+      "loss": 0.5946,
+      "step": 32540
+    },
+    {
+      "epoch": 3.255,
+      "grad_norm": 51.85481643676758,
+      "learning_rate": 3.4934000000000003e-06,
+      "loss": 0.6992,
+      "step": 32550
+    },
+    {
+      "epoch": 3.2560000000000002,
+      "grad_norm": 29.262605667114258,
+      "learning_rate": 3.4914e-06,
+      "loss": 0.7368,
+      "step": 32560
+    },
+    {
+      "epoch": 3.257,
+      "grad_norm": 60.6501579284668,
+      "learning_rate": 3.4894e-06,
+      "loss": 0.7476,
+      "step": 32570
+    },
+    {
+      "epoch": 3.258,
+      "grad_norm": 9.376657485961914,
+      "learning_rate": 3.4874000000000003e-06,
+      "loss": 0.5938,
+      "step": 32580
+    },
+    {
+      "epoch": 3.259,
+      "grad_norm": 20.48821449279785,
+      "learning_rate": 3.4854000000000005e-06,
+      "loss": 0.6259,
+      "step": 32590
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 28.9674015045166,
+      "learning_rate": 3.4834e-06,
+      "loss": 0.5057,
+      "step": 32600
+    },
+    {
+      "epoch": 3.261,
+      "grad_norm": 32.977699279785156,
+      "learning_rate": 3.4814000000000002e-06,
+      "loss": 0.5966,
+      "step": 32610
+    },
+    {
+      "epoch": 3.262,
+      "grad_norm": 9.944561004638672,
+      "learning_rate": 3.4794000000000005e-06,
+      "loss": 0.7086,
+      "step": 32620
+    },
+    {
+      "epoch": 3.263,
+      "grad_norm": 26.56491470336914,
+      "learning_rate": 3.4774000000000004e-06,
+      "loss": 0.6401,
+      "step": 32630
+    },
+    {
+      "epoch": 3.2640000000000002,
+      "grad_norm": 36.066864013671875,
+      "learning_rate": 3.4754000000000006e-06,
+      "loss": 0.6594,
+      "step": 32640
+    },
+    {
+      "epoch": 3.265,
+      "grad_norm": 37.5883903503418,
+      "learning_rate": 3.4734e-06,
+      "loss": 0.5561,
+      "step": 32650
+    },
+    {
+      "epoch": 3.266,
+      "grad_norm": 4.712746620178223,
+      "learning_rate": 3.4714000000000003e-06,
+      "loss": 0.4307,
+      "step": 32660
+    },
+    {
+      "epoch": 3.267,
+      "grad_norm": 28.391803741455078,
+      "learning_rate": 3.4694000000000006e-06,
+      "loss": 0.834,
+      "step": 32670
+    },
+    {
+      "epoch": 3.268,
+      "grad_norm": 33.248355865478516,
+      "learning_rate": 3.4674e-06,
+      "loss": 0.4551,
+      "step": 32680
+    },
+    {
+      "epoch": 3.269,
+      "grad_norm": 25.97675895690918,
+      "learning_rate": 3.4654000000000003e-06,
+      "loss": 0.5444,
+      "step": 32690
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 19.075571060180664,
+      "learning_rate": 3.4634e-06,
+      "loss": 0.553,
+      "step": 32700
+    },
+    {
+      "epoch": 3.271,
+      "grad_norm": 22.07334327697754,
+      "learning_rate": 3.4614000000000004e-06,
+      "loss": 0.5503,
+      "step": 32710
+    },
+    {
+      "epoch": 3.2720000000000002,
+      "grad_norm": 63.43888854980469,
+      "learning_rate": 3.4594000000000007e-06,
+      "loss": 0.4606,
+      "step": 32720
+    },
+    {
+      "epoch": 3.273,
+      "grad_norm": 40.647926330566406,
+      "learning_rate": 3.4574e-06,
+      "loss": 0.6632,
+      "step": 32730
+    },
+    {
+      "epoch": 3.274,
+      "grad_norm": 29.407672882080078,
+      "learning_rate": 3.4554000000000004e-06,
+      "loss": 0.7169,
+      "step": 32740
+    },
+    {
+      "epoch": 3.275,
+      "grad_norm": 25.562196731567383,
+      "learning_rate": 3.4534000000000003e-06,
+      "loss": 0.5418,
+      "step": 32750
+    },
+    {
+      "epoch": 3.276,
+      "grad_norm": 35.23017883300781,
+      "learning_rate": 3.4514e-06,
+      "loss": 0.5916,
+      "step": 32760
+    },
+    {
+      "epoch": 3.277,
+      "grad_norm": 5.2610392570495605,
+      "learning_rate": 3.4494000000000004e-06,
+      "loss": 0.6641,
+      "step": 32770
+    },
+    {
+      "epoch": 3.278,
+      "grad_norm": 10.967763900756836,
+      "learning_rate": 3.4474000000000002e-06,
+      "loss": 0.4844,
+      "step": 32780
+    },
+    {
+      "epoch": 3.279,
+      "grad_norm": 26.412208557128906,
+      "learning_rate": 3.4454000000000005e-06,
+      "loss": 0.788,
+      "step": 32790
+    },
+    {
+      "epoch": 3.2800000000000002,
+      "grad_norm": 69.90411376953125,
+      "learning_rate": 3.4434e-06,
+      "loss": 1.0829,
+      "step": 32800
+    },
+    {
+      "epoch": 3.281,
+      "grad_norm": 39.49662780761719,
+      "learning_rate": 3.4414e-06,
+      "loss": 0.7883,
+      "step": 32810
+    },
+    {
+      "epoch": 3.282,
+      "grad_norm": 21.244108200073242,
+      "learning_rate": 3.4394000000000005e-06,
+      "loss": 0.6491,
+      "step": 32820
+    },
+    {
+      "epoch": 3.283,
+      "grad_norm": 21.366939544677734,
+      "learning_rate": 3.4374000000000003e-06,
+      "loss": 0.5096,
+      "step": 32830
+    },
+    {
+      "epoch": 3.284,
+      "grad_norm": 3.5856432914733887,
+      "learning_rate": 3.4354000000000006e-06,
+      "loss": 0.551,
+      "step": 32840
+    },
+    {
+      "epoch": 3.285,
+      "grad_norm": 22.398399353027344,
+      "learning_rate": 3.4334e-06,
+      "loss": 0.7356,
+      "step": 32850
+    },
+    {
+      "epoch": 3.286,
+      "grad_norm": 47.277099609375,
+      "learning_rate": 3.4314000000000003e-06,
+      "loss": 0.6598,
+      "step": 32860
+    },
+    {
+      "epoch": 3.287,
+      "grad_norm": 45.505672454833984,
+      "learning_rate": 3.4294000000000006e-06,
+      "loss": 0.5803,
+      "step": 32870
+    },
+    {
+      "epoch": 3.288,
+      "grad_norm": 31.135108947753906,
+      "learning_rate": 3.4274e-06,
+      "loss": 0.7393,
+      "step": 32880
+    },
+    {
+      "epoch": 3.289,
+      "grad_norm": 32.842384338378906,
+      "learning_rate": 3.4254000000000003e-06,
+      "loss": 0.5627,
+      "step": 32890
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 42.18638610839844,
+      "learning_rate": 3.4234e-06,
+      "loss": 0.5442,
+      "step": 32900
+    },
+    {
+      "epoch": 3.291,
+      "grad_norm": 41.7022705078125,
+      "learning_rate": 3.4214000000000004e-06,
+      "loss": 0.6849,
+      "step": 32910
+    },
+    {
+      "epoch": 3.292,
+      "grad_norm": 13.518721580505371,
+      "learning_rate": 3.4194000000000007e-06,
+      "loss": 0.3897,
+      "step": 32920
+    },
+    {
+      "epoch": 3.293,
+      "grad_norm": 48.264923095703125,
+      "learning_rate": 3.4174e-06,
+      "loss": 0.9303,
+      "step": 32930
+    },
+    {
+      "epoch": 3.294,
+      "grad_norm": 7.329990386962891,
+      "learning_rate": 3.4154000000000004e-06,
+      "loss": 0.4068,
+      "step": 32940
+    },
+    {
+      "epoch": 3.295,
+      "grad_norm": 24.180370330810547,
+      "learning_rate": 3.4134000000000002e-06,
+      "loss": 0.5774,
+      "step": 32950
+    },
+    {
+      "epoch": 3.296,
+      "grad_norm": 17.27824592590332,
+      "learning_rate": 3.4114e-06,
+      "loss": 0.5578,
+      "step": 32960
+    },
+    {
+      "epoch": 3.297,
+      "grad_norm": 40.2707633972168,
+      "learning_rate": 3.4094000000000003e-06,
+      "loss": 0.5842,
+      "step": 32970
+    },
+    {
+      "epoch": 3.298,
+      "grad_norm": 28.49213409423828,
+      "learning_rate": 3.4074e-06,
+      "loss": 0.6671,
+      "step": 32980
+    },
+    {
+      "epoch": 3.299,
+      "grad_norm": 18.07645606994629,
+      "learning_rate": 3.4054000000000005e-06,
+      "loss": 0.4864,
+      "step": 32990
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 34.3114128112793,
+      "learning_rate": 3.4034e-06,
+      "loss": 0.6038,
+      "step": 33000
+    },
+    {
+      "epoch": 3.301,
+      "grad_norm": 5.593450546264648,
+      "learning_rate": 3.4014e-06,
+      "loss": 0.6182,
+      "step": 33010
+    },
+    {
+      "epoch": 3.302,
+      "grad_norm": 40.547847747802734,
+      "learning_rate": 3.3994000000000004e-06,
+      "loss": 0.611,
+      "step": 33020
+    },
+    {
+      "epoch": 3.303,
+      "grad_norm": 45.57594299316406,
+      "learning_rate": 3.3974000000000003e-06,
+      "loss": 0.4329,
+      "step": 33030
+    },
+    {
+      "epoch": 3.304,
+      "grad_norm": 37.91822814941406,
+      "learning_rate": 3.3954000000000006e-06,
+      "loss": 0.4367,
+      "step": 33040
+    },
+    {
+      "epoch": 3.305,
+      "grad_norm": 25.43527603149414,
+      "learning_rate": 3.3934e-06,
+      "loss": 0.8501,
+      "step": 33050
+    },
+    {
+      "epoch": 3.306,
+      "grad_norm": 19.600317001342773,
+      "learning_rate": 3.3914000000000003e-06,
+      "loss": 0.554,
+      "step": 33060
+    },
+    {
+      "epoch": 3.307,
+      "grad_norm": 47.621376037597656,
+      "learning_rate": 3.3894000000000005e-06,
+      "loss": 0.5234,
+      "step": 33070
+    },
+    {
+      "epoch": 3.308,
+      "grad_norm": 2.732773542404175,
+      "learning_rate": 3.3874e-06,
+      "loss": 0.6959,
+      "step": 33080
+    },
+    {
+      "epoch": 3.309,
+      "grad_norm": 27.551868438720703,
+      "learning_rate": 3.3854000000000002e-06,
+      "loss": 0.7955,
+      "step": 33090
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 7.095355987548828,
+      "learning_rate": 3.3834e-06,
+      "loss": 0.4998,
+      "step": 33100
+    },
+    {
+      "epoch": 3.311,
+      "grad_norm": 26.43609619140625,
+      "learning_rate": 3.3814000000000004e-06,
+      "loss": 0.4003,
+      "step": 33110
+    },
+    {
+      "epoch": 3.312,
+      "grad_norm": 26.44409942626953,
+      "learning_rate": 3.3794000000000006e-06,
+      "loss": 0.6276,
+      "step": 33120
+    },
+    {
+      "epoch": 3.313,
+      "grad_norm": 20.006427764892578,
+      "learning_rate": 3.3774e-06,
+      "loss": 0.713,
+      "step": 33130
+    },
+    {
+      "epoch": 3.314,
+      "grad_norm": 28.21528434753418,
+      "learning_rate": 3.3754000000000003e-06,
+      "loss": 0.3301,
+      "step": 33140
+    },
+    {
+      "epoch": 3.315,
+      "grad_norm": 67.65629577636719,
+      "learning_rate": 3.3734e-06,
+      "loss": 0.5409,
+      "step": 33150
+    },
+    {
+      "epoch": 3.316,
+      "grad_norm": 27.215896606445312,
+      "learning_rate": 3.3714e-06,
+      "loss": 0.789,
+      "step": 33160
+    },
+    {
+      "epoch": 3.317,
+      "grad_norm": 33.35668182373047,
+      "learning_rate": 3.3694000000000003e-06,
+      "loss": 0.7049,
+      "step": 33170
+    },
+    {
+      "epoch": 3.318,
+      "grad_norm": 27.733373641967773,
+      "learning_rate": 3.3674e-06,
+      "loss": 0.5918,
+      "step": 33180
+    },
+    {
+      "epoch": 3.319,
+      "grad_norm": 21.185148239135742,
+      "learning_rate": 3.3654000000000004e-06,
+      "loss": 0.6805,
+      "step": 33190
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 24.466445922851562,
+      "learning_rate": 3.3634e-06,
+      "loss": 0.5639,
+      "step": 33200
+    },
+    {
+      "epoch": 3.321,
+      "grad_norm": 36.54484558105469,
+      "learning_rate": 3.3614e-06,
+      "loss": 0.5826,
+      "step": 33210
+    },
+    {
+      "epoch": 3.322,
+      "grad_norm": 49.59003829956055,
+      "learning_rate": 3.3594000000000004e-06,
+      "loss": 0.6006,
+      "step": 33220
+    },
+    {
+      "epoch": 3.323,
+      "grad_norm": 28.964061737060547,
+      "learning_rate": 3.3574000000000002e-06,
+      "loss": 0.621,
+      "step": 33230
+    },
+    {
+      "epoch": 3.324,
+      "grad_norm": 41.858428955078125,
+      "learning_rate": 3.3554000000000005e-06,
+      "loss": 0.7396,
+      "step": 33240
+    },
+    {
+      "epoch": 3.325,
+      "grad_norm": 47.466148376464844,
+      "learning_rate": 3.3534000000000004e-06,
+      "loss": 0.774,
+      "step": 33250
+    },
+    {
+      "epoch": 3.326,
+      "grad_norm": 34.58932113647461,
+      "learning_rate": 3.3514000000000002e-06,
+      "loss": 0.5768,
+      "step": 33260
+    },
+    {
+      "epoch": 3.327,
+      "grad_norm": 30.386682510375977,
+      "learning_rate": 3.3494000000000005e-06,
+      "loss": 0.4744,
+      "step": 33270
+    },
+    {
+      "epoch": 3.328,
+      "grad_norm": 32.04347610473633,
+      "learning_rate": 3.3474e-06,
+      "loss": 0.5459,
+      "step": 33280
+    },
+    {
+      "epoch": 3.329,
+      "grad_norm": 34.02959442138672,
+      "learning_rate": 3.3454e-06,
+      "loss": 0.5957,
+      "step": 33290
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 28.056381225585938,
+      "learning_rate": 3.3434000000000005e-06,
+      "loss": 0.4959,
+      "step": 33300
+    },
+    {
+      "epoch": 3.331,
+      "grad_norm": 11.723163604736328,
+      "learning_rate": 3.3414000000000003e-06,
+      "loss": 0.4785,
+      "step": 33310
+    },
+    {
+      "epoch": 3.332,
+      "grad_norm": 45.393795013427734,
+      "learning_rate": 3.3394000000000006e-06,
+      "loss": 0.6107,
+      "step": 33320
+    },
+    {
+      "epoch": 3.333,
+      "grad_norm": 195.7186737060547,
+      "learning_rate": 3.3374e-06,
+      "loss": 0.557,
+      "step": 33330
+    },
+    {
+      "epoch": 3.334,
+      "grad_norm": 36.78384017944336,
+      "learning_rate": 3.3354000000000003e-06,
+      "loss": 0.568,
+      "step": 33340
+    },
+    {
+      "epoch": 3.335,
+      "grad_norm": 37.96222686767578,
+      "learning_rate": 3.3334000000000006e-06,
+      "loss": 0.5528,
+      "step": 33350
+    },
+    {
+      "epoch": 3.336,
+      "grad_norm": 38.002418518066406,
+      "learning_rate": 3.3314e-06,
+      "loss": 0.5147,
+      "step": 33360
+    },
+    {
+      "epoch": 3.337,
+      "grad_norm": 35.465614318847656,
+      "learning_rate": 3.3294000000000003e-06,
+      "loss": 0.4127,
+      "step": 33370
+    },
+    {
+      "epoch": 3.338,
+      "grad_norm": 15.506857872009277,
+      "learning_rate": 3.3274e-06,
+      "loss": 0.4986,
+      "step": 33380
+    },
+    {
+      "epoch": 3.339,
+      "grad_norm": 29.845312118530273,
+      "learning_rate": 3.3254000000000004e-06,
+      "loss": 0.4565,
+      "step": 33390
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 6.27945613861084,
+      "learning_rate": 3.3234000000000007e-06,
+      "loss": 0.62,
+      "step": 33400
+    },
+    {
+      "epoch": 3.341,
+      "grad_norm": 51.38041687011719,
+      "learning_rate": 3.3214e-06,
+      "loss": 0.8341,
+      "step": 33410
+    },
+    {
+      "epoch": 3.342,
+      "grad_norm": 37.613433837890625,
+      "learning_rate": 3.3194000000000004e-06,
+      "loss": 0.6525,
+      "step": 33420
+    },
+    {
+      "epoch": 3.343,
+      "grad_norm": 6.7415266036987305,
+      "learning_rate": 3.3174e-06,
+      "loss": 0.5533,
+      "step": 33430
+    },
+    {
+      "epoch": 3.344,
+      "grad_norm": 4.774569988250732,
+      "learning_rate": 3.3154000000000005e-06,
+      "loss": 0.5075,
+      "step": 33440
+    },
+    {
+      "epoch": 3.3449999999999998,
+      "grad_norm": 26.686466217041016,
+      "learning_rate": 3.3134000000000003e-06,
+      "loss": 0.6392,
+      "step": 33450
+    },
+    {
+      "epoch": 3.346,
+      "grad_norm": 28.92704200744629,
+      "learning_rate": 3.3114e-06,
+      "loss": 0.6978,
+      "step": 33460
+    },
+    {
+      "epoch": 3.347,
+      "grad_norm": 26.489171981811523,
+      "learning_rate": 3.3094000000000005e-06,
+      "loss": 0.5123,
+      "step": 33470
+    },
+    {
+      "epoch": 3.348,
+      "grad_norm": 69.60842895507812,
+      "learning_rate": 3.3074e-06,
+      "loss": 0.7418,
+      "step": 33480
+    },
+    {
+      "epoch": 3.349,
+      "grad_norm": 42.1660270690918,
+      "learning_rate": 3.3054e-06,
+      "loss": 0.7195,
+      "step": 33490
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 53.14607620239258,
+      "learning_rate": 3.3034000000000004e-06,
+      "loss": 0.4782,
+      "step": 33500
+    },
+    {
+      "epoch": 3.351,
+      "grad_norm": 39.90643310546875,
+      "learning_rate": 3.3014000000000003e-06,
+      "loss": 0.6492,
+      "step": 33510
+    },
+    {
+      "epoch": 3.352,
+      "grad_norm": 37.26142501831055,
+      "learning_rate": 3.2994000000000005e-06,
+      "loss": 0.4781,
+      "step": 33520
+    },
+    {
+      "epoch": 3.3529999999999998,
+      "grad_norm": 59.464988708496094,
+      "learning_rate": 3.2974e-06,
+      "loss": 0.684,
+      "step": 33530
+    },
+    {
+      "epoch": 3.354,
+      "grad_norm": 43.060523986816406,
+      "learning_rate": 3.2954000000000002e-06,
+      "loss": 0.5223,
+      "step": 33540
+    },
+    {
+      "epoch": 3.355,
+      "grad_norm": 23.580224990844727,
+      "learning_rate": 3.2934000000000005e-06,
+      "loss": 0.4565,
+      "step": 33550
+    },
+    {
+      "epoch": 3.356,
+      "grad_norm": 54.410423278808594,
+      "learning_rate": 3.2914e-06,
+      "loss": 0.586,
+      "step": 33560
+    },
+    {
+      "epoch": 3.357,
+      "grad_norm": 8.084722518920898,
+      "learning_rate": 3.2894000000000002e-06,
+      "loss": 0.442,
+      "step": 33570
+    },
+    {
+      "epoch": 3.358,
+      "grad_norm": 21.548755645751953,
+      "learning_rate": 3.2874e-06,
+      "loss": 0.5678,
+      "step": 33580
+    },
+    {
+      "epoch": 3.359,
+      "grad_norm": 14.505980491638184,
+      "learning_rate": 3.2854000000000003e-06,
+      "loss": 0.539,
+      "step": 33590
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 4.456478118896484,
+      "learning_rate": 3.2834000000000006e-06,
+      "loss": 0.4719,
+      "step": 33600
+    },
+    {
+      "epoch": 3.3609999999999998,
+      "grad_norm": 58.42336654663086,
+      "learning_rate": 3.2814e-06,
+      "loss": 0.5288,
+      "step": 33610
+    },
+    {
+      "epoch": 3.362,
+      "grad_norm": 36.164154052734375,
+      "learning_rate": 3.2794000000000003e-06,
+      "loss": 0.5947,
+      "step": 33620
+    },
+    {
+      "epoch": 3.363,
+      "grad_norm": 53.009727478027344,
+      "learning_rate": 3.2774e-06,
+      "loss": 0.4736,
+      "step": 33630
+    },
+    {
+      "epoch": 3.364,
+      "grad_norm": 33.75554656982422,
+      "learning_rate": 3.2754000000000004e-06,
+      "loss": 0.6622,
+      "step": 33640
+    },
+    {
+      "epoch": 3.365,
+      "grad_norm": 38.6405143737793,
+      "learning_rate": 3.2734000000000003e-06,
+      "loss": 0.5776,
+      "step": 33650
+    },
+    {
+      "epoch": 3.366,
+      "grad_norm": 5.115553379058838,
+      "learning_rate": 3.2714e-06,
+      "loss": 0.6407,
+      "step": 33660
+    },
+    {
+      "epoch": 3.367,
+      "grad_norm": 33.69319152832031,
+      "learning_rate": 3.2694000000000004e-06,
+      "loss": 0.5197,
+      "step": 33670
+    },
+    {
+      "epoch": 3.368,
+      "grad_norm": 37.610408782958984,
+      "learning_rate": 3.2674e-06,
+      "loss": 0.794,
+      "step": 33680
+    },
+    {
+      "epoch": 3.3689999999999998,
+      "grad_norm": 69.22075653076172,
+      "learning_rate": 3.2654e-06,
+      "loss": 0.4071,
+      "step": 33690
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 33.58649444580078,
+      "learning_rate": 3.2634000000000004e-06,
+      "loss": 0.5019,
+      "step": 33700
+    },
+    {
+      "epoch": 3.371,
+      "grad_norm": 67.5018539428711,
+      "learning_rate": 3.2614000000000002e-06,
+      "loss": 0.5652,
+      "step": 33710
+    },
+    {
+      "epoch": 3.372,
+      "grad_norm": 45.005184173583984,
+      "learning_rate": 3.2594000000000005e-06,
+      "loss": 0.4798,
+      "step": 33720
+    },
+    {
+      "epoch": 3.373,
+      "grad_norm": 30.43639373779297,
+      "learning_rate": 3.2574e-06,
+      "loss": 0.559,
+      "step": 33730
+    },
+    {
+      "epoch": 3.374,
+      "grad_norm": 45.54741287231445,
+      "learning_rate": 3.2554e-06,
+      "loss": 0.6114,
+      "step": 33740
+    },
+    {
+      "epoch": 3.375,
+      "grad_norm": 16.796628952026367,
+      "learning_rate": 3.2534000000000005e-06,
+      "loss": 0.7101,
+      "step": 33750
+    },
+    {
+      "epoch": 3.376,
+      "grad_norm": 13.3946533203125,
+      "learning_rate": 3.2514e-06,
+      "loss": 0.6903,
+      "step": 33760
+    },
+    {
+      "epoch": 3.377,
+      "grad_norm": 69.04631805419922,
+      "learning_rate": 3.2494e-06,
+      "loss": 0.6845,
+      "step": 33770
+    },
+    {
+      "epoch": 3.378,
+      "grad_norm": 29.6506404876709,
+      "learning_rate": 3.2474e-06,
+      "loss": 0.9378,
+      "step": 33780
+    },
+    {
+      "epoch": 3.379,
+      "grad_norm": 47.310462951660156,
+      "learning_rate": 3.2454000000000003e-06,
+      "loss": 0.4969,
+      "step": 33790
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 41.4119987487793,
+      "learning_rate": 3.2434000000000006e-06,
+      "loss": 0.479,
+      "step": 33800
+    },
+    {
+      "epoch": 3.3810000000000002,
+      "grad_norm": 51.03746032714844,
+      "learning_rate": 3.2414e-06,
+      "loss": 0.6456,
+      "step": 33810
+    },
+    {
+      "epoch": 3.382,
+      "grad_norm": 21.19719886779785,
+      "learning_rate": 3.2394000000000003e-06,
+      "loss": 0.4372,
+      "step": 33820
+    },
+    {
+      "epoch": 3.383,
+      "grad_norm": 6.403212547302246,
+      "learning_rate": 3.2374000000000005e-06,
+      "loss": 0.4273,
+      "step": 33830
+    },
+    {
+      "epoch": 3.384,
+      "grad_norm": 37.43207550048828,
+      "learning_rate": 3.2354000000000004e-06,
+      "loss": 0.6179,
+      "step": 33840
+    },
+    {
+      "epoch": 3.385,
+      "grad_norm": 1.2079825401306152,
+      "learning_rate": 3.2334000000000002e-06,
+      "loss": 0.2904,
+      "step": 33850
+    },
+    {
+      "epoch": 3.386,
+      "grad_norm": 29.27356719970703,
+      "learning_rate": 3.2314e-06,
+      "loss": 0.6627,
+      "step": 33860
+    },
+    {
+      "epoch": 3.387,
+      "grad_norm": 45.79487228393555,
+      "learning_rate": 3.2294000000000004e-06,
+      "loss": 0.8661,
+      "step": 33870
+    },
+    {
+      "epoch": 3.388,
+      "grad_norm": 17.66958999633789,
+      "learning_rate": 3.2274000000000006e-06,
+      "loss": 0.3627,
+      "step": 33880
+    },
+    {
+      "epoch": 3.3890000000000002,
+      "grad_norm": 35.9542236328125,
+      "learning_rate": 3.2254e-06,
+      "loss": 0.9139,
+      "step": 33890
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 11.291265487670898,
+      "learning_rate": 3.2234000000000003e-06,
+      "loss": 0.4528,
+      "step": 33900
+    },
+    {
+      "epoch": 3.391,
+      "grad_norm": 78.31896209716797,
+      "learning_rate": 3.2214e-06,
+      "loss": 0.8227,
+      "step": 33910
+    },
+    {
+      "epoch": 3.392,
+      "grad_norm": 24.657066345214844,
+      "learning_rate": 3.2194000000000005e-06,
+      "loss": 0.4872,
+      "step": 33920
+    },
+    {
+      "epoch": 3.393,
+      "grad_norm": 45.37953567504883,
+      "learning_rate": 3.2174000000000003e-06,
+      "loss": 0.7539,
+      "step": 33930
+    },
+    {
+      "epoch": 3.394,
+      "grad_norm": 42.956459045410156,
+      "learning_rate": 3.2154e-06,
+      "loss": 0.8764,
+      "step": 33940
+    },
+    {
+      "epoch": 3.395,
+      "grad_norm": 9.4888277053833,
+      "learning_rate": 3.2134000000000004e-06,
+      "loss": 0.6115,
+      "step": 33950
+    },
+    {
+      "epoch": 3.396,
+      "grad_norm": 23.10115623474121,
+      "learning_rate": 3.2114e-06,
+      "loss": 0.6539,
+      "step": 33960
+    },
+    {
+      "epoch": 3.3970000000000002,
+      "grad_norm": 17.13572883605957,
+      "learning_rate": 3.2094e-06,
+      "loss": 0.5708,
+      "step": 33970
+    },
+    {
+      "epoch": 3.398,
+      "grad_norm": 15.064404487609863,
+      "learning_rate": 3.2074000000000004e-06,
+      "loss": 0.4718,
+      "step": 33980
+    },
+    {
+      "epoch": 3.399,
+      "grad_norm": 63.72896194458008,
+      "learning_rate": 3.2054000000000003e-06,
+      "loss": 0.4698,
+      "step": 33990
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 12.187103271484375,
+      "learning_rate": 3.2034000000000005e-06,
+      "loss": 0.4326,
+      "step": 34000
+    },
+    {
+      "epoch": 3.401,
+      "grad_norm": 67.7624740600586,
+      "learning_rate": 3.2014e-06,
+      "loss": 0.6557,
+      "step": 34010
+    },
+    {
+      "epoch": 3.402,
+      "grad_norm": 34.88738250732422,
+      "learning_rate": 3.1994000000000002e-06,
+      "loss": 0.5402,
+      "step": 34020
+    },
+    {
+      "epoch": 3.403,
+      "grad_norm": 50.04561996459961,
+      "learning_rate": 3.1974000000000005e-06,
+      "loss": 0.6956,
+      "step": 34030
+    },
+    {
+      "epoch": 3.404,
+      "grad_norm": 48.32029342651367,
+      "learning_rate": 3.1954000000000004e-06,
+      "loss": 0.6118,
+      "step": 34040
+    },
+    {
+      "epoch": 3.4050000000000002,
+      "grad_norm": 60.040218353271484,
+      "learning_rate": 3.1934e-06,
+      "loss": 0.5796,
+      "step": 34050
+    },
+    {
+      "epoch": 3.406,
+      "grad_norm": 41.5383415222168,
+      "learning_rate": 3.1914e-06,
+      "loss": 0.5164,
+      "step": 34060
+    },
+    {
+      "epoch": 3.407,
+      "grad_norm": 22.694530487060547,
+      "learning_rate": 3.1894000000000003e-06,
+      "loss": 0.4927,
+      "step": 34070
+    },
+    {
+      "epoch": 3.408,
+      "grad_norm": 45.76729202270508,
+      "learning_rate": 3.1874000000000006e-06,
+      "loss": 0.6338,
+      "step": 34080
+    },
+    {
+      "epoch": 3.409,
+      "grad_norm": 2.770658493041992,
+      "learning_rate": 3.1854e-06,
+      "loss": 0.8599,
+      "step": 34090
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 12.581628799438477,
+      "learning_rate": 3.1834000000000003e-06,
+      "loss": 0.4446,
+      "step": 34100
+    },
+    {
+      "epoch": 3.411,
+      "grad_norm": 15.565549850463867,
+      "learning_rate": 3.1814e-06,
+      "loss": 0.5537,
+      "step": 34110
+    },
+    {
+      "epoch": 3.412,
+      "grad_norm": 24.086557388305664,
+      "learning_rate": 3.1794000000000004e-06,
+      "loss": 0.5949,
+      "step": 34120
+    },
+    {
+      "epoch": 3.413,
+      "grad_norm": 18.560720443725586,
+      "learning_rate": 3.1774000000000003e-06,
+      "loss": 0.5971,
+      "step": 34130
+    },
+    {
+      "epoch": 3.414,
+      "grad_norm": 40.32215881347656,
+      "learning_rate": 3.1754e-06,
+      "loss": 0.5366,
+      "step": 34140
+    },
+    {
+      "epoch": 3.415,
+      "grad_norm": 42.73981857299805,
+      "learning_rate": 3.1734000000000004e-06,
+      "loss": 0.4257,
+      "step": 34150
+    },
+    {
+      "epoch": 3.416,
+      "grad_norm": 27.142242431640625,
+      "learning_rate": 3.1714e-06,
+      "loss": 0.7793,
+      "step": 34160
+    },
+    {
+      "epoch": 3.417,
+      "grad_norm": 31.951887130737305,
+      "learning_rate": 3.1694e-06,
+      "loss": 0.3895,
+      "step": 34170
+    },
+    {
+      "epoch": 3.418,
+      "grad_norm": 34.47846603393555,
+      "learning_rate": 3.1674000000000004e-06,
+      "loss": 0.6155,
+      "step": 34180
+    },
+    {
+      "epoch": 3.419,
+      "grad_norm": 33.36940383911133,
+      "learning_rate": 3.1654000000000002e-06,
+      "loss": 0.416,
+      "step": 34190
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 71.54866027832031,
+      "learning_rate": 3.1634000000000005e-06,
+      "loss": 0.8407,
+      "step": 34200
+    },
+    {
+      "epoch": 3.421,
+      "grad_norm": 2.4726550579071045,
+      "learning_rate": 3.1614e-06,
+      "loss": 0.7122,
+      "step": 34210
+    },
+    {
+      "epoch": 3.422,
+      "grad_norm": 55.77376174926758,
+      "learning_rate": 3.1594e-06,
+      "loss": 0.543,
+      "step": 34220
+    },
+    {
+      "epoch": 3.423,
+      "grad_norm": 30.04120635986328,
+      "learning_rate": 3.1574000000000005e-06,
+      "loss": 0.7113,
+      "step": 34230
+    },
+    {
+      "epoch": 3.424,
+      "grad_norm": 7.0647993087768555,
+      "learning_rate": 3.1554000000000003e-06,
+      "loss": 0.6297,
+      "step": 34240
+    },
+    {
+      "epoch": 3.425,
+      "grad_norm": 44.25429916381836,
+      "learning_rate": 3.1534e-06,
+      "loss": 0.6442,
+      "step": 34250
+    },
+    {
+      "epoch": 3.426,
+      "grad_norm": 8.99540901184082,
+      "learning_rate": 3.1514e-06,
+      "loss": 0.63,
+      "step": 34260
+    },
+    {
+      "epoch": 3.427,
+      "grad_norm": 5.797199249267578,
+      "learning_rate": 3.1494000000000003e-06,
+      "loss": 0.3925,
+      "step": 34270
+    },
+    {
+      "epoch": 3.428,
+      "grad_norm": 33.43487548828125,
+      "learning_rate": 3.1474000000000006e-06,
+      "loss": 0.3708,
+      "step": 34280
+    },
+    {
+      "epoch": 3.429,
+      "grad_norm": 20.71714210510254,
+      "learning_rate": 3.1454e-06,
+      "loss": 0.6929,
+      "step": 34290
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 32.383670806884766,
+      "learning_rate": 3.1434000000000003e-06,
+      "loss": 0.5878,
+      "step": 34300
+    },
+    {
+      "epoch": 3.431,
+      "grad_norm": 56.46757888793945,
+      "learning_rate": 3.1414e-06,
+      "loss": 0.6765,
+      "step": 34310
+    },
+    {
+      "epoch": 3.432,
+      "grad_norm": 5.454736709594727,
+      "learning_rate": 3.1394000000000004e-06,
+      "loss": 0.7627,
+      "step": 34320
+    },
+    {
+      "epoch": 3.433,
+      "grad_norm": 51.933231353759766,
+      "learning_rate": 3.1374000000000002e-06,
+      "loss": 0.4866,
+      "step": 34330
+    },
+    {
+      "epoch": 3.434,
+      "grad_norm": 41.372867584228516,
+      "learning_rate": 3.1354e-06,
+      "loss": 0.5965,
+      "step": 34340
+    },
+    {
+      "epoch": 3.435,
+      "grad_norm": 56.871681213378906,
+      "learning_rate": 3.1334000000000004e-06,
+      "loss": 0.6184,
+      "step": 34350
+    },
+    {
+      "epoch": 3.436,
+      "grad_norm": 6.995004177093506,
+      "learning_rate": 3.1314e-06,
+      "loss": 0.7549,
+      "step": 34360
+    },
+    {
+      "epoch": 3.437,
+      "grad_norm": 46.217041015625,
+      "learning_rate": 3.1294e-06,
+      "loss": 0.7235,
+      "step": 34370
+    },
+    {
+      "epoch": 3.438,
+      "grad_norm": 29.683107376098633,
+      "learning_rate": 3.1274000000000003e-06,
+      "loss": 0.7433,
+      "step": 34380
+    },
+    {
+      "epoch": 3.439,
+      "grad_norm": 26.741121292114258,
+      "learning_rate": 3.1254e-06,
+      "loss": 0.6973,
+      "step": 34390
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 65.56108856201172,
+      "learning_rate": 3.1234000000000005e-06,
+      "loss": 0.5786,
+      "step": 34400
+    },
+    {
+      "epoch": 3.441,
+      "grad_norm": 13.837389945983887,
+      "learning_rate": 3.1214e-06,
+      "loss": 0.6994,
+      "step": 34410
+    },
+    {
+      "epoch": 3.442,
+      "grad_norm": 24.370201110839844,
+      "learning_rate": 3.1194e-06,
+      "loss": 0.5863,
+      "step": 34420
+    },
+    {
+      "epoch": 3.443,
+      "grad_norm": 1.192421555519104,
+      "learning_rate": 3.1174000000000004e-06,
+      "loss": 0.6617,
+      "step": 34430
+    },
+    {
+      "epoch": 3.444,
+      "grad_norm": 61.917198181152344,
+      "learning_rate": 3.1154000000000003e-06,
+      "loss": 0.5432,
+      "step": 34440
+    },
+    {
+      "epoch": 3.445,
+      "grad_norm": 26.635774612426758,
+      "learning_rate": 3.1134e-06,
+      "loss": 0.6587,
+      "step": 34450
+    },
+    {
+      "epoch": 3.446,
+      "grad_norm": 20.741262435913086,
+      "learning_rate": 3.1114000000000004e-06,
+      "loss": 0.5384,
+      "step": 34460
+    },
+    {
+      "epoch": 3.447,
+      "grad_norm": 56.38450622558594,
+      "learning_rate": 3.1094000000000002e-06,
+      "loss": 0.8204,
+      "step": 34470
+    },
+    {
+      "epoch": 3.448,
+      "grad_norm": 58.865074157714844,
+      "learning_rate": 3.1074000000000005e-06,
+      "loss": 0.679,
+      "step": 34480
+    },
+    {
+      "epoch": 3.449,
+      "grad_norm": 79.67637634277344,
+      "learning_rate": 3.1054e-06,
+      "loss": 0.6898,
+      "step": 34490
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 52.565040588378906,
+      "learning_rate": 3.1034000000000002e-06,
+      "loss": 0.7095,
+      "step": 34500
+    },
+    {
+      "epoch": 3.451,
+      "grad_norm": 47.64340591430664,
+      "learning_rate": 3.1014000000000005e-06,
+      "loss": 0.5069,
+      "step": 34510
+    },
+    {
+      "epoch": 3.452,
+      "grad_norm": 54.49386978149414,
+      "learning_rate": 3.0994000000000003e-06,
+      "loss": 0.7183,
+      "step": 34520
+    },
+    {
+      "epoch": 3.453,
+      "grad_norm": 38.74276351928711,
+      "learning_rate": 3.0974000000000006e-06,
+      "loss": 0.4679,
+      "step": 34530
+    },
+    {
+      "epoch": 3.454,
+      "grad_norm": 21.229961395263672,
+      "learning_rate": 3.0954e-06,
+      "loss": 0.7752,
+      "step": 34540
+    },
+    {
+      "epoch": 3.455,
+      "grad_norm": 72.66622924804688,
+      "learning_rate": 3.0934000000000003e-06,
+      "loss": 0.5673,
+      "step": 34550
+    },
+    {
+      "epoch": 3.456,
+      "grad_norm": 40.880306243896484,
+      "learning_rate": 3.0914000000000006e-06,
+      "loss": 0.4422,
+      "step": 34560
+    },
+    {
+      "epoch": 3.457,
+      "grad_norm": 40.722320556640625,
+      "learning_rate": 3.0894e-06,
+      "loss": 0.6761,
+      "step": 34570
+    },
+    {
+      "epoch": 3.458,
+      "grad_norm": 63.91718292236328,
+      "learning_rate": 3.0874000000000003e-06,
+      "loss": 0.6881,
+      "step": 34580
+    },
+    {
+      "epoch": 3.459,
+      "grad_norm": 79.54600524902344,
+      "learning_rate": 3.0854e-06,
+      "loss": 0.5785,
+      "step": 34590
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 38.31000518798828,
+      "learning_rate": 3.0834000000000004e-06,
+      "loss": 0.8471,
+      "step": 34600
+    },
+    {
+      "epoch": 3.461,
+      "grad_norm": 42.8505859375,
+      "learning_rate": 3.0814000000000007e-06,
+      "loss": 0.7023,
+      "step": 34610
+    },
+    {
+      "epoch": 3.462,
+      "grad_norm": 27.422082901000977,
+      "learning_rate": 3.0794e-06,
+      "loss": 0.7369,
+      "step": 34620
+    },
+    {
+      "epoch": 3.463,
+      "grad_norm": 45.18264389038086,
+      "learning_rate": 3.0774000000000004e-06,
+      "loss": 0.4083,
+      "step": 34630
+    },
+    {
+      "epoch": 3.464,
+      "grad_norm": 32.00255584716797,
+      "learning_rate": 3.0754000000000002e-06,
+      "loss": 0.6028,
+      "step": 34640
+    },
+    {
+      "epoch": 3.465,
+      "grad_norm": 10.15850830078125,
+      "learning_rate": 3.0734e-06,
+      "loss": 0.5721,
+      "step": 34650
+    },
+    {
+      "epoch": 3.466,
+      "grad_norm": 12.090837478637695,
+      "learning_rate": 3.0714000000000004e-06,
+      "loss": 0.6242,
+      "step": 34660
+    },
+    {
+      "epoch": 3.467,
+      "grad_norm": 55.292484283447266,
+      "learning_rate": 3.0694e-06,
+      "loss": 0.4978,
+      "step": 34670
+    },
+    {
+      "epoch": 3.468,
+      "grad_norm": 68.95906066894531,
+      "learning_rate": 3.0674000000000005e-06,
+      "loss": 0.7369,
+      "step": 34680
+    },
+    {
+      "epoch": 3.469,
+      "grad_norm": 53.9699821472168,
+      "learning_rate": 3.0654e-06,
+      "loss": 0.68,
+      "step": 34690
+    },
+    {
+      "epoch": 3.4699999999999998,
+      "grad_norm": 60.666229248046875,
+      "learning_rate": 3.0634e-06,
+      "loss": 0.855,
+      "step": 34700
+    },
+    {
+      "epoch": 3.471,
+      "grad_norm": 17.940872192382812,
+      "learning_rate": 3.0614000000000005e-06,
+      "loss": 0.3829,
+      "step": 34710
+    },
+    {
+      "epoch": 3.472,
+      "grad_norm": 38.6558723449707,
+      "learning_rate": 3.0594000000000003e-06,
+      "loss": 0.5485,
+      "step": 34720
+    },
+    {
+      "epoch": 3.473,
+      "grad_norm": 31.948421478271484,
+      "learning_rate": 3.0574000000000006e-06,
+      "loss": 0.664,
+      "step": 34730
+    },
+    {
+      "epoch": 3.474,
+      "grad_norm": 33.4412727355957,
+      "learning_rate": 3.0554e-06,
+      "loss": 0.5501,
+      "step": 34740
+    },
+    {
+      "epoch": 3.475,
+      "grad_norm": 42.890480041503906,
+      "learning_rate": 3.0534000000000003e-06,
+      "loss": 0.4292,
+      "step": 34750
+    },
+    {
+      "epoch": 3.476,
+      "grad_norm": 52.22254180908203,
+      "learning_rate": 3.0514000000000005e-06,
+      "loss": 0.6644,
+      "step": 34760
+    },
+    {
+      "epoch": 3.477,
+      "grad_norm": 3.8460352420806885,
+      "learning_rate": 3.0494e-06,
+      "loss": 0.6048,
+      "step": 34770
+    },
+    {
+      "epoch": 3.4779999999999998,
+      "grad_norm": 38.23897171020508,
+      "learning_rate": 3.0476e-06,
+      "loss": 0.6073,
+      "step": 34780
+    },
+    {
+      "epoch": 3.479,
+      "grad_norm": 39.942691802978516,
+      "learning_rate": 3.0456e-06,
+      "loss": 0.4298,
+      "step": 34790
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 63.8907470703125,
+      "learning_rate": 3.0436000000000004e-06,
+      "loss": 0.6932,
+      "step": 34800
+    },
+    {
+      "epoch": 3.481,
+      "grad_norm": 30.509557723999023,
+      "learning_rate": 3.0416000000000002e-06,
+      "loss": 0.6961,
+      "step": 34810
+    },
+    {
+      "epoch": 3.482,
+      "grad_norm": 31.000228881835938,
+      "learning_rate": 3.0396000000000005e-06,
+      "loss": 0.5928,
+      "step": 34820
+    },
+    {
+      "epoch": 3.483,
+      "grad_norm": 15.010289192199707,
+      "learning_rate": 3.0376e-06,
+      "loss": 0.4602,
+      "step": 34830
+    },
+    {
+      "epoch": 3.484,
+      "grad_norm": 57.35770797729492,
+      "learning_rate": 3.0356e-06,
+      "loss": 0.6064,
+      "step": 34840
+    },
+    {
+      "epoch": 3.485,
+      "grad_norm": 28.138120651245117,
+      "learning_rate": 3.0336000000000005e-06,
+      "loss": 0.7436,
+      "step": 34850
+    },
+    {
+      "epoch": 3.4859999999999998,
+      "grad_norm": 44.76555252075195,
+      "learning_rate": 3.0316e-06,
+      "loss": 0.7035,
+      "step": 34860
+    },
+    {
+      "epoch": 3.487,
+      "grad_norm": 15.496353149414062,
+      "learning_rate": 3.0296e-06,
+      "loss": 0.5574,
+      "step": 34870
+    },
+    {
+      "epoch": 3.488,
+      "grad_norm": 9.24044132232666,
+      "learning_rate": 3.0276e-06,
+      "loss": 0.5043,
+      "step": 34880
+    },
+    {
+      "epoch": 3.489,
+      "grad_norm": 10.469659805297852,
+      "learning_rate": 3.0256000000000003e-06,
+      "loss": 0.4792,
+      "step": 34890
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 23.2648983001709,
+      "learning_rate": 3.0236000000000006e-06,
+      "loss": 0.606,
+      "step": 34900
+    },
+    {
+      "epoch": 3.491,
+      "grad_norm": 60.732452392578125,
+      "learning_rate": 3.0216e-06,
+      "loss": 0.934,
+      "step": 34910
+    },
+    {
+      "epoch": 3.492,
+      "grad_norm": 10.405770301818848,
+      "learning_rate": 3.0196000000000003e-06,
+      "loss": 0.4321,
+      "step": 34920
+    },
+    {
+      "epoch": 3.493,
+      "grad_norm": 40.18657302856445,
+      "learning_rate": 3.0176e-06,
+      "loss": 0.4344,
+      "step": 34930
+    },
+    {
+      "epoch": 3.4939999999999998,
+      "grad_norm": 23.268260955810547,
+      "learning_rate": 3.0156000000000004e-06,
+      "loss": 0.4978,
+      "step": 34940
+    },
+    {
+      "epoch": 3.495,
+      "grad_norm": 24.141082763671875,
+      "learning_rate": 3.0136000000000003e-06,
+      "loss": 0.5395,
+      "step": 34950
+    },
+    {
+      "epoch": 3.496,
+      "grad_norm": 29.806568145751953,
+      "learning_rate": 3.0116e-06,
+      "loss": 0.3322,
+      "step": 34960
+    },
+    {
+      "epoch": 3.497,
+      "grad_norm": 51.73660659790039,
+      "learning_rate": 3.0096000000000004e-06,
+      "loss": 0.719,
+      "step": 34970
+    },
+    {
+      "epoch": 3.498,
+      "grad_norm": 15.9256591796875,
+      "learning_rate": 3.0076000000000006e-06,
+      "loss": 0.6364,
+      "step": 34980
+    },
+    {
+      "epoch": 3.499,
+      "grad_norm": 6.268895626068115,
+      "learning_rate": 3.0056e-06,
+      "loss": 0.6966,
+      "step": 34990
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 64.48417663574219,
+      "learning_rate": 3.0036000000000003e-06,
+      "loss": 0.557,
+      "step": 35000
+    },
+    {
+      "epoch": 3.501,
+      "grad_norm": 26.548616409301758,
+      "learning_rate": 3.0016e-06,
+      "loss": 0.7495,
+      "step": 35010
+    },
+    {
+      "epoch": 3.502,
+      "grad_norm": 43.20073699951172,
+      "learning_rate": 2.9996000000000005e-06,
+      "loss": 0.3843,
+      "step": 35020
+    },
+    {
+      "epoch": 3.503,
+      "grad_norm": 29.13274383544922,
+      "learning_rate": 2.9976000000000003e-06,
+      "loss": 0.4718,
+      "step": 35030
+    },
+    {
+      "epoch": 3.504,
+      "grad_norm": 54.19246292114258,
+      "learning_rate": 2.9956e-06,
+      "loss": 0.6532,
+      "step": 35040
+    },
+    {
+      "epoch": 3.505,
+      "grad_norm": 42.29218673706055,
+      "learning_rate": 2.9936000000000004e-06,
+      "loss": 0.5487,
+      "step": 35050
+    },
+    {
+      "epoch": 3.5060000000000002,
+      "grad_norm": 42.01010513305664,
+      "learning_rate": 2.9916e-06,
+      "loss": 0.6537,
+      "step": 35060
+    },
+    {
+      "epoch": 3.507,
+      "grad_norm": 22.00259017944336,
+      "learning_rate": 2.9896e-06,
+      "loss": 0.5337,
+      "step": 35070
+    },
+    {
+      "epoch": 3.508,
+      "grad_norm": 47.87183380126953,
+      "learning_rate": 2.9876000000000004e-06,
+      "loss": 0.5742,
+      "step": 35080
+    },
+    {
+      "epoch": 3.509,
+      "grad_norm": 39.39622497558594,
+      "learning_rate": 2.9856000000000003e-06,
+      "loss": 0.4731,
+      "step": 35090
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 12.731461524963379,
+      "learning_rate": 2.9836000000000005e-06,
+      "loss": 0.5601,
+      "step": 35100
+    },
+    {
+      "epoch": 3.511,
+      "grad_norm": 80.69575500488281,
+      "learning_rate": 2.9816e-06,
+      "loss": 1.0194,
+      "step": 35110
+    },
+    {
+      "epoch": 3.512,
+      "grad_norm": 26.40245246887207,
+      "learning_rate": 2.9796000000000002e-06,
+      "loss": 0.5556,
+      "step": 35120
+    },
+    {
+      "epoch": 3.513,
+      "grad_norm": 38.29823684692383,
+      "learning_rate": 2.9776000000000005e-06,
+      "loss": 0.5171,
+      "step": 35130
+    },
+    {
+      "epoch": 3.5140000000000002,
+      "grad_norm": 26.385711669921875,
+      "learning_rate": 2.9756000000000004e-06,
+      "loss": 0.6902,
+      "step": 35140
+    },
+    {
+      "epoch": 3.515,
+      "grad_norm": 46.439876556396484,
+      "learning_rate": 2.9736e-06,
+      "loss": 0.6092,
+      "step": 35150
+    },
+    {
+      "epoch": 3.516,
+      "grad_norm": 16.673818588256836,
+      "learning_rate": 2.9716e-06,
+      "loss": 0.5462,
+      "step": 35160
+    },
+    {
+      "epoch": 3.517,
+      "grad_norm": 23.730607986450195,
+      "learning_rate": 2.9696000000000003e-06,
+      "loss": 0.6426,
+      "step": 35170
+    },
+    {
+      "epoch": 3.518,
+      "grad_norm": 27.162260055541992,
+      "learning_rate": 2.9676000000000006e-06,
+      "loss": 0.5589,
+      "step": 35180
+    },
+    {
+      "epoch": 3.519,
+      "grad_norm": 47.120635986328125,
+      "learning_rate": 2.9656e-06,
+      "loss": 0.5825,
+      "step": 35190
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 42.60955047607422,
+      "learning_rate": 2.9636000000000003e-06,
+      "loss": 0.4523,
+      "step": 35200
+    },
+    {
+      "epoch": 3.521,
+      "grad_norm": 60.848388671875,
+      "learning_rate": 2.9616e-06,
+      "loss": 0.7928,
+      "step": 35210
+    },
+    {
+      "epoch": 3.5220000000000002,
+      "grad_norm": 10.591123580932617,
+      "learning_rate": 2.9596000000000004e-06,
+      "loss": 0.3901,
+      "step": 35220
+    },
+    {
+      "epoch": 3.523,
+      "grad_norm": 57.188079833984375,
+      "learning_rate": 2.9576000000000003e-06,
+      "loss": 0.6111,
+      "step": 35230
+    },
+    {
+      "epoch": 3.524,
+      "grad_norm": 45.49052047729492,
+      "learning_rate": 2.9556e-06,
+      "loss": 0.6336,
+      "step": 35240
+    },
+    {
+      "epoch": 3.525,
+      "grad_norm": 59.25286102294922,
+      "learning_rate": 2.9536000000000004e-06,
+      "loss": 0.4022,
+      "step": 35250
+    },
+    {
+      "epoch": 3.526,
+      "grad_norm": 45.68498992919922,
+      "learning_rate": 2.9516e-06,
+      "loss": 1.1201,
+      "step": 35260
+    },
+    {
+      "epoch": 3.527,
+      "grad_norm": 12.463373184204102,
+      "learning_rate": 2.9496e-06,
+      "loss": 0.595,
+      "step": 35270
+    },
+    {
+      "epoch": 3.528,
+      "grad_norm": 50.46893310546875,
+      "learning_rate": 2.9476000000000004e-06,
+      "loss": 0.7465,
+      "step": 35280
+    },
+    {
+      "epoch": 3.529,
+      "grad_norm": 1.739295244216919,
+      "learning_rate": 2.9456000000000002e-06,
+      "loss": 0.6406,
+      "step": 35290
+    },
+    {
+      "epoch": 3.5300000000000002,
+      "grad_norm": 64.8057861328125,
+      "learning_rate": 2.9436000000000005e-06,
+      "loss": 0.6426,
+      "step": 35300
+    },
+    {
+      "epoch": 3.531,
+      "grad_norm": 7.33566427230835,
+      "learning_rate": 2.9416e-06,
+      "loss": 0.4964,
+      "step": 35310
+    },
+    {
+      "epoch": 3.532,
+      "grad_norm": 29.327688217163086,
+      "learning_rate": 2.9396e-06,
+      "loss": 0.5833,
+      "step": 35320
+    },
+    {
+      "epoch": 3.533,
+      "grad_norm": 56.58113479614258,
+      "learning_rate": 2.9376000000000005e-06,
+      "loss": 0.444,
+      "step": 35330
+    },
+    {
+      "epoch": 3.534,
+      "grad_norm": 23.588838577270508,
+      "learning_rate": 2.9356000000000003e-06,
+      "loss": 0.5028,
+      "step": 35340
+    },
+    {
+      "epoch": 3.535,
+      "grad_norm": 14.40968132019043,
+      "learning_rate": 2.9336e-06,
+      "loss": 0.2349,
+      "step": 35350
+    },
+    {
+      "epoch": 3.536,
+      "grad_norm": 58.149803161621094,
+      "learning_rate": 2.9316e-06,
+      "loss": 0.7023,
+      "step": 35360
+    },
+    {
+      "epoch": 3.537,
+      "grad_norm": 46.038902282714844,
+      "learning_rate": 2.9296000000000003e-06,
+      "loss": 0.7448,
+      "step": 35370
+    },
+    {
+      "epoch": 3.5380000000000003,
+      "grad_norm": 45.67920684814453,
+      "learning_rate": 2.9276000000000006e-06,
+      "loss": 0.6292,
+      "step": 35380
+    },
+    {
+      "epoch": 3.539,
+      "grad_norm": 14.963251113891602,
+      "learning_rate": 2.9256e-06,
+      "loss": 0.726,
+      "step": 35390
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 30.30369758605957,
+      "learning_rate": 2.9236000000000003e-06,
+      "loss": 0.6129,
+      "step": 35400
+    },
+    {
+      "epoch": 3.541,
+      "grad_norm": 21.017148971557617,
+      "learning_rate": 2.9216e-06,
+      "loss": 0.8496,
+      "step": 35410
+    },
+    {
+      "epoch": 3.542,
+      "grad_norm": 61.78471374511719,
+      "learning_rate": 2.9196000000000004e-06,
+      "loss": 0.5446,
+      "step": 35420
+    },
+    {
+      "epoch": 3.543,
+      "grad_norm": 21.3669376373291,
+      "learning_rate": 2.9176000000000002e-06,
+      "loss": 0.3647,
+      "step": 35430
+    },
+    {
+      "epoch": 3.544,
+      "grad_norm": 19.953702926635742,
+      "learning_rate": 2.9156e-06,
+      "loss": 0.385,
+      "step": 35440
+    },
+    {
+      "epoch": 3.545,
+      "grad_norm": 46.433876037597656,
+      "learning_rate": 2.9136000000000004e-06,
+      "loss": 0.676,
+      "step": 35450
+    },
+    {
+      "epoch": 3.5460000000000003,
+      "grad_norm": 62.306941986083984,
+      "learning_rate": 2.9115999999999998e-06,
+      "loss": 0.6844,
+      "step": 35460
+    },
+    {
+      "epoch": 3.547,
+      "grad_norm": 42.35296630859375,
+      "learning_rate": 2.9096e-06,
+      "loss": 0.6106,
+      "step": 35470
+    },
+    {
+      "epoch": 3.548,
+      "grad_norm": 37.69320297241211,
+      "learning_rate": 2.9076000000000003e-06,
+      "loss": 0.6195,
+      "step": 35480
+    },
+    {
+      "epoch": 3.549,
+      "grad_norm": 46.660640716552734,
+      "learning_rate": 2.9056e-06,
+      "loss": 0.6841,
+      "step": 35490
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 73.18340301513672,
+      "learning_rate": 2.9036000000000005e-06,
+      "loss": 0.6472,
+      "step": 35500
+    },
+    {
+      "epoch": 3.551,
+      "grad_norm": 44.00477600097656,
+      "learning_rate": 2.9016e-06,
+      "loss": 0.7367,
+      "step": 35510
+    },
+    {
+      "epoch": 3.552,
+      "grad_norm": 140.4091339111328,
+      "learning_rate": 2.8996e-06,
+      "loss": 0.6896,
+      "step": 35520
+    },
+    {
+      "epoch": 3.553,
+      "grad_norm": 9.06778621673584,
+      "learning_rate": 2.8976000000000004e-06,
+      "loss": 0.2333,
+      "step": 35530
+    },
+    {
+      "epoch": 3.5540000000000003,
+      "grad_norm": 48.09809494018555,
+      "learning_rate": 2.8956000000000003e-06,
+      "loss": 0.6038,
+      "step": 35540
+    },
+    {
+      "epoch": 3.555,
+      "grad_norm": 36.15306854248047,
+      "learning_rate": 2.8936e-06,
+      "loss": 0.4335,
+      "step": 35550
+    },
+    {
+      "epoch": 3.556,
+      "grad_norm": 9.26518726348877,
+      "learning_rate": 2.8916e-06,
+      "loss": 0.4004,
+      "step": 35560
+    },
+    {
+      "epoch": 3.557,
+      "grad_norm": 81.55866241455078,
+      "learning_rate": 2.8896000000000003e-06,
+      "loss": 0.5066,
+      "step": 35570
+    },
+    {
+      "epoch": 3.558,
+      "grad_norm": 8.793334007263184,
+      "learning_rate": 2.8876000000000005e-06,
+      "loss": 0.5561,
+      "step": 35580
+    },
+    {
+      "epoch": 3.559,
+      "grad_norm": 28.917333602905273,
+      "learning_rate": 2.8856e-06,
+      "loss": 0.7647,
+      "step": 35590
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 36.710777282714844,
+      "learning_rate": 2.8836000000000002e-06,
+      "loss": 0.7894,
+      "step": 35600
+    },
+    {
+      "epoch": 3.561,
+      "grad_norm": 65.8742446899414,
+      "learning_rate": 2.8816000000000005e-06,
+      "loss": 0.7871,
+      "step": 35610
+    },
+    {
+      "epoch": 3.5620000000000003,
+      "grad_norm": 48.99974060058594,
+      "learning_rate": 2.8796000000000003e-06,
+      "loss": 0.6941,
+      "step": 35620
+    },
+    {
+      "epoch": 3.5629999999999997,
+      "grad_norm": 53.21778106689453,
+      "learning_rate": 2.8776e-06,
+      "loss": 0.8488,
+      "step": 35630
+    },
+    {
+      "epoch": 3.564,
+      "grad_norm": 18.24955177307129,
+      "learning_rate": 2.8756e-06,
+      "loss": 0.7039,
+      "step": 35640
+    },
+    {
+      "epoch": 3.565,
+      "grad_norm": 11.103310585021973,
+      "learning_rate": 2.8736000000000003e-06,
+      "loss": 0.5776,
+      "step": 35650
+    },
+    {
+      "epoch": 3.566,
+      "grad_norm": 21.356523513793945,
+      "learning_rate": 2.8716000000000006e-06,
+      "loss": 0.76,
+      "step": 35660
+    },
+    {
+      "epoch": 3.567,
+      "grad_norm": 61.79388427734375,
+      "learning_rate": 2.8696e-06,
+      "loss": 0.4611,
+      "step": 35670
+    },
+    {
+      "epoch": 3.568,
+      "grad_norm": 40.665496826171875,
+      "learning_rate": 2.8676000000000003e-06,
+      "loss": 0.4536,
+      "step": 35680
+    },
+    {
+      "epoch": 3.569,
+      "grad_norm": 42.91267776489258,
+      "learning_rate": 2.8656e-06,
+      "loss": 0.7923,
+      "step": 35690
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 47.55757522583008,
+      "learning_rate": 2.8636000000000004e-06,
+      "loss": 0.924,
+      "step": 35700
+    },
+    {
+      "epoch": 3.5709999999999997,
+      "grad_norm": 51.175811767578125,
+      "learning_rate": 2.8616000000000007e-06,
+      "loss": 0.9656,
+      "step": 35710
+    },
+    {
+      "epoch": 3.572,
+      "grad_norm": 68.177734375,
+      "learning_rate": 2.8596e-06,
+      "loss": 0.645,
+      "step": 35720
+    },
+    {
+      "epoch": 3.573,
+      "grad_norm": 0.8722525835037231,
+      "learning_rate": 2.8576000000000004e-06,
+      "loss": 0.6464,
+      "step": 35730
+    },
+    {
+      "epoch": 3.574,
+      "grad_norm": 61.88508605957031,
+      "learning_rate": 2.8556000000000002e-06,
+      "loss": 0.6672,
+      "step": 35740
+    },
+    {
+      "epoch": 3.575,
+      "grad_norm": 31.792068481445312,
+      "learning_rate": 2.8536e-06,
+      "loss": 0.7106,
+      "step": 35750
+    },
+    {
+      "epoch": 3.576,
+      "grad_norm": 34.22478485107422,
+      "learning_rate": 2.8516000000000004e-06,
+      "loss": 0.5765,
+      "step": 35760
+    },
+    {
+      "epoch": 3.577,
+      "grad_norm": 31.679950714111328,
+      "learning_rate": 2.8496e-06,
+      "loss": 0.5261,
+      "step": 35770
+    },
+    {
+      "epoch": 3.578,
+      "grad_norm": 17.40015983581543,
+      "learning_rate": 2.8476000000000005e-06,
+      "loss": 0.5874,
+      "step": 35780
+    },
+    {
+      "epoch": 3.5789999999999997,
+      "grad_norm": 15.593846321105957,
+      "learning_rate": 2.8456e-06,
+      "loss": 0.4362,
+      "step": 35790
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 27.332752227783203,
+      "learning_rate": 2.8436e-06,
+      "loss": 0.6304,
+      "step": 35800
+    },
+    {
+      "epoch": 3.581,
+      "grad_norm": 26.46154022216797,
+      "learning_rate": 2.8416000000000005e-06,
+      "loss": 0.6574,
+      "step": 35810
+    },
+    {
+      "epoch": 3.582,
+      "grad_norm": 22.900833129882812,
+      "learning_rate": 2.8396000000000003e-06,
+      "loss": 0.6414,
+      "step": 35820
+    },
+    {
+      "epoch": 3.583,
+      "grad_norm": 40.177005767822266,
+      "learning_rate": 2.8376e-06,
+      "loss": 0.6131,
+      "step": 35830
+    },
+    {
+      "epoch": 3.584,
+      "grad_norm": 29.940670013427734,
+      "learning_rate": 2.8356e-06,
+      "loss": 0.7146,
+      "step": 35840
+    },
+    {
+      "epoch": 3.585,
+      "grad_norm": 40.88385772705078,
+      "learning_rate": 2.8336000000000003e-06,
+      "loss": 0.4976,
+      "step": 35850
+    },
+    {
+      "epoch": 3.586,
+      "grad_norm": 20.683185577392578,
+      "learning_rate": 2.8316000000000006e-06,
+      "loss": 0.9411,
+      "step": 35860
+    },
+    {
+      "epoch": 3.5869999999999997,
+      "grad_norm": 4.140332221984863,
+      "learning_rate": 2.8296e-06,
+      "loss": 0.9062,
+      "step": 35870
+    },
+    {
+      "epoch": 3.588,
+      "grad_norm": 46.50490188598633,
+      "learning_rate": 2.8276000000000003e-06,
+      "loss": 0.7577,
+      "step": 35880
+    },
+    {
+      "epoch": 3.589,
+      "grad_norm": 38.72293472290039,
+      "learning_rate": 2.8256e-06,
+      "loss": 0.8583,
+      "step": 35890
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 37.587318420410156,
+      "learning_rate": 2.8236000000000004e-06,
+      "loss": 0.6687,
+      "step": 35900
+    },
+    {
+      "epoch": 3.591,
+      "grad_norm": 42.93622970581055,
+      "learning_rate": 2.8216000000000006e-06,
+      "loss": 0.7726,
+      "step": 35910
+    },
+    {
+      "epoch": 3.592,
+      "grad_norm": 5.569044589996338,
+      "learning_rate": 2.8196e-06,
+      "loss": 0.4084,
+      "step": 35920
+    },
+    {
+      "epoch": 3.593,
+      "grad_norm": 18.586912155151367,
+      "learning_rate": 2.8176000000000003e-06,
+      "loss": 0.6104,
+      "step": 35930
+    },
+    {
+      "epoch": 3.594,
+      "grad_norm": 50.00092697143555,
+      "learning_rate": 2.8156e-06,
+      "loss": 0.5622,
+      "step": 35940
+    },
+    {
+      "epoch": 3.5949999999999998,
+      "grad_norm": 30.04667091369629,
+      "learning_rate": 2.8136e-06,
+      "loss": 0.5719,
+      "step": 35950
+    },
+    {
+      "epoch": 3.596,
+      "grad_norm": 13.743998527526855,
+      "learning_rate": 2.8116000000000003e-06,
+      "loss": 0.5752,
+      "step": 35960
+    },
+    {
+      "epoch": 3.597,
+      "grad_norm": 39.8201904296875,
+      "learning_rate": 2.8096e-06,
+      "loss": 0.4261,
+      "step": 35970
+    },
+    {
+      "epoch": 3.598,
+      "grad_norm": 31.560199737548828,
+      "learning_rate": 2.8076000000000004e-06,
+      "loss": 0.5454,
+      "step": 35980
+    },
+    {
+      "epoch": 3.599,
+      "grad_norm": 55.95412063598633,
+      "learning_rate": 2.8056e-06,
+      "loss": 0.7822,
+      "step": 35990
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 27.65061378479004,
+      "learning_rate": 2.8036e-06,
+      "loss": 0.5083,
+      "step": 36000
+    },
+    {
+      "epoch": 3.601,
+      "grad_norm": 33.482330322265625,
+      "learning_rate": 2.8016000000000004e-06,
+      "loss": 0.5487,
+      "step": 36010
+    },
+    {
+      "epoch": 3.602,
+      "grad_norm": 56.06214904785156,
+      "learning_rate": 2.7996000000000003e-06,
+      "loss": 0.6015,
+      "step": 36020
+    },
+    {
+      "epoch": 3.6029999999999998,
+      "grad_norm": 26.832622528076172,
+      "learning_rate": 2.7976e-06,
+      "loss": 0.5467,
+      "step": 36030
+    },
+    {
+      "epoch": 3.604,
+      "grad_norm": 9.77216625213623,
+      "learning_rate": 2.7956e-06,
+      "loss": 0.3149,
+      "step": 36040
+    },
+    {
+      "epoch": 3.605,
+      "grad_norm": 45.25501251220703,
+      "learning_rate": 2.7936000000000002e-06,
+      "loss": 0.5729,
+      "step": 36050
+    },
+    {
+      "epoch": 3.606,
+      "grad_norm": 47.72624969482422,
+      "learning_rate": 2.7916000000000005e-06,
+      "loss": 0.7686,
+      "step": 36060
+    },
+    {
+      "epoch": 3.607,
+      "grad_norm": 37.49579620361328,
+      "learning_rate": 2.7896e-06,
+      "loss": 0.473,
+      "step": 36070
+    },
+    {
+      "epoch": 3.608,
+      "grad_norm": 34.11646270751953,
+      "learning_rate": 2.7876000000000002e-06,
+      "loss": 0.5037,
+      "step": 36080
+    },
+    {
+      "epoch": 3.609,
+      "grad_norm": 7.480967044830322,
+      "learning_rate": 2.7856e-06,
+      "loss": 0.4404,
+      "step": 36090
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 12.523974418640137,
+      "learning_rate": 2.7836000000000003e-06,
+      "loss": 0.5428,
+      "step": 36100
+    },
+    {
+      "epoch": 3.6109999999999998,
+      "grad_norm": 27.516239166259766,
+      "learning_rate": 2.7816000000000006e-06,
+      "loss": 0.4199,
+      "step": 36110
+    },
+    {
+      "epoch": 3.612,
+      "grad_norm": 14.997575759887695,
+      "learning_rate": 2.7796e-06,
+      "loss": 0.3202,
+      "step": 36120
+    },
+    {
+      "epoch": 3.613,
+      "grad_norm": 23.749460220336914,
+      "learning_rate": 2.7776000000000003e-06,
+      "loss": 0.5419,
+      "step": 36130
+    },
+    {
+      "epoch": 3.614,
+      "grad_norm": 18.59351348876953,
+      "learning_rate": 2.7756e-06,
+      "loss": 0.5096,
+      "step": 36140
+    },
+    {
+      "epoch": 3.615,
+      "grad_norm": 36.554222106933594,
+      "learning_rate": 2.7736e-06,
+      "loss": 0.4627,
+      "step": 36150
+    },
+    {
+      "epoch": 3.616,
+      "grad_norm": 21.801746368408203,
+      "learning_rate": 2.7716000000000003e-06,
+      "loss": 0.5231,
+      "step": 36160
+    },
+    {
+      "epoch": 3.617,
+      "grad_norm": 9.660021781921387,
+      "learning_rate": 2.7696e-06,
+      "loss": 0.6033,
+      "step": 36170
+    },
+    {
+      "epoch": 3.618,
+      "grad_norm": 52.21128845214844,
+      "learning_rate": 2.7676000000000004e-06,
+      "loss": 0.679,
+      "step": 36180
+    },
+    {
+      "epoch": 3.6189999999999998,
+      "grad_norm": 56.615543365478516,
+      "learning_rate": 2.7656e-06,
+      "loss": 0.4759,
+      "step": 36190
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 17.087657928466797,
+      "learning_rate": 2.7636e-06,
+      "loss": 0.6594,
+      "step": 36200
+    },
+    {
+      "epoch": 3.621,
+      "grad_norm": 1.8077119588851929,
+      "learning_rate": 2.7616000000000004e-06,
+      "loss": 0.456,
+      "step": 36210
+    },
+    {
+      "epoch": 3.622,
+      "grad_norm": 53.12010192871094,
+      "learning_rate": 2.7596000000000002e-06,
+      "loss": 0.3854,
+      "step": 36220
+    },
+    {
+      "epoch": 3.623,
+      "grad_norm": 46.78509521484375,
+      "learning_rate": 2.7576e-06,
+      "loss": 0.5449,
+      "step": 36230
+    },
+    {
+      "epoch": 3.624,
+      "grad_norm": 50.42572021484375,
+      "learning_rate": 2.7556000000000003e-06,
+      "loss": 0.6184,
+      "step": 36240
+    },
+    {
+      "epoch": 3.625,
+      "grad_norm": 40.22068405151367,
+      "learning_rate": 2.7536e-06,
+      "loss": 0.5953,
+      "step": 36250
+    },
+    {
+      "epoch": 3.626,
+      "grad_norm": 29.53036117553711,
+      "learning_rate": 2.7516000000000005e-06,
+      "loss": 0.499,
+      "step": 36260
+    },
+    {
+      "epoch": 3.627,
+      "grad_norm": 18.926544189453125,
+      "learning_rate": 2.7496e-06,
+      "loss": 0.3777,
+      "step": 36270
+    },
+    {
+      "epoch": 3.628,
+      "grad_norm": 40.2806282043457,
+      "learning_rate": 2.7476e-06,
+      "loss": 0.6982,
+      "step": 36280
+    },
+    {
+      "epoch": 3.629,
+      "grad_norm": 20.797760009765625,
+      "learning_rate": 2.7456000000000004e-06,
+      "loss": 0.7362,
+      "step": 36290
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 6.7826032638549805,
+      "learning_rate": 2.7436000000000003e-06,
+      "loss": 0.5419,
+      "step": 36300
+    },
+    {
+      "epoch": 3.6310000000000002,
+      "grad_norm": 6.89370059967041,
+      "learning_rate": 2.7416000000000006e-06,
+      "loss": 0.6738,
+      "step": 36310
+    },
+    {
+      "epoch": 3.632,
+      "grad_norm": 18.179012298583984,
+      "learning_rate": 2.7396e-06,
+      "loss": 0.5437,
+      "step": 36320
+    },
+    {
+      "epoch": 3.633,
+      "grad_norm": 60.44831848144531,
+      "learning_rate": 2.7376000000000003e-06,
+      "loss": 0.4036,
+      "step": 36330
+    },
+    {
+      "epoch": 3.634,
+      "grad_norm": 26.54794692993164,
+      "learning_rate": 2.7356000000000005e-06,
+      "loss": 0.3875,
+      "step": 36340
+    },
+    {
+      "epoch": 3.635,
+      "grad_norm": 36.836124420166016,
+      "learning_rate": 2.7336e-06,
+      "loss": 0.5515,
+      "step": 36350
+    },
+    {
+      "epoch": 3.636,
+      "grad_norm": 66.89555358886719,
+      "learning_rate": 2.7316000000000002e-06,
+      "loss": 0.5732,
+      "step": 36360
+    },
+    {
+      "epoch": 3.637,
+      "grad_norm": 43.293033599853516,
+      "learning_rate": 2.7296e-06,
+      "loss": 0.5602,
+      "step": 36370
+    },
+    {
+      "epoch": 3.638,
+      "grad_norm": 31.94192886352539,
+      "learning_rate": 2.7276000000000004e-06,
+      "loss": 0.522,
+      "step": 36380
+    },
+    {
+      "epoch": 3.6390000000000002,
+      "grad_norm": 31.9433650970459,
+      "learning_rate": 2.7256000000000006e-06,
+      "loss": 0.4046,
+      "step": 36390
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 49.40196990966797,
+      "learning_rate": 2.7236e-06,
+      "loss": 0.7042,
+      "step": 36400
+    },
+    {
+      "epoch": 3.641,
+      "grad_norm": 33.30634307861328,
+      "learning_rate": 2.7216000000000003e-06,
+      "loss": 0.6307,
+      "step": 36410
+    },
+    {
+      "epoch": 3.642,
+      "grad_norm": 55.43235778808594,
+      "learning_rate": 2.7196e-06,
+      "loss": 0.5498,
+      "step": 36420
+    },
+    {
+      "epoch": 3.643,
+      "grad_norm": 45.17135238647461,
+      "learning_rate": 2.7176e-06,
+      "loss": 0.5377,
+      "step": 36430
+    },
+    {
+      "epoch": 3.644,
+      "grad_norm": 52.483306884765625,
+      "learning_rate": 2.7156000000000003e-06,
+      "loss": 0.5028,
+      "step": 36440
+    },
+    {
+      "epoch": 3.645,
+      "grad_norm": 48.63310623168945,
+      "learning_rate": 2.7136e-06,
+      "loss": 0.6501,
+      "step": 36450
+    },
+    {
+      "epoch": 3.646,
+      "grad_norm": 32.639774322509766,
+      "learning_rate": 2.7116000000000004e-06,
+      "loss": 0.7266,
+      "step": 36460
+    },
+    {
+      "epoch": 3.6470000000000002,
+      "grad_norm": 12.3084135055542,
+      "learning_rate": 2.7096e-06,
+      "loss": 0.56,
+      "step": 36470
+    },
+    {
+      "epoch": 3.648,
+      "grad_norm": 3.4930202960968018,
+      "learning_rate": 2.7076e-06,
+      "loss": 0.6945,
+      "step": 36480
+    },
+    {
+      "epoch": 3.649,
+      "grad_norm": 55.61498260498047,
+      "learning_rate": 2.7056000000000004e-06,
+      "loss": 0.3165,
+      "step": 36490
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 56.448055267333984,
+      "learning_rate": 2.7036000000000003e-06,
+      "loss": 0.3924,
+      "step": 36500
+    },
+    {
+      "epoch": 3.651,
+      "grad_norm": 7.915106773376465,
+      "learning_rate": 2.7016000000000005e-06,
+      "loss": 0.4978,
+      "step": 36510
+    },
+    {
+      "epoch": 3.652,
+      "grad_norm": 13.365342140197754,
+      "learning_rate": 2.6996e-06,
+      "loss": 0.5635,
+      "step": 36520
+    },
+    {
+      "epoch": 3.653,
+      "grad_norm": 19.216312408447266,
+      "learning_rate": 2.6976000000000002e-06,
+      "loss": 0.4758,
+      "step": 36530
+    },
+    {
+      "epoch": 3.654,
+      "grad_norm": 45.420921325683594,
+      "learning_rate": 2.6956000000000005e-06,
+      "loss": 0.7895,
+      "step": 36540
+    },
+    {
+      "epoch": 3.6550000000000002,
+      "grad_norm": 78.061279296875,
+      "learning_rate": 2.6936e-06,
+      "loss": 0.4275,
+      "step": 36550
+    },
+    {
+      "epoch": 3.656,
+      "grad_norm": 1.6079028844833374,
+      "learning_rate": 2.6916e-06,
+      "loss": 0.5539,
+      "step": 36560
+    },
+    {
+      "epoch": 3.657,
+      "grad_norm": 58.142704010009766,
+      "learning_rate": 2.6896e-06,
+      "loss": 0.9741,
+      "step": 36570
+    },
+    {
+      "epoch": 3.658,
+      "grad_norm": 8.05836009979248,
+      "learning_rate": 2.6876000000000003e-06,
+      "loss": 0.4898,
+      "step": 36580
+    },
+    {
+      "epoch": 3.659,
+      "grad_norm": 31.095792770385742,
+      "learning_rate": 2.6856000000000006e-06,
+      "loss": 0.5192,
+      "step": 36590
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 29.871849060058594,
+      "learning_rate": 2.6836e-06,
+      "loss": 0.7613,
+      "step": 36600
+    },
+    {
+      "epoch": 3.661,
+      "grad_norm": 63.34565353393555,
+      "learning_rate": 2.6816000000000003e-06,
+      "loss": 0.5743,
+      "step": 36610
+    },
+    {
+      "epoch": 3.662,
+      "grad_norm": 7.080972671508789,
+      "learning_rate": 2.6796e-06,
+      "loss": 0.7588,
+      "step": 36620
+    },
+    {
+      "epoch": 3.6630000000000003,
+      "grad_norm": 28.010923385620117,
+      "learning_rate": 2.6776e-06,
+      "loss": 0.4268,
+      "step": 36630
+    },
+    {
+      "epoch": 3.664,
+      "grad_norm": 50.9543342590332,
+      "learning_rate": 2.6756000000000003e-06,
+      "loss": 0.6525,
+      "step": 36640
+    },
+    {
+      "epoch": 3.665,
+      "grad_norm": 2.2711169719696045,
+      "learning_rate": 2.6736e-06,
+      "loss": 0.4292,
+      "step": 36650
+    },
+    {
+      "epoch": 3.666,
+      "grad_norm": 58.64751052856445,
+      "learning_rate": 2.6716000000000004e-06,
+      "loss": 0.7755,
+      "step": 36660
+    },
+    {
+      "epoch": 3.667,
+      "grad_norm": 40.33562088012695,
+      "learning_rate": 2.6696e-06,
+      "loss": 0.6762,
+      "step": 36670
+    },
+    {
+      "epoch": 3.668,
+      "grad_norm": 8.384110450744629,
+      "learning_rate": 2.6676e-06,
+      "loss": 0.5171,
+      "step": 36680
+    },
+    {
+      "epoch": 3.669,
+      "grad_norm": 29.39927101135254,
+      "learning_rate": 2.6656000000000004e-06,
+      "loss": 0.5208,
+      "step": 36690
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 35.47958755493164,
+      "learning_rate": 2.6636000000000002e-06,
+      "loss": 0.4768,
+      "step": 36700
+    },
+    {
+      "epoch": 3.6710000000000003,
+      "grad_norm": 49.37161636352539,
+      "learning_rate": 2.6616000000000005e-06,
+      "loss": 0.7873,
+      "step": 36710
+    },
+    {
+      "epoch": 3.672,
+      "grad_norm": 53.93315887451172,
+      "learning_rate": 2.6596e-06,
+      "loss": 0.559,
+      "step": 36720
+    },
+    {
+      "epoch": 3.673,
+      "grad_norm": 32.21564483642578,
+      "learning_rate": 2.6576e-06,
+      "loss": 0.5429,
+      "step": 36730
+    },
+    {
+      "epoch": 3.674,
+      "grad_norm": 0.6642956137657166,
+      "learning_rate": 2.6556000000000005e-06,
+      "loss": 0.5247,
+      "step": 36740
+    },
+    {
+      "epoch": 3.675,
+      "grad_norm": 43.1329460144043,
+      "learning_rate": 2.6536e-06,
+      "loss": 0.6105,
+      "step": 36750
+    },
+    {
+      "epoch": 3.676,
+      "grad_norm": 71.51045989990234,
+      "learning_rate": 2.6516e-06,
+      "loss": 0.6286,
+      "step": 36760
+    },
+    {
+      "epoch": 3.677,
+      "grad_norm": 35.833473205566406,
+      "learning_rate": 2.6496e-06,
+      "loss": 0.3915,
+      "step": 36770
+    },
+    {
+      "epoch": 3.678,
+      "grad_norm": 35.758827209472656,
+      "learning_rate": 2.6476000000000003e-06,
+      "loss": 0.4121,
+      "step": 36780
+    },
+    {
+      "epoch": 3.6790000000000003,
+      "grad_norm": 38.435794830322266,
+      "learning_rate": 2.6456000000000006e-06,
+      "loss": 0.6117,
+      "step": 36790
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 66.92293548583984,
+      "learning_rate": 2.6436e-06,
+      "loss": 0.6519,
+      "step": 36800
+    },
+    {
+      "epoch": 3.681,
+      "grad_norm": 60.6037712097168,
+      "learning_rate": 2.6416000000000003e-06,
+      "loss": 0.5819,
+      "step": 36810
+    },
+    {
+      "epoch": 3.682,
+      "grad_norm": 79.33817291259766,
+      "learning_rate": 2.6396e-06,
+      "loss": 0.7536,
+      "step": 36820
+    },
+    {
+      "epoch": 3.683,
+      "grad_norm": 28.001657485961914,
+      "learning_rate": 2.6376e-06,
+      "loss": 0.5932,
+      "step": 36830
+    },
+    {
+      "epoch": 3.684,
+      "grad_norm": 25.109481811523438,
+      "learning_rate": 2.6356000000000002e-06,
+      "loss": 0.5357,
+      "step": 36840
+    },
+    {
+      "epoch": 3.685,
+      "grad_norm": 36.05546188354492,
+      "learning_rate": 2.6336e-06,
+      "loss": 0.7685,
+      "step": 36850
+    },
+    {
+      "epoch": 3.686,
+      "grad_norm": 14.421683311462402,
+      "learning_rate": 2.6316000000000004e-06,
+      "loss": 0.7285,
+      "step": 36860
+    },
+    {
+      "epoch": 3.6870000000000003,
+      "grad_norm": 41.31953048706055,
+      "learning_rate": 2.6296000000000006e-06,
+      "loss": 0.7947,
+      "step": 36870
+    },
+    {
+      "epoch": 3.6879999999999997,
+      "grad_norm": 18.700454711914062,
+      "learning_rate": 2.6276e-06,
+      "loss": 0.8291,
+      "step": 36880
+    },
+    {
+      "epoch": 3.689,
+      "grad_norm": 47.96348190307617,
+      "learning_rate": 2.6256000000000003e-06,
+      "loss": 0.5966,
+      "step": 36890
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 26.686315536499023,
+      "learning_rate": 2.6236e-06,
+      "loss": 0.6098,
+      "step": 36900
+    },
+    {
+      "epoch": 3.691,
+      "grad_norm": 47.52505111694336,
+      "learning_rate": 2.6216000000000004e-06,
+      "loss": 0.7313,
+      "step": 36910
+    },
+    {
+      "epoch": 3.692,
+      "grad_norm": 75.91033935546875,
+      "learning_rate": 2.6196000000000003e-06,
+      "loss": 0.6876,
+      "step": 36920
+    },
+    {
+      "epoch": 3.693,
+      "grad_norm": 57.06541442871094,
+      "learning_rate": 2.6176e-06,
+      "loss": 0.5015,
+      "step": 36930
+    },
+    {
+      "epoch": 3.694,
+      "grad_norm": 2.7463643550872803,
+      "learning_rate": 2.6156000000000004e-06,
+      "loss": 0.5357,
+      "step": 36940
+    },
+    {
+      "epoch": 3.695,
+      "grad_norm": 92.55400848388672,
+      "learning_rate": 2.6136e-06,
+      "loss": 0.4267,
+      "step": 36950
+    },
+    {
+      "epoch": 3.6959999999999997,
+      "grad_norm": 47.57724380493164,
+      "learning_rate": 2.6116e-06,
+      "loss": 0.7616,
+      "step": 36960
+    },
+    {
+      "epoch": 3.697,
+      "grad_norm": 72.47662353515625,
+      "learning_rate": 2.6096000000000004e-06,
+      "loss": 0.6118,
+      "step": 36970
+    },
+    {
+      "epoch": 3.698,
+      "grad_norm": 66.8961410522461,
+      "learning_rate": 2.6076000000000002e-06,
+      "loss": 0.7714,
+      "step": 36980
+    },
+    {
+      "epoch": 3.699,
+      "grad_norm": 7.666031360626221,
+      "learning_rate": 2.6056000000000005e-06,
+      "loss": 0.9331,
+      "step": 36990
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 6.739872932434082,
+      "learning_rate": 2.6036e-06,
+      "loss": 0.6918,
+      "step": 37000
+    },
+    {
+      "epoch": 3.701,
+      "grad_norm": 30.467918395996094,
+      "learning_rate": 2.6016000000000002e-06,
+      "loss": 0.5595,
+      "step": 37010
+    },
+    {
+      "epoch": 3.702,
+      "grad_norm": 6.347413063049316,
+      "learning_rate": 2.5996000000000005e-06,
+      "loss": 0.3831,
+      "step": 37020
+    },
+    {
+      "epoch": 3.703,
+      "grad_norm": 37.63015365600586,
+      "learning_rate": 2.5976e-06,
+      "loss": 0.4942,
+      "step": 37030
+    },
+    {
+      "epoch": 3.7039999999999997,
+      "grad_norm": 28.430194854736328,
+      "learning_rate": 2.5956e-06,
+      "loss": 0.7804,
+      "step": 37040
+    },
+    {
+      "epoch": 3.705,
+      "grad_norm": 18.33820343017578,
+      "learning_rate": 2.5936e-06,
+      "loss": 0.5023,
+      "step": 37050
+    },
+    {
+      "epoch": 3.706,
+      "grad_norm": 40.74871826171875,
+      "learning_rate": 2.5916000000000003e-06,
+      "loss": 0.6542,
+      "step": 37060
+    },
+    {
+      "epoch": 3.707,
+      "grad_norm": 77.77648162841797,
+      "learning_rate": 2.5896000000000006e-06,
+      "loss": 0.5121,
+      "step": 37070
+    },
+    {
+      "epoch": 3.708,
+      "grad_norm": 70.61822509765625,
+      "learning_rate": 2.5876e-06,
+      "loss": 0.6368,
+      "step": 37080
+    },
+    {
+      "epoch": 3.709,
+      "grad_norm": 34.1216926574707,
+      "learning_rate": 2.5856000000000003e-06,
+      "loss": 0.5764,
+      "step": 37090
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 9.746734619140625,
+      "learning_rate": 2.5836e-06,
+      "loss": 0.6669,
+      "step": 37100
+    },
+    {
+      "epoch": 3.711,
+      "grad_norm": 31.46854591369629,
+      "learning_rate": 2.5816000000000004e-06,
+      "loss": 0.6934,
+      "step": 37110
+    },
+    {
+      "epoch": 3.7119999999999997,
+      "grad_norm": 28.925506591796875,
+      "learning_rate": 2.5796000000000003e-06,
+      "loss": 0.3579,
+      "step": 37120
+    },
+    {
+      "epoch": 3.713,
+      "grad_norm": 48.56563186645508,
+      "learning_rate": 2.5776e-06,
+      "loss": 0.5464,
+      "step": 37130
+    },
+    {
+      "epoch": 3.714,
+      "grad_norm": 48.421627044677734,
+      "learning_rate": 2.5756000000000004e-06,
+      "loss": 0.999,
+      "step": 37140
+    },
+    {
+      "epoch": 3.715,
+      "grad_norm": 35.519554138183594,
+      "learning_rate": 2.5736e-06,
+      "loss": 0.5991,
+      "step": 37150
+    },
+    {
+      "epoch": 3.716,
+      "grad_norm": 49.56758117675781,
+      "learning_rate": 2.5716e-06,
+      "loss": 0.6012,
+      "step": 37160
+    },
+    {
+      "epoch": 3.717,
+      "grad_norm": 32.166297912597656,
+      "learning_rate": 2.5696000000000004e-06,
+      "loss": 0.6095,
+      "step": 37170
+    },
+    {
+      "epoch": 3.718,
+      "grad_norm": 4.212522506713867,
+      "learning_rate": 2.5676e-06,
+      "loss": 0.5991,
+      "step": 37180
+    },
+    {
+      "epoch": 3.719,
+      "grad_norm": 23.69797706604004,
+      "learning_rate": 2.5656000000000005e-06,
+      "loss": 0.8162,
+      "step": 37190
+    },
+    {
+      "epoch": 3.7199999999999998,
+      "grad_norm": 36.765499114990234,
+      "learning_rate": 2.5636e-06,
+      "loss": 0.535,
+      "step": 37200
+    },
+    {
+      "epoch": 3.721,
+      "grad_norm": 27.32730484008789,
+      "learning_rate": 2.5616e-06,
+      "loss": 0.558,
+      "step": 37210
+    },
+    {
+      "epoch": 3.722,
+      "grad_norm": 0.6880661845207214,
+      "learning_rate": 2.5596000000000004e-06,
+      "loss": 0.5504,
+      "step": 37220
+    },
+    {
+      "epoch": 3.723,
+      "grad_norm": 37.06109619140625,
+      "learning_rate": 2.5576e-06,
+      "loss": 0.5269,
+      "step": 37230
+    },
+    {
+      "epoch": 3.724,
+      "grad_norm": 63.31182861328125,
+      "learning_rate": 2.5556e-06,
+      "loss": 0.627,
+      "step": 37240
+    },
+    {
+      "epoch": 3.725,
+      "grad_norm": 2.2652759552001953,
+      "learning_rate": 2.5536e-06,
+      "loss": 0.5961,
+      "step": 37250
+    },
+    {
+      "epoch": 3.726,
+      "grad_norm": 34.18130874633789,
+      "learning_rate": 2.5516000000000003e-06,
+      "loss": 0.5139,
+      "step": 37260
+    },
+    {
+      "epoch": 3.727,
+      "grad_norm": 17.9378662109375,
+      "learning_rate": 2.5496000000000005e-06,
+      "loss": 0.4794,
+      "step": 37270
+    },
+    {
+      "epoch": 3.7279999999999998,
+      "grad_norm": 5.379673004150391,
+      "learning_rate": 2.5476e-06,
+      "loss": 0.3321,
+      "step": 37280
+    },
+    {
+      "epoch": 3.729,
+      "grad_norm": 14.593806266784668,
+      "learning_rate": 2.5456000000000002e-06,
+      "loss": 0.4052,
+      "step": 37290
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 33.24942398071289,
+      "learning_rate": 2.5436e-06,
+      "loss": 0.8694,
+      "step": 37300
+    },
+    {
+      "epoch": 3.731,
+      "grad_norm": 11.167141914367676,
+      "learning_rate": 2.5416000000000004e-06,
+      "loss": 0.4076,
+      "step": 37310
+    },
+    {
+      "epoch": 3.732,
+      "grad_norm": 6.8190388679504395,
+      "learning_rate": 2.5396000000000002e-06,
+      "loss": 0.6243,
+      "step": 37320
+    },
+    {
+      "epoch": 3.733,
+      "grad_norm": 49.4874382019043,
+      "learning_rate": 2.5376e-06,
+      "loss": 0.3881,
+      "step": 37330
+    },
+    {
+      "epoch": 3.734,
+      "grad_norm": 10.311324119567871,
+      "learning_rate": 2.5356000000000003e-06,
+      "loss": 0.264,
+      "step": 37340
+    },
+    {
+      "epoch": 3.735,
+      "grad_norm": 10.175442695617676,
+      "learning_rate": 2.5335999999999998e-06,
+      "loss": 0.5226,
+      "step": 37350
+    },
+    {
+      "epoch": 3.7359999999999998,
+      "grad_norm": 25.542966842651367,
+      "learning_rate": 2.5316e-06,
+      "loss": 0.6076,
+      "step": 37360
+    },
+    {
+      "epoch": 3.737,
+      "grad_norm": 6.893240928649902,
+      "learning_rate": 2.5296000000000003e-06,
+      "loss": 0.4515,
+      "step": 37370
+    },
+    {
+      "epoch": 3.738,
+      "grad_norm": 12.86868667602539,
+      "learning_rate": 2.5276e-06,
+      "loss": 0.778,
+      "step": 37380
+    },
+    {
+      "epoch": 3.739,
+      "grad_norm": 51.52862548828125,
+      "learning_rate": 2.5256000000000004e-06,
+      "loss": 0.7111,
+      "step": 37390
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 50.969703674316406,
+      "learning_rate": 2.5236e-06,
+      "loss": 0.5213,
+      "step": 37400
+    },
+    {
+      "epoch": 3.741,
+      "grad_norm": 40.725242614746094,
+      "learning_rate": 2.5216e-06,
+      "loss": 0.5364,
+      "step": 37410
+    },
+    {
+      "epoch": 3.742,
+      "grad_norm": 36.41386795043945,
+      "learning_rate": 2.5196000000000004e-06,
+      "loss": 0.6574,
+      "step": 37420
+    },
+    {
+      "epoch": 3.743,
+      "grad_norm": 23.86066246032715,
+      "learning_rate": 2.5176e-06,
+      "loss": 0.3526,
+      "step": 37430
+    },
+    {
+      "epoch": 3.7439999999999998,
+      "grad_norm": 26.9005069732666,
+      "learning_rate": 2.5156e-06,
+      "loss": 0.5263,
+      "step": 37440
+    },
+    {
+      "epoch": 3.745,
+      "grad_norm": 53.60567855834961,
+      "learning_rate": 2.5136e-06,
+      "loss": 0.5625,
+      "step": 37450
+    },
+    {
+      "epoch": 3.746,
+      "grad_norm": 39.05552291870117,
+      "learning_rate": 2.5116000000000002e-06,
+      "loss": 1.0419,
+      "step": 37460
+    },
+    {
+      "epoch": 3.747,
+      "grad_norm": 57.375396728515625,
+      "learning_rate": 2.5096000000000005e-06,
+      "loss": 0.6052,
+      "step": 37470
+    },
+    {
+      "epoch": 3.748,
+      "grad_norm": 61.77552795410156,
+      "learning_rate": 2.5076e-06,
+      "loss": 0.6023,
+      "step": 37480
+    },
+    {
+      "epoch": 3.749,
+      "grad_norm": 4.251457691192627,
+      "learning_rate": 2.5056e-06,
+      "loss": 0.7081,
+      "step": 37490
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 10.436452865600586,
+      "learning_rate": 2.5036000000000005e-06,
+      "loss": 0.5283,
+      "step": 37500
+    },
+    {
+      "epoch": 3.751,
+      "grad_norm": 6.220841884613037,
+      "learning_rate": 2.5016000000000003e-06,
+      "loss": 0.8396,
+      "step": 37510
+    },
+    {
+      "epoch": 3.752,
+      "grad_norm": 46.980194091796875,
+      "learning_rate": 2.4996e-06,
+      "loss": 0.5082,
+      "step": 37520
+    },
+    {
+      "epoch": 3.753,
+      "grad_norm": 26.87305450439453,
+      "learning_rate": 2.4976000000000004e-06,
+      "loss": 0.4748,
+      "step": 37530
+    },
+    {
+      "epoch": 3.754,
+      "grad_norm": 118.64867401123047,
+      "learning_rate": 2.4956000000000003e-06,
+      "loss": 0.7019,
+      "step": 37540
+    },
+    {
+      "epoch": 3.755,
+      "grad_norm": 33.75884246826172,
+      "learning_rate": 2.4936e-06,
+      "loss": 0.6893,
+      "step": 37550
+    },
+    {
+      "epoch": 3.7560000000000002,
+      "grad_norm": 6.014034271240234,
+      "learning_rate": 2.4916e-06,
+      "loss": 0.4001,
+      "step": 37560
+    },
+    {
+      "epoch": 3.757,
+      "grad_norm": 66.10906982421875,
+      "learning_rate": 2.4896000000000003e-06,
+      "loss": 0.6762,
+      "step": 37570
+    },
+    {
+      "epoch": 3.758,
+      "grad_norm": 55.02159881591797,
+      "learning_rate": 2.4876e-06,
+      "loss": 0.6555,
+      "step": 37580
+    },
+    {
+      "epoch": 3.759,
+      "grad_norm": 5.82243013381958,
+      "learning_rate": 2.4856000000000004e-06,
+      "loss": 0.4801,
+      "step": 37590
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 50.23249053955078,
+      "learning_rate": 2.4836000000000002e-06,
+      "loss": 0.5579,
+      "step": 37600
+    },
+    {
+      "epoch": 3.761,
+      "grad_norm": 5.414667129516602,
+      "learning_rate": 2.4816e-06,
+      "loss": 0.7534,
+      "step": 37610
+    },
+    {
+      "epoch": 3.762,
+      "grad_norm": 47.884521484375,
+      "learning_rate": 2.4796000000000004e-06,
+      "loss": 0.9884,
+      "step": 37620
+    },
+    {
+      "epoch": 3.763,
+      "grad_norm": 59.99955749511719,
+      "learning_rate": 2.4776000000000002e-06,
+      "loss": 0.8165,
+      "step": 37630
+    },
+    {
+      "epoch": 3.7640000000000002,
+      "grad_norm": 30.009929656982422,
+      "learning_rate": 2.4756e-06,
+      "loss": 0.5666,
+      "step": 37640
+    },
+    {
+      "epoch": 3.765,
+      "grad_norm": 57.524356842041016,
+      "learning_rate": 2.4736000000000003e-06,
+      "loss": 0.6281,
+      "step": 37650
+    },
+    {
+      "epoch": 3.766,
+      "grad_norm": 5.745800018310547,
+      "learning_rate": 2.4716e-06,
+      "loss": 0.6068,
+      "step": 37660
+    },
+    {
+      "epoch": 3.767,
+      "grad_norm": 52.89751052856445,
+      "learning_rate": 2.4696000000000005e-06,
+      "loss": 0.4946,
+      "step": 37670
+    },
+    {
+      "epoch": 3.768,
+      "grad_norm": 60.40765380859375,
+      "learning_rate": 2.4676000000000003e-06,
+      "loss": 0.5985,
+      "step": 37680
+    },
+    {
+      "epoch": 3.769,
+      "grad_norm": 33.4071159362793,
+      "learning_rate": 2.4656e-06,
+      "loss": 0.4445,
+      "step": 37690
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 7.831329345703125,
+      "learning_rate": 2.4636e-06,
+      "loss": 0.4106,
+      "step": 37700
+    },
+    {
+      "epoch": 3.771,
+      "grad_norm": 31.423812866210938,
+      "learning_rate": 2.4616000000000003e-06,
+      "loss": 0.6481,
+      "step": 37710
+    },
+    {
+      "epoch": 3.7720000000000002,
+      "grad_norm": 58.97570037841797,
+      "learning_rate": 2.4596e-06,
+      "loss": 0.6293,
+      "step": 37720
+    },
+    {
+      "epoch": 3.773,
+      "grad_norm": 45.02798843383789,
+      "learning_rate": 2.4576000000000004e-06,
+      "loss": 0.4715,
+      "step": 37730
+    },
+    {
+      "epoch": 3.774,
+      "grad_norm": 44.902523040771484,
+      "learning_rate": 2.4556000000000003e-06,
+      "loss": 0.6748,
+      "step": 37740
+    },
+    {
+      "epoch": 3.775,
+      "grad_norm": 25.11823081970215,
+      "learning_rate": 2.4536e-06,
+      "loss": 0.5346,
+      "step": 37750
+    },
+    {
+      "epoch": 3.776,
+      "grad_norm": 24.013734817504883,
+      "learning_rate": 2.4516e-06,
+      "loss": 0.776,
+      "step": 37760
+    },
+    {
+      "epoch": 3.777,
+      "grad_norm": 38.328773498535156,
+      "learning_rate": 2.4496000000000002e-06,
+      "loss": 0.7096,
+      "step": 37770
+    },
+    {
+      "epoch": 3.778,
+      "grad_norm": 27.20368003845215,
+      "learning_rate": 2.4476e-06,
+      "loss": 0.7903,
+      "step": 37780
+    },
+    {
+      "epoch": 3.779,
+      "grad_norm": 20.354225158691406,
+      "learning_rate": 2.4456000000000004e-06,
+      "loss": 0.9121,
+      "step": 37790
+    },
+    {
+      "epoch": 3.7800000000000002,
+      "grad_norm": 32.40660095214844,
+      "learning_rate": 2.4436e-06,
+      "loss": 0.5785,
+      "step": 37800
+    },
+    {
+      "epoch": 3.781,
+      "grad_norm": 47.86864471435547,
+      "learning_rate": 2.4416e-06,
+      "loss": 0.7947,
+      "step": 37810
+    },
+    {
+      "epoch": 3.782,
+      "grad_norm": 3.703575372695923,
+      "learning_rate": 2.4396000000000003e-06,
+      "loss": 0.5196,
+      "step": 37820
+    },
+    {
+      "epoch": 3.783,
+      "grad_norm": 44.8717041015625,
+      "learning_rate": 2.4376e-06,
+      "loss": 0.6596,
+      "step": 37830
+    },
+    {
+      "epoch": 3.784,
+      "grad_norm": 37.06995391845703,
+      "learning_rate": 2.4356e-06,
+      "loss": 0.7149,
+      "step": 37840
+    },
+    {
+      "epoch": 3.785,
+      "grad_norm": 21.56605339050293,
+      "learning_rate": 2.4336000000000003e-06,
+      "loss": 0.4211,
+      "step": 37850
+    },
+    {
+      "epoch": 3.786,
+      "grad_norm": 21.40250015258789,
+      "learning_rate": 2.4316e-06,
+      "loss": 0.5447,
+      "step": 37860
+    },
+    {
+      "epoch": 3.787,
+      "grad_norm": 64.63874053955078,
+      "learning_rate": 2.4296000000000004e-06,
+      "loss": 0.4935,
+      "step": 37870
+    },
+    {
+      "epoch": 3.7880000000000003,
+      "grad_norm": 50.10771179199219,
+      "learning_rate": 2.4276000000000003e-06,
+      "loss": 0.6575,
+      "step": 37880
+    },
+    {
+      "epoch": 3.789,
+      "grad_norm": 21.860170364379883,
+      "learning_rate": 2.4256e-06,
+      "loss": 0.6575,
+      "step": 37890
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 28.551694869995117,
+      "learning_rate": 2.4236e-06,
+      "loss": 0.5472,
+      "step": 37900
+    },
+    {
+      "epoch": 3.791,
+      "grad_norm": 51.53792190551758,
+      "learning_rate": 2.4216000000000002e-06,
+      "loss": 0.3918,
+      "step": 37910
+    },
+    {
+      "epoch": 3.792,
+      "grad_norm": 2.6405768394470215,
+      "learning_rate": 2.4196e-06,
+      "loss": 0.5887,
+      "step": 37920
+    },
+    {
+      "epoch": 3.793,
+      "grad_norm": 37.74142837524414,
+      "learning_rate": 2.4176000000000004e-06,
+      "loss": 0.9124,
+      "step": 37930
+    },
+    {
+      "epoch": 3.794,
+      "grad_norm": 1.126764178276062,
+      "learning_rate": 2.4156000000000002e-06,
+      "loss": 0.3395,
+      "step": 37940
+    },
+    {
+      "epoch": 3.795,
+      "grad_norm": 50.959747314453125,
+      "learning_rate": 2.4136e-06,
+      "loss": 0.5296,
+      "step": 37950
+    },
+    {
+      "epoch": 3.7960000000000003,
+      "grad_norm": 53.771331787109375,
+      "learning_rate": 2.4116000000000003e-06,
+      "loss": 0.6108,
+      "step": 37960
+    },
+    {
+      "epoch": 3.797,
+      "grad_norm": 75.44929504394531,
+      "learning_rate": 2.4096e-06,
+      "loss": 1.0028,
+      "step": 37970
+    },
+    {
+      "epoch": 3.798,
+      "grad_norm": 14.876928329467773,
+      "learning_rate": 2.4076e-06,
+      "loss": 0.4708,
+      "step": 37980
+    },
+    {
+      "epoch": 3.799,
+      "grad_norm": 56.54835510253906,
+      "learning_rate": 2.4056000000000003e-06,
+      "loss": 0.4433,
+      "step": 37990
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 55.80618667602539,
+      "learning_rate": 2.4036e-06,
+      "loss": 0.4473,
+      "step": 38000
+    },
+    {
+      "epoch": 3.801,
+      "grad_norm": 32.187435150146484,
+      "learning_rate": 2.4016000000000004e-06,
+      "loss": 0.6997,
+      "step": 38010
+    },
+    {
+      "epoch": 3.802,
+      "grad_norm": 80.04641723632812,
+      "learning_rate": 2.3996000000000003e-06,
+      "loss": 0.604,
+      "step": 38020
+    },
+    {
+      "epoch": 3.803,
+      "grad_norm": 45.173240661621094,
+      "learning_rate": 2.3976e-06,
+      "loss": 0.5006,
+      "step": 38030
+    },
+    {
+      "epoch": 3.8040000000000003,
+      "grad_norm": 65.73089599609375,
+      "learning_rate": 2.3956e-06,
+      "loss": 1.1262,
+      "step": 38040
+    },
+    {
+      "epoch": 3.805,
+      "grad_norm": 39.5977897644043,
+      "learning_rate": 2.3936000000000003e-06,
+      "loss": 0.5203,
+      "step": 38050
+    },
+    {
+      "epoch": 3.806,
+      "grad_norm": 13.670785903930664,
+      "learning_rate": 2.3916e-06,
+      "loss": 0.4645,
+      "step": 38060
+    },
+    {
+      "epoch": 3.807,
+      "grad_norm": 5.6472039222717285,
+      "learning_rate": 2.3896000000000004e-06,
+      "loss": 0.7319,
+      "step": 38070
+    },
+    {
+      "epoch": 3.808,
+      "grad_norm": 28.19403839111328,
+      "learning_rate": 2.3876000000000002e-06,
+      "loss": 0.6628,
+      "step": 38080
+    },
+    {
+      "epoch": 3.809,
+      "grad_norm": 19.916555404663086,
+      "learning_rate": 2.3856e-06,
+      "loss": 0.5419,
+      "step": 38090
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 55.237937927246094,
+      "learning_rate": 2.3836e-06,
+      "loss": 0.4992,
+      "step": 38100
+    },
+    {
+      "epoch": 3.811,
+      "grad_norm": 8.9008150100708,
+      "learning_rate": 2.3816e-06,
+      "loss": 0.6025,
+      "step": 38110
+    },
+    {
+      "epoch": 3.8120000000000003,
+      "grad_norm": 10.064616203308105,
+      "learning_rate": 2.3796e-06,
+      "loss": 0.5929,
+      "step": 38120
+    },
+    {
+      "epoch": 3.8129999999999997,
+      "grad_norm": 40.20355224609375,
+      "learning_rate": 2.3776000000000003e-06,
+      "loss": 0.6161,
+      "step": 38130
+    },
+    {
+      "epoch": 3.814,
+      "grad_norm": 87.54853057861328,
+      "learning_rate": 2.3756e-06,
+      "loss": 0.8959,
+      "step": 38140
+    },
+    {
+      "epoch": 3.815,
+      "grad_norm": 20.800928115844727,
+      "learning_rate": 2.3736e-06,
+      "loss": 0.5493,
+      "step": 38150
+    },
+    {
+      "epoch": 3.816,
+      "grad_norm": 45.83055877685547,
+      "learning_rate": 2.3716000000000003e-06,
+      "loss": 0.5517,
+      "step": 38160
+    },
+    {
+      "epoch": 3.817,
+      "grad_norm": 37.758338928222656,
+      "learning_rate": 2.3696e-06,
+      "loss": 0.7703,
+      "step": 38170
+    },
+    {
+      "epoch": 3.818,
+      "grad_norm": 43.727909088134766,
+      "learning_rate": 2.3676e-06,
+      "loss": 0.752,
+      "step": 38180
+    },
+    {
+      "epoch": 3.819,
+      "grad_norm": 15.53500747680664,
+      "learning_rate": 2.3656000000000003e-06,
+      "loss": 0.7634,
+      "step": 38190
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 15.645066261291504,
+      "learning_rate": 2.3636e-06,
+      "loss": 0.8673,
+      "step": 38200
+    },
+    {
+      "epoch": 3.8209999999999997,
+      "grad_norm": 24.320234298706055,
+      "learning_rate": 2.3616000000000004e-06,
+      "loss": 0.525,
+      "step": 38210
+    },
+    {
+      "epoch": 3.822,
+      "grad_norm": 48.00284957885742,
+      "learning_rate": 2.3596000000000002e-06,
+      "loss": 0.4945,
+      "step": 38220
+    },
+    {
+      "epoch": 3.823,
+      "grad_norm": 104.06059265136719,
+      "learning_rate": 2.3576e-06,
+      "loss": 0.5866,
+      "step": 38230
+    },
+    {
+      "epoch": 3.824,
+      "grad_norm": 47.05179214477539,
+      "learning_rate": 2.3556e-06,
+      "loss": 0.6431,
+      "step": 38240
+    },
+    {
+      "epoch": 3.825,
+      "grad_norm": 26.590356826782227,
+      "learning_rate": 2.3536000000000002e-06,
+      "loss": 0.6083,
+      "step": 38250
+    },
+    {
+      "epoch": 3.826,
+      "grad_norm": 6.2458295822143555,
+      "learning_rate": 2.3516e-06,
+      "loss": 0.7281,
+      "step": 38260
+    },
+    {
+      "epoch": 3.827,
+      "grad_norm": 33.54754638671875,
+      "learning_rate": 2.3496000000000003e-06,
+      "loss": 0.5375,
+      "step": 38270
+    },
+    {
+      "epoch": 3.828,
+      "grad_norm": 55.06610107421875,
+      "learning_rate": 2.3476e-06,
+      "loss": 0.398,
+      "step": 38280
+    },
+    {
+      "epoch": 3.8289999999999997,
+      "grad_norm": 61.530181884765625,
+      "learning_rate": 2.3456e-06,
+      "loss": 0.9226,
+      "step": 38290
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 67.4300537109375,
+      "learning_rate": 2.3436000000000003e-06,
+      "loss": 0.7849,
+      "step": 38300
+    },
+    {
+      "epoch": 3.831,
+      "grad_norm": 17.95315170288086,
+      "learning_rate": 2.3416e-06,
+      "loss": 0.574,
+      "step": 38310
+    },
+    {
+      "epoch": 3.832,
+      "grad_norm": 26.313135147094727,
+      "learning_rate": 2.3396e-06,
+      "loss": 0.4189,
+      "step": 38320
+    },
+    {
+      "epoch": 3.833,
+      "grad_norm": 35.44434356689453,
+      "learning_rate": 2.3376000000000003e-06,
+      "loss": 0.7136,
+      "step": 38330
+    },
+    {
+      "epoch": 3.834,
+      "grad_norm": 100.7297592163086,
+      "learning_rate": 2.3356e-06,
+      "loss": 1.1248,
+      "step": 38340
+    },
+    {
+      "epoch": 3.835,
+      "grad_norm": 42.63737106323242,
+      "learning_rate": 2.3336000000000004e-06,
+      "loss": 0.8059,
+      "step": 38350
+    },
+    {
+      "epoch": 3.836,
+      "grad_norm": 28.802104949951172,
+      "learning_rate": 2.3316000000000003e-06,
+      "loss": 0.4204,
+      "step": 38360
+    },
+    {
+      "epoch": 3.8369999999999997,
+      "grad_norm": 36.844635009765625,
+      "learning_rate": 2.3296e-06,
+      "loss": 0.6372,
+      "step": 38370
+    },
+    {
+      "epoch": 3.838,
+      "grad_norm": 38.515411376953125,
+      "learning_rate": 2.3276e-06,
+      "loss": 0.4195,
+      "step": 38380
+    },
+    {
+      "epoch": 3.839,
+      "grad_norm": 44.74578857421875,
+      "learning_rate": 2.3256000000000002e-06,
+      "loss": 0.6128,
+      "step": 38390
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 48.676204681396484,
+      "learning_rate": 2.3236000000000005e-06,
+      "loss": 0.4594,
+      "step": 38400
+    },
+    {
+      "epoch": 3.841,
+      "grad_norm": 32.13726043701172,
+      "learning_rate": 2.3216000000000004e-06,
+      "loss": 0.4666,
+      "step": 38410
+    },
+    {
+      "epoch": 3.842,
+      "grad_norm": 54.24015426635742,
+      "learning_rate": 2.3196e-06,
+      "loss": 0.578,
+      "step": 38420
+    },
+    {
+      "epoch": 3.843,
+      "grad_norm": 53.41969299316406,
+      "learning_rate": 2.3176e-06,
+      "loss": 0.7449,
+      "step": 38430
+    },
+    {
+      "epoch": 3.844,
+      "grad_norm": 57.150970458984375,
+      "learning_rate": 2.3156e-06,
+      "loss": 0.6208,
+      "step": 38440
+    },
+    {
+      "epoch": 3.8449999999999998,
+      "grad_norm": 29.20799446105957,
+      "learning_rate": 2.3136e-06,
+      "loss": 0.6802,
+      "step": 38450
+    },
+    {
+      "epoch": 3.846,
+      "grad_norm": 49.80082702636719,
+      "learning_rate": 2.3116e-06,
+      "loss": 0.4248,
+      "step": 38460
+    },
+    {
+      "epoch": 3.847,
+      "grad_norm": 38.34299087524414,
+      "learning_rate": 2.3096000000000003e-06,
+      "loss": 0.4581,
+      "step": 38470
+    },
+    {
+      "epoch": 3.848,
+      "grad_norm": 39.137451171875,
+      "learning_rate": 2.3076e-06,
+      "loss": 0.5434,
+      "step": 38480
+    },
+    {
+      "epoch": 3.849,
+      "grad_norm": 2.7317628860473633,
+      "learning_rate": 2.3056e-06,
+      "loss": 0.3184,
+      "step": 38490
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 24.794809341430664,
+      "learning_rate": 2.3036000000000003e-06,
+      "loss": 0.4862,
+      "step": 38500
+    },
+    {
+      "epoch": 3.851,
+      "grad_norm": 43.81166076660156,
+      "learning_rate": 2.3016e-06,
+      "loss": 0.6984,
+      "step": 38510
+    },
+    {
+      "epoch": 3.852,
+      "grad_norm": 46.706764221191406,
+      "learning_rate": 2.2996e-06,
+      "loss": 0.6245,
+      "step": 38520
+    },
+    {
+      "epoch": 3.8529999999999998,
+      "grad_norm": 32.64492416381836,
+      "learning_rate": 2.2976000000000002e-06,
+      "loss": 0.5363,
+      "step": 38530
+    },
+    {
+      "epoch": 3.854,
+      "grad_norm": 43.35952377319336,
+      "learning_rate": 2.2956e-06,
+      "loss": 0.9169,
+      "step": 38540
+    },
+    {
+      "epoch": 3.855,
+      "grad_norm": 56.83952713012695,
+      "learning_rate": 2.2936000000000004e-06,
+      "loss": 0.5871,
+      "step": 38550
+    },
+    {
+      "epoch": 3.856,
+      "grad_norm": 24.13530921936035,
+      "learning_rate": 2.2916000000000002e-06,
+      "loss": 0.7488,
+      "step": 38560
+    },
+    {
+      "epoch": 3.857,
+      "grad_norm": 4.121044635772705,
+      "learning_rate": 2.2896e-06,
+      "loss": 0.687,
+      "step": 38570
+    },
+    {
+      "epoch": 3.858,
+      "grad_norm": 3.334587812423706,
+      "learning_rate": 2.2876e-06,
+      "loss": 0.764,
+      "step": 38580
+    },
+    {
+      "epoch": 3.859,
+      "grad_norm": 15.22492790222168,
+      "learning_rate": 2.2856e-06,
+      "loss": 0.5746,
+      "step": 38590
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 39.12681198120117,
+      "learning_rate": 2.2836000000000005e-06,
+      "loss": 0.5068,
+      "step": 38600
+    },
+    {
+      "epoch": 3.8609999999999998,
+      "grad_norm": 59.579742431640625,
+      "learning_rate": 2.2816000000000003e-06,
+      "loss": 0.7578,
+      "step": 38610
+    },
+    {
+      "epoch": 3.862,
+      "grad_norm": 4.269464015960693,
+      "learning_rate": 2.2796e-06,
+      "loss": 0.5282,
+      "step": 38620
+    },
+    {
+      "epoch": 3.863,
+      "grad_norm": 11.499958992004395,
+      "learning_rate": 2.2776e-06,
+      "loss": 0.6187,
+      "step": 38630
+    },
+    {
+      "epoch": 3.864,
+      "grad_norm": 15.392281532287598,
+      "learning_rate": 2.2756000000000003e-06,
+      "loss": 0.5032,
+      "step": 38640
+    },
+    {
+      "epoch": 3.865,
+      "grad_norm": 35.44303512573242,
+      "learning_rate": 2.2736e-06,
+      "loss": 0.4503,
+      "step": 38650
+    },
+    {
+      "epoch": 3.866,
+      "grad_norm": 38.298221588134766,
+      "learning_rate": 2.2716e-06,
+      "loss": 0.6979,
+      "step": 38660
+    },
+    {
+      "epoch": 3.867,
+      "grad_norm": 40.04734420776367,
+      "learning_rate": 2.2696000000000003e-06,
+      "loss": 0.8594,
+      "step": 38670
+    },
+    {
+      "epoch": 3.868,
+      "grad_norm": 22.192415237426758,
+      "learning_rate": 2.2676e-06,
+      "loss": 0.641,
+      "step": 38680
+    },
+    {
+      "epoch": 3.8689999999999998,
+      "grad_norm": 41.32742691040039,
+      "learning_rate": 2.2656000000000004e-06,
+      "loss": 0.5788,
+      "step": 38690
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 8.416460037231445,
+      "learning_rate": 2.2636000000000002e-06,
+      "loss": 0.4709,
+      "step": 38700
+    },
+    {
+      "epoch": 3.871,
+      "grad_norm": 16.503379821777344,
+      "learning_rate": 2.2616e-06,
+      "loss": 0.5078,
+      "step": 38710
+    },
+    {
+      "epoch": 3.872,
+      "grad_norm": 12.669565200805664,
+      "learning_rate": 2.2596e-06,
+      "loss": 0.5737,
+      "step": 38720
+    },
+    {
+      "epoch": 3.873,
+      "grad_norm": 16.520822525024414,
+      "learning_rate": 2.2576e-06,
+      "loss": 0.4941,
+      "step": 38730
+    },
+    {
+      "epoch": 3.874,
+      "grad_norm": 29.02339744567871,
+      "learning_rate": 2.2556000000000005e-06,
+      "loss": 0.6283,
+      "step": 38740
+    },
+    {
+      "epoch": 3.875,
+      "grad_norm": 4.977586269378662,
+      "learning_rate": 2.2536000000000003e-06,
+      "loss": 0.4775,
+      "step": 38750
+    },
+    {
+      "epoch": 3.876,
+      "grad_norm": 64.50590515136719,
+      "learning_rate": 2.2516e-06,
+      "loss": 0.6762,
+      "step": 38760
+    },
+    {
+      "epoch": 3.877,
+      "grad_norm": 41.48358917236328,
+      "learning_rate": 2.2496e-06,
+      "loss": 0.8202,
+      "step": 38770
+    },
+    {
+      "epoch": 3.878,
+      "grad_norm": 54.37251663208008,
+      "learning_rate": 2.2476e-06,
+      "loss": 0.6263,
+      "step": 38780
+    },
+    {
+      "epoch": 3.879,
+      "grad_norm": 37.51151657104492,
+      "learning_rate": 2.2456e-06,
+      "loss": 0.5234,
+      "step": 38790
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 8.200417518615723,
+      "learning_rate": 2.2436000000000004e-06,
+      "loss": 0.4962,
+      "step": 38800
+    },
+    {
+      "epoch": 3.8810000000000002,
+      "grad_norm": 5.233102798461914,
+      "learning_rate": 2.2416000000000003e-06,
+      "loss": 0.6761,
+      "step": 38810
+    },
+    {
+      "epoch": 3.882,
+      "grad_norm": 1.343250036239624,
+      "learning_rate": 2.2396e-06,
+      "loss": 0.4689,
+      "step": 38820
+    },
+    {
+      "epoch": 3.883,
+      "grad_norm": 42.027427673339844,
+      "learning_rate": 2.2376e-06,
+      "loss": 0.8863,
+      "step": 38830
+    },
+    {
+      "epoch": 3.884,
+      "grad_norm": 7.038177967071533,
+      "learning_rate": 2.2356000000000002e-06,
+      "loss": 0.8442,
+      "step": 38840
+    },
+    {
+      "epoch": 3.885,
+      "grad_norm": 23.81315803527832,
+      "learning_rate": 2.2336e-06,
+      "loss": 0.5624,
+      "step": 38850
+    },
+    {
+      "epoch": 3.886,
+      "grad_norm": 43.30754470825195,
+      "learning_rate": 2.2316e-06,
+      "loss": 0.5765,
+      "step": 38860
+    },
+    {
+      "epoch": 3.887,
+      "grad_norm": 34.442176818847656,
+      "learning_rate": 2.2296000000000002e-06,
+      "loss": 0.9356,
+      "step": 38870
+    },
+    {
+      "epoch": 3.888,
+      "grad_norm": 41.87175369262695,
+      "learning_rate": 2.2276000000000005e-06,
+      "loss": 1.1155,
+      "step": 38880
+    },
+    {
+      "epoch": 3.8890000000000002,
+      "grad_norm": 34.54471206665039,
+      "learning_rate": 2.2256000000000003e-06,
+      "loss": 0.4604,
+      "step": 38890
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 50.42923355102539,
+      "learning_rate": 2.2236e-06,
+      "loss": 0.6437,
+      "step": 38900
+    },
+    {
+      "epoch": 3.891,
+      "grad_norm": 39.492305755615234,
+      "learning_rate": 2.2216e-06,
+      "loss": 0.4618,
+      "step": 38910
+    },
+    {
+      "epoch": 3.892,
+      "grad_norm": 9.663975715637207,
+      "learning_rate": 2.2196e-06,
+      "loss": 0.5767,
+      "step": 38920
+    },
+    {
+      "epoch": 3.893,
+      "grad_norm": 46.55671310424805,
+      "learning_rate": 2.2176e-06,
+      "loss": 0.513,
+      "step": 38930
+    },
+    {
+      "epoch": 3.894,
+      "grad_norm": 65.58306121826172,
+      "learning_rate": 2.2156000000000004e-06,
+      "loss": 0.6941,
+      "step": 38940
+    },
+    {
+      "epoch": 3.895,
+      "grad_norm": 78.95509338378906,
+      "learning_rate": 2.2136000000000003e-06,
+      "loss": 0.7419,
+      "step": 38950
+    },
+    {
+      "epoch": 3.896,
+      "grad_norm": 16.70928955078125,
+      "learning_rate": 2.2116e-06,
+      "loss": 0.736,
+      "step": 38960
+    },
+    {
+      "epoch": 3.8970000000000002,
+      "grad_norm": 3.3220157623291016,
+      "learning_rate": 2.2096e-06,
+      "loss": 0.2733,
+      "step": 38970
+    },
+    {
+      "epoch": 3.898,
+      "grad_norm": 14.191826820373535,
+      "learning_rate": 2.2076000000000003e-06,
+      "loss": 0.5306,
+      "step": 38980
+    },
+    {
+      "epoch": 3.899,
+      "grad_norm": 35.20226287841797,
+      "learning_rate": 2.2056e-06,
+      "loss": 0.2515,
+      "step": 38990
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 26.15009880065918,
+      "learning_rate": 2.2036000000000004e-06,
+      "loss": 0.596,
+      "step": 39000
+    },
+    {
+      "epoch": 3.901,
+      "grad_norm": 47.9098014831543,
+      "learning_rate": 2.2016000000000002e-06,
+      "loss": 0.4469,
+      "step": 39010
+    },
+    {
+      "epoch": 3.902,
+      "grad_norm": 68.36515045166016,
+      "learning_rate": 2.1996e-06,
+      "loss": 0.6961,
+      "step": 39020
+    },
+    {
+      "epoch": 3.903,
+      "grad_norm": 18.675844192504883,
+      "learning_rate": 2.1976000000000004e-06,
+      "loss": 0.4955,
+      "step": 39030
+    },
+    {
+      "epoch": 3.904,
+      "grad_norm": 2.813723087310791,
+      "learning_rate": 2.1956e-06,
+      "loss": 0.6415,
+      "step": 39040
+    },
+    {
+      "epoch": 3.9050000000000002,
+      "grad_norm": 40.3653564453125,
+      "learning_rate": 2.1936e-06,
+      "loss": 0.7362,
+      "step": 39050
+    },
+    {
+      "epoch": 3.906,
+      "grad_norm": 43.43606185913086,
+      "learning_rate": 2.1916e-06,
+      "loss": 0.7509,
+      "step": 39060
+    },
+    {
+      "epoch": 3.907,
+      "grad_norm": 40.229148864746094,
+      "learning_rate": 2.1896e-06,
+      "loss": 0.54,
+      "step": 39070
+    },
+    {
+      "epoch": 3.908,
+      "grad_norm": 2.370300531387329,
+      "learning_rate": 2.1876000000000005e-06,
+      "loss": 0.4631,
+      "step": 39080
+    },
+    {
+      "epoch": 3.909,
+      "grad_norm": 66.48926544189453,
+      "learning_rate": 2.1856000000000003e-06,
+      "loss": 0.8684,
+      "step": 39090
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 71.65203094482422,
+      "learning_rate": 2.1836e-06,
+      "loss": 0.5653,
+      "step": 39100
+    },
+    {
+      "epoch": 3.911,
+      "grad_norm": 67.22386169433594,
+      "learning_rate": 2.1816e-06,
+      "loss": 0.577,
+      "step": 39110
+    },
+    {
+      "epoch": 3.912,
+      "grad_norm": 11.789389610290527,
+      "learning_rate": 2.1796e-06,
+      "loss": 0.6232,
+      "step": 39120
+    },
+    {
+      "epoch": 3.9130000000000003,
+      "grad_norm": 22.935331344604492,
+      "learning_rate": 2.1776e-06,
+      "loss": 0.4742,
+      "step": 39130
+    },
+    {
+      "epoch": 3.914,
+      "grad_norm": 49.52204513549805,
+      "learning_rate": 2.1756000000000004e-06,
+      "loss": 0.6813,
+      "step": 39140
+    },
+    {
+      "epoch": 3.915,
+      "grad_norm": 54.53312683105469,
+      "learning_rate": 2.1736000000000002e-06,
+      "loss": 0.6947,
+      "step": 39150
+    },
+    {
+      "epoch": 3.916,
+      "grad_norm": 23.98177146911621,
+      "learning_rate": 2.1716e-06,
+      "loss": 0.4229,
+      "step": 39160
+    },
+    {
+      "epoch": 3.917,
+      "grad_norm": 39.71488571166992,
+      "learning_rate": 2.1696e-06,
+      "loss": 0.3979,
+      "step": 39170
+    },
+    {
+      "epoch": 3.918,
+      "grad_norm": 80.45480346679688,
+      "learning_rate": 2.1676000000000002e-06,
+      "loss": 0.6924,
+      "step": 39180
+    },
+    {
+      "epoch": 3.919,
+      "grad_norm": 10.103964805603027,
+      "learning_rate": 2.1656e-06,
+      "loss": 0.7003,
+      "step": 39190
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 27.01572036743164,
+      "learning_rate": 2.1636000000000003e-06,
+      "loss": 0.6469,
+      "step": 39200
+    },
+    {
+      "epoch": 3.9210000000000003,
+      "grad_norm": 61.562255859375,
+      "learning_rate": 2.1616e-06,
+      "loss": 0.5936,
+      "step": 39210
+    },
+    {
+      "epoch": 3.922,
+      "grad_norm": 48.38149642944336,
+      "learning_rate": 2.1596000000000005e-06,
+      "loss": 0.4467,
+      "step": 39220
+    },
+    {
+      "epoch": 3.923,
+      "grad_norm": 39.5386962890625,
+      "learning_rate": 2.1576000000000003e-06,
+      "loss": 0.4228,
+      "step": 39230
+    },
+    {
+      "epoch": 3.924,
+      "grad_norm": 30.5782527923584,
+      "learning_rate": 2.1556e-06,
+      "loss": 0.6259,
+      "step": 39240
+    },
+    {
+      "epoch": 3.925,
+      "grad_norm": 41.15312576293945,
+      "learning_rate": 2.1536e-06,
+      "loss": 0.5045,
+      "step": 39250
+    },
+    {
+      "epoch": 3.926,
+      "grad_norm": 18.565710067749023,
+      "learning_rate": 2.1516000000000003e-06,
+      "loss": 0.5476,
+      "step": 39260
+    },
+    {
+      "epoch": 3.927,
+      "grad_norm": 39.46208953857422,
+      "learning_rate": 2.1496e-06,
+      "loss": 0.3473,
+      "step": 39270
+    },
+    {
+      "epoch": 3.928,
+      "grad_norm": 28.1165771484375,
+      "learning_rate": 2.1478e-06,
+      "loss": 0.8228,
+      "step": 39280
+    },
+    {
+      "epoch": 3.9290000000000003,
+      "grad_norm": 19.74481964111328,
+      "learning_rate": 2.1458000000000003e-06,
+      "loss": 0.6381,
+      "step": 39290
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 36.65999221801758,
+      "learning_rate": 2.1438e-06,
+      "loss": 0.7628,
+      "step": 39300
+    },
+    {
+      "epoch": 3.931,
+      "grad_norm": 28.818679809570312,
+      "learning_rate": 2.1418000000000004e-06,
+      "loss": 0.4013,
+      "step": 39310
+    },
+    {
+      "epoch": 3.932,
+      "grad_norm": 2.3065967559814453,
+      "learning_rate": 2.1398000000000002e-06,
+      "loss": 0.4747,
+      "step": 39320
+    },
+    {
+      "epoch": 3.933,
+      "grad_norm": 66.51754760742188,
+      "learning_rate": 2.1378e-06,
+      "loss": 0.59,
+      "step": 39330
+    },
+    {
+      "epoch": 3.934,
+      "grad_norm": 18.615644454956055,
+      "learning_rate": 2.1358e-06,
+      "loss": 0.4391,
+      "step": 39340
+    },
+    {
+      "epoch": 3.935,
+      "grad_norm": 37.48296356201172,
+      "learning_rate": 2.1338000000000002e-06,
+      "loss": 0.5004,
+      "step": 39350
+    },
+    {
+      "epoch": 3.936,
+      "grad_norm": 24.665191650390625,
+      "learning_rate": 2.1318e-06,
+      "loss": 0.6554,
+      "step": 39360
+    },
+    {
+      "epoch": 3.9370000000000003,
+      "grad_norm": 44.627471923828125,
+      "learning_rate": 2.1298000000000003e-06,
+      "loss": 0.5226,
+      "step": 39370
+    },
+    {
+      "epoch": 3.9379999999999997,
+      "grad_norm": 42.4708137512207,
+      "learning_rate": 2.1278e-06,
+      "loss": 0.503,
+      "step": 39380
+    },
+    {
+      "epoch": 3.939,
+      "grad_norm": 69.4271469116211,
+      "learning_rate": 2.1258e-06,
+      "loss": 0.8558,
+      "step": 39390
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 54.04352569580078,
+      "learning_rate": 2.1238000000000003e-06,
+      "loss": 0.6508,
+      "step": 39400
+    },
+    {
+      "epoch": 3.941,
+      "grad_norm": 52.962120056152344,
+      "learning_rate": 2.1218e-06,
+      "loss": 0.4722,
+      "step": 39410
+    },
+    {
+      "epoch": 3.942,
+      "grad_norm": 55.27985382080078,
+      "learning_rate": 2.1198e-06,
+      "loss": 0.7672,
+      "step": 39420
+    },
+    {
+      "epoch": 3.943,
+      "grad_norm": 45.97237014770508,
+      "learning_rate": 2.1178000000000003e-06,
+      "loss": 0.7393,
+      "step": 39430
+    },
+    {
+      "epoch": 3.944,
+      "grad_norm": 37.7562255859375,
+      "learning_rate": 2.1158e-06,
+      "loss": 0.5531,
+      "step": 39440
+    },
+    {
+      "epoch": 3.945,
+      "grad_norm": 2.4542977809906006,
+      "learning_rate": 2.1138000000000004e-06,
+      "loss": 0.3792,
+      "step": 39450
+    },
+    {
+      "epoch": 3.9459999999999997,
+      "grad_norm": 60.25086975097656,
+      "learning_rate": 2.1118000000000003e-06,
+      "loss": 0.7512,
+      "step": 39460
+    },
+    {
+      "epoch": 3.947,
+      "grad_norm": 47.01914978027344,
+      "learning_rate": 2.1098e-06,
+      "loss": 0.5358,
+      "step": 39470
+    },
+    {
+      "epoch": 3.948,
+      "grad_norm": 7.820398807525635,
+      "learning_rate": 2.1078e-06,
+      "loss": 0.7304,
+      "step": 39480
+    },
+    {
+      "epoch": 3.949,
+      "grad_norm": 4.501594066619873,
+      "learning_rate": 2.1058000000000002e-06,
+      "loss": 0.7448,
+      "step": 39490
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 21.622995376586914,
+      "learning_rate": 2.1038e-06,
+      "loss": 0.5995,
+      "step": 39500
+    },
+    {
+      "epoch": 3.951,
+      "grad_norm": 9.188790321350098,
+      "learning_rate": 2.1018000000000004e-06,
+      "loss": 0.4282,
+      "step": 39510
+    },
+    {
+      "epoch": 3.952,
+      "grad_norm": 46.63243103027344,
+      "learning_rate": 2.0998e-06,
+      "loss": 0.9215,
+      "step": 39520
+    },
+    {
+      "epoch": 3.953,
+      "grad_norm": 32.74306106567383,
+      "learning_rate": 2.0978e-06,
+      "loss": 0.5652,
+      "step": 39530
+    },
+    {
+      "epoch": 3.9539999999999997,
+      "grad_norm": 14.57941722869873,
+      "learning_rate": 2.0958e-06,
+      "loss": 0.4371,
+      "step": 39540
+    },
+    {
+      "epoch": 3.955,
+      "grad_norm": 31.52217674255371,
+      "learning_rate": 2.0938e-06,
+      "loss": 0.3675,
+      "step": 39550
+    },
+    {
+      "epoch": 3.956,
+      "grad_norm": 2.353820323944092,
+      "learning_rate": 2.0918e-06,
+      "loss": 0.4438,
+      "step": 39560
+    },
+    {
+      "epoch": 3.957,
+      "grad_norm": 6.106788635253906,
+      "learning_rate": 2.0898000000000003e-06,
+      "loss": 0.5282,
+      "step": 39570
+    },
+    {
+      "epoch": 3.958,
+      "grad_norm": 34.10426712036133,
+      "learning_rate": 2.0878e-06,
+      "loss": 0.3473,
+      "step": 39580
+    },
+    {
+      "epoch": 3.959,
+      "grad_norm": 3.547370672225952,
+      "learning_rate": 2.0858e-06,
+      "loss": 0.8163,
+      "step": 39590
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 11.41083812713623,
+      "learning_rate": 2.0838000000000003e-06,
+      "loss": 0.4964,
+      "step": 39600
+    },
+    {
+      "epoch": 3.961,
+      "grad_norm": 20.70975112915039,
+      "learning_rate": 2.0818e-06,
+      "loss": 0.6299,
+      "step": 39610
+    },
+    {
+      "epoch": 3.9619999999999997,
+      "grad_norm": 40.6129264831543,
+      "learning_rate": 2.0798e-06,
+      "loss": 0.6077,
+      "step": 39620
+    },
+    {
+      "epoch": 3.963,
+      "grad_norm": 49.65641403198242,
+      "learning_rate": 2.0778000000000002e-06,
+      "loss": 0.7575,
+      "step": 39630
+    },
+    {
+      "epoch": 3.964,
+      "grad_norm": 43.09720230102539,
+      "learning_rate": 2.0758e-06,
+      "loss": 0.5197,
+      "step": 39640
+    },
+    {
+      "epoch": 3.965,
+      "grad_norm": 4.485574245452881,
+      "learning_rate": 2.0738000000000004e-06,
+      "loss": 0.8253,
+      "step": 39650
+    },
+    {
+      "epoch": 3.966,
+      "grad_norm": 24.993141174316406,
+      "learning_rate": 2.0718000000000002e-06,
+      "loss": 0.5464,
+      "step": 39660
+    },
+    {
+      "epoch": 3.967,
+      "grad_norm": 63.31475830078125,
+      "learning_rate": 2.0698e-06,
+      "loss": 0.7333,
+      "step": 39670
+    },
+    {
+      "epoch": 3.968,
+      "grad_norm": 53.776859283447266,
+      "learning_rate": 2.0678e-06,
+      "loss": 0.5568,
+      "step": 39680
+    },
+    {
+      "epoch": 3.969,
+      "grad_norm": 18.64484214782715,
+      "learning_rate": 2.0658e-06,
+      "loss": 0.3912,
+      "step": 39690
+    },
+    {
+      "epoch": 3.9699999999999998,
+      "grad_norm": 68.29088592529297,
+      "learning_rate": 2.0638e-06,
+      "loss": 0.5753,
+      "step": 39700
+    },
+    {
+      "epoch": 3.971,
+      "grad_norm": 53.469207763671875,
+      "learning_rate": 2.0618000000000003e-06,
+      "loss": 0.4962,
+      "step": 39710
+    },
+    {
+      "epoch": 3.972,
+      "grad_norm": 22.82310676574707,
+      "learning_rate": 2.0598e-06,
+      "loss": 0.7009,
+      "step": 39720
+    },
+    {
+      "epoch": 3.973,
+      "grad_norm": 47.31079864501953,
+      "learning_rate": 2.0578e-06,
+      "loss": 0.8921,
+      "step": 39730
+    },
+    {
+      "epoch": 3.974,
+      "grad_norm": 12.895943641662598,
+      "learning_rate": 2.0558000000000003e-06,
+      "loss": 0.736,
+      "step": 39740
+    },
+    {
+      "epoch": 3.975,
+      "grad_norm": 21.22369384765625,
+      "learning_rate": 2.0538e-06,
+      "loss": 0.4136,
+      "step": 39750
+    },
+    {
+      "epoch": 3.976,
+      "grad_norm": 1.816238522529602,
+      "learning_rate": 2.0518e-06,
+      "loss": 0.3451,
+      "step": 39760
+    },
+    {
+      "epoch": 3.977,
+      "grad_norm": 22.588109970092773,
+      "learning_rate": 2.0498000000000003e-06,
+      "loss": 0.7551,
+      "step": 39770
+    },
+    {
+      "epoch": 3.9779999999999998,
+      "grad_norm": 15.299209594726562,
+      "learning_rate": 2.0478e-06,
+      "loss": 0.79,
+      "step": 39780
+    },
+    {
+      "epoch": 3.979,
+      "grad_norm": 49.50896453857422,
+      "learning_rate": 2.0458000000000004e-06,
+      "loss": 0.6644,
+      "step": 39790
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 21.569507598876953,
+      "learning_rate": 2.0438000000000002e-06,
+      "loss": 0.6802,
+      "step": 39800
+    },
+    {
+      "epoch": 3.981,
+      "grad_norm": 45.40610885620117,
+      "learning_rate": 2.0418e-06,
+      "loss": 1.0853,
+      "step": 39810
+    },
+    {
+      "epoch": 3.982,
+      "grad_norm": 17.77871322631836,
+      "learning_rate": 2.0398e-06,
+      "loss": 0.4756,
+      "step": 39820
+    },
+    {
+      "epoch": 3.983,
+      "grad_norm": 33.64976119995117,
+      "learning_rate": 2.0378e-06,
+      "loss": 0.9198,
+      "step": 39830
+    },
+    {
+      "epoch": 3.984,
+      "grad_norm": 51.46430587768555,
+      "learning_rate": 2.0358000000000005e-06,
+      "loss": 0.7289,
+      "step": 39840
+    },
+    {
+      "epoch": 3.985,
+      "grad_norm": 49.01710510253906,
+      "learning_rate": 2.0338000000000003e-06,
+      "loss": 0.7091,
+      "step": 39850
+    },
+    {
+      "epoch": 3.9859999999999998,
+      "grad_norm": 7.222930431365967,
+      "learning_rate": 2.0318e-06,
+      "loss": 0.5029,
+      "step": 39860
+    },
+    {
+      "epoch": 3.987,
+      "grad_norm": 50.691001892089844,
+      "learning_rate": 2.0298e-06,
+      "loss": 0.8107,
+      "step": 39870
+    },
+    {
+      "epoch": 3.988,
+      "grad_norm": 6.308470249176025,
+      "learning_rate": 2.0278e-06,
+      "loss": 0.4074,
+      "step": 39880
+    },
+    {
+      "epoch": 3.989,
+      "grad_norm": 12.49143123626709,
+      "learning_rate": 2.0258e-06,
+      "loss": 0.6216,
+      "step": 39890
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 44.38469696044922,
+      "learning_rate": 2.0238e-06,
+      "loss": 0.6535,
+      "step": 39900
+    },
+    {
+      "epoch": 3.991,
+      "grad_norm": 26.98875617980957,
+      "learning_rate": 2.0218000000000003e-06,
+      "loss": 0.3996,
+      "step": 39910
+    },
+    {
+      "epoch": 3.992,
+      "grad_norm": 78.4300537109375,
+      "learning_rate": 2.0198e-06,
+      "loss": 0.5802,
+      "step": 39920
+    },
+    {
+      "epoch": 3.993,
+      "grad_norm": 49.365081787109375,
+      "learning_rate": 2.0178e-06,
+      "loss": 0.6089,
+      "step": 39930
+    },
+    {
+      "epoch": 3.9939999999999998,
+      "grad_norm": 49.79063415527344,
+      "learning_rate": 2.0158000000000002e-06,
+      "loss": 0.7299,
+      "step": 39940
+    },
+    {
+      "epoch": 3.995,
+      "grad_norm": 56.70253372192383,
+      "learning_rate": 2.0138e-06,
+      "loss": 0.5325,
+      "step": 39950
+    },
+    {
+      "epoch": 3.996,
+      "grad_norm": 35.089202880859375,
+      "learning_rate": 2.0118e-06,
+      "loss": 0.5027,
+      "step": 39960
+    },
+    {
+      "epoch": 3.997,
+      "grad_norm": 31.567665100097656,
+      "learning_rate": 2.0098000000000002e-06,
+      "loss": 0.7014,
+      "step": 39970
+    },
+    {
+      "epoch": 3.998,
+      "grad_norm": 14.97442626953125,
+      "learning_rate": 2.0078e-06,
+      "loss": 0.6241,
+      "step": 39980
+    },
+    {
+      "epoch": 3.999,
+      "grad_norm": 20.558170318603516,
+      "learning_rate": 2.0058000000000003e-06,
+      "loss": 0.6377,
+      "step": 39990
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 25.69771385192871,
+      "learning_rate": 2.0038e-06,
+      "loss": 0.5495,
+      "step": 40000
+    },
+    {
+      "epoch": 4.001,
+      "grad_norm": 2.9470462799072266,
+      "learning_rate": 2.0018e-06,
+      "loss": 0.624,
+      "step": 40010
+    },
+    {
+      "epoch": 4.002,
+      "grad_norm": 7.803587436676025,
+      "learning_rate": 1.9998e-06,
+      "loss": 0.6188,
+      "step": 40020
+    },
+    {
+      "epoch": 4.003,
+      "grad_norm": 58.450340270996094,
+      "learning_rate": 1.9978e-06,
+      "loss": 0.6628,
+      "step": 40030
+    },
+    {
+      "epoch": 4.004,
+      "grad_norm": 2.4178032875061035,
+      "learning_rate": 1.9958000000000004e-06,
+      "loss": 0.3826,
+      "step": 40040
+    },
+    {
+      "epoch": 4.005,
+      "grad_norm": 28.441543579101562,
+      "learning_rate": 1.9938000000000003e-06,
+      "loss": 0.6182,
+      "step": 40050
+    },
+    {
+      "epoch": 4.006,
+      "grad_norm": 4.209710121154785,
+      "learning_rate": 1.9918e-06,
+      "loss": 0.5498,
+      "step": 40060
+    },
+    {
+      "epoch": 4.007,
+      "grad_norm": 6.908935070037842,
+      "learning_rate": 1.9898e-06,
+      "loss": 0.6298,
+      "step": 40070
+    },
+    {
+      "epoch": 4.008,
+      "grad_norm": 49.947784423828125,
+      "learning_rate": 1.9878000000000003e-06,
+      "loss": 0.6013,
+      "step": 40080
+    },
+    {
+      "epoch": 4.009,
+      "grad_norm": 32.38127517700195,
+      "learning_rate": 1.9858e-06,
+      "loss": 0.3015,
+      "step": 40090
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 47.0320930480957,
+      "learning_rate": 1.9838e-06,
+      "loss": 0.5804,
+      "step": 40100
+    },
+    {
+      "epoch": 4.011,
+      "grad_norm": 56.43113708496094,
+      "learning_rate": 1.9818000000000002e-06,
+      "loss": 0.7144,
+      "step": 40110
+    },
+    {
+      "epoch": 4.012,
+      "grad_norm": 41.424720764160156,
+      "learning_rate": 1.9798e-06,
+      "loss": 0.6213,
+      "step": 40120
+    },
+    {
+      "epoch": 4.013,
+      "grad_norm": 35.13981628417969,
+      "learning_rate": 1.9778000000000004e-06,
+      "loss": 0.5925,
+      "step": 40130
+    },
+    {
+      "epoch": 4.014,
+      "grad_norm": 23.2755126953125,
+      "learning_rate": 1.9758e-06,
+      "loss": 0.7576,
+      "step": 40140
+    },
+    {
+      "epoch": 4.015,
+      "grad_norm": 41.35113525390625,
+      "learning_rate": 1.9738e-06,
+      "loss": 0.7327,
+      "step": 40150
+    },
+    {
+      "epoch": 4.016,
+      "grad_norm": 33.476600646972656,
+      "learning_rate": 1.9718e-06,
+      "loss": 0.5042,
+      "step": 40160
+    },
+    {
+      "epoch": 4.017,
+      "grad_norm": 47.169395446777344,
+      "learning_rate": 1.9698e-06,
+      "loss": 0.7718,
+      "step": 40170
+    },
+    {
+      "epoch": 4.018,
+      "grad_norm": 34.22056198120117,
+      "learning_rate": 1.9678000000000005e-06,
+      "loss": 0.6289,
+      "step": 40180
+    },
+    {
+      "epoch": 4.019,
+      "grad_norm": 14.578975677490234,
+      "learning_rate": 1.9658000000000003e-06,
+      "loss": 0.4055,
+      "step": 40190
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 18.142715454101562,
+      "learning_rate": 1.9638e-06,
+      "loss": 0.4437,
+      "step": 40200
+    },
+    {
+      "epoch": 4.021,
+      "grad_norm": 56.41261291503906,
+      "learning_rate": 1.9618e-06,
+      "loss": 0.684,
+      "step": 40210
+    },
+    {
+      "epoch": 4.022,
+      "grad_norm": 95.6523666381836,
+      "learning_rate": 1.9598e-06,
+      "loss": 0.8494,
+      "step": 40220
+    },
+    {
+      "epoch": 4.023,
+      "grad_norm": 21.764108657836914,
+      "learning_rate": 1.9578e-06,
+      "loss": 0.501,
+      "step": 40230
+    },
+    {
+      "epoch": 4.024,
+      "grad_norm": 26.407747268676758,
+      "learning_rate": 1.9558000000000004e-06,
+      "loss": 0.5673,
+      "step": 40240
+    },
+    {
+      "epoch": 4.025,
+      "grad_norm": 37.53788757324219,
+      "learning_rate": 1.9538000000000003e-06,
+      "loss": 0.5227,
+      "step": 40250
+    },
+    {
+      "epoch": 4.026,
+      "grad_norm": 24.256620407104492,
+      "learning_rate": 1.9518e-06,
+      "loss": 0.4052,
+      "step": 40260
+    },
+    {
+      "epoch": 4.027,
+      "grad_norm": 39.48225021362305,
+      "learning_rate": 1.9498e-06,
+      "loss": 0.405,
+      "step": 40270
+    },
+    {
+      "epoch": 4.028,
+      "grad_norm": 14.27397346496582,
+      "learning_rate": 1.9478000000000002e-06,
+      "loss": 0.6717,
+      "step": 40280
+    },
+    {
+      "epoch": 4.029,
+      "grad_norm": 2.7721076011657715,
+      "learning_rate": 1.9458e-06,
+      "loss": 0.2324,
+      "step": 40290
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 22.514490127563477,
+      "learning_rate": 1.9438e-06,
+      "loss": 0.7317,
+      "step": 40300
+    },
+    {
+      "epoch": 4.031,
+      "grad_norm": 41.21731948852539,
+      "learning_rate": 1.9418e-06,
+      "loss": 0.4756,
+      "step": 40310
+    },
+    {
+      "epoch": 4.032,
+      "grad_norm": 27.539142608642578,
+      "learning_rate": 1.9398000000000005e-06,
+      "loss": 0.4482,
+      "step": 40320
+    },
+    {
+      "epoch": 4.033,
+      "grad_norm": 11.524066925048828,
+      "learning_rate": 1.9378000000000003e-06,
+      "loss": 0.4907,
+      "step": 40330
+    },
+    {
+      "epoch": 4.034,
+      "grad_norm": 30.499555587768555,
+      "learning_rate": 1.9358e-06,
+      "loss": 0.4553,
+      "step": 40340
+    },
+    {
+      "epoch": 4.035,
+      "grad_norm": 1.9976459741592407,
+      "learning_rate": 1.9338e-06,
+      "loss": 0.4783,
+      "step": 40350
+    },
+    {
+      "epoch": 4.036,
+      "grad_norm": 29.816242218017578,
+      "learning_rate": 1.9318e-06,
+      "loss": 0.6985,
+      "step": 40360
+    },
+    {
+      "epoch": 4.037,
+      "grad_norm": 46.56698989868164,
+      "learning_rate": 1.9298e-06,
+      "loss": 0.5885,
+      "step": 40370
+    },
+    {
+      "epoch": 4.038,
+      "grad_norm": 52.0035514831543,
+      "learning_rate": 1.9278000000000004e-06,
+      "loss": 0.6332,
+      "step": 40380
+    },
+    {
+      "epoch": 4.039,
+      "grad_norm": 40.463199615478516,
+      "learning_rate": 1.9258000000000003e-06,
+      "loss": 0.8216,
+      "step": 40390
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 7.643418788909912,
+      "learning_rate": 1.9238e-06,
+      "loss": 0.578,
+      "step": 40400
+    },
+    {
+      "epoch": 4.041,
+      "grad_norm": 4.566254615783691,
+      "learning_rate": 1.9218e-06,
+      "loss": 0.4617,
+      "step": 40410
+    },
+    {
+      "epoch": 4.042,
+      "grad_norm": 6.617149353027344,
+      "learning_rate": 1.9198000000000002e-06,
+      "loss": 0.4563,
+      "step": 40420
+    },
+    {
+      "epoch": 4.043,
+      "grad_norm": 12.791312217712402,
+      "learning_rate": 1.9178e-06,
+      "loss": 0.5765,
+      "step": 40430
+    },
+    {
+      "epoch": 4.044,
+      "grad_norm": 4.998269081115723,
+      "learning_rate": 1.9158000000000004e-06,
+      "loss": 0.3831,
+      "step": 40440
+    },
+    {
+      "epoch": 4.045,
+      "grad_norm": 100.47882080078125,
+      "learning_rate": 1.9138e-06,
+      "loss": 0.7214,
+      "step": 40450
+    },
+    {
+      "epoch": 4.046,
+      "grad_norm": 30.22802734375,
+      "learning_rate": 1.9118e-06,
+      "loss": 0.6184,
+      "step": 40460
+    },
+    {
+      "epoch": 4.047,
+      "grad_norm": 25.04363441467285,
+      "learning_rate": 1.9098000000000003e-06,
+      "loss": 0.5158,
+      "step": 40470
+    },
+    {
+      "epoch": 4.048,
+      "grad_norm": 21.190914154052734,
+      "learning_rate": 1.9078e-06,
+      "loss": 0.5553,
+      "step": 40480
+    },
+    {
+      "epoch": 4.049,
+      "grad_norm": 68.58704376220703,
+      "learning_rate": 1.9058000000000002e-06,
+      "loss": 0.4504,
+      "step": 40490
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 59.41545486450195,
+      "learning_rate": 1.9038e-06,
+      "loss": 0.33,
+      "step": 40500
+    },
+    {
+      "epoch": 4.051,
+      "grad_norm": 44.25145721435547,
+      "learning_rate": 1.9018e-06,
+      "loss": 0.5571,
+      "step": 40510
+    },
+    {
+      "epoch": 4.052,
+      "grad_norm": 47.63776779174805,
+      "learning_rate": 1.8998000000000002e-06,
+      "loss": 0.6659,
+      "step": 40520
+    },
+    {
+      "epoch": 4.053,
+      "grad_norm": 3.7169432640075684,
+      "learning_rate": 1.8978000000000003e-06,
+      "loss": 0.6236,
+      "step": 40530
+    },
+    {
+      "epoch": 4.054,
+      "grad_norm": 28.970352172851562,
+      "learning_rate": 1.8958000000000001e-06,
+      "loss": 0.371,
+      "step": 40540
+    },
+    {
+      "epoch": 4.055,
+      "grad_norm": 20.295238494873047,
+      "learning_rate": 1.8938e-06,
+      "loss": 0.2769,
+      "step": 40550
+    },
+    {
+      "epoch": 4.056,
+      "grad_norm": 55.8317756652832,
+      "learning_rate": 1.8918e-06,
+      "loss": 0.4959,
+      "step": 40560
+    },
+    {
+      "epoch": 4.057,
+      "grad_norm": 42.73490524291992,
+      "learning_rate": 1.8898000000000003e-06,
+      "loss": 0.5813,
+      "step": 40570
+    },
+    {
+      "epoch": 4.058,
+      "grad_norm": 82.32331848144531,
+      "learning_rate": 1.8878000000000002e-06,
+      "loss": 0.7,
+      "step": 40580
+    },
+    {
+      "epoch": 4.059,
+      "grad_norm": 61.270931243896484,
+      "learning_rate": 1.8858000000000002e-06,
+      "loss": 0.5079,
+      "step": 40590
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 2.542825937271118,
+      "learning_rate": 1.8838e-06,
+      "loss": 0.4273,
+      "step": 40600
+    },
+    {
+      "epoch": 4.061,
+      "grad_norm": 35.910396575927734,
+      "learning_rate": 1.8818e-06,
+      "loss": 0.6405,
+      "step": 40610
+    },
+    {
+      "epoch": 4.062,
+      "grad_norm": 67.07048797607422,
+      "learning_rate": 1.8798000000000002e-06,
+      "loss": 0.2777,
+      "step": 40620
+    },
+    {
+      "epoch": 4.063,
+      "grad_norm": 42.202327728271484,
+      "learning_rate": 1.8778000000000003e-06,
+      "loss": 0.4228,
+      "step": 40630
+    },
+    {
+      "epoch": 4.064,
+      "grad_norm": 34.79640579223633,
+      "learning_rate": 1.8758000000000001e-06,
+      "loss": 0.8618,
+      "step": 40640
+    },
+    {
+      "epoch": 4.065,
+      "grad_norm": 19.536710739135742,
+      "learning_rate": 1.8738000000000002e-06,
+      "loss": 0.5667,
+      "step": 40650
+    },
+    {
+      "epoch": 4.066,
+      "grad_norm": 45.35587692260742,
+      "learning_rate": 1.8718000000000002e-06,
+      "loss": 0.7089,
+      "step": 40660
+    },
+    {
+      "epoch": 4.067,
+      "grad_norm": 56.838645935058594,
+      "learning_rate": 1.8698000000000003e-06,
+      "loss": 0.5213,
+      "step": 40670
+    },
+    {
+      "epoch": 4.068,
+      "grad_norm": 33.80918884277344,
+      "learning_rate": 1.8678000000000001e-06,
+      "loss": 0.8448,
+      "step": 40680
+    },
+    {
+      "epoch": 4.069,
+      "grad_norm": 10.72282600402832,
+      "learning_rate": 1.8658000000000002e-06,
+      "loss": 0.5146,
+      "step": 40690
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 31.536436080932617,
+      "learning_rate": 1.8638e-06,
+      "loss": 0.4951,
+      "step": 40700
+    },
+    {
+      "epoch": 4.071,
+      "grad_norm": 51.59791946411133,
+      "learning_rate": 1.8618000000000003e-06,
+      "loss": 0.4358,
+      "step": 40710
+    },
+    {
+      "epoch": 4.072,
+      "grad_norm": 71.39259338378906,
+      "learning_rate": 1.8598000000000002e-06,
+      "loss": 0.6669,
+      "step": 40720
+    },
+    {
+      "epoch": 4.073,
+      "grad_norm": 13.469985008239746,
+      "learning_rate": 1.8578000000000002e-06,
+      "loss": 0.5264,
+      "step": 40730
+    },
+    {
+      "epoch": 4.074,
+      "grad_norm": 45.90078353881836,
+      "learning_rate": 1.8558e-06,
+      "loss": 0.9989,
+      "step": 40740
+    },
+    {
+      "epoch": 4.075,
+      "grad_norm": 71.98507690429688,
+      "learning_rate": 1.8538000000000001e-06,
+      "loss": 0.8605,
+      "step": 40750
+    },
+    {
+      "epoch": 4.076,
+      "grad_norm": 42.841304779052734,
+      "learning_rate": 1.8518000000000002e-06,
+      "loss": 0.7456,
+      "step": 40760
+    },
+    {
+      "epoch": 4.077,
+      "grad_norm": 43.334720611572266,
+      "learning_rate": 1.8498000000000003e-06,
+      "loss": 0.6266,
+      "step": 40770
+    },
+    {
+      "epoch": 4.078,
+      "grad_norm": 37.730350494384766,
+      "learning_rate": 1.8478000000000001e-06,
+      "loss": 0.5659,
+      "step": 40780
+    },
+    {
+      "epoch": 4.079,
+      "grad_norm": 40.119144439697266,
+      "learning_rate": 1.8458000000000002e-06,
+      "loss": 0.4387,
+      "step": 40790
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 3.083294630050659,
+      "learning_rate": 1.8438e-06,
+      "loss": 0.6261,
+      "step": 40800
+    },
+    {
+      "epoch": 4.081,
+      "grad_norm": 28.532621383666992,
+      "learning_rate": 1.8418000000000003e-06,
+      "loss": 0.5223,
+      "step": 40810
+    },
+    {
+      "epoch": 4.082,
+      "grad_norm": 41.42976379394531,
+      "learning_rate": 1.8398000000000002e-06,
+      "loss": 0.6217,
+      "step": 40820
+    },
+    {
+      "epoch": 4.083,
+      "grad_norm": 17.437150955200195,
+      "learning_rate": 1.8378000000000002e-06,
+      "loss": 0.7407,
+      "step": 40830
+    },
+    {
+      "epoch": 4.084,
+      "grad_norm": 20.878398895263672,
+      "learning_rate": 1.8358e-06,
+      "loss": 0.6634,
+      "step": 40840
+    },
+    {
+      "epoch": 4.085,
+      "grad_norm": 74.67936706542969,
+      "learning_rate": 1.8338000000000001e-06,
+      "loss": 0.8132,
+      "step": 40850
+    },
+    {
+      "epoch": 4.086,
+      "grad_norm": 24.71967124938965,
+      "learning_rate": 1.8318000000000002e-06,
+      "loss": 0.8061,
+      "step": 40860
+    },
+    {
+      "epoch": 4.087,
+      "grad_norm": 76.58692932128906,
+      "learning_rate": 1.8298000000000003e-06,
+      "loss": 0.5159,
+      "step": 40870
+    },
+    {
+      "epoch": 4.088,
+      "grad_norm": 49.60830307006836,
+      "learning_rate": 1.8278e-06,
+      "loss": 0.5036,
+      "step": 40880
+    },
+    {
+      "epoch": 4.089,
+      "grad_norm": 33.77963638305664,
+      "learning_rate": 1.8258000000000002e-06,
+      "loss": 0.6621,
+      "step": 40890
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 25.869617462158203,
+      "learning_rate": 1.8238e-06,
+      "loss": 0.7384,
+      "step": 40900
+    },
+    {
+      "epoch": 4.091,
+      "grad_norm": 35.43751907348633,
+      "learning_rate": 1.8218000000000003e-06,
+      "loss": 0.7112,
+      "step": 40910
+    },
+    {
+      "epoch": 4.092,
+      "grad_norm": 66.21663665771484,
+      "learning_rate": 1.8198000000000001e-06,
+      "loss": 0.5357,
+      "step": 40920
+    },
+    {
+      "epoch": 4.093,
+      "grad_norm": 18.179227828979492,
+      "learning_rate": 1.8178000000000002e-06,
+      "loss": 0.4763,
+      "step": 40930
+    },
+    {
+      "epoch": 4.094,
+      "grad_norm": 17.802780151367188,
+      "learning_rate": 1.8158e-06,
+      "loss": 0.5537,
+      "step": 40940
+    },
+    {
+      "epoch": 4.095,
+      "grad_norm": 53.57586669921875,
+      "learning_rate": 1.8138000000000003e-06,
+      "loss": 0.6408,
+      "step": 40950
+    },
+    {
+      "epoch": 4.096,
+      "grad_norm": 51.94672775268555,
+      "learning_rate": 1.8118000000000002e-06,
+      "loss": 0.7398,
+      "step": 40960
+    },
+    {
+      "epoch": 4.097,
+      "grad_norm": 15.93326473236084,
+      "learning_rate": 1.8098000000000002e-06,
+      "loss": 0.6094,
+      "step": 40970
+    },
+    {
+      "epoch": 4.098,
+      "grad_norm": 27.715545654296875,
+      "learning_rate": 1.8078e-06,
+      "loss": 0.4746,
+      "step": 40980
+    },
+    {
+      "epoch": 4.099,
+      "grad_norm": 15.6454496383667,
+      "learning_rate": 1.8058000000000001e-06,
+      "loss": 0.5093,
+      "step": 40990
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 26.22260093688965,
+      "learning_rate": 1.8038000000000002e-06,
+      "loss": 0.6028,
+      "step": 41000
+    },
+    {
+      "epoch": 4.101,
+      "grad_norm": 1.7048978805541992,
+      "learning_rate": 1.8018000000000003e-06,
+      "loss": 0.7598,
+      "step": 41010
+    },
+    {
+      "epoch": 4.102,
+      "grad_norm": 31.586240768432617,
+      "learning_rate": 1.7998000000000001e-06,
+      "loss": 0.4546,
+      "step": 41020
+    },
+    {
+      "epoch": 4.103,
+      "grad_norm": 63.79600143432617,
+      "learning_rate": 1.7978000000000002e-06,
+      "loss": 0.9011,
+      "step": 41030
+    },
+    {
+      "epoch": 4.104,
+      "grad_norm": 55.758968353271484,
+      "learning_rate": 1.7958e-06,
+      "loss": 0.5311,
+      "step": 41040
+    },
+    {
+      "epoch": 4.105,
+      "grad_norm": 13.822869300842285,
+      "learning_rate": 1.7938000000000003e-06,
+      "loss": 0.6063,
+      "step": 41050
+    },
+    {
+      "epoch": 4.106,
+      "grad_norm": 24.2291202545166,
+      "learning_rate": 1.7918000000000001e-06,
+      "loss": 0.662,
+      "step": 41060
+    },
+    {
+      "epoch": 4.107,
+      "grad_norm": 3.7248871326446533,
+      "learning_rate": 1.7898000000000002e-06,
+      "loss": 0.7849,
+      "step": 41070
+    },
+    {
+      "epoch": 4.108,
+      "grad_norm": 89.2628173828125,
+      "learning_rate": 1.7878e-06,
+      "loss": 0.6894,
+      "step": 41080
+    },
+    {
+      "epoch": 4.109,
+      "grad_norm": 41.21138381958008,
+      "learning_rate": 1.7858000000000001e-06,
+      "loss": 0.6289,
+      "step": 41090
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 43.779544830322266,
+      "learning_rate": 1.7838000000000002e-06,
+      "loss": 0.7203,
+      "step": 41100
+    },
+    {
+      "epoch": 4.111,
+      "grad_norm": 57.70969772338867,
+      "learning_rate": 1.7818000000000002e-06,
+      "loss": 0.5379,
+      "step": 41110
+    },
+    {
+      "epoch": 4.112,
+      "grad_norm": 18.52546501159668,
+      "learning_rate": 1.7798e-06,
+      "loss": 0.6473,
+      "step": 41120
+    },
+    {
+      "epoch": 4.113,
+      "grad_norm": 27.89804458618164,
+      "learning_rate": 1.7778000000000002e-06,
+      "loss": 0.5624,
+      "step": 41130
+    },
+    {
+      "epoch": 4.114,
+      "grad_norm": 29.954504013061523,
+      "learning_rate": 1.7758e-06,
+      "loss": 0.6228,
+      "step": 41140
+    },
+    {
+      "epoch": 4.115,
+      "grad_norm": 13.705012321472168,
+      "learning_rate": 1.7738000000000003e-06,
+      "loss": 0.4084,
+      "step": 41150
+    },
+    {
+      "epoch": 4.116,
+      "grad_norm": 3.7398502826690674,
+      "learning_rate": 1.7718000000000001e-06,
+      "loss": 0.3762,
+      "step": 41160
+    },
+    {
+      "epoch": 4.117,
+      "grad_norm": 20.722627639770508,
+      "learning_rate": 1.7698000000000002e-06,
+      "loss": 0.4139,
+      "step": 41170
+    },
+    {
+      "epoch": 4.118,
+      "grad_norm": 6.001708507537842,
+      "learning_rate": 1.7678e-06,
+      "loss": 0.6918,
+      "step": 41180
+    },
+    {
+      "epoch": 4.119,
+      "grad_norm": 47.202796936035156,
+      "learning_rate": 1.7658e-06,
+      "loss": 0.3182,
+      "step": 41190
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 59.22598648071289,
+      "learning_rate": 1.7638000000000002e-06,
+      "loss": 0.4179,
+      "step": 41200
+    },
+    {
+      "epoch": 4.121,
+      "grad_norm": 25.26884651184082,
+      "learning_rate": 1.7618000000000002e-06,
+      "loss": 0.4623,
+      "step": 41210
+    },
+    {
+      "epoch": 4.122,
+      "grad_norm": 21.561010360717773,
+      "learning_rate": 1.7598e-06,
+      "loss": 0.4474,
+      "step": 41220
+    },
+    {
+      "epoch": 4.123,
+      "grad_norm": 38.01866149902344,
+      "learning_rate": 1.7578000000000001e-06,
+      "loss": 0.6089,
+      "step": 41230
+    },
+    {
+      "epoch": 4.124,
+      "grad_norm": 23.74958038330078,
+      "learning_rate": 1.7558000000000002e-06,
+      "loss": 0.5146,
+      "step": 41240
+    },
+    {
+      "epoch": 4.125,
+      "grad_norm": 42.2169075012207,
+      "learning_rate": 1.7538000000000003e-06,
+      "loss": 0.6283,
+      "step": 41250
+    },
+    {
+      "epoch": 4.126,
+      "grad_norm": 30.261451721191406,
+      "learning_rate": 1.7518000000000001e-06,
+      "loss": 0.3849,
+      "step": 41260
+    },
+    {
+      "epoch": 4.127,
+      "grad_norm": 16.222307205200195,
+      "learning_rate": 1.7498000000000002e-06,
+      "loss": 0.477,
+      "step": 41270
+    },
+    {
+      "epoch": 4.128,
+      "grad_norm": 54.456966400146484,
+      "learning_rate": 1.7478e-06,
+      "loss": 0.9572,
+      "step": 41280
+    },
+    {
+      "epoch": 4.129,
+      "grad_norm": 8.039139747619629,
+      "learning_rate": 1.7458000000000003e-06,
+      "loss": 0.5364,
+      "step": 41290
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 24.38113784790039,
+      "learning_rate": 1.7438000000000001e-06,
+      "loss": 0.7153,
+      "step": 41300
+    },
+    {
+      "epoch": 4.131,
+      "grad_norm": 53.22614669799805,
+      "learning_rate": 1.7418000000000002e-06,
+      "loss": 0.6344,
+      "step": 41310
+    },
+    {
+      "epoch": 4.132,
+      "grad_norm": 38.1357536315918,
+      "learning_rate": 1.7398e-06,
+      "loss": 0.7408,
+      "step": 41320
+    },
+    {
+      "epoch": 4.133,
+      "grad_norm": 5.696017265319824,
+      "learning_rate": 1.7378000000000001e-06,
+      "loss": 0.4738,
+      "step": 41330
+    },
+    {
+      "epoch": 4.134,
+      "grad_norm": 43.506187438964844,
+      "learning_rate": 1.7358000000000002e-06,
+      "loss": 0.6412,
+      "step": 41340
+    },
+    {
+      "epoch": 4.135,
+      "grad_norm": 19.607364654541016,
+      "learning_rate": 1.7338000000000002e-06,
+      "loss": 0.4156,
+      "step": 41350
+    },
+    {
+      "epoch": 4.136,
+      "grad_norm": 9.471464157104492,
+      "learning_rate": 1.7318e-06,
+      "loss": 0.5737,
+      "step": 41360
+    },
+    {
+      "epoch": 4.1370000000000005,
+      "grad_norm": 44.34270095825195,
+      "learning_rate": 1.7298000000000002e-06,
+      "loss": 0.7362,
+      "step": 41370
+    },
+    {
+      "epoch": 4.138,
+      "grad_norm": 17.764965057373047,
+      "learning_rate": 1.7278e-06,
+      "loss": 0.3977,
+      "step": 41380
+    },
+    {
+      "epoch": 4.139,
+      "grad_norm": 17.33642578125,
+      "learning_rate": 1.7258000000000003e-06,
+      "loss": 0.5518,
+      "step": 41390
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 56.447975158691406,
+      "learning_rate": 1.7238000000000001e-06,
+      "loss": 0.7362,
+      "step": 41400
+    },
+    {
+      "epoch": 4.141,
+      "grad_norm": 40.42461395263672,
+      "learning_rate": 1.7218000000000002e-06,
+      "loss": 0.4879,
+      "step": 41410
+    },
+    {
+      "epoch": 4.142,
+      "grad_norm": 44.50244140625,
+      "learning_rate": 1.7198e-06,
+      "loss": 0.4516,
+      "step": 41420
+    },
+    {
+      "epoch": 4.143,
+      "grad_norm": 60.044456481933594,
+      "learning_rate": 1.7178e-06,
+      "loss": 0.7069,
+      "step": 41430
+    },
+    {
+      "epoch": 4.144,
+      "grad_norm": 9.243979454040527,
+      "learning_rate": 1.7158000000000002e-06,
+      "loss": 0.5722,
+      "step": 41440
+    },
+    {
+      "epoch": 4.145,
+      "grad_norm": 29.480884552001953,
+      "learning_rate": 1.7138000000000002e-06,
+      "loss": 0.5234,
+      "step": 41450
+    },
+    {
+      "epoch": 4.146,
+      "grad_norm": 23.5086669921875,
+      "learning_rate": 1.7118e-06,
+      "loss": 0.5544,
+      "step": 41460
+    },
+    {
+      "epoch": 4.147,
+      "grad_norm": 34.623836517333984,
+      "learning_rate": 1.7098000000000001e-06,
+      "loss": 0.6038,
+      "step": 41470
+    },
+    {
+      "epoch": 4.148,
+      "grad_norm": 14.842521667480469,
+      "learning_rate": 1.7078e-06,
+      "loss": 0.7382,
+      "step": 41480
+    },
+    {
+      "epoch": 4.149,
+      "grad_norm": 38.99134826660156,
+      "learning_rate": 1.7058000000000003e-06,
+      "loss": 0.6176,
+      "step": 41490
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 20.720136642456055,
+      "learning_rate": 1.7038e-06,
+      "loss": 0.7344,
+      "step": 41500
+    },
+    {
+      "epoch": 4.151,
+      "grad_norm": 42.731204986572266,
+      "learning_rate": 1.7018000000000002e-06,
+      "loss": 0.8725,
+      "step": 41510
+    },
+    {
+      "epoch": 4.152,
+      "grad_norm": 35.921173095703125,
+      "learning_rate": 1.6998e-06,
+      "loss": 0.7057,
+      "step": 41520
+    },
+    {
+      "epoch": 4.153,
+      "grad_norm": 33.33821105957031,
+      "learning_rate": 1.6978e-06,
+      "loss": 0.6767,
+      "step": 41530
+    },
+    {
+      "epoch": 4.154,
+      "grad_norm": 51.7625617980957,
+      "learning_rate": 1.6958000000000001e-06,
+      "loss": 0.5458,
+      "step": 41540
+    },
+    {
+      "epoch": 4.155,
+      "grad_norm": 48.38890838623047,
+      "learning_rate": 1.6938000000000002e-06,
+      "loss": 0.4643,
+      "step": 41550
+    },
+    {
+      "epoch": 4.156,
+      "grad_norm": 3.2673356533050537,
+      "learning_rate": 1.6918e-06,
+      "loss": 0.7652,
+      "step": 41560
+    },
+    {
+      "epoch": 4.157,
+      "grad_norm": 3.0207319259643555,
+      "learning_rate": 1.6898000000000001e-06,
+      "loss": 0.3428,
+      "step": 41570
+    },
+    {
+      "epoch": 4.158,
+      "grad_norm": 8.048981666564941,
+      "learning_rate": 1.6878000000000002e-06,
+      "loss": 0.674,
+      "step": 41580
+    },
+    {
+      "epoch": 4.159,
+      "grad_norm": 41.031864166259766,
+      "learning_rate": 1.6858000000000002e-06,
+      "loss": 0.6165,
+      "step": 41590
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 20.935832977294922,
+      "learning_rate": 1.6838e-06,
+      "loss": 0.3379,
+      "step": 41600
+    },
+    {
+      "epoch": 4.161,
+      "grad_norm": 5.029835224151611,
+      "learning_rate": 1.6818000000000001e-06,
+      "loss": 0.3947,
+      "step": 41610
+    },
+    {
+      "epoch": 4.162,
+      "grad_norm": 36.44365692138672,
+      "learning_rate": 1.6798e-06,
+      "loss": 0.6268,
+      "step": 41620
+    },
+    {
+      "epoch": 4.163,
+      "grad_norm": 13.238295555114746,
+      "learning_rate": 1.6778000000000003e-06,
+      "loss": 0.8674,
+      "step": 41630
+    },
+    {
+      "epoch": 4.164,
+      "grad_norm": 30.991918563842773,
+      "learning_rate": 1.6758000000000001e-06,
+      "loss": 0.6755,
+      "step": 41640
+    },
+    {
+      "epoch": 4.165,
+      "grad_norm": 14.726704597473145,
+      "learning_rate": 1.6738000000000002e-06,
+      "loss": 0.3431,
+      "step": 41650
+    },
+    {
+      "epoch": 4.166,
+      "grad_norm": 43.18964385986328,
+      "learning_rate": 1.6718e-06,
+      "loss": 0.9822,
+      "step": 41660
+    },
+    {
+      "epoch": 4.167,
+      "grad_norm": 7.514678478240967,
+      "learning_rate": 1.6698e-06,
+      "loss": 0.3684,
+      "step": 41670
+    },
+    {
+      "epoch": 4.168,
+      "grad_norm": 42.86118698120117,
+      "learning_rate": 1.6678000000000002e-06,
+      "loss": 0.6661,
+      "step": 41680
+    },
+    {
+      "epoch": 4.169,
+      "grad_norm": 46.86203384399414,
+      "learning_rate": 1.6658000000000002e-06,
+      "loss": 0.3551,
+      "step": 41690
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 49.551971435546875,
+      "learning_rate": 1.6638e-06,
+      "loss": 0.7988,
+      "step": 41700
+    },
+    {
+      "epoch": 4.171,
+      "grad_norm": 48.88741683959961,
+      "learning_rate": 1.6618000000000001e-06,
+      "loss": 0.4471,
+      "step": 41710
+    },
+    {
+      "epoch": 4.172,
+      "grad_norm": 11.72259521484375,
+      "learning_rate": 1.6598e-06,
+      "loss": 0.4281,
+      "step": 41720
+    },
+    {
+      "epoch": 4.173,
+      "grad_norm": 8.203263282775879,
+      "learning_rate": 1.6578000000000002e-06,
+      "loss": 0.3705,
+      "step": 41730
+    },
+    {
+      "epoch": 4.174,
+      "grad_norm": 87.56128692626953,
+      "learning_rate": 1.6558e-06,
+      "loss": 0.8434,
+      "step": 41740
+    },
+    {
+      "epoch": 4.175,
+      "grad_norm": 34.857051849365234,
+      "learning_rate": 1.6538000000000002e-06,
+      "loss": 0.5267,
+      "step": 41750
+    },
+    {
+      "epoch": 4.176,
+      "grad_norm": 7.201389789581299,
+      "learning_rate": 1.6518e-06,
+      "loss": 0.3331,
+      "step": 41760
+    },
+    {
+      "epoch": 4.177,
+      "grad_norm": 26.428733825683594,
+      "learning_rate": 1.6498e-06,
+      "loss": 0.616,
+      "step": 41770
+    },
+    {
+      "epoch": 4.178,
+      "grad_norm": 39.79397201538086,
+      "learning_rate": 1.6478000000000001e-06,
+      "loss": 0.4973,
+      "step": 41780
+    },
+    {
+      "epoch": 4.179,
+      "grad_norm": 41.15972900390625,
+      "learning_rate": 1.6458000000000002e-06,
+      "loss": 0.4637,
+      "step": 41790
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 91.5743179321289,
+      "learning_rate": 1.6438e-06,
+      "loss": 0.7188,
+      "step": 41800
+    },
+    {
+      "epoch": 4.181,
+      "grad_norm": 35.621795654296875,
+      "learning_rate": 1.6418e-06,
+      "loss": 0.565,
+      "step": 41810
+    },
+    {
+      "epoch": 4.182,
+      "grad_norm": 27.934715270996094,
+      "learning_rate": 1.6398e-06,
+      "loss": 0.6133,
+      "step": 41820
+    },
+    {
+      "epoch": 4.183,
+      "grad_norm": 42.21739959716797,
+      "learning_rate": 1.6378000000000002e-06,
+      "loss": 0.479,
+      "step": 41830
+    },
+    {
+      "epoch": 4.184,
+      "grad_norm": 50.77339172363281,
+      "learning_rate": 1.6358e-06,
+      "loss": 0.4843,
+      "step": 41840
+    },
+    {
+      "epoch": 4.185,
+      "grad_norm": 10.044069290161133,
+      "learning_rate": 1.6338000000000001e-06,
+      "loss": 0.4999,
+      "step": 41850
+    },
+    {
+      "epoch": 4.186,
+      "grad_norm": 10.109586715698242,
+      "learning_rate": 1.6318e-06,
+      "loss": 0.4976,
+      "step": 41860
+    },
+    {
+      "epoch": 4.187,
+      "grad_norm": 58.6328125,
+      "learning_rate": 1.6298000000000003e-06,
+      "loss": 0.4399,
+      "step": 41870
+    },
+    {
+      "epoch": 4.188,
+      "grad_norm": 24.95362663269043,
+      "learning_rate": 1.6278000000000001e-06,
+      "loss": 0.5809,
+      "step": 41880
+    },
+    {
+      "epoch": 4.189,
+      "grad_norm": 40.617855072021484,
+      "learning_rate": 1.6258000000000002e-06,
+      "loss": 0.431,
+      "step": 41890
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 37.19096374511719,
+      "learning_rate": 1.6238e-06,
+      "loss": 0.5244,
+      "step": 41900
+    },
+    {
+      "epoch": 4.191,
+      "grad_norm": 53.91977310180664,
+      "learning_rate": 1.6218e-06,
+      "loss": 0.6796,
+      "step": 41910
+    },
+    {
+      "epoch": 4.192,
+      "grad_norm": 3.6687369346618652,
+      "learning_rate": 1.6198000000000004e-06,
+      "loss": 0.6903,
+      "step": 41920
+    },
+    {
+      "epoch": 4.193,
+      "grad_norm": 22.804624557495117,
+      "learning_rate": 1.6178000000000002e-06,
+      "loss": 0.4569,
+      "step": 41930
+    },
+    {
+      "epoch": 4.194,
+      "grad_norm": 56.276893615722656,
+      "learning_rate": 1.6158e-06,
+      "loss": 0.6385,
+      "step": 41940
+    },
+    {
+      "epoch": 4.195,
+      "grad_norm": 41.6296272277832,
+      "learning_rate": 1.6138000000000001e-06,
+      "loss": 0.3963,
+      "step": 41950
+    },
+    {
+      "epoch": 4.196,
+      "grad_norm": 66.05530548095703,
+      "learning_rate": 1.6118e-06,
+      "loss": 0.8673,
+      "step": 41960
+    },
+    {
+      "epoch": 4.197,
+      "grad_norm": 47.94367218017578,
+      "learning_rate": 1.6098000000000002e-06,
+      "loss": 0.815,
+      "step": 41970
+    },
+    {
+      "epoch": 4.198,
+      "grad_norm": 0.4343576431274414,
+      "learning_rate": 1.6078e-06,
+      "loss": 0.39,
+      "step": 41980
+    },
+    {
+      "epoch": 4.199,
+      "grad_norm": 8.591513633728027,
+      "learning_rate": 1.6058000000000002e-06,
+      "loss": 0.66,
+      "step": 41990
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 66.63551330566406,
+      "learning_rate": 1.6038e-06,
+      "loss": 0.4365,
+      "step": 42000
+    },
+    {
+      "epoch": 4.201,
+      "grad_norm": 34.489219665527344,
+      "learning_rate": 1.6018e-06,
+      "loss": 0.6739,
+      "step": 42010
+    },
+    {
+      "epoch": 4.202,
+      "grad_norm": 45.61314392089844,
+      "learning_rate": 1.5998000000000003e-06,
+      "loss": 0.848,
+      "step": 42020
+    },
+    {
+      "epoch": 4.203,
+      "grad_norm": 24.58756446838379,
+      "learning_rate": 1.5978000000000002e-06,
+      "loss": 0.6043,
+      "step": 42030
+    },
+    {
+      "epoch": 4.204,
+      "grad_norm": 22.763195037841797,
+      "learning_rate": 1.5958e-06,
+      "loss": 0.4617,
+      "step": 42040
+    },
+    {
+      "epoch": 4.205,
+      "grad_norm": 6.496039867401123,
+      "learning_rate": 1.5938e-06,
+      "loss": 0.6126,
+      "step": 42050
+    },
+    {
+      "epoch": 4.206,
+      "grad_norm": 36.29051971435547,
+      "learning_rate": 1.5918e-06,
+      "loss": 0.5598,
+      "step": 42060
+    },
+    {
+      "epoch": 4.207,
+      "grad_norm": 28.738428115844727,
+      "learning_rate": 1.5898000000000002e-06,
+      "loss": 0.7017,
+      "step": 42070
+    },
+    {
+      "epoch": 4.208,
+      "grad_norm": 60.276710510253906,
+      "learning_rate": 1.5878e-06,
+      "loss": 0.8548,
+      "step": 42080
+    },
+    {
+      "epoch": 4.209,
+      "grad_norm": 44.32135772705078,
+      "learning_rate": 1.5858000000000001e-06,
+      "loss": 0.6744,
+      "step": 42090
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 35.0414924621582,
+      "learning_rate": 1.5838e-06,
+      "loss": 0.6921,
+      "step": 42100
+    },
+    {
+      "epoch": 4.211,
+      "grad_norm": 44.21099090576172,
+      "learning_rate": 1.5818e-06,
+      "loss": 0.5771,
+      "step": 42110
+    },
+    {
+      "epoch": 4.212,
+      "grad_norm": 38.59548568725586,
+      "learning_rate": 1.5798000000000003e-06,
+      "loss": 0.4153,
+      "step": 42120
+    },
+    {
+      "epoch": 4.213,
+      "grad_norm": 30.115324020385742,
+      "learning_rate": 1.5778000000000002e-06,
+      "loss": 0.363,
+      "step": 42130
+    },
+    {
+      "epoch": 4.214,
+      "grad_norm": 31.779865264892578,
+      "learning_rate": 1.5758e-06,
+      "loss": 0.7121,
+      "step": 42140
+    },
+    {
+      "epoch": 4.215,
+      "grad_norm": 55.575439453125,
+      "learning_rate": 1.5738e-06,
+      "loss": 0.7864,
+      "step": 42150
+    },
+    {
+      "epoch": 4.216,
+      "grad_norm": 49.964656829833984,
+      "learning_rate": 1.5718e-06,
+      "loss": 0.5532,
+      "step": 42160
+    },
+    {
+      "epoch": 4.217,
+      "grad_norm": 72.00129699707031,
+      "learning_rate": 1.5698000000000002e-06,
+      "loss": 0.5505,
+      "step": 42170
+    },
+    {
+      "epoch": 4.218,
+      "grad_norm": 49.91716003417969,
+      "learning_rate": 1.5678e-06,
+      "loss": 0.5834,
+      "step": 42180
+    },
+    {
+      "epoch": 4.219,
+      "grad_norm": 36.31462860107422,
+      "learning_rate": 1.5658000000000001e-06,
+      "loss": 0.478,
+      "step": 42190
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 47.1279182434082,
+      "learning_rate": 1.5638e-06,
+      "loss": 0.6494,
+      "step": 42200
+    },
+    {
+      "epoch": 4.221,
+      "grad_norm": 7.2586517333984375,
+      "learning_rate": 1.5618000000000002e-06,
+      "loss": 0.5058,
+      "step": 42210
+    },
+    {
+      "epoch": 4.222,
+      "grad_norm": 56.2879638671875,
+      "learning_rate": 1.5598000000000003e-06,
+      "loss": 0.5478,
+      "step": 42220
+    },
+    {
+      "epoch": 4.223,
+      "grad_norm": 32.43406295776367,
+      "learning_rate": 1.5578000000000001e-06,
+      "loss": 0.5427,
+      "step": 42230
+    },
+    {
+      "epoch": 4.224,
+      "grad_norm": 23.466623306274414,
+      "learning_rate": 1.5558e-06,
+      "loss": 0.5276,
+      "step": 42240
+    },
+    {
+      "epoch": 4.225,
+      "grad_norm": 42.27043533325195,
+      "learning_rate": 1.5538e-06,
+      "loss": 0.5317,
+      "step": 42250
+    },
+    {
+      "epoch": 4.226,
+      "grad_norm": 60.1855583190918,
+      "learning_rate": 1.5518000000000003e-06,
+      "loss": 0.7668,
+      "step": 42260
+    },
+    {
+      "epoch": 4.227,
+      "grad_norm": 58.870296478271484,
+      "learning_rate": 1.5498000000000002e-06,
+      "loss": 0.6913,
+      "step": 42270
+    },
+    {
+      "epoch": 4.228,
+      "grad_norm": 13.989805221557617,
+      "learning_rate": 1.5478000000000002e-06,
+      "loss": 0.4032,
+      "step": 42280
+    },
+    {
+      "epoch": 4.229,
+      "grad_norm": 18.71600914001465,
+      "learning_rate": 1.5458e-06,
+      "loss": 0.542,
+      "step": 42290
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 11.906790733337402,
+      "learning_rate": 1.5438e-06,
+      "loss": 0.6872,
+      "step": 42300
+    },
+    {
+      "epoch": 4.231,
+      "grad_norm": 19.292734146118164,
+      "learning_rate": 1.5418000000000002e-06,
+      "loss": 0.4605,
+      "step": 42310
+    },
+    {
+      "epoch": 4.232,
+      "grad_norm": 65.98567199707031,
+      "learning_rate": 1.5398000000000003e-06,
+      "loss": 0.4226,
+      "step": 42320
+    },
+    {
+      "epoch": 4.233,
+      "grad_norm": 12.095321655273438,
+      "learning_rate": 1.5378000000000001e-06,
+      "loss": 0.8225,
+      "step": 42330
+    },
+    {
+      "epoch": 4.234,
+      "grad_norm": 15.595905303955078,
+      "learning_rate": 1.5358e-06,
+      "loss": 0.4639,
+      "step": 42340
+    },
+    {
+      "epoch": 4.235,
+      "grad_norm": 48.75836944580078,
+      "learning_rate": 1.5338e-06,
+      "loss": 0.2764,
+      "step": 42350
+    },
+    {
+      "epoch": 4.236,
+      "grad_norm": 31.910316467285156,
+      "learning_rate": 1.5318000000000003e-06,
+      "loss": 0.6411,
+      "step": 42360
+    },
+    {
+      "epoch": 4.237,
+      "grad_norm": 47.41012954711914,
+      "learning_rate": 1.5298000000000002e-06,
+      "loss": 0.8059,
+      "step": 42370
+    },
+    {
+      "epoch": 4.2379999999999995,
+      "grad_norm": 47.09851837158203,
+      "learning_rate": 1.5278000000000002e-06,
+      "loss": 0.685,
+      "step": 42380
+    },
+    {
+      "epoch": 4.239,
+      "grad_norm": 61.549922943115234,
+      "learning_rate": 1.5258e-06,
+      "loss": 0.7116,
+      "step": 42390
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 41.90022659301758,
+      "learning_rate": 1.5238e-06,
+      "loss": 0.8182,
+      "step": 42400
+    },
+    {
+      "epoch": 4.241,
+      "grad_norm": 19.784992218017578,
+      "learning_rate": 1.5218000000000002e-06,
+      "loss": 0.6319,
+      "step": 42410
+    },
+    {
+      "epoch": 4.242,
+      "grad_norm": 14.872299194335938,
+      "learning_rate": 1.5198000000000003e-06,
+      "loss": 0.9768,
+      "step": 42420
+    },
+    {
+      "epoch": 4.243,
+      "grad_norm": 14.719264030456543,
+      "learning_rate": 1.5178e-06,
+      "loss": 0.5052,
+      "step": 42430
+    },
+    {
+      "epoch": 4.244,
+      "grad_norm": 24.612985610961914,
+      "learning_rate": 1.5158e-06,
+      "loss": 0.5563,
+      "step": 42440
+    },
+    {
+      "epoch": 4.245,
+      "grad_norm": 60.18811798095703,
+      "learning_rate": 1.5138e-06,
+      "loss": 0.5981,
+      "step": 42450
+    },
+    {
+      "epoch": 4.246,
+      "grad_norm": 80.66825103759766,
+      "learning_rate": 1.5118000000000003e-06,
+      "loss": 0.4633,
+      "step": 42460
+    },
+    {
+      "epoch": 4.247,
+      "grad_norm": 13.229582786560059,
+      "learning_rate": 1.5098000000000001e-06,
+      "loss": 0.6477,
+      "step": 42470
+    },
+    {
+      "epoch": 4.248,
+      "grad_norm": 2.7987141609191895,
+      "learning_rate": 1.5078000000000002e-06,
+      "loss": 0.3514,
+      "step": 42480
+    },
+    {
+      "epoch": 4.249,
+      "grad_norm": 44.73332214355469,
+      "learning_rate": 1.5058e-06,
+      "loss": 0.5872,
+      "step": 42490
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 51.61971664428711,
+      "learning_rate": 1.5038000000000003e-06,
+      "loss": 0.4737,
+      "step": 42500
+    },
+    {
+      "epoch": 4.251,
+      "grad_norm": 5.696041107177734,
+      "learning_rate": 1.5018000000000002e-06,
+      "loss": 0.6114,
+      "step": 42510
+    },
+    {
+      "epoch": 4.252,
+      "grad_norm": 19.12189483642578,
+      "learning_rate": 1.4998000000000002e-06,
+      "loss": 0.64,
+      "step": 42520
+    },
+    {
+      "epoch": 4.253,
+      "grad_norm": 54.96479797363281,
+      "learning_rate": 1.4978e-06,
+      "loss": 0.628,
+      "step": 42530
+    },
+    {
+      "epoch": 4.254,
+      "grad_norm": 13.414827346801758,
+      "learning_rate": 1.4958e-06,
+      "loss": 0.7258,
+      "step": 42540
+    },
+    {
+      "epoch": 4.255,
+      "grad_norm": 20.74048614501953,
+      "learning_rate": 1.4938000000000002e-06,
+      "loss": 0.6399,
+      "step": 42550
+    },
+    {
+      "epoch": 4.256,
+      "grad_norm": 17.05641746520996,
+      "learning_rate": 1.4918000000000003e-06,
+      "loss": 0.6818,
+      "step": 42560
+    },
+    {
+      "epoch": 4.257,
+      "grad_norm": 3.655008554458618,
+      "learning_rate": 1.4898000000000001e-06,
+      "loss": 0.6916,
+      "step": 42570
+    },
+    {
+      "epoch": 4.258,
+      "grad_norm": 29.011837005615234,
+      "learning_rate": 1.4878000000000002e-06,
+      "loss": 0.5478,
+      "step": 42580
+    },
+    {
+      "epoch": 4.259,
+      "grad_norm": 6.549126148223877,
+      "learning_rate": 1.4858e-06,
+      "loss": 0.3863,
+      "step": 42590
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 59.96338653564453,
+      "learning_rate": 1.4838000000000003e-06,
+      "loss": 0.7019,
+      "step": 42600
+    },
+    {
+      "epoch": 4.261,
+      "grad_norm": 3.527003526687622,
+      "learning_rate": 1.4818000000000002e-06,
+      "loss": 0.5075,
+      "step": 42610
+    },
+    {
+      "epoch": 4.2620000000000005,
+      "grad_norm": 66.73096466064453,
+      "learning_rate": 1.4798000000000002e-06,
+      "loss": 0.6057,
+      "step": 42620
+    },
+    {
+      "epoch": 4.263,
+      "grad_norm": 51.33205795288086,
+      "learning_rate": 1.4778e-06,
+      "loss": 0.5798,
+      "step": 42630
+    },
+    {
+      "epoch": 4.264,
+      "grad_norm": 61.584869384765625,
+      "learning_rate": 1.4758e-06,
+      "loss": 0.4595,
+      "step": 42640
+    },
+    {
+      "epoch": 4.265,
+      "grad_norm": 43.708984375,
+      "learning_rate": 1.4738000000000002e-06,
+      "loss": 0.5689,
+      "step": 42650
+    },
+    {
+      "epoch": 4.266,
+      "grad_norm": 16.016572952270508,
+      "learning_rate": 1.4718000000000002e-06,
+      "loss": 0.5122,
+      "step": 42660
+    },
+    {
+      "epoch": 4.267,
+      "grad_norm": 42.98859405517578,
+      "learning_rate": 1.4698e-06,
+      "loss": 0.4487,
+      "step": 42670
+    },
+    {
+      "epoch": 4.268,
+      "grad_norm": 48.663185119628906,
+      "learning_rate": 1.4678000000000002e-06,
+      "loss": 0.4685,
+      "step": 42680
+    },
+    {
+      "epoch": 4.269,
+      "grad_norm": 55.1309700012207,
+      "learning_rate": 1.4658e-06,
+      "loss": 0.4818,
+      "step": 42690
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 46.402252197265625,
+      "learning_rate": 1.4638000000000003e-06,
+      "loss": 0.52,
+      "step": 42700
+    },
+    {
+      "epoch": 4.271,
+      "grad_norm": 19.3892822265625,
+      "learning_rate": 1.4618000000000001e-06,
+      "loss": 0.5317,
+      "step": 42710
+    },
+    {
+      "epoch": 4.272,
+      "grad_norm": 40.837039947509766,
+      "learning_rate": 1.4598000000000002e-06,
+      "loss": 0.5411,
+      "step": 42720
+    },
+    {
+      "epoch": 4.273,
+      "grad_norm": 31.86037254333496,
+      "learning_rate": 1.4578e-06,
+      "loss": 0.5173,
+      "step": 42730
+    },
+    {
+      "epoch": 4.274,
+      "grad_norm": 62.265621185302734,
+      "learning_rate": 1.4557999999999999e-06,
+      "loss": 0.6121,
+      "step": 42740
+    },
+    {
+      "epoch": 4.275,
+      "grad_norm": 44.93138122558594,
+      "learning_rate": 1.4538000000000002e-06,
+      "loss": 0.54,
+      "step": 42750
+    },
+    {
+      "epoch": 4.276,
+      "grad_norm": 67.72014617919922,
+      "learning_rate": 1.4518000000000002e-06,
+      "loss": 0.784,
+      "step": 42760
+    },
+    {
+      "epoch": 4.277,
+      "grad_norm": 40.16845703125,
+      "learning_rate": 1.4498e-06,
+      "loss": 0.7528,
+      "step": 42770
+    },
+    {
+      "epoch": 4.2780000000000005,
+      "grad_norm": 61.454490661621094,
+      "learning_rate": 1.4478000000000001e-06,
+      "loss": 0.5798,
+      "step": 42780
+    },
+    {
+      "epoch": 4.279,
+      "grad_norm": 55.03683853149414,
+      "learning_rate": 1.4458e-06,
+      "loss": 0.756,
+      "step": 42790
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 42.04933547973633,
+      "learning_rate": 1.4438000000000003e-06,
+      "loss": 0.7275,
+      "step": 42800
+    },
+    {
+      "epoch": 4.281,
+      "grad_norm": 56.68568801879883,
+      "learning_rate": 1.4418000000000001e-06,
+      "loss": 0.5234,
+      "step": 42810
+    },
+    {
+      "epoch": 4.282,
+      "grad_norm": 6.141942501068115,
+      "learning_rate": 1.4398000000000002e-06,
+      "loss": 0.311,
+      "step": 42820
+    },
+    {
+      "epoch": 4.283,
+      "grad_norm": 36.63520050048828,
+      "learning_rate": 1.4378e-06,
+      "loss": 0.6934,
+      "step": 42830
+    },
+    {
+      "epoch": 4.284,
+      "grad_norm": 40.52085494995117,
+      "learning_rate": 1.4358000000000003e-06,
+      "loss": 0.5693,
+      "step": 42840
+    },
+    {
+      "epoch": 4.285,
+      "grad_norm": 25.857606887817383,
+      "learning_rate": 1.4338000000000001e-06,
+      "loss": 0.4119,
+      "step": 42850
+    },
+    {
+      "epoch": 4.286,
+      "grad_norm": 34.83437728881836,
+      "learning_rate": 1.4318000000000002e-06,
+      "loss": 0.6163,
+      "step": 42860
+    },
+    {
+      "epoch": 4.287,
+      "grad_norm": 9.676780700683594,
+      "learning_rate": 1.4298e-06,
+      "loss": 0.5552,
+      "step": 42870
+    },
+    {
+      "epoch": 4.288,
+      "grad_norm": 7.326340198516846,
+      "learning_rate": 1.4278000000000001e-06,
+      "loss": 0.4313,
+      "step": 42880
+    },
+    {
+      "epoch": 4.289,
+      "grad_norm": 45.25532913208008,
+      "learning_rate": 1.4258000000000002e-06,
+      "loss": 0.7117,
+      "step": 42890
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 9.764056205749512,
+      "learning_rate": 1.4238000000000002e-06,
+      "loss": 0.3815,
+      "step": 42900
+    },
+    {
+      "epoch": 4.291,
+      "grad_norm": 20.11690902709961,
+      "learning_rate": 1.4218e-06,
+      "loss": 0.5441,
+      "step": 42910
+    },
+    {
+      "epoch": 4.292,
+      "grad_norm": 26.8229923248291,
+      "learning_rate": 1.4198000000000002e-06,
+      "loss": 0.7206,
+      "step": 42920
+    },
+    {
+      "epoch": 4.293,
+      "grad_norm": 30.303131103515625,
+      "learning_rate": 1.4178e-06,
+      "loss": 0.3032,
+      "step": 42930
+    },
+    {
+      "epoch": 4.294,
+      "grad_norm": 47.463016510009766,
+      "learning_rate": 1.4158000000000003e-06,
+      "loss": 0.4759,
+      "step": 42940
+    },
+    {
+      "epoch": 4.295,
+      "grad_norm": 57.69182586669922,
+      "learning_rate": 1.4138000000000001e-06,
+      "loss": 0.6244,
+      "step": 42950
+    },
+    {
+      "epoch": 4.296,
+      "grad_norm": 40.416038513183594,
+      "learning_rate": 1.4118000000000002e-06,
+      "loss": 0.5296,
+      "step": 42960
+    },
+    {
+      "epoch": 4.297,
+      "grad_norm": 36.48316955566406,
+      "learning_rate": 1.4098e-06,
+      "loss": 0.6653,
+      "step": 42970
+    },
+    {
+      "epoch": 4.298,
+      "grad_norm": 4.3902764320373535,
+      "learning_rate": 1.4078e-06,
+      "loss": 0.5357,
+      "step": 42980
+    },
+    {
+      "epoch": 4.299,
+      "grad_norm": 11.479517936706543,
+      "learning_rate": 1.4058000000000002e-06,
+      "loss": 0.5215,
+      "step": 42990
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 26.62324333190918,
+      "learning_rate": 1.4038000000000002e-06,
+      "loss": 0.6729,
+      "step": 43000
+    },
+    {
+      "epoch": 4.301,
+      "grad_norm": 79.1596908569336,
+      "learning_rate": 1.4018e-06,
+      "loss": 0.7622,
+      "step": 43010
+    },
+    {
+      "epoch": 4.302,
+      "grad_norm": 43.866886138916016,
+      "learning_rate": 1.3998000000000001e-06,
+      "loss": 0.5798,
+      "step": 43020
+    },
+    {
+      "epoch": 4.303,
+      "grad_norm": 75.57432556152344,
+      "learning_rate": 1.3978e-06,
+      "loss": 0.7021,
+      "step": 43030
+    },
+    {
+      "epoch": 4.304,
+      "grad_norm": 55.48227310180664,
+      "learning_rate": 1.3958000000000003e-06,
+      "loss": 0.5015,
+      "step": 43040
+    },
+    {
+      "epoch": 4.305,
+      "grad_norm": 64.40435791015625,
+      "learning_rate": 1.3938000000000001e-06,
+      "loss": 0.787,
+      "step": 43050
+    },
+    {
+      "epoch": 4.306,
+      "grad_norm": 32.22929000854492,
+      "learning_rate": 1.3918000000000002e-06,
+      "loss": 0.6516,
+      "step": 43060
+    },
+    {
+      "epoch": 4.307,
+      "grad_norm": 11.515064239501953,
+      "learning_rate": 1.3898e-06,
+      "loss": 0.3711,
+      "step": 43070
+    },
+    {
+      "epoch": 4.308,
+      "grad_norm": 68.9782485961914,
+      "learning_rate": 1.3878e-06,
+      "loss": 0.522,
+      "step": 43080
+    },
+    {
+      "epoch": 4.309,
+      "grad_norm": 35.55900573730469,
+      "learning_rate": 1.3858000000000001e-06,
+      "loss": 0.4953,
+      "step": 43090
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 49.99945831298828,
+      "learning_rate": 1.3838000000000002e-06,
+      "loss": 0.5207,
+      "step": 43100
+    },
+    {
+      "epoch": 4.311,
+      "grad_norm": 3.150254726409912,
+      "learning_rate": 1.3818e-06,
+      "loss": 0.4207,
+      "step": 43110
+    },
+    {
+      "epoch": 4.312,
+      "grad_norm": 6.824671745300293,
+      "learning_rate": 1.3798000000000001e-06,
+      "loss": 0.66,
+      "step": 43120
+    },
+    {
+      "epoch": 4.313,
+      "grad_norm": 7.719470500946045,
+      "learning_rate": 1.3778000000000002e-06,
+      "loss": 0.5685,
+      "step": 43130
+    },
+    {
+      "epoch": 4.314,
+      "grad_norm": 10.207037925720215,
+      "learning_rate": 1.3758000000000002e-06,
+      "loss": 0.5784,
+      "step": 43140
+    },
+    {
+      "epoch": 4.315,
+      "grad_norm": 43.468528747558594,
+      "learning_rate": 1.3738e-06,
+      "loss": 0.4621,
+      "step": 43150
+    },
+    {
+      "epoch": 4.316,
+      "grad_norm": 35.88478469848633,
+      "learning_rate": 1.3718000000000001e-06,
+      "loss": 0.3395,
+      "step": 43160
+    },
+    {
+      "epoch": 4.317,
+      "grad_norm": 4.721527576446533,
+      "learning_rate": 1.3698e-06,
+      "loss": 0.3421,
+      "step": 43170
+    },
+    {
+      "epoch": 4.318,
+      "grad_norm": 60.953182220458984,
+      "learning_rate": 1.3678000000000003e-06,
+      "loss": 0.6947,
+      "step": 43180
+    },
+    {
+      "epoch": 4.319,
+      "grad_norm": 11.123573303222656,
+      "learning_rate": 1.3658000000000001e-06,
+      "loss": 0.5041,
+      "step": 43190
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 66.29830932617188,
+      "learning_rate": 1.3638000000000002e-06,
+      "loss": 0.9367,
+      "step": 43200
+    },
+    {
+      "epoch": 4.321,
+      "grad_norm": 24.80149269104004,
+      "learning_rate": 1.3618e-06,
+      "loss": 0.5205,
+      "step": 43210
+    },
+    {
+      "epoch": 4.322,
+      "grad_norm": 41.04637145996094,
+      "learning_rate": 1.3598e-06,
+      "loss": 0.4256,
+      "step": 43220
+    },
+    {
+      "epoch": 4.323,
+      "grad_norm": 52.53520202636719,
+      "learning_rate": 1.3578000000000002e-06,
+      "loss": 0.6093,
+      "step": 43230
+    },
+    {
+      "epoch": 4.324,
+      "grad_norm": 2.6135032176971436,
+      "learning_rate": 1.3558000000000002e-06,
+      "loss": 0.2796,
+      "step": 43240
+    },
+    {
+      "epoch": 4.325,
+      "grad_norm": 59.64257049560547,
+      "learning_rate": 1.3538e-06,
+      "loss": 0.7455,
+      "step": 43250
+    },
+    {
+      "epoch": 4.326,
+      "grad_norm": 26.32259750366211,
+      "learning_rate": 1.3518000000000001e-06,
+      "loss": 0.3787,
+      "step": 43260
+    },
+    {
+      "epoch": 4.327,
+      "grad_norm": 7.690916538238525,
+      "learning_rate": 1.3498e-06,
+      "loss": 0.5234,
+      "step": 43270
+    },
+    {
+      "epoch": 4.328,
+      "grad_norm": 41.93429946899414,
+      "learning_rate": 1.3478000000000003e-06,
+      "loss": 0.5841,
+      "step": 43280
+    },
+    {
+      "epoch": 4.329,
+      "grad_norm": 15.757827758789062,
+      "learning_rate": 1.3458e-06,
+      "loss": 0.3595,
+      "step": 43290
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 15.618181228637695,
+      "learning_rate": 1.3438000000000002e-06,
+      "loss": 0.6405,
+      "step": 43300
+    },
+    {
+      "epoch": 4.331,
+      "grad_norm": 35.363834381103516,
+      "learning_rate": 1.3418e-06,
+      "loss": 0.5346,
+      "step": 43310
+    },
+    {
+      "epoch": 4.332,
+      "grad_norm": 18.901430130004883,
+      "learning_rate": 1.3398e-06,
+      "loss": 0.7658,
+      "step": 43320
+    },
+    {
+      "epoch": 4.333,
+      "grad_norm": 15.545169830322266,
+      "learning_rate": 1.3378000000000001e-06,
+      "loss": 0.3699,
+      "step": 43330
+    },
+    {
+      "epoch": 4.334,
+      "grad_norm": 16.735557556152344,
+      "learning_rate": 1.3358000000000002e-06,
+      "loss": 0.3499,
+      "step": 43340
+    },
+    {
+      "epoch": 4.335,
+      "grad_norm": 13.97559928894043,
+      "learning_rate": 1.3338e-06,
+      "loss": 0.522,
+      "step": 43350
+    },
+    {
+      "epoch": 4.336,
+      "grad_norm": 58.62332534790039,
+      "learning_rate": 1.3318000000000001e-06,
+      "loss": 0.6854,
+      "step": 43360
+    },
+    {
+      "epoch": 4.337,
+      "grad_norm": 38.00271224975586,
+      "learning_rate": 1.3298e-06,
+      "loss": 0.4987,
+      "step": 43370
+    },
+    {
+      "epoch": 4.338,
+      "grad_norm": 48.7909049987793,
+      "learning_rate": 1.3278000000000002e-06,
+      "loss": 0.536,
+      "step": 43380
+    },
+    {
+      "epoch": 4.339,
+      "grad_norm": 10.43138599395752,
+      "learning_rate": 1.3258e-06,
+      "loss": 0.568,
+      "step": 43390
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 47.01791000366211,
+      "learning_rate": 1.324e-06,
+      "loss": 0.4428,
+      "step": 43400
+    },
+    {
+      "epoch": 4.341,
+      "grad_norm": 2.758685350418091,
+      "learning_rate": 1.3220000000000002e-06,
+      "loss": 0.4351,
+      "step": 43410
+    },
+    {
+      "epoch": 4.342,
+      "grad_norm": 9.140447616577148,
+      "learning_rate": 1.32e-06,
+      "loss": 0.3583,
+      "step": 43420
+    },
+    {
+      "epoch": 4.343,
+      "grad_norm": 21.019405364990234,
+      "learning_rate": 1.3180000000000001e-06,
+      "loss": 0.6468,
+      "step": 43430
+    },
+    {
+      "epoch": 4.344,
+      "grad_norm": 10.030232429504395,
+      "learning_rate": 1.316e-06,
+      "loss": 0.6114,
+      "step": 43440
+    },
+    {
+      "epoch": 4.345,
+      "grad_norm": 21.26093101501465,
+      "learning_rate": 1.314e-06,
+      "loss": 0.3142,
+      "step": 43450
+    },
+    {
+      "epoch": 4.346,
+      "grad_norm": 31.151533126831055,
+      "learning_rate": 1.3120000000000003e-06,
+      "loss": 0.3533,
+      "step": 43460
+    },
+    {
+      "epoch": 4.3469999999999995,
+      "grad_norm": 28.893701553344727,
+      "learning_rate": 1.3100000000000002e-06,
+      "loss": 0.614,
+      "step": 43470
+    },
+    {
+      "epoch": 4.348,
+      "grad_norm": 67.40365600585938,
+      "learning_rate": 1.308e-06,
+      "loss": 0.4243,
+      "step": 43480
+    },
+    {
+      "epoch": 4.349,
+      "grad_norm": 28.138734817504883,
+      "learning_rate": 1.306e-06,
+      "loss": 0.4,
+      "step": 43490
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 36.38271713256836,
+      "learning_rate": 1.304e-06,
+      "loss": 0.7892,
+      "step": 43500
+    },
+    {
+      "epoch": 4.351,
+      "grad_norm": 36.84843826293945,
+      "learning_rate": 1.3020000000000002e-06,
+      "loss": 0.5398,
+      "step": 43510
+    },
+    {
+      "epoch": 4.352,
+      "grad_norm": 50.559635162353516,
+      "learning_rate": 1.3e-06,
+      "loss": 0.7597,
+      "step": 43520
+    },
+    {
+      "epoch": 4.353,
+      "grad_norm": 18.063173294067383,
+      "learning_rate": 1.2980000000000001e-06,
+      "loss": 0.5857,
+      "step": 43530
+    },
+    {
+      "epoch": 4.354,
+      "grad_norm": 55.33283615112305,
+      "learning_rate": 1.296e-06,
+      "loss": 0.6993,
+      "step": 43540
+    },
+    {
+      "epoch": 4.355,
+      "grad_norm": 48.159889221191406,
+      "learning_rate": 1.294e-06,
+      "loss": 0.6422,
+      "step": 43550
+    },
+    {
+      "epoch": 4.356,
+      "grad_norm": 56.55753707885742,
+      "learning_rate": 1.2920000000000003e-06,
+      "loss": 0.5918,
+      "step": 43560
+    },
+    {
+      "epoch": 4.357,
+      "grad_norm": 28.358430862426758,
+      "learning_rate": 1.2900000000000001e-06,
+      "loss": 0.5238,
+      "step": 43570
+    },
+    {
+      "epoch": 4.358,
+      "grad_norm": 16.447559356689453,
+      "learning_rate": 1.288e-06,
+      "loss": 0.8666,
+      "step": 43580
+    },
+    {
+      "epoch": 4.359,
+      "grad_norm": 22.63037872314453,
+      "learning_rate": 1.286e-06,
+      "loss": 0.5798,
+      "step": 43590
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 6.575098991394043,
+      "learning_rate": 1.284e-06,
+      "loss": 0.5513,
+      "step": 43600
+    },
+    {
+      "epoch": 4.361,
+      "grad_norm": 30.44253158569336,
+      "learning_rate": 1.2820000000000002e-06,
+      "loss": 0.651,
+      "step": 43610
+    },
+    {
+      "epoch": 4.362,
+      "grad_norm": 62.634090423583984,
+      "learning_rate": 1.28e-06,
+      "loss": 0.9522,
+      "step": 43620
+    },
+    {
+      "epoch": 4.3629999999999995,
+      "grad_norm": 42.4155158996582,
+      "learning_rate": 1.278e-06,
+      "loss": 0.681,
+      "step": 43630
+    },
+    {
+      "epoch": 4.364,
+      "grad_norm": 7.64183235168457,
+      "learning_rate": 1.276e-06,
+      "loss": 0.3704,
+      "step": 43640
+    },
+    {
+      "epoch": 4.365,
+      "grad_norm": 55.91519546508789,
+      "learning_rate": 1.2740000000000002e-06,
+      "loss": 0.956,
+      "step": 43650
+    },
+    {
+      "epoch": 4.366,
+      "grad_norm": 43.432830810546875,
+      "learning_rate": 1.2720000000000003e-06,
+      "loss": 0.5726,
+      "step": 43660
+    },
+    {
+      "epoch": 4.367,
+      "grad_norm": 5.294198513031006,
+      "learning_rate": 1.2700000000000001e-06,
+      "loss": 0.5203,
+      "step": 43670
+    },
+    {
+      "epoch": 4.368,
+      "grad_norm": 49.72700881958008,
+      "learning_rate": 1.268e-06,
+      "loss": 0.5669,
+      "step": 43680
+    },
+    {
+      "epoch": 4.369,
+      "grad_norm": 25.87734031677246,
+      "learning_rate": 1.266e-06,
+      "loss": 0.7345,
+      "step": 43690
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 34.031837463378906,
+      "learning_rate": 1.2640000000000003e-06,
+      "loss": 0.6073,
+      "step": 43700
+    },
+    {
+      "epoch": 4.371,
+      "grad_norm": 8.156420707702637,
+      "learning_rate": 1.2620000000000002e-06,
+      "loss": 0.548,
+      "step": 43710
+    },
+    {
+      "epoch": 4.372,
+      "grad_norm": 28.363479614257812,
+      "learning_rate": 1.26e-06,
+      "loss": 0.699,
+      "step": 43720
+    },
+    {
+      "epoch": 4.373,
+      "grad_norm": 27.95302391052246,
+      "learning_rate": 1.258e-06,
+      "loss": 0.5499,
+      "step": 43730
+    },
+    {
+      "epoch": 4.374,
+      "grad_norm": 28.509695053100586,
+      "learning_rate": 1.256e-06,
+      "loss": 0.4613,
+      "step": 43740
+    },
+    {
+      "epoch": 4.375,
+      "grad_norm": 23.42409324645996,
+      "learning_rate": 1.2540000000000002e-06,
+      "loss": 0.5041,
+      "step": 43750
+    },
+    {
+      "epoch": 4.376,
+      "grad_norm": 75.357666015625,
+      "learning_rate": 1.2520000000000003e-06,
+      "loss": 0.8561,
+      "step": 43760
+    },
+    {
+      "epoch": 4.377,
+      "grad_norm": 19.810129165649414,
+      "learning_rate": 1.25e-06,
+      "loss": 0.5149,
+      "step": 43770
+    },
+    {
+      "epoch": 4.378,
+      "grad_norm": 49.86960983276367,
+      "learning_rate": 1.248e-06,
+      "loss": 1.0917,
+      "step": 43780
+    },
+    {
+      "epoch": 4.379,
+      "grad_norm": 48.763427734375,
+      "learning_rate": 1.2460000000000002e-06,
+      "loss": 0.5727,
+      "step": 43790
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 40.76609420776367,
+      "learning_rate": 1.244e-06,
+      "loss": 0.6641,
+      "step": 43800
+    },
+    {
+      "epoch": 4.381,
+      "grad_norm": 24.340974807739258,
+      "learning_rate": 1.2420000000000001e-06,
+      "loss": 0.3564,
+      "step": 43810
+    },
+    {
+      "epoch": 4.382,
+      "grad_norm": 20.576501846313477,
+      "learning_rate": 1.2400000000000002e-06,
+      "loss": 0.4448,
+      "step": 43820
+    },
+    {
+      "epoch": 4.383,
+      "grad_norm": 33.2127571105957,
+      "learning_rate": 1.238e-06,
+      "loss": 0.7988,
+      "step": 43830
+    },
+    {
+      "epoch": 4.384,
+      "grad_norm": 62.10398864746094,
+      "learning_rate": 1.2360000000000001e-06,
+      "loss": 0.4453,
+      "step": 43840
+    },
+    {
+      "epoch": 4.385,
+      "grad_norm": 33.38579177856445,
+      "learning_rate": 1.234e-06,
+      "loss": 0.5619,
+      "step": 43850
+    },
+    {
+      "epoch": 4.386,
+      "grad_norm": 17.548891067504883,
+      "learning_rate": 1.2320000000000002e-06,
+      "loss": 0.3925,
+      "step": 43860
+    },
+    {
+      "epoch": 4.3870000000000005,
+      "grad_norm": 12.536221504211426,
+      "learning_rate": 1.23e-06,
+      "loss": 0.2319,
+      "step": 43870
+    },
+    {
+      "epoch": 4.388,
+      "grad_norm": 12.191761016845703,
+      "learning_rate": 1.2280000000000001e-06,
+      "loss": 0.7507,
+      "step": 43880
+    },
+    {
+      "epoch": 4.389,
+      "grad_norm": 26.444080352783203,
+      "learning_rate": 1.2260000000000002e-06,
+      "loss": 0.5662,
+      "step": 43890
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 52.09244918823242,
+      "learning_rate": 1.224e-06,
+      "loss": 0.5521,
+      "step": 43900
+    },
+    {
+      "epoch": 4.391,
+      "grad_norm": 42.6751594543457,
+      "learning_rate": 1.2220000000000001e-06,
+      "loss": 0.5683,
+      "step": 43910
+    },
+    {
+      "epoch": 4.392,
+      "grad_norm": 102.61528015136719,
+      "learning_rate": 1.2200000000000002e-06,
+      "loss": 0.5493,
+      "step": 43920
+    },
+    {
+      "epoch": 4.393,
+      "grad_norm": 37.700775146484375,
+      "learning_rate": 1.2180000000000002e-06,
+      "loss": 0.4303,
+      "step": 43930
+    },
+    {
+      "epoch": 4.394,
+      "grad_norm": 12.019837379455566,
+      "learning_rate": 1.216e-06,
+      "loss": 0.3392,
+      "step": 43940
+    },
+    {
+      "epoch": 4.395,
+      "grad_norm": 16.28959846496582,
+      "learning_rate": 1.214e-06,
+      "loss": 0.5402,
+      "step": 43950
+    },
+    {
+      "epoch": 4.396,
+      "grad_norm": 27.6002197265625,
+      "learning_rate": 1.2120000000000002e-06,
+      "loss": 0.7572,
+      "step": 43960
+    },
+    {
+      "epoch": 4.397,
+      "grad_norm": 74.15684509277344,
+      "learning_rate": 1.21e-06,
+      "loss": 0.4435,
+      "step": 43970
+    },
+    {
+      "epoch": 4.398,
+      "grad_norm": 7.384476661682129,
+      "learning_rate": 1.2080000000000001e-06,
+      "loss": 0.3989,
+      "step": 43980
+    },
+    {
+      "epoch": 4.399,
+      "grad_norm": 21.462980270385742,
+      "learning_rate": 1.2060000000000002e-06,
+      "loss": 0.4035,
+      "step": 43990
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 50.33262252807617,
+      "learning_rate": 1.204e-06,
+      "loss": 0.5787,
+      "step": 44000
+    },
+    {
+      "epoch": 4.401,
+      "grad_norm": 28.74320411682129,
+      "learning_rate": 1.202e-06,
+      "loss": 0.4512,
+      "step": 44010
+    },
+    {
+      "epoch": 4.402,
+      "grad_norm": 6.188956260681152,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.5807,
+      "step": 44020
+    },
+    {
+      "epoch": 4.4030000000000005,
+      "grad_norm": 16.039443969726562,
+      "learning_rate": 1.1980000000000002e-06,
+      "loss": 0.3878,
+      "step": 44030
+    },
+    {
+      "epoch": 4.404,
+      "grad_norm": 4.788004398345947,
+      "learning_rate": 1.196e-06,
+      "loss": 0.7235,
+      "step": 44040
+    },
+    {
+      "epoch": 4.405,
+      "grad_norm": 34.19270324707031,
+      "learning_rate": 1.1940000000000001e-06,
+      "loss": 0.2703,
+      "step": 44050
+    },
+    {
+      "epoch": 4.406,
+      "grad_norm": 17.581157684326172,
+      "learning_rate": 1.1920000000000002e-06,
+      "loss": 0.6753,
+      "step": 44060
+    },
+    {
+      "epoch": 4.407,
+      "grad_norm": 37.96449661254883,
+      "learning_rate": 1.19e-06,
+      "loss": 0.4106,
+      "step": 44070
+    },
+    {
+      "epoch": 4.408,
+      "grad_norm": 35.01892852783203,
+      "learning_rate": 1.188e-06,
+      "loss": 0.4215,
+      "step": 44080
+    },
+    {
+      "epoch": 4.409,
+      "grad_norm": 7.498485565185547,
+      "learning_rate": 1.1860000000000002e-06,
+      "loss": 0.6555,
+      "step": 44090
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 43.318233489990234,
+      "learning_rate": 1.1840000000000002e-06,
+      "loss": 0.8606,
+      "step": 44100
+    },
+    {
+      "epoch": 4.411,
+      "grad_norm": 45.79195022583008,
+      "learning_rate": 1.182e-06,
+      "loss": 0.4382,
+      "step": 44110
+    },
+    {
+      "epoch": 4.412,
+      "grad_norm": 55.97035598754883,
+      "learning_rate": 1.1800000000000001e-06,
+      "loss": 0.6438,
+      "step": 44120
+    },
+    {
+      "epoch": 4.413,
+      "grad_norm": 11.8862943649292,
+      "learning_rate": 1.1780000000000002e-06,
+      "loss": 0.3783,
+      "step": 44130
+    },
+    {
+      "epoch": 4.414,
+      "grad_norm": 53.719871520996094,
+      "learning_rate": 1.176e-06,
+      "loss": 0.476,
+      "step": 44140
+    },
+    {
+      "epoch": 4.415,
+      "grad_norm": 47.69070053100586,
+      "learning_rate": 1.1740000000000001e-06,
+      "loss": 0.4777,
+      "step": 44150
+    },
+    {
+      "epoch": 4.416,
+      "grad_norm": 60.3067626953125,
+      "learning_rate": 1.1720000000000002e-06,
+      "loss": 0.782,
+      "step": 44160
+    },
+    {
+      "epoch": 4.417,
+      "grad_norm": 53.17386245727539,
+      "learning_rate": 1.1700000000000002e-06,
+      "loss": 0.5574,
+      "step": 44170
+    },
+    {
+      "epoch": 4.418,
+      "grad_norm": 71.64693450927734,
+      "learning_rate": 1.168e-06,
+      "loss": 0.7901,
+      "step": 44180
+    },
+    {
+      "epoch": 4.419,
+      "grad_norm": 7.084906101226807,
+      "learning_rate": 1.1660000000000001e-06,
+      "loss": 0.5609,
+      "step": 44190
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 42.32343673706055,
+      "learning_rate": 1.1640000000000002e-06,
+      "loss": 0.6591,
+      "step": 44200
+    },
+    {
+      "epoch": 4.421,
+      "grad_norm": 1.9073995351791382,
+      "learning_rate": 1.162e-06,
+      "loss": 0.643,
+      "step": 44210
+    },
+    {
+      "epoch": 4.422,
+      "grad_norm": 3.4417290687561035,
+      "learning_rate": 1.1600000000000001e-06,
+      "loss": 0.4076,
+      "step": 44220
+    },
+    {
+      "epoch": 4.423,
+      "grad_norm": 5.756375789642334,
+      "learning_rate": 1.1580000000000002e-06,
+      "loss": 0.644,
+      "step": 44230
+    },
+    {
+      "epoch": 4.424,
+      "grad_norm": 2.408170461654663,
+      "learning_rate": 1.156e-06,
+      "loss": 0.7076,
+      "step": 44240
+    },
+    {
+      "epoch": 4.425,
+      "grad_norm": 47.58955383300781,
+      "learning_rate": 1.154e-06,
+      "loss": 0.532,
+      "step": 44250
+    },
+    {
+      "epoch": 4.426,
+      "grad_norm": 43.46930694580078,
+      "learning_rate": 1.1520000000000002e-06,
+      "loss": 0.636,
+      "step": 44260
+    },
+    {
+      "epoch": 4.427,
+      "grad_norm": 76.24018859863281,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 0.5307,
+      "step": 44270
+    },
+    {
+      "epoch": 4.428,
+      "grad_norm": 60.64769744873047,
+      "learning_rate": 1.148e-06,
+      "loss": 0.5436,
+      "step": 44280
+    },
+    {
+      "epoch": 4.429,
+      "grad_norm": 8.244017601013184,
+      "learning_rate": 1.1460000000000001e-06,
+      "loss": 0.5549,
+      "step": 44290
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 5.509671211242676,
+      "learning_rate": 1.1440000000000002e-06,
+      "loss": 0.4393,
+      "step": 44300
+    },
+    {
+      "epoch": 4.431,
+      "grad_norm": 41.25537109375,
+      "learning_rate": 1.142e-06,
+      "loss": 0.5594,
+      "step": 44310
+    },
+    {
+      "epoch": 4.432,
+      "grad_norm": 64.00753784179688,
+      "learning_rate": 1.14e-06,
+      "loss": 0.8429,
+      "step": 44320
+    },
+    {
+      "epoch": 4.433,
+      "grad_norm": 13.171112060546875,
+      "learning_rate": 1.1380000000000002e-06,
+      "loss": 0.715,
+      "step": 44330
+    },
+    {
+      "epoch": 4.434,
+      "grad_norm": 21.44270896911621,
+      "learning_rate": 1.1360000000000002e-06,
+      "loss": 0.4565,
+      "step": 44340
+    },
+    {
+      "epoch": 4.435,
+      "grad_norm": 17.770322799682617,
+      "learning_rate": 1.134e-06,
+      "loss": 0.5274,
+      "step": 44350
+    },
+    {
+      "epoch": 4.436,
+      "grad_norm": 1.8393038511276245,
+      "learning_rate": 1.1320000000000001e-06,
+      "loss": 0.4659,
+      "step": 44360
+    },
+    {
+      "epoch": 4.437,
+      "grad_norm": 2.2791614532470703,
+      "learning_rate": 1.1300000000000002e-06,
+      "loss": 0.3917,
+      "step": 44370
+    },
+    {
+      "epoch": 4.438,
+      "grad_norm": 99.7815170288086,
+      "learning_rate": 1.128e-06,
+      "loss": 0.6234,
+      "step": 44380
+    },
+    {
+      "epoch": 4.439,
+      "grad_norm": 59.29553985595703,
+      "learning_rate": 1.126e-06,
+      "loss": 0.6075,
+      "step": 44390
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 19.124174118041992,
+      "learning_rate": 1.1240000000000002e-06,
+      "loss": 0.8834,
+      "step": 44400
+    },
+    {
+      "epoch": 4.441,
+      "grad_norm": 17.70714569091797,
+      "learning_rate": 1.122e-06,
+      "loss": 0.3932,
+      "step": 44410
+    },
+    {
+      "epoch": 4.442,
+      "grad_norm": 59.30381774902344,
+      "learning_rate": 1.12e-06,
+      "loss": 0.8754,
+      "step": 44420
+    },
+    {
+      "epoch": 4.443,
+      "grad_norm": 17.972822189331055,
+      "learning_rate": 1.1180000000000001e-06,
+      "loss": 0.4991,
+      "step": 44430
+    },
+    {
+      "epoch": 4.444,
+      "grad_norm": 5.084202289581299,
+      "learning_rate": 1.1160000000000002e-06,
+      "loss": 0.9272,
+      "step": 44440
+    },
+    {
+      "epoch": 4.445,
+      "grad_norm": 21.268983840942383,
+      "learning_rate": 1.114e-06,
+      "loss": 0.5014,
+      "step": 44450
+    },
+    {
+      "epoch": 4.446,
+      "grad_norm": 46.41943359375,
+      "learning_rate": 1.1120000000000001e-06,
+      "loss": 0.7518,
+      "step": 44460
+    },
+    {
+      "epoch": 4.447,
+      "grad_norm": 25.36971664428711,
+      "learning_rate": 1.1100000000000002e-06,
+      "loss": 0.776,
+      "step": 44470
+    },
+    {
+      "epoch": 4.448,
+      "grad_norm": 57.427738189697266,
+      "learning_rate": 1.108e-06,
+      "loss": 0.469,
+      "step": 44480
+    },
+    {
+      "epoch": 4.449,
+      "grad_norm": 17.008895874023438,
+      "learning_rate": 1.106e-06,
+      "loss": 0.6733,
+      "step": 44490
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 13.199875831604004,
+      "learning_rate": 1.1040000000000001e-06,
+      "loss": 0.7869,
+      "step": 44500
+    },
+    {
+      "epoch": 4.451,
+      "grad_norm": 24.979137420654297,
+      "learning_rate": 1.1020000000000002e-06,
+      "loss": 0.5413,
+      "step": 44510
+    },
+    {
+      "epoch": 4.452,
+      "grad_norm": 34.9896354675293,
+      "learning_rate": 1.1e-06,
+      "loss": 0.6921,
+      "step": 44520
+    },
+    {
+      "epoch": 4.453,
+      "grad_norm": 34.93574523925781,
+      "learning_rate": 1.0980000000000001e-06,
+      "loss": 0.418,
+      "step": 44530
+    },
+    {
+      "epoch": 4.454,
+      "grad_norm": 2.021084785461426,
+      "learning_rate": 1.0960000000000002e-06,
+      "loss": 0.4103,
+      "step": 44540
+    },
+    {
+      "epoch": 4.455,
+      "grad_norm": 5.255341529846191,
+      "learning_rate": 1.094e-06,
+      "loss": 0.4357,
+      "step": 44550
+    },
+    {
+      "epoch": 4.456,
+      "grad_norm": 60.97416687011719,
+      "learning_rate": 1.092e-06,
+      "loss": 0.4744,
+      "step": 44560
+    },
+    {
+      "epoch": 4.457,
+      "grad_norm": 3.359767198562622,
+      "learning_rate": 1.0900000000000002e-06,
+      "loss": 0.5682,
+      "step": 44570
+    },
+    {
+      "epoch": 4.458,
+      "grad_norm": 6.942392826080322,
+      "learning_rate": 1.088e-06,
+      "loss": 0.7497,
+      "step": 44580
+    },
+    {
+      "epoch": 4.459,
+      "grad_norm": 35.56052780151367,
+      "learning_rate": 1.086e-06,
+      "loss": 0.903,
+      "step": 44590
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 53.783538818359375,
+      "learning_rate": 1.0840000000000001e-06,
+      "loss": 0.7616,
+      "step": 44600
+    },
+    {
+      "epoch": 4.461,
+      "grad_norm": 26.523746490478516,
+      "learning_rate": 1.0820000000000002e-06,
+      "loss": 0.5566,
+      "step": 44610
+    },
+    {
+      "epoch": 4.462,
+      "grad_norm": 41.27775573730469,
+      "learning_rate": 1.08e-06,
+      "loss": 0.8386,
+      "step": 44620
+    },
+    {
+      "epoch": 4.463,
+      "grad_norm": 15.443517684936523,
+      "learning_rate": 1.078e-06,
+      "loss": 0.4062,
+      "step": 44630
+    },
+    {
+      "epoch": 4.464,
+      "grad_norm": 39.41666793823242,
+      "learning_rate": 1.0760000000000002e-06,
+      "loss": 0.5806,
+      "step": 44640
+    },
+    {
+      "epoch": 4.465,
+      "grad_norm": 19.193965911865234,
+      "learning_rate": 1.074e-06,
+      "loss": 0.5718,
+      "step": 44650
+    },
+    {
+      "epoch": 4.466,
+      "grad_norm": 3.6427814960479736,
+      "learning_rate": 1.072e-06,
+      "loss": 0.4269,
+      "step": 44660
+    },
+    {
+      "epoch": 4.467,
+      "grad_norm": 50.77950668334961,
+      "learning_rate": 1.0700000000000001e-06,
+      "loss": 0.6791,
+      "step": 44670
+    },
+    {
+      "epoch": 4.468,
+      "grad_norm": 29.322980880737305,
+      "learning_rate": 1.0680000000000002e-06,
+      "loss": 0.4933,
+      "step": 44680
+    },
+    {
+      "epoch": 4.469,
+      "grad_norm": 86.63690185546875,
+      "learning_rate": 1.066e-06,
+      "loss": 0.666,
+      "step": 44690
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 11.375699996948242,
+      "learning_rate": 1.064e-06,
+      "loss": 0.4718,
+      "step": 44700
+    },
+    {
+      "epoch": 4.471,
+      "grad_norm": 12.194612503051758,
+      "learning_rate": 1.0620000000000002e-06,
+      "loss": 0.5804,
+      "step": 44710
+    },
+    {
+      "epoch": 4.4719999999999995,
+      "grad_norm": 33.018497467041016,
+      "learning_rate": 1.06e-06,
+      "loss": 0.6052,
+      "step": 44720
+    },
+    {
+      "epoch": 4.473,
+      "grad_norm": 62.320247650146484,
+      "learning_rate": 1.058e-06,
+      "loss": 0.7786,
+      "step": 44730
+    },
+    {
+      "epoch": 4.474,
+      "grad_norm": 2.1804471015930176,
+      "learning_rate": 1.0560000000000001e-06,
+      "loss": 0.3413,
+      "step": 44740
+    },
+    {
+      "epoch": 4.475,
+      "grad_norm": 78.71873474121094,
+      "learning_rate": 1.054e-06,
+      "loss": 0.5368,
+      "step": 44750
+    },
+    {
+      "epoch": 4.476,
+      "grad_norm": 64.87627410888672,
+      "learning_rate": 1.052e-06,
+      "loss": 0.5431,
+      "step": 44760
+    },
+    {
+      "epoch": 4.477,
+      "grad_norm": 47.796199798583984,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 0.6959,
+      "step": 44770
+    },
+    {
+      "epoch": 4.478,
+      "grad_norm": 44.98781204223633,
+      "learning_rate": 1.0480000000000002e-06,
+      "loss": 0.5795,
+      "step": 44780
+    },
+    {
+      "epoch": 4.479,
+      "grad_norm": 22.857219696044922,
+      "learning_rate": 1.046e-06,
+      "loss": 0.9602,
+      "step": 44790
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 32.4987678527832,
+      "learning_rate": 1.044e-06,
+      "loss": 0.941,
+      "step": 44800
+    },
+    {
+      "epoch": 4.481,
+      "grad_norm": 42.770912170410156,
+      "learning_rate": 1.0420000000000001e-06,
+      "loss": 0.4813,
+      "step": 44810
+    },
+    {
+      "epoch": 4.482,
+      "grad_norm": 2.932715654373169,
+      "learning_rate": 1.04e-06,
+      "loss": 0.4139,
+      "step": 44820
+    },
+    {
+      "epoch": 4.483,
+      "grad_norm": 20.752546310424805,
+      "learning_rate": 1.038e-06,
+      "loss": 0.6361,
+      "step": 44830
+    },
+    {
+      "epoch": 4.484,
+      "grad_norm": 56.61863708496094,
+      "learning_rate": 1.0360000000000001e-06,
+      "loss": 0.8215,
+      "step": 44840
+    },
+    {
+      "epoch": 4.485,
+      "grad_norm": 59.40135192871094,
+      "learning_rate": 1.0340000000000002e-06,
+      "loss": 0.297,
+      "step": 44850
+    },
+    {
+      "epoch": 4.486,
+      "grad_norm": 14.24195384979248,
+      "learning_rate": 1.032e-06,
+      "loss": 0.577,
+      "step": 44860
+    },
+    {
+      "epoch": 4.487,
+      "grad_norm": 75.15922546386719,
+      "learning_rate": 1.03e-06,
+      "loss": 0.6376,
+      "step": 44870
+    },
+    {
+      "epoch": 4.4879999999999995,
+      "grad_norm": 71.25492095947266,
+      "learning_rate": 1.0280000000000002e-06,
+      "loss": 0.6669,
+      "step": 44880
+    },
+    {
+      "epoch": 4.489,
+      "grad_norm": 6.253817081451416,
+      "learning_rate": 1.026e-06,
+      "loss": 0.4921,
+      "step": 44890
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 44.84489822387695,
+      "learning_rate": 1.024e-06,
+      "loss": 0.6678,
+      "step": 44900
+    },
+    {
+      "epoch": 4.491,
+      "grad_norm": 54.61061477661133,
+      "learning_rate": 1.0220000000000001e-06,
+      "loss": 0.6607,
+      "step": 44910
+    },
+    {
+      "epoch": 4.492,
+      "grad_norm": 49.33956527709961,
+      "learning_rate": 1.02e-06,
+      "loss": 0.7338,
+      "step": 44920
+    },
+    {
+      "epoch": 4.493,
+      "grad_norm": 11.369890213012695,
+      "learning_rate": 1.018e-06,
+      "loss": 0.5798,
+      "step": 44930
+    },
+    {
+      "epoch": 4.494,
+      "grad_norm": 1.8291288614273071,
+      "learning_rate": 1.016e-06,
+      "loss": 0.5389,
+      "step": 44940
+    },
+    {
+      "epoch": 4.495,
+      "grad_norm": 42.062808990478516,
+      "learning_rate": 1.0140000000000002e-06,
+      "loss": 0.6113,
+      "step": 44950
+    },
+    {
+      "epoch": 4.496,
+      "grad_norm": 21.281036376953125,
+      "learning_rate": 1.012e-06,
+      "loss": 0.4842,
+      "step": 44960
+    },
+    {
+      "epoch": 4.497,
+      "grad_norm": 19.57158660888672,
+      "learning_rate": 1.01e-06,
+      "loss": 0.4564,
+      "step": 44970
+    },
+    {
+      "epoch": 4.498,
+      "grad_norm": 16.572309494018555,
+      "learning_rate": 1.0080000000000001e-06,
+      "loss": 0.4437,
+      "step": 44980
+    },
+    {
+      "epoch": 4.499,
+      "grad_norm": 52.86716079711914,
+      "learning_rate": 1.006e-06,
+      "loss": 0.6462,
+      "step": 44990
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 13.026033401489258,
+      "learning_rate": 1.004e-06,
+      "loss": 0.59,
+      "step": 45000
+    },
+    {
+      "epoch": 4.501,
+      "grad_norm": 22.181806564331055,
+      "learning_rate": 1.002e-06,
+      "loss": 0.5236,
+      "step": 45010
+    },
+    {
+      "epoch": 4.502,
+      "grad_norm": 32.47010803222656,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.7592,
+      "step": 45020
+    },
+    {
+      "epoch": 4.503,
+      "grad_norm": 4.673216819763184,
+      "learning_rate": 9.98e-07,
+      "loss": 0.3559,
+      "step": 45030
+    },
+    {
+      "epoch": 4.504,
+      "grad_norm": 29.3690128326416,
+      "learning_rate": 9.96e-07,
+      "loss": 0.5105,
+      "step": 45040
+    },
+    {
+      "epoch": 4.505,
+      "grad_norm": 13.767579078674316,
+      "learning_rate": 9.940000000000001e-07,
+      "loss": 0.5032,
+      "step": 45050
+    },
+    {
+      "epoch": 4.506,
+      "grad_norm": 42.104862213134766,
+      "learning_rate": 9.92e-07,
+      "loss": 0.7156,
+      "step": 45060
+    },
+    {
+      "epoch": 4.507,
+      "grad_norm": 42.20527648925781,
+      "learning_rate": 9.9e-07,
+      "loss": 0.5525,
+      "step": 45070
+    },
+    {
+      "epoch": 4.508,
+      "grad_norm": 34.467933654785156,
+      "learning_rate": 9.880000000000001e-07,
+      "loss": 0.5533,
+      "step": 45080
+    },
+    {
+      "epoch": 4.509,
+      "grad_norm": 3.9194490909576416,
+      "learning_rate": 9.86e-07,
+      "loss": 0.3542,
+      "step": 45090
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 71.32640838623047,
+      "learning_rate": 9.84e-07,
+      "loss": 0.4193,
+      "step": 45100
+    },
+    {
+      "epoch": 4.511,
+      "grad_norm": 56.496826171875,
+      "learning_rate": 9.82e-07,
+      "loss": 0.591,
+      "step": 45110
+    },
+    {
+      "epoch": 4.5120000000000005,
+      "grad_norm": 17.558574676513672,
+      "learning_rate": 9.800000000000001e-07,
+      "loss": 0.4785,
+      "step": 45120
+    },
+    {
+      "epoch": 4.513,
+      "grad_norm": 32.05054473876953,
+      "learning_rate": 9.78e-07,
+      "loss": 0.3853,
+      "step": 45130
+    },
+    {
+      "epoch": 4.514,
+      "grad_norm": 22.20078468322754,
+      "learning_rate": 9.76e-07,
+      "loss": 0.6342,
+      "step": 45140
+    },
+    {
+      "epoch": 4.515,
+      "grad_norm": 44.59579086303711,
+      "learning_rate": 9.740000000000001e-07,
+      "loss": 0.4882,
+      "step": 45150
+    },
+    {
+      "epoch": 4.516,
+      "grad_norm": 45.14775466918945,
+      "learning_rate": 9.72e-07,
+      "loss": 0.5903,
+      "step": 45160
+    },
+    {
+      "epoch": 4.517,
+      "grad_norm": 117.20407104492188,
+      "learning_rate": 9.7e-07,
+      "loss": 0.7166,
+      "step": 45170
+    },
+    {
+      "epoch": 4.518,
+      "grad_norm": 29.686355590820312,
+      "learning_rate": 9.68e-07,
+      "loss": 0.6031,
+      "step": 45180
+    },
+    {
+      "epoch": 4.519,
+      "grad_norm": 44.74457550048828,
+      "learning_rate": 9.660000000000002e-07,
+      "loss": 0.5089,
+      "step": 45190
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 22.07326889038086,
+      "learning_rate": 9.64e-07,
+      "loss": 0.6083,
+      "step": 45200
+    },
+    {
+      "epoch": 4.521,
+      "grad_norm": 49.83152389526367,
+      "learning_rate": 9.62e-07,
+      "loss": 0.562,
+      "step": 45210
+    },
+    {
+      "epoch": 4.522,
+      "grad_norm": 50.51664733886719,
+      "learning_rate": 9.600000000000001e-07,
+      "loss": 0.6047,
+      "step": 45220
+    },
+    {
+      "epoch": 4.523,
+      "grad_norm": 56.542938232421875,
+      "learning_rate": 9.58e-07,
+      "loss": 0.5488,
+      "step": 45230
+    },
+    {
+      "epoch": 4.524,
+      "grad_norm": 46.843406677246094,
+      "learning_rate": 9.56e-07,
+      "loss": 0.7209,
+      "step": 45240
+    },
+    {
+      "epoch": 4.525,
+      "grad_norm": 6.081303119659424,
+      "learning_rate": 9.54e-07,
+      "loss": 0.4724,
+      "step": 45250
+    },
+    {
+      "epoch": 4.526,
+      "grad_norm": 28.085397720336914,
+      "learning_rate": 9.520000000000002e-07,
+      "loss": 0.7269,
+      "step": 45260
+    },
+    {
+      "epoch": 4.527,
+      "grad_norm": 52.40395736694336,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 0.6334,
+      "step": 45270
+    },
+    {
+      "epoch": 4.5280000000000005,
+      "grad_norm": 49.768287658691406,
+      "learning_rate": 9.480000000000001e-07,
+      "loss": 0.3782,
+      "step": 45280
+    },
+    {
+      "epoch": 4.529,
+      "grad_norm": 10.42574405670166,
+      "learning_rate": 9.460000000000001e-07,
+      "loss": 0.8404,
+      "step": 45290
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 13.148209571838379,
+      "learning_rate": 9.440000000000001e-07,
+      "loss": 0.4263,
+      "step": 45300
+    },
+    {
+      "epoch": 4.531,
+      "grad_norm": 56.37141799926758,
+      "learning_rate": 9.420000000000002e-07,
+      "loss": 0.6954,
+      "step": 45310
+    },
+    {
+      "epoch": 4.532,
+      "grad_norm": 41.969112396240234,
+      "learning_rate": 9.400000000000001e-07,
+      "loss": 0.4243,
+      "step": 45320
+    },
+    {
+      "epoch": 4.533,
+      "grad_norm": 61.941429138183594,
+      "learning_rate": 9.380000000000001e-07,
+      "loss": 0.8364,
+      "step": 45330
+    },
+    {
+      "epoch": 4.534,
+      "grad_norm": 7.106180667877197,
+      "learning_rate": 9.360000000000001e-07,
+      "loss": 0.6544,
+      "step": 45340
+    },
+    {
+      "epoch": 4.535,
+      "grad_norm": 31.053434371948242,
+      "learning_rate": 9.340000000000001e-07,
+      "loss": 0.4281,
+      "step": 45350
+    },
+    {
+      "epoch": 4.536,
+      "grad_norm": 38.16424560546875,
+      "learning_rate": 9.320000000000001e-07,
+      "loss": 0.6195,
+      "step": 45360
+    },
+    {
+      "epoch": 4.537,
+      "grad_norm": 57.80653381347656,
+      "learning_rate": 9.300000000000001e-07,
+      "loss": 0.6443,
+      "step": 45370
+    },
+    {
+      "epoch": 4.538,
+      "grad_norm": 7.248365879058838,
+      "learning_rate": 9.28e-07,
+      "loss": 0.6943,
+      "step": 45380
+    },
+    {
+      "epoch": 4.539,
+      "grad_norm": 2.168619394302368,
+      "learning_rate": 9.260000000000001e-07,
+      "loss": 0.4539,
+      "step": 45390
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 9.784174919128418,
+      "learning_rate": 9.240000000000001e-07,
+      "loss": 0.4954,
+      "step": 45400
+    },
+    {
+      "epoch": 4.541,
+      "grad_norm": 27.470226287841797,
+      "learning_rate": 9.220000000000001e-07,
+      "loss": 0.9503,
+      "step": 45410
+    },
+    {
+      "epoch": 4.542,
+      "grad_norm": 24.790925979614258,
+      "learning_rate": 9.200000000000001e-07,
+      "loss": 0.3663,
+      "step": 45420
+    },
+    {
+      "epoch": 4.543,
+      "grad_norm": 37.76428985595703,
+      "learning_rate": 9.180000000000001e-07,
+      "loss": 0.7152,
+      "step": 45430
+    },
+    {
+      "epoch": 4.5440000000000005,
+      "grad_norm": 29.315391540527344,
+      "learning_rate": 9.160000000000001e-07,
+      "loss": 0.3819,
+      "step": 45440
+    },
+    {
+      "epoch": 4.545,
+      "grad_norm": 19.65518569946289,
+      "learning_rate": 9.140000000000001e-07,
+      "loss": 0.3005,
+      "step": 45450
+    },
+    {
+      "epoch": 4.546,
+      "grad_norm": 4.487301349639893,
+      "learning_rate": 9.120000000000001e-07,
+      "loss": 0.4622,
+      "step": 45460
+    },
+    {
+      "epoch": 4.547,
+      "grad_norm": 41.73360824584961,
+      "learning_rate": 9.100000000000001e-07,
+      "loss": 0.5676,
+      "step": 45470
+    },
+    {
+      "epoch": 4.548,
+      "grad_norm": 48.73320770263672,
+      "learning_rate": 9.080000000000001e-07,
+      "loss": 0.8195,
+      "step": 45480
+    },
+    {
+      "epoch": 4.549,
+      "grad_norm": 12.341236114501953,
+      "learning_rate": 9.062000000000001e-07,
+      "loss": 0.3925,
+      "step": 45490
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 4.098908424377441,
+      "learning_rate": 9.042000000000001e-07,
+      "loss": 0.4159,
+      "step": 45500
+    },
+    {
+      "epoch": 4.551,
+      "grad_norm": 22.557924270629883,
+      "learning_rate": 9.022e-07,
+      "loss": 0.2809,
+      "step": 45510
+    },
+    {
+      "epoch": 4.552,
+      "grad_norm": 19.807369232177734,
+      "learning_rate": 9.002000000000001e-07,
+      "loss": 0.6193,
+      "step": 45520
+    },
+    {
+      "epoch": 4.553,
+      "grad_norm": 30.91956901550293,
+      "learning_rate": 8.982e-07,
+      "loss": 0.9155,
+      "step": 45530
+    },
+    {
+      "epoch": 4.554,
+      "grad_norm": 30.81283187866211,
+      "learning_rate": 8.962000000000001e-07,
+      "loss": 0.5091,
+      "step": 45540
+    },
+    {
+      "epoch": 4.555,
+      "grad_norm": 39.03916549682617,
+      "learning_rate": 8.942e-07,
+      "loss": 0.4792,
+      "step": 45550
+    },
+    {
+      "epoch": 4.556,
+      "grad_norm": 30.683202743530273,
+      "learning_rate": 8.922e-07,
+      "loss": 0.5482,
+      "step": 45560
+    },
+    {
+      "epoch": 4.557,
+      "grad_norm": 36.451107025146484,
+      "learning_rate": 8.902000000000001e-07,
+      "loss": 0.5982,
+      "step": 45570
+    },
+    {
+      "epoch": 4.558,
+      "grad_norm": 19.348159790039062,
+      "learning_rate": 8.882e-07,
+      "loss": 0.4162,
+      "step": 45580
+    },
+    {
+      "epoch": 4.559,
+      "grad_norm": 69.45014953613281,
+      "learning_rate": 8.862000000000001e-07,
+      "loss": 0.6971,
+      "step": 45590
+    },
+    {
+      "epoch": 4.5600000000000005,
+      "grad_norm": 21.53099250793457,
+      "learning_rate": 8.842e-07,
+      "loss": 0.5507,
+      "step": 45600
+    },
+    {
+      "epoch": 4.561,
+      "grad_norm": 25.485855102539062,
+      "learning_rate": 8.822000000000001e-07,
+      "loss": 0.6958,
+      "step": 45610
+    },
+    {
+      "epoch": 4.562,
+      "grad_norm": 43.30933380126953,
+      "learning_rate": 8.802000000000001e-07,
+      "loss": 0.5381,
+      "step": 45620
+    },
+    {
+      "epoch": 4.563,
+      "grad_norm": 29.24176597595215,
+      "learning_rate": 8.782e-07,
+      "loss": 0.5499,
+      "step": 45630
+    },
+    {
+      "epoch": 4.564,
+      "grad_norm": 34.37154006958008,
+      "learning_rate": 8.762000000000001e-07,
+      "loss": 0.2587,
+      "step": 45640
+    },
+    {
+      "epoch": 4.5649999999999995,
+      "grad_norm": 31.237136840820312,
+      "learning_rate": 8.742e-07,
+      "loss": 0.6016,
+      "step": 45650
+    },
+    {
+      "epoch": 4.566,
+      "grad_norm": 35.17666244506836,
+      "learning_rate": 8.722000000000001e-07,
+      "loss": 0.4353,
+      "step": 45660
+    },
+    {
+      "epoch": 4.567,
+      "grad_norm": 24.52681541442871,
+      "learning_rate": 8.702e-07,
+      "loss": 0.8129,
+      "step": 45670
+    },
+    {
+      "epoch": 4.568,
+      "grad_norm": 73.54619598388672,
+      "learning_rate": 8.682e-07,
+      "loss": 0.4345,
+      "step": 45680
+    },
+    {
+      "epoch": 4.569,
+      "grad_norm": 51.41396713256836,
+      "learning_rate": 8.662000000000001e-07,
+      "loss": 0.5862,
+      "step": 45690
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 22.1870059967041,
+      "learning_rate": 8.642e-07,
+      "loss": 0.8158,
+      "step": 45700
+    },
+    {
+      "epoch": 4.571,
+      "grad_norm": 53.16621398925781,
+      "learning_rate": 8.622000000000001e-07,
+      "loss": 0.5901,
+      "step": 45710
+    },
+    {
+      "epoch": 4.572,
+      "grad_norm": 52.27548599243164,
+      "learning_rate": 8.602e-07,
+      "loss": 0.5686,
+      "step": 45720
+    },
+    {
+      "epoch": 4.573,
+      "grad_norm": 30.82957649230957,
+      "learning_rate": 8.582e-07,
+      "loss": 0.4284,
+      "step": 45730
+    },
+    {
+      "epoch": 4.574,
+      "grad_norm": 4.987204551696777,
+      "learning_rate": 8.562000000000001e-07,
+      "loss": 0.5962,
+      "step": 45740
+    },
+    {
+      "epoch": 4.575,
+      "grad_norm": 44.10116195678711,
+      "learning_rate": 8.542e-07,
+      "loss": 0.8412,
+      "step": 45750
+    },
+    {
+      "epoch": 4.576,
+      "grad_norm": 18.577295303344727,
+      "learning_rate": 8.522000000000001e-07,
+      "loss": 0.7083,
+      "step": 45760
+    },
+    {
+      "epoch": 4.577,
+      "grad_norm": 30.12908363342285,
+      "learning_rate": 8.502e-07,
+      "loss": 0.7062,
+      "step": 45770
+    },
+    {
+      "epoch": 4.578,
+      "grad_norm": 5.816198825836182,
+      "learning_rate": 8.482000000000001e-07,
+      "loss": 0.8032,
+      "step": 45780
+    },
+    {
+      "epoch": 4.579,
+      "grad_norm": 5.501286506652832,
+      "learning_rate": 8.462e-07,
+      "loss": 0.3029,
+      "step": 45790
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 41.75872039794922,
+      "learning_rate": 8.442e-07,
+      "loss": 0.975,
+      "step": 45800
+    },
+    {
+      "epoch": 4.5809999999999995,
+      "grad_norm": 28.56104278564453,
+      "learning_rate": 8.422000000000001e-07,
+      "loss": 0.6924,
+      "step": 45810
+    },
+    {
+      "epoch": 4.582,
+      "grad_norm": 64.78409576416016,
+      "learning_rate": 8.402e-07,
+      "loss": 0.6091,
+      "step": 45820
+    },
+    {
+      "epoch": 4.583,
+      "grad_norm": 25.520673751831055,
+      "learning_rate": 8.382000000000001e-07,
+      "loss": 0.6849,
+      "step": 45830
+    },
+    {
+      "epoch": 4.584,
+      "grad_norm": 15.34619140625,
+      "learning_rate": 8.362e-07,
+      "loss": 0.5021,
+      "step": 45840
+    },
+    {
+      "epoch": 4.585,
+      "grad_norm": 42.24744415283203,
+      "learning_rate": 8.342e-07,
+      "loss": 0.4582,
+      "step": 45850
+    },
+    {
+      "epoch": 4.586,
+      "grad_norm": 24.966739654541016,
+      "learning_rate": 8.322000000000001e-07,
+      "loss": 0.4246,
+      "step": 45860
+    },
+    {
+      "epoch": 4.587,
+      "grad_norm": 22.995031356811523,
+      "learning_rate": 8.302e-07,
+      "loss": 0.7483,
+      "step": 45870
+    },
+    {
+      "epoch": 4.588,
+      "grad_norm": 36.11094284057617,
+      "learning_rate": 8.282000000000001e-07,
+      "loss": 0.4904,
+      "step": 45880
+    },
+    {
+      "epoch": 4.589,
+      "grad_norm": 40.967899322509766,
+      "learning_rate": 8.262e-07,
+      "loss": 0.4971,
+      "step": 45890
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 40.29697799682617,
+      "learning_rate": 8.242e-07,
+      "loss": 0.8347,
+      "step": 45900
+    },
+    {
+      "epoch": 4.591,
+      "grad_norm": 28.283031463623047,
+      "learning_rate": 8.222e-07,
+      "loss": 0.658,
+      "step": 45910
+    },
+    {
+      "epoch": 4.592,
+      "grad_norm": 25.174957275390625,
+      "learning_rate": 8.202e-07,
+      "loss": 0.6491,
+      "step": 45920
+    },
+    {
+      "epoch": 4.593,
+      "grad_norm": 62.129547119140625,
+      "learning_rate": 8.184000000000001e-07,
+      "loss": 0.6112,
+      "step": 45930
+    },
+    {
+      "epoch": 4.594,
+      "grad_norm": 26.676013946533203,
+      "learning_rate": 8.164000000000001e-07,
+      "loss": 0.4537,
+      "step": 45940
+    },
+    {
+      "epoch": 4.595,
+      "grad_norm": 66.71521759033203,
+      "learning_rate": 8.144000000000001e-07,
+      "loss": 0.6426,
+      "step": 45950
+    },
+    {
+      "epoch": 4.596,
+      "grad_norm": 52.26560974121094,
+      "learning_rate": 8.124000000000002e-07,
+      "loss": 0.5763,
+      "step": 45960
+    },
+    {
+      "epoch": 4.5969999999999995,
+      "grad_norm": 29.50721549987793,
+      "learning_rate": 8.104000000000001e-07,
+      "loss": 0.6662,
+      "step": 45970
+    },
+    {
+      "epoch": 4.598,
+      "grad_norm": 28.411151885986328,
+      "learning_rate": 8.084000000000001e-07,
+      "loss": 0.4994,
+      "step": 45980
+    },
+    {
+      "epoch": 4.599,
+      "grad_norm": 19.696985244750977,
+      "learning_rate": 8.064000000000001e-07,
+      "loss": 0.6327,
+      "step": 45990
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 34.80906295776367,
+      "learning_rate": 8.044000000000001e-07,
+      "loss": 0.3584,
+      "step": 46000
+    },
+    {
+      "epoch": 4.601,
+      "grad_norm": 40.034202575683594,
+      "learning_rate": 8.024000000000001e-07,
+      "loss": 0.4287,
+      "step": 46010
+    },
+    {
+      "epoch": 4.602,
+      "grad_norm": 65.08027648925781,
+      "learning_rate": 8.004000000000001e-07,
+      "loss": 0.531,
+      "step": 46020
+    },
+    {
+      "epoch": 4.603,
+      "grad_norm": 46.98908615112305,
+      "learning_rate": 7.984000000000001e-07,
+      "loss": 0.469,
+      "step": 46030
+    },
+    {
+      "epoch": 4.604,
+      "grad_norm": 35.33990478515625,
+      "learning_rate": 7.964000000000001e-07,
+      "loss": 0.532,
+      "step": 46040
+    },
+    {
+      "epoch": 4.605,
+      "grad_norm": 25.324203491210938,
+      "learning_rate": 7.944000000000001e-07,
+      "loss": 0.571,
+      "step": 46050
+    },
+    {
+      "epoch": 4.606,
+      "grad_norm": 7.645961284637451,
+      "learning_rate": 7.924000000000001e-07,
+      "loss": 0.4075,
+      "step": 46060
+    },
+    {
+      "epoch": 4.607,
+      "grad_norm": 12.345396041870117,
+      "learning_rate": 7.904000000000001e-07,
+      "loss": 0.5891,
+      "step": 46070
+    },
+    {
+      "epoch": 4.608,
+      "grad_norm": 8.199934959411621,
+      "learning_rate": 7.884e-07,
+      "loss": 0.6039,
+      "step": 46080
+    },
+    {
+      "epoch": 4.609,
+      "grad_norm": 3.992199420928955,
+      "learning_rate": 7.864000000000001e-07,
+      "loss": 0.6168,
+      "step": 46090
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 82.65238189697266,
+      "learning_rate": 7.844000000000001e-07,
+      "loss": 0.9055,
+      "step": 46100
+    },
+    {
+      "epoch": 4.611,
+      "grad_norm": 3.730809450149536,
+      "learning_rate": 7.824000000000001e-07,
+      "loss": 0.5675,
+      "step": 46110
+    },
+    {
+      "epoch": 4.612,
+      "grad_norm": 48.9563102722168,
+      "learning_rate": 7.804000000000001e-07,
+      "loss": 0.6902,
+      "step": 46120
+    },
+    {
+      "epoch": 4.6129999999999995,
+      "grad_norm": 7.581906795501709,
+      "learning_rate": 7.784000000000001e-07,
+      "loss": 0.8163,
+      "step": 46130
+    },
+    {
+      "epoch": 4.614,
+      "grad_norm": 2.6554839611053467,
+      "learning_rate": 7.764000000000001e-07,
+      "loss": 0.4846,
+      "step": 46140
+    },
+    {
+      "epoch": 4.615,
+      "grad_norm": 40.605262756347656,
+      "learning_rate": 7.744e-07,
+      "loss": 1.097,
+      "step": 46150
+    },
+    {
+      "epoch": 4.616,
+      "grad_norm": 30.826297760009766,
+      "learning_rate": 7.724000000000001e-07,
+      "loss": 0.5738,
+      "step": 46160
+    },
+    {
+      "epoch": 4.617,
+      "grad_norm": 8.789215087890625,
+      "learning_rate": 7.704000000000001e-07,
+      "loss": 0.3848,
+      "step": 46170
+    },
+    {
+      "epoch": 4.618,
+      "grad_norm": 7.239010810852051,
+      "learning_rate": 7.684000000000001e-07,
+      "loss": 0.5266,
+      "step": 46180
+    },
+    {
+      "epoch": 4.619,
+      "grad_norm": 62.47308349609375,
+      "learning_rate": 7.664000000000001e-07,
+      "loss": 0.8456,
+      "step": 46190
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 38.9660758972168,
+      "learning_rate": 7.644e-07,
+      "loss": 0.7502,
+      "step": 46200
+    },
+    {
+      "epoch": 4.621,
+      "grad_norm": 59.97346878051758,
+      "learning_rate": 7.624000000000001e-07,
+      "loss": 0.4927,
+      "step": 46210
+    },
+    {
+      "epoch": 4.622,
+      "grad_norm": 6.728061676025391,
+      "learning_rate": 7.604000000000001e-07,
+      "loss": 0.4951,
+      "step": 46220
+    },
+    {
+      "epoch": 4.623,
+      "grad_norm": 25.506572723388672,
+      "learning_rate": 7.584000000000001e-07,
+      "loss": 0.4714,
+      "step": 46230
+    },
+    {
+      "epoch": 4.624,
+      "grad_norm": 85.05602264404297,
+      "learning_rate": 7.564000000000001e-07,
+      "loss": 0.5414,
+      "step": 46240
+    },
+    {
+      "epoch": 4.625,
+      "grad_norm": 31.7155818939209,
+      "learning_rate": 7.544e-07,
+      "loss": 0.5379,
+      "step": 46250
+    },
+    {
+      "epoch": 4.626,
+      "grad_norm": 6.111644268035889,
+      "learning_rate": 7.524000000000001e-07,
+      "loss": 0.4328,
+      "step": 46260
+    },
+    {
+      "epoch": 4.627,
+      "grad_norm": 41.613582611083984,
+      "learning_rate": 7.504e-07,
+      "loss": 0.7449,
+      "step": 46270
+    },
+    {
+      "epoch": 4.628,
+      "grad_norm": 62.398040771484375,
+      "learning_rate": 7.484000000000001e-07,
+      "loss": 0.4983,
+      "step": 46280
+    },
+    {
+      "epoch": 4.629,
+      "grad_norm": 71.82337951660156,
+      "learning_rate": 7.464000000000001e-07,
+      "loss": 0.6349,
+      "step": 46290
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 3.016054153442383,
+      "learning_rate": 7.444000000000001e-07,
+      "loss": 0.6469,
+      "step": 46300
+    },
+    {
+      "epoch": 4.631,
+      "grad_norm": 64.6484603881836,
+      "learning_rate": 7.424000000000001e-07,
+      "loss": 0.504,
+      "step": 46310
+    },
+    {
+      "epoch": 4.632,
+      "grad_norm": 66.25682067871094,
+      "learning_rate": 7.404e-07,
+      "loss": 0.6037,
+      "step": 46320
+    },
+    {
+      "epoch": 4.633,
+      "grad_norm": 16.990617752075195,
+      "learning_rate": 7.384000000000001e-07,
+      "loss": 0.4708,
+      "step": 46330
+    },
+    {
+      "epoch": 4.634,
+      "grad_norm": 42.48113250732422,
+      "learning_rate": 7.364000000000001e-07,
+      "loss": 0.8633,
+      "step": 46340
+    },
+    {
+      "epoch": 4.635,
+      "grad_norm": 20.38407325744629,
+      "learning_rate": 7.344000000000001e-07,
+      "loss": 0.6119,
+      "step": 46350
+    },
+    {
+      "epoch": 4.636,
+      "grad_norm": 60.075557708740234,
+      "learning_rate": 7.324000000000001e-07,
+      "loss": 0.5929,
+      "step": 46360
+    },
+    {
+      "epoch": 4.6370000000000005,
+      "grad_norm": 5.794075012207031,
+      "learning_rate": 7.304e-07,
+      "loss": 0.5081,
+      "step": 46370
+    },
+    {
+      "epoch": 4.638,
+      "grad_norm": 15.373174667358398,
+      "learning_rate": 7.284000000000001e-07,
+      "loss": 0.4686,
+      "step": 46380
+    },
+    {
+      "epoch": 4.639,
+      "grad_norm": 16.02683448791504,
+      "learning_rate": 7.264e-07,
+      "loss": 0.3858,
+      "step": 46390
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 35.6048583984375,
+      "learning_rate": 7.244000000000001e-07,
+      "loss": 0.4706,
+      "step": 46400
+    },
+    {
+      "epoch": 4.641,
+      "grad_norm": 16.356292724609375,
+      "learning_rate": 7.224000000000001e-07,
+      "loss": 0.3739,
+      "step": 46410
+    },
+    {
+      "epoch": 4.642,
+      "grad_norm": 43.27982711791992,
+      "learning_rate": 7.204000000000001e-07,
+      "loss": 0.529,
+      "step": 46420
+    },
+    {
+      "epoch": 4.643,
+      "grad_norm": 38.51179122924805,
+      "learning_rate": 7.184000000000001e-07,
+      "loss": 0.6962,
+      "step": 46430
+    },
+    {
+      "epoch": 4.644,
+      "grad_norm": 20.879756927490234,
+      "learning_rate": 7.164e-07,
+      "loss": 0.6176,
+      "step": 46440
+    },
+    {
+      "epoch": 4.645,
+      "grad_norm": 45.5706672668457,
+      "learning_rate": 7.144000000000001e-07,
+      "loss": 0.8066,
+      "step": 46450
+    },
+    {
+      "epoch": 4.646,
+      "grad_norm": 54.51177215576172,
+      "learning_rate": 7.124e-07,
+      "loss": 0.4615,
+      "step": 46460
+    },
+    {
+      "epoch": 4.647,
+      "grad_norm": 87.40930938720703,
+      "learning_rate": 7.104000000000001e-07,
+      "loss": 0.4578,
+      "step": 46470
+    },
+    {
+      "epoch": 4.648,
+      "grad_norm": 7.970445156097412,
+      "learning_rate": 7.084000000000001e-07,
+      "loss": 0.3683,
+      "step": 46480
+    },
+    {
+      "epoch": 4.649,
+      "grad_norm": 45.14849853515625,
+      "learning_rate": 7.064e-07,
+      "loss": 0.8526,
+      "step": 46490
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 7.996401786804199,
+      "learning_rate": 7.044000000000001e-07,
+      "loss": 0.2735,
+      "step": 46500
+    },
+    {
+      "epoch": 4.651,
+      "grad_norm": 27.318073272705078,
+      "learning_rate": 7.024e-07,
+      "loss": 0.6858,
+      "step": 46510
+    },
+    {
+      "epoch": 4.652,
+      "grad_norm": 28.923954010009766,
+      "learning_rate": 7.004000000000001e-07,
+      "loss": 0.2842,
+      "step": 46520
+    },
+    {
+      "epoch": 4.6530000000000005,
+      "grad_norm": 33.51847457885742,
+      "learning_rate": 6.984000000000001e-07,
+      "loss": 0.5165,
+      "step": 46530
+    },
+    {
+      "epoch": 4.654,
+      "grad_norm": 52.75628662109375,
+      "learning_rate": 6.964e-07,
+      "loss": 0.6128,
+      "step": 46540
+    },
+    {
+      "epoch": 4.655,
+      "grad_norm": 42.89748764038086,
+      "learning_rate": 6.944000000000001e-07,
+      "loss": 0.4989,
+      "step": 46550
+    },
+    {
+      "epoch": 4.656,
+      "grad_norm": 46.054161071777344,
+      "learning_rate": 6.924e-07,
+      "loss": 0.6215,
+      "step": 46560
+    },
+    {
+      "epoch": 4.657,
+      "grad_norm": 12.866451263427734,
+      "learning_rate": 6.904000000000001e-07,
+      "loss": 0.4183,
+      "step": 46570
+    },
+    {
+      "epoch": 4.658,
+      "grad_norm": 13.437576293945312,
+      "learning_rate": 6.884e-07,
+      "loss": 0.5324,
+      "step": 46580
+    },
+    {
+      "epoch": 4.659,
+      "grad_norm": 7.054804801940918,
+      "learning_rate": 6.864000000000001e-07,
+      "loss": 0.697,
+      "step": 46590
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 6.7603373527526855,
+      "learning_rate": 6.844000000000001e-07,
+      "loss": 0.3793,
+      "step": 46600
+    },
+    {
+      "epoch": 4.661,
+      "grad_norm": 5.329585075378418,
+      "learning_rate": 6.824e-07,
+      "loss": 0.7276,
+      "step": 46610
+    },
+    {
+      "epoch": 4.662,
+      "grad_norm": 24.512773513793945,
+      "learning_rate": 6.804000000000001e-07,
+      "loss": 0.4644,
+      "step": 46620
+    },
+    {
+      "epoch": 4.663,
+      "grad_norm": 5.408405780792236,
+      "learning_rate": 6.784e-07,
+      "loss": 0.3011,
+      "step": 46630
+    },
+    {
+      "epoch": 4.664,
+      "grad_norm": 56.339935302734375,
+      "learning_rate": 6.764000000000001e-07,
+      "loss": 0.6349,
+      "step": 46640
+    },
+    {
+      "epoch": 4.665,
+      "grad_norm": 1.9910184144973755,
+      "learning_rate": 6.744000000000001e-07,
+      "loss": 0.7266,
+      "step": 46650
+    },
+    {
+      "epoch": 4.666,
+      "grad_norm": 47.08732986450195,
+      "learning_rate": 6.724e-07,
+      "loss": 0.5365,
+      "step": 46660
+    },
+    {
+      "epoch": 4.667,
+      "grad_norm": 77.28034973144531,
+      "learning_rate": 6.704000000000001e-07,
+      "loss": 0.6119,
+      "step": 46670
+    },
+    {
+      "epoch": 4.668,
+      "grad_norm": 61.18629837036133,
+      "learning_rate": 6.684e-07,
+      "loss": 0.6465,
+      "step": 46680
+    },
+    {
+      "epoch": 4.6690000000000005,
+      "grad_norm": 37.235626220703125,
+      "learning_rate": 6.664000000000001e-07,
+      "loss": 0.504,
+      "step": 46690
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 11.716021537780762,
+      "learning_rate": 6.644e-07,
+      "loss": 0.5464,
+      "step": 46700
+    },
+    {
+      "epoch": 4.671,
+      "grad_norm": 40.31194305419922,
+      "learning_rate": 6.624e-07,
+      "loss": 0.2915,
+      "step": 46710
+    },
+    {
+      "epoch": 4.672,
+      "grad_norm": 41.208595275878906,
+      "learning_rate": 6.604000000000001e-07,
+      "loss": 0.759,
+      "step": 46720
+    },
+    {
+      "epoch": 4.673,
+      "grad_norm": 37.54507827758789,
+      "learning_rate": 6.584e-07,
+      "loss": 0.6278,
+      "step": 46730
+    },
+    {
+      "epoch": 4.674,
+      "grad_norm": 52.48876953125,
+      "learning_rate": 6.564000000000001e-07,
+      "loss": 0.6083,
+      "step": 46740
+    },
+    {
+      "epoch": 4.675,
+      "grad_norm": 20.885011672973633,
+      "learning_rate": 6.544e-07,
+      "loss": 0.6174,
+      "step": 46750
+    },
+    {
+      "epoch": 4.676,
+      "grad_norm": 41.223793029785156,
+      "learning_rate": 6.524000000000001e-07,
+      "loss": 0.6118,
+      "step": 46760
+    },
+    {
+      "epoch": 4.677,
+      "grad_norm": 28.714216232299805,
+      "learning_rate": 6.504000000000001e-07,
+      "loss": 0.54,
+      "step": 46770
+    },
+    {
+      "epoch": 4.678,
+      "grad_norm": 23.326507568359375,
+      "learning_rate": 6.484e-07,
+      "loss": 0.6858,
+      "step": 46780
+    },
+    {
+      "epoch": 4.679,
+      "grad_norm": 5.641720771789551,
+      "learning_rate": 6.464000000000001e-07,
+      "loss": 0.5215,
+      "step": 46790
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 12.176937103271484,
+      "learning_rate": 6.444e-07,
+      "loss": 0.8599,
+      "step": 46800
+    },
+    {
+      "epoch": 4.681,
+      "grad_norm": 15.131599426269531,
+      "learning_rate": 6.424000000000001e-07,
+      "loss": 0.5364,
+      "step": 46810
+    },
+    {
+      "epoch": 4.682,
+      "grad_norm": 17.37408447265625,
+      "learning_rate": 6.404e-07,
+      "loss": 0.6698,
+      "step": 46820
+    },
+    {
+      "epoch": 4.683,
+      "grad_norm": 49.73809814453125,
+      "learning_rate": 6.384e-07,
+      "loss": 0.7712,
+      "step": 46830
+    },
+    {
+      "epoch": 4.684,
+      "grad_norm": 36.00857925415039,
+      "learning_rate": 6.364000000000001e-07,
+      "loss": 0.6566,
+      "step": 46840
+    },
+    {
+      "epoch": 4.6850000000000005,
+      "grad_norm": 41.920127868652344,
+      "learning_rate": 6.344e-07,
+      "loss": 0.5428,
+      "step": 46850
+    },
+    {
+      "epoch": 4.686,
+      "grad_norm": 45.1785888671875,
+      "learning_rate": 6.324000000000001e-07,
+      "loss": 0.4806,
+      "step": 46860
+    },
+    {
+      "epoch": 4.687,
+      "grad_norm": 44.14488220214844,
+      "learning_rate": 6.304e-07,
+      "loss": 0.6205,
+      "step": 46870
+    },
+    {
+      "epoch": 4.688,
+      "grad_norm": 1.7023221254348755,
+      "learning_rate": 6.284e-07,
+      "loss": 0.4456,
+      "step": 46880
+    },
+    {
+      "epoch": 4.689,
+      "grad_norm": 13.15688419342041,
+      "learning_rate": 6.266e-07,
+      "loss": 0.6845,
+      "step": 46890
+    },
+    {
+      "epoch": 4.6899999999999995,
+      "grad_norm": 15.498842239379883,
+      "learning_rate": 6.246e-07,
+      "loss": 0.5764,
+      "step": 46900
+    },
+    {
+      "epoch": 4.691,
+      "grad_norm": 23.742435455322266,
+      "learning_rate": 6.226000000000001e-07,
+      "loss": 0.5667,
+      "step": 46910
+    },
+    {
+      "epoch": 4.692,
+      "grad_norm": 16.880615234375,
+      "learning_rate": 6.206e-07,
+      "loss": 0.3689,
+      "step": 46920
+    },
+    {
+      "epoch": 4.693,
+      "grad_norm": 34.34412384033203,
+      "learning_rate": 6.186e-07,
+      "loss": 0.5903,
+      "step": 46930
+    },
+    {
+      "epoch": 4.694,
+      "grad_norm": 41.17317199707031,
+      "learning_rate": 6.166000000000001e-07,
+      "loss": 0.4317,
+      "step": 46940
+    },
+    {
+      "epoch": 4.695,
+      "grad_norm": 40.64748001098633,
+      "learning_rate": 6.146e-07,
+      "loss": 0.4934,
+      "step": 46950
+    },
+    {
+      "epoch": 4.696,
+      "grad_norm": 49.303955078125,
+      "learning_rate": 6.126000000000001e-07,
+      "loss": 0.5757,
+      "step": 46960
+    },
+    {
+      "epoch": 4.697,
+      "grad_norm": 43.190826416015625,
+      "learning_rate": 6.106000000000001e-07,
+      "loss": 0.6163,
+      "step": 46970
+    },
+    {
+      "epoch": 4.698,
+      "grad_norm": 9.210566520690918,
+      "learning_rate": 6.086e-07,
+      "loss": 0.3262,
+      "step": 46980
+    },
+    {
+      "epoch": 4.699,
+      "grad_norm": 41.45174026489258,
+      "learning_rate": 6.066e-07,
+      "loss": 0.6636,
+      "step": 46990
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 42.5493049621582,
+      "learning_rate": 6.046e-07,
+      "loss": 0.7651,
+      "step": 47000
+    },
+    {
+      "epoch": 4.701,
+      "grad_norm": 14.347480773925781,
+      "learning_rate": 6.026000000000001e-07,
+      "loss": 0.7254,
+      "step": 47010
+    },
+    {
+      "epoch": 4.702,
+      "grad_norm": 55.891910552978516,
+      "learning_rate": 6.006000000000001e-07,
+      "loss": 0.6508,
+      "step": 47020
+    },
+    {
+      "epoch": 4.703,
+      "grad_norm": 66.14480590820312,
+      "learning_rate": 5.986e-07,
+      "loss": 0.5825,
+      "step": 47030
+    },
+    {
+      "epoch": 4.704,
+      "grad_norm": 29.104063034057617,
+      "learning_rate": 5.966e-07,
+      "loss": 0.6674,
+      "step": 47040
+    },
+    {
+      "epoch": 4.705,
+      "grad_norm": 12.830475807189941,
+      "learning_rate": 5.946e-07,
+      "loss": 0.3653,
+      "step": 47050
+    },
+    {
+      "epoch": 4.7059999999999995,
+      "grad_norm": 30.010398864746094,
+      "learning_rate": 5.926e-07,
+      "loss": 0.8521,
+      "step": 47060
+    },
+    {
+      "epoch": 4.707,
+      "grad_norm": 31.613298416137695,
+      "learning_rate": 5.906000000000001e-07,
+      "loss": 0.7231,
+      "step": 47070
+    },
+    {
+      "epoch": 4.708,
+      "grad_norm": 17.527433395385742,
+      "learning_rate": 5.886000000000001e-07,
+      "loss": 0.4612,
+      "step": 47080
+    },
+    {
+      "epoch": 4.709,
+      "grad_norm": 55.67793655395508,
+      "learning_rate": 5.866e-07,
+      "loss": 0.373,
+      "step": 47090
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 40.79391098022461,
+      "learning_rate": 5.846e-07,
+      "loss": 0.5253,
+      "step": 47100
+    },
+    {
+      "epoch": 4.711,
+      "grad_norm": 18.90387725830078,
+      "learning_rate": 5.826e-07,
+      "loss": 0.6227,
+      "step": 47110
+    },
+    {
+      "epoch": 4.712,
+      "grad_norm": 83.02630615234375,
+      "learning_rate": 5.806000000000001e-07,
+      "loss": 0.3808,
+      "step": 47120
+    },
+    {
+      "epoch": 4.713,
+      "grad_norm": 11.049756050109863,
+      "learning_rate": 5.786000000000001e-07,
+      "loss": 0.6286,
+      "step": 47130
+    },
+    {
+      "epoch": 4.714,
+      "grad_norm": 46.445465087890625,
+      "learning_rate": 5.766000000000001e-07,
+      "loss": 0.5005,
+      "step": 47140
+    },
+    {
+      "epoch": 4.715,
+      "grad_norm": 58.55226135253906,
+      "learning_rate": 5.746e-07,
+      "loss": 0.708,
+      "step": 47150
+    },
+    {
+      "epoch": 4.716,
+      "grad_norm": 37.50386047363281,
+      "learning_rate": 5.726e-07,
+      "loss": 0.5185,
+      "step": 47160
+    },
+    {
+      "epoch": 4.717,
+      "grad_norm": 18.741193771362305,
+      "learning_rate": 5.706000000000001e-07,
+      "loss": 0.7934,
+      "step": 47170
+    },
+    {
+      "epoch": 4.718,
+      "grad_norm": 31.67978858947754,
+      "learning_rate": 5.686e-07,
+      "loss": 0.5104,
+      "step": 47180
+    },
+    {
+      "epoch": 4.719,
+      "grad_norm": 58.85725021362305,
+      "learning_rate": 5.666000000000001e-07,
+      "loss": 0.5489,
+      "step": 47190
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 42.56383514404297,
+      "learning_rate": 5.646000000000001e-07,
+      "loss": 0.3513,
+      "step": 47200
+    },
+    {
+      "epoch": 4.721,
+      "grad_norm": 21.690288543701172,
+      "learning_rate": 5.626e-07,
+      "loss": 0.4654,
+      "step": 47210
+    },
+    {
+      "epoch": 4.7219999999999995,
+      "grad_norm": 42.254913330078125,
+      "learning_rate": 5.606000000000001e-07,
+      "loss": 0.6836,
+      "step": 47220
+    },
+    {
+      "epoch": 4.723,
+      "grad_norm": 52.095916748046875,
+      "learning_rate": 5.586e-07,
+      "loss": 0.4583,
+      "step": 47230
+    },
+    {
+      "epoch": 4.724,
+      "grad_norm": 74.2481918334961,
+      "learning_rate": 5.566000000000001e-07,
+      "loss": 0.4986,
+      "step": 47240
+    },
+    {
+      "epoch": 4.725,
+      "grad_norm": 27.45410919189453,
+      "learning_rate": 5.546000000000001e-07,
+      "loss": 0.7195,
+      "step": 47250
+    },
+    {
+      "epoch": 4.726,
+      "grad_norm": 14.02073860168457,
+      "learning_rate": 5.526e-07,
+      "loss": 0.5896,
+      "step": 47260
+    },
+    {
+      "epoch": 4.727,
+      "grad_norm": 27.584619522094727,
+      "learning_rate": 5.506000000000001e-07,
+      "loss": 0.5507,
+      "step": 47270
+    },
+    {
+      "epoch": 4.728,
+      "grad_norm": 23.282672882080078,
+      "learning_rate": 5.486e-07,
+      "loss": 0.4576,
+      "step": 47280
+    },
+    {
+      "epoch": 4.729,
+      "grad_norm": 40.43513107299805,
+      "learning_rate": 5.466000000000001e-07,
+      "loss": 0.7413,
+      "step": 47290
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 43.14529800415039,
+      "learning_rate": 5.446e-07,
+      "loss": 0.5918,
+      "step": 47300
+    },
+    {
+      "epoch": 4.731,
+      "grad_norm": 41.957130432128906,
+      "learning_rate": 5.426000000000001e-07,
+      "loss": 0.365,
+      "step": 47310
+    },
+    {
+      "epoch": 4.732,
+      "grad_norm": 41.200340270996094,
+      "learning_rate": 5.406000000000001e-07,
+      "loss": 0.5841,
+      "step": 47320
+    },
+    {
+      "epoch": 4.733,
+      "grad_norm": 2.041828155517578,
+      "learning_rate": 5.386e-07,
+      "loss": 0.5923,
+      "step": 47330
+    },
+    {
+      "epoch": 4.734,
+      "grad_norm": 44.379859924316406,
+      "learning_rate": 5.366000000000001e-07,
+      "loss": 0.4744,
+      "step": 47340
+    },
+    {
+      "epoch": 4.735,
+      "grad_norm": 16.02285385131836,
+      "learning_rate": 5.346e-07,
+      "loss": 0.5911,
+      "step": 47350
+    },
+    {
+      "epoch": 4.736,
+      "grad_norm": 47.35900115966797,
+      "learning_rate": 5.326000000000001e-07,
+      "loss": 0.4788,
+      "step": 47360
+    },
+    {
+      "epoch": 4.737,
+      "grad_norm": 8.870173454284668,
+      "learning_rate": 5.306e-07,
+      "loss": 0.3585,
+      "step": 47370
+    },
+    {
+      "epoch": 4.7379999999999995,
+      "grad_norm": 27.033184051513672,
+      "learning_rate": 5.286e-07,
+      "loss": 0.3327,
+      "step": 47380
+    },
+    {
+      "epoch": 4.739,
+      "grad_norm": 11.23887825012207,
+      "learning_rate": 5.266000000000001e-07,
+      "loss": 0.7588,
+      "step": 47390
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 25.569849014282227,
+      "learning_rate": 5.246e-07,
+      "loss": 0.3565,
+      "step": 47400
+    },
+    {
+      "epoch": 4.741,
+      "grad_norm": 59.0633430480957,
+      "learning_rate": 5.226000000000001e-07,
+      "loss": 1.0683,
+      "step": 47410
+    },
+    {
+      "epoch": 4.742,
+      "grad_norm": 7.970805644989014,
+      "learning_rate": 5.206e-07,
+      "loss": 0.7565,
+      "step": 47420
+    },
+    {
+      "epoch": 4.743,
+      "grad_norm": 24.26109504699707,
+      "learning_rate": 5.186000000000001e-07,
+      "loss": 0.7171,
+      "step": 47430
+    },
+    {
+      "epoch": 4.744,
+      "grad_norm": 63.284576416015625,
+      "learning_rate": 5.166000000000001e-07,
+      "loss": 0.6909,
+      "step": 47440
+    },
+    {
+      "epoch": 4.745,
+      "grad_norm": 46.69160842895508,
+      "learning_rate": 5.146e-07,
+      "loss": 0.777,
+      "step": 47450
+    },
+    {
+      "epoch": 4.746,
+      "grad_norm": 33.398040771484375,
+      "learning_rate": 5.126000000000001e-07,
+      "loss": 0.5602,
+      "step": 47460
+    },
+    {
+      "epoch": 4.747,
+      "grad_norm": 1.8984227180480957,
+      "learning_rate": 5.106e-07,
+      "loss": 0.4943,
+      "step": 47470
+    },
+    {
+      "epoch": 4.748,
+      "grad_norm": 35.92313003540039,
+      "learning_rate": 5.086000000000001e-07,
+      "loss": 0.4842,
+      "step": 47480
+    },
+    {
+      "epoch": 4.749,
+      "grad_norm": 33.2078742980957,
+      "learning_rate": 5.066e-07,
+      "loss": 0.4065,
+      "step": 47490
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 3.545969247817993,
+      "learning_rate": 5.046e-07,
+      "loss": 0.4536,
+      "step": 47500
+    },
+    {
+      "epoch": 4.751,
+      "grad_norm": 27.299518585205078,
+      "learning_rate": 5.026000000000001e-07,
+      "loss": 0.5854,
+      "step": 47510
+    },
+    {
+      "epoch": 4.752,
+      "grad_norm": 57.00627517700195,
+      "learning_rate": 5.006e-07,
+      "loss": 0.6149,
+      "step": 47520
+    },
+    {
+      "epoch": 4.753,
+      "grad_norm": 11.110939979553223,
+      "learning_rate": 4.986000000000001e-07,
+      "loss": 0.5755,
+      "step": 47530
+    },
+    {
+      "epoch": 4.754,
+      "grad_norm": 31.530773162841797,
+      "learning_rate": 4.966e-07,
+      "loss": 0.5945,
+      "step": 47540
+    },
+    {
+      "epoch": 4.755,
+      "grad_norm": 44.44669723510742,
+      "learning_rate": 4.946e-07,
+      "loss": 0.5209,
+      "step": 47550
+    },
+    {
+      "epoch": 4.756,
+      "grad_norm": 2.758739948272705,
+      "learning_rate": 4.926000000000001e-07,
+      "loss": 0.3809,
+      "step": 47560
+    },
+    {
+      "epoch": 4.757,
+      "grad_norm": 49.22467041015625,
+      "learning_rate": 4.906e-07,
+      "loss": 0.8543,
+      "step": 47570
+    },
+    {
+      "epoch": 4.758,
+      "grad_norm": 7.413259983062744,
+      "learning_rate": 4.886000000000001e-07,
+      "loss": 0.5036,
+      "step": 47580
+    },
+    {
+      "epoch": 4.759,
+      "grad_norm": 16.42161750793457,
+      "learning_rate": 4.866e-07,
+      "loss": 0.5253,
+      "step": 47590
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 55.729469299316406,
+      "learning_rate": 4.846000000000001e-07,
+      "loss": 0.5325,
+      "step": 47600
+    },
+    {
+      "epoch": 4.761,
+      "grad_norm": 13.986032485961914,
+      "learning_rate": 4.826e-07,
+      "loss": 0.7039,
+      "step": 47610
+    },
+    {
+      "epoch": 4.7620000000000005,
+      "grad_norm": 41.571712493896484,
+      "learning_rate": 4.806e-07,
+      "loss": 0.6829,
+      "step": 47620
+    },
+    {
+      "epoch": 4.763,
+      "grad_norm": 29.633747100830078,
+      "learning_rate": 4.786000000000001e-07,
+      "loss": 0.8681,
+      "step": 47630
+    },
+    {
+      "epoch": 4.764,
+      "grad_norm": 16.89261817932129,
+      "learning_rate": 4.7660000000000007e-07,
+      "loss": 0.4153,
+      "step": 47640
+    },
+    {
+      "epoch": 4.765,
+      "grad_norm": 21.218664169311523,
+      "learning_rate": 4.746000000000001e-07,
+      "loss": 0.467,
+      "step": 47650
+    },
+    {
+      "epoch": 4.766,
+      "grad_norm": 14.788098335266113,
+      "learning_rate": 4.726000000000001e-07,
+      "loss": 0.5418,
+      "step": 47660
+    },
+    {
+      "epoch": 4.767,
+      "grad_norm": 8.380966186523438,
+      "learning_rate": 4.706e-07,
+      "loss": 0.7172,
+      "step": 47670
+    },
+    {
+      "epoch": 4.768,
+      "grad_norm": 27.72803497314453,
+      "learning_rate": 4.6860000000000005e-07,
+      "loss": 0.6403,
+      "step": 47680
+    },
+    {
+      "epoch": 4.769,
+      "grad_norm": 48.69464874267578,
+      "learning_rate": 4.6660000000000006e-07,
+      "loss": 0.628,
+      "step": 47690
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 16.649084091186523,
+      "learning_rate": 4.6460000000000007e-07,
+      "loss": 0.5426,
+      "step": 47700
+    },
+    {
+      "epoch": 4.771,
+      "grad_norm": 47.99180603027344,
+      "learning_rate": 4.626000000000001e-07,
+      "loss": 0.5476,
+      "step": 47710
+    },
+    {
+      "epoch": 4.772,
+      "grad_norm": 9.165487289428711,
+      "learning_rate": 4.606e-07,
+      "loss": 0.6779,
+      "step": 47720
+    },
+    {
+      "epoch": 4.773,
+      "grad_norm": 36.032318115234375,
+      "learning_rate": 4.5860000000000004e-07,
+      "loss": 0.5837,
+      "step": 47730
+    },
+    {
+      "epoch": 4.774,
+      "grad_norm": 76.74220275878906,
+      "learning_rate": 4.5660000000000005e-07,
+      "loss": 0.7682,
+      "step": 47740
+    },
+    {
+      "epoch": 4.775,
+      "grad_norm": 76.99014282226562,
+      "learning_rate": 4.5460000000000006e-07,
+      "loss": 0.4876,
+      "step": 47750
+    },
+    {
+      "epoch": 4.776,
+      "grad_norm": 65.31263732910156,
+      "learning_rate": 4.5260000000000007e-07,
+      "loss": 0.7052,
+      "step": 47760
+    },
+    {
+      "epoch": 4.777,
+      "grad_norm": 86.4690170288086,
+      "learning_rate": 4.506000000000001e-07,
+      "loss": 0.654,
+      "step": 47770
+    },
+    {
+      "epoch": 4.7780000000000005,
+      "grad_norm": 46.257225036621094,
+      "learning_rate": 4.4860000000000003e-07,
+      "loss": 0.9597,
+      "step": 47780
+    },
+    {
+      "epoch": 4.779,
+      "grad_norm": 18.706884384155273,
+      "learning_rate": 4.4660000000000004e-07,
+      "loss": 0.3825,
+      "step": 47790
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 35.63112258911133,
+      "learning_rate": 4.4460000000000005e-07,
+      "loss": 0.7898,
+      "step": 47800
+    },
+    {
+      "epoch": 4.781,
+      "grad_norm": 91.87278747558594,
+      "learning_rate": 4.4260000000000006e-07,
+      "loss": 0.9079,
+      "step": 47810
+    },
+    {
+      "epoch": 4.782,
+      "grad_norm": 3.9658284187316895,
+      "learning_rate": 4.4060000000000007e-07,
+      "loss": 0.5102,
+      "step": 47820
+    },
+    {
+      "epoch": 4.783,
+      "grad_norm": 50.287960052490234,
+      "learning_rate": 4.386000000000001e-07,
+      "loss": 0.5358,
+      "step": 47830
+    },
+    {
+      "epoch": 4.784,
+      "grad_norm": 50.575645446777344,
+      "learning_rate": 4.3660000000000003e-07,
+      "loss": 0.559,
+      "step": 47840
+    },
+    {
+      "epoch": 4.785,
+      "grad_norm": 73.50508880615234,
+      "learning_rate": 4.3460000000000004e-07,
+      "loss": 0.5853,
+      "step": 47850
+    },
+    {
+      "epoch": 4.786,
+      "grad_norm": 40.10487747192383,
+      "learning_rate": 4.3260000000000005e-07,
+      "loss": 0.6971,
+      "step": 47860
+    },
+    {
+      "epoch": 4.787,
+      "grad_norm": 48.18107223510742,
+      "learning_rate": 4.3060000000000006e-07,
+      "loss": 0.5523,
+      "step": 47870
+    },
+    {
+      "epoch": 4.788,
+      "grad_norm": 44.405696868896484,
+      "learning_rate": 4.2860000000000007e-07,
+      "loss": 0.6391,
+      "step": 47880
+    },
+    {
+      "epoch": 4.789,
+      "grad_norm": 42.72483444213867,
+      "learning_rate": 4.266e-07,
+      "loss": 0.7918,
+      "step": 47890
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 70.26227569580078,
+      "learning_rate": 4.2460000000000003e-07,
+      "loss": 0.8501,
+      "step": 47900
+    },
+    {
+      "epoch": 4.791,
+      "grad_norm": 31.3819522857666,
+      "learning_rate": 4.2260000000000004e-07,
+      "loss": 0.4523,
+      "step": 47910
+    },
+    {
+      "epoch": 4.792,
+      "grad_norm": 4.014766216278076,
+      "learning_rate": 4.2060000000000005e-07,
+      "loss": 0.7206,
+      "step": 47920
+    },
+    {
+      "epoch": 4.793,
+      "grad_norm": 43.470123291015625,
+      "learning_rate": 4.1860000000000006e-07,
+      "loss": 0.6754,
+      "step": 47930
+    },
+    {
+      "epoch": 4.7940000000000005,
+      "grad_norm": 29.402721405029297,
+      "learning_rate": 4.1660000000000006e-07,
+      "loss": 0.4421,
+      "step": 47940
+    },
+    {
+      "epoch": 4.795,
+      "grad_norm": 15.759544372558594,
+      "learning_rate": 4.146e-07,
+      "loss": 0.6538,
+      "step": 47950
+    },
+    {
+      "epoch": 4.796,
+      "grad_norm": 59.475364685058594,
+      "learning_rate": 4.1260000000000003e-07,
+      "loss": 0.4141,
+      "step": 47960
+    },
+    {
+      "epoch": 4.797,
+      "grad_norm": 15.005043983459473,
+      "learning_rate": 4.1060000000000004e-07,
+      "loss": 0.4547,
+      "step": 47970
+    },
+    {
+      "epoch": 4.798,
+      "grad_norm": 38.049049377441406,
+      "learning_rate": 4.0860000000000005e-07,
+      "loss": 0.5253,
+      "step": 47980
+    },
+    {
+      "epoch": 4.799,
+      "grad_norm": 59.97571563720703,
+      "learning_rate": 4.0660000000000005e-07,
+      "loss": 0.4392,
+      "step": 47990
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 17.64630699157715,
+      "learning_rate": 4.0460000000000006e-07,
+      "loss": 0.9641,
+      "step": 48000
+    },
+    {
+      "epoch": 4.801,
+      "grad_norm": 1.9898121356964111,
+      "learning_rate": 4.026e-07,
+      "loss": 0.4899,
+      "step": 48010
+    },
+    {
+      "epoch": 4.802,
+      "grad_norm": 11.052675247192383,
+      "learning_rate": 4.0060000000000003e-07,
+      "loss": 0.4204,
+      "step": 48020
+    },
+    {
+      "epoch": 4.803,
+      "grad_norm": 17.712017059326172,
+      "learning_rate": 3.9860000000000004e-07,
+      "loss": 0.3256,
+      "step": 48030
+    },
+    {
+      "epoch": 4.804,
+      "grad_norm": 51.75345230102539,
+      "learning_rate": 3.9660000000000004e-07,
+      "loss": 0.59,
+      "step": 48040
+    },
+    {
+      "epoch": 4.805,
+      "grad_norm": 10.139244079589844,
+      "learning_rate": 3.9460000000000005e-07,
+      "loss": 0.4246,
+      "step": 48050
+    },
+    {
+      "epoch": 4.806,
+      "grad_norm": 4.2397565841674805,
+      "learning_rate": 3.9260000000000006e-07,
+      "loss": 0.6121,
+      "step": 48060
+    },
+    {
+      "epoch": 4.807,
+      "grad_norm": 54.398658752441406,
+      "learning_rate": 3.906e-07,
+      "loss": 0.5002,
+      "step": 48070
+    },
+    {
+      "epoch": 4.808,
+      "grad_norm": 7.418992519378662,
+      "learning_rate": 3.886e-07,
+      "loss": 0.6649,
+      "step": 48080
+    },
+    {
+      "epoch": 4.809,
+      "grad_norm": 47.92473220825195,
+      "learning_rate": 3.8660000000000003e-07,
+      "loss": 0.8182,
+      "step": 48090
+    },
+    {
+      "epoch": 4.8100000000000005,
+      "grad_norm": 4.537411212921143,
+      "learning_rate": 3.8460000000000004e-07,
+      "loss": 0.7614,
+      "step": 48100
+    },
+    {
+      "epoch": 4.811,
+      "grad_norm": 48.04644012451172,
+      "learning_rate": 3.8260000000000005e-07,
+      "loss": 0.3634,
+      "step": 48110
+    },
+    {
+      "epoch": 4.812,
+      "grad_norm": 59.322444915771484,
+      "learning_rate": 3.806e-07,
+      "loss": 0.5423,
+      "step": 48120
+    },
+    {
+      "epoch": 4.813,
+      "grad_norm": 33.344459533691406,
+      "learning_rate": 3.786e-07,
+      "loss": 0.5785,
+      "step": 48130
+    },
+    {
+      "epoch": 4.814,
+      "grad_norm": 12.657391548156738,
+      "learning_rate": 3.766e-07,
+      "loss": 0.2472,
+      "step": 48140
+    },
+    {
+      "epoch": 4.8149999999999995,
+      "grad_norm": 47.85975646972656,
+      "learning_rate": 3.7460000000000003e-07,
+      "loss": 0.7532,
+      "step": 48150
+    },
+    {
+      "epoch": 4.816,
+      "grad_norm": 24.278907775878906,
+      "learning_rate": 3.7260000000000004e-07,
+      "loss": 0.3976,
+      "step": 48160
+    },
+    {
+      "epoch": 4.817,
+      "grad_norm": 54.600223541259766,
+      "learning_rate": 3.7060000000000005e-07,
+      "loss": 0.6928,
+      "step": 48170
+    },
+    {
+      "epoch": 4.818,
+      "grad_norm": 8.999923706054688,
+      "learning_rate": 3.686e-07,
+      "loss": 0.3901,
+      "step": 48180
+    },
+    {
+      "epoch": 4.819,
+      "grad_norm": 37.98063278198242,
+      "learning_rate": 3.666e-07,
+      "loss": 0.7222,
+      "step": 48190
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 1.4624940156936646,
+      "learning_rate": 3.646e-07,
+      "loss": 0.3561,
+      "step": 48200
+    },
+    {
+      "epoch": 4.821,
+      "grad_norm": 60.897274017333984,
+      "learning_rate": 3.6260000000000003e-07,
+      "loss": 0.6086,
+      "step": 48210
+    },
+    {
+      "epoch": 4.822,
+      "grad_norm": 7.063051223754883,
+      "learning_rate": 3.6060000000000004e-07,
+      "loss": 0.7798,
+      "step": 48220
+    },
+    {
+      "epoch": 4.823,
+      "grad_norm": 9.376193046569824,
+      "learning_rate": 3.5860000000000005e-07,
+      "loss": 0.4789,
+      "step": 48230
+    },
+    {
+      "epoch": 4.824,
+      "grad_norm": 19.10816192626953,
+      "learning_rate": 3.566e-07,
+      "loss": 0.5622,
+      "step": 48240
+    },
+    {
+      "epoch": 4.825,
+      "grad_norm": 11.21962833404541,
+      "learning_rate": 3.546e-07,
+      "loss": 0.635,
+      "step": 48250
+    },
+    {
+      "epoch": 4.826,
+      "grad_norm": 37.95383071899414,
+      "learning_rate": 3.526e-07,
+      "loss": 0.4713,
+      "step": 48260
+    },
+    {
+      "epoch": 4.827,
+      "grad_norm": 22.839855194091797,
+      "learning_rate": 3.5060000000000003e-07,
+      "loss": 0.4713,
+      "step": 48270
+    },
+    {
+      "epoch": 4.828,
+      "grad_norm": 11.70213794708252,
+      "learning_rate": 3.4860000000000004e-07,
+      "loss": 0.5201,
+      "step": 48280
+    },
+    {
+      "epoch": 4.829,
+      "grad_norm": 45.10593795776367,
+      "learning_rate": 3.4660000000000005e-07,
+      "loss": 0.3367,
+      "step": 48290
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 36.21401596069336,
+      "learning_rate": 3.446e-07,
+      "loss": 0.8444,
+      "step": 48300
+    },
+    {
+      "epoch": 4.8309999999999995,
+      "grad_norm": 75.2431640625,
+      "learning_rate": 3.426e-07,
+      "loss": 0.5507,
+      "step": 48310
+    },
+    {
+      "epoch": 4.832,
+      "grad_norm": 29.88875961303711,
+      "learning_rate": 3.406e-07,
+      "loss": 0.5728,
+      "step": 48320
+    },
+    {
+      "epoch": 4.833,
+      "grad_norm": 11.833975791931152,
+      "learning_rate": 3.3860000000000003e-07,
+      "loss": 0.6646,
+      "step": 48330
+    },
+    {
+      "epoch": 4.834,
+      "grad_norm": 4.960333347320557,
+      "learning_rate": 3.3660000000000004e-07,
+      "loss": 0.427,
+      "step": 48340
+    },
+    {
+      "epoch": 4.835,
+      "grad_norm": 49.04979705810547,
+      "learning_rate": 3.346e-07,
+      "loss": 0.4861,
+      "step": 48350
+    },
+    {
+      "epoch": 4.836,
+      "grad_norm": 56.9226188659668,
+      "learning_rate": 3.326e-07,
+      "loss": 0.4323,
+      "step": 48360
+    },
+    {
+      "epoch": 4.837,
+      "grad_norm": 21.75094223022461,
+      "learning_rate": 3.306e-07,
+      "loss": 0.333,
+      "step": 48370
+    },
+    {
+      "epoch": 4.838,
+      "grad_norm": 10.19582748413086,
+      "learning_rate": 3.286e-07,
+      "loss": 0.4192,
+      "step": 48380
+    },
+    {
+      "epoch": 4.839,
+      "grad_norm": 32.832767486572266,
+      "learning_rate": 3.266e-07,
+      "loss": 0.3247,
+      "step": 48390
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 47.89490509033203,
+      "learning_rate": 3.2460000000000003e-07,
+      "loss": 0.6204,
+      "step": 48400
+    },
+    {
+      "epoch": 4.841,
+      "grad_norm": 64.20399475097656,
+      "learning_rate": 3.226e-07,
+      "loss": 0.4072,
+      "step": 48410
+    },
+    {
+      "epoch": 4.842,
+      "grad_norm": 42.400634765625,
+      "learning_rate": 3.206e-07,
+      "loss": 0.5873,
+      "step": 48420
+    },
+    {
+      "epoch": 4.843,
+      "grad_norm": 7.792730331420898,
+      "learning_rate": 3.186e-07,
+      "loss": 0.5349,
+      "step": 48430
+    },
+    {
+      "epoch": 4.844,
+      "grad_norm": 40.12997817993164,
+      "learning_rate": 3.166e-07,
+      "loss": 0.7333,
+      "step": 48440
+    },
+    {
+      "epoch": 4.845,
+      "grad_norm": 14.662808418273926,
+      "learning_rate": 3.146e-07,
+      "loss": 0.8452,
+      "step": 48450
+    },
+    {
+      "epoch": 4.846,
+      "grad_norm": 11.02692985534668,
+      "learning_rate": 3.1260000000000003e-07,
+      "loss": 0.6037,
+      "step": 48460
+    },
+    {
+      "epoch": 4.8469999999999995,
+      "grad_norm": 107.82157897949219,
+      "learning_rate": 3.1060000000000004e-07,
+      "loss": 0.8563,
+      "step": 48470
+    },
+    {
+      "epoch": 4.848,
+      "grad_norm": 6.1017961502075195,
+      "learning_rate": 3.086e-07,
+      "loss": 0.4407,
+      "step": 48480
+    },
+    {
+      "epoch": 4.849,
+      "grad_norm": 32.714542388916016,
+      "learning_rate": 3.066e-07,
+      "loss": 0.4017,
+      "step": 48490
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 3.4416143894195557,
+      "learning_rate": 3.046e-07,
+      "loss": 0.4439,
+      "step": 48500
+    },
+    {
+      "epoch": 4.851,
+      "grad_norm": 45.20467758178711,
+      "learning_rate": 3.026e-07,
+      "loss": 0.4796,
+      "step": 48510
+    },
+    {
+      "epoch": 4.852,
+      "grad_norm": 44.18875503540039,
+      "learning_rate": 3.0060000000000003e-07,
+      "loss": 0.5077,
+      "step": 48520
+    },
+    {
+      "epoch": 4.853,
+      "grad_norm": 44.58082962036133,
+      "learning_rate": 2.9860000000000004e-07,
+      "loss": 0.5945,
+      "step": 48530
+    },
+    {
+      "epoch": 4.854,
+      "grad_norm": 18.132604598999023,
+      "learning_rate": 2.966e-07,
+      "loss": 0.8124,
+      "step": 48540
+    },
+    {
+      "epoch": 4.855,
+      "grad_norm": 35.660037994384766,
+      "learning_rate": 2.946e-07,
+      "loss": 0.3541,
+      "step": 48550
+    },
+    {
+      "epoch": 4.856,
+      "grad_norm": 37.676273345947266,
+      "learning_rate": 2.926e-07,
+      "loss": 0.4191,
+      "step": 48560
+    },
+    {
+      "epoch": 4.857,
+      "grad_norm": 18.786176681518555,
+      "learning_rate": 2.906e-07,
+      "loss": 0.3393,
+      "step": 48570
+    },
+    {
+      "epoch": 4.858,
+      "grad_norm": 35.841712951660156,
+      "learning_rate": 2.8860000000000003e-07,
+      "loss": 0.7147,
+      "step": 48580
+    },
+    {
+      "epoch": 4.859,
+      "grad_norm": 4.490417003631592,
+      "learning_rate": 2.8660000000000004e-07,
+      "loss": 0.441,
+      "step": 48590
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 27.341760635375977,
+      "learning_rate": 2.846e-07,
+      "loss": 0.7622,
+      "step": 48600
+    },
+    {
+      "epoch": 4.861,
+      "grad_norm": 43.09422302246094,
+      "learning_rate": 2.826e-07,
+      "loss": 0.6165,
+      "step": 48610
+    },
+    {
+      "epoch": 4.862,
+      "grad_norm": 38.81334686279297,
+      "learning_rate": 2.8060000000000006e-07,
+      "loss": 0.5359,
+      "step": 48620
+    },
+    {
+      "epoch": 4.8629999999999995,
+      "grad_norm": 5.982778549194336,
+      "learning_rate": 2.786e-07,
+      "loss": 0.5695,
+      "step": 48630
+    },
+    {
+      "epoch": 4.864,
+      "grad_norm": 33.21123504638672,
+      "learning_rate": 2.7660000000000003e-07,
+      "loss": 0.4943,
+      "step": 48640
+    },
+    {
+      "epoch": 4.865,
+      "grad_norm": 61.49201965332031,
+      "learning_rate": 2.746e-07,
+      "loss": 0.6732,
+      "step": 48650
+    },
+    {
+      "epoch": 4.866,
+      "grad_norm": 34.918212890625,
+      "learning_rate": 2.726e-07,
+      "loss": 0.7058,
+      "step": 48660
+    },
+    {
+      "epoch": 4.867,
+      "grad_norm": 42.28673553466797,
+      "learning_rate": 2.7060000000000005e-07,
+      "loss": 0.4972,
+      "step": 48670
+    },
+    {
+      "epoch": 4.868,
+      "grad_norm": 12.002147674560547,
+      "learning_rate": 2.686e-07,
+      "loss": 0.8624,
+      "step": 48680
+    },
+    {
+      "epoch": 4.869,
+      "grad_norm": 62.372684478759766,
+      "learning_rate": 2.666e-07,
+      "loss": 0.6257,
+      "step": 48690
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 39.709686279296875,
+      "learning_rate": 2.6460000000000003e-07,
+      "loss": 0.6162,
+      "step": 48700
+    },
+    {
+      "epoch": 4.871,
+      "grad_norm": 16.147838592529297,
+      "learning_rate": 2.626e-07,
+      "loss": 0.8548,
+      "step": 48710
+    },
+    {
+      "epoch": 4.872,
+      "grad_norm": 4.075834274291992,
+      "learning_rate": 2.6060000000000004e-07,
+      "loss": 0.382,
+      "step": 48720
+    },
+    {
+      "epoch": 4.873,
+      "grad_norm": 19.27889060974121,
+      "learning_rate": 2.5860000000000005e-07,
+      "loss": 0.641,
+      "step": 48730
+    },
+    {
+      "epoch": 4.874,
+      "grad_norm": 48.61284255981445,
+      "learning_rate": 2.566e-07,
+      "loss": 0.3655,
+      "step": 48740
+    },
+    {
+      "epoch": 4.875,
+      "grad_norm": 56.545310974121094,
+      "learning_rate": 2.546e-07,
+      "loss": 0.6569,
+      "step": 48750
+    },
+    {
+      "epoch": 4.876,
+      "grad_norm": 36.132381439208984,
+      "learning_rate": 2.526e-07,
+      "loss": 0.4681,
+      "step": 48760
+    },
+    {
+      "epoch": 4.877,
+      "grad_norm": 51.53099822998047,
+      "learning_rate": 2.5060000000000003e-07,
+      "loss": 0.6395,
+      "step": 48770
+    },
+    {
+      "epoch": 4.878,
+      "grad_norm": 46.316402435302734,
+      "learning_rate": 2.4860000000000004e-07,
+      "loss": 0.4726,
+      "step": 48780
+    },
+    {
+      "epoch": 4.879,
+      "grad_norm": 6.3659210205078125,
+      "learning_rate": 2.4660000000000005e-07,
+      "loss": 0.8035,
+      "step": 48790
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 16.26994514465332,
+      "learning_rate": 2.446e-07,
+      "loss": 0.4792,
+      "step": 48800
+    },
+    {
+      "epoch": 4.881,
+      "grad_norm": 46.80910110473633,
+      "learning_rate": 2.426e-07,
+      "loss": 0.9064,
+      "step": 48810
+    },
+    {
+      "epoch": 4.882,
+      "grad_norm": 19.964271545410156,
+      "learning_rate": 2.406e-07,
+      "loss": 0.8111,
+      "step": 48820
+    },
+    {
+      "epoch": 4.883,
+      "grad_norm": 72.45056915283203,
+      "learning_rate": 2.3860000000000003e-07,
+      "loss": 0.6046,
+      "step": 48830
+    },
+    {
+      "epoch": 4.884,
+      "grad_norm": 33.62139892578125,
+      "learning_rate": 2.3660000000000001e-07,
+      "loss": 0.2831,
+      "step": 48840
+    },
+    {
+      "epoch": 4.885,
+      "grad_norm": 6.268035888671875,
+      "learning_rate": 2.3460000000000002e-07,
+      "loss": 0.2903,
+      "step": 48850
+    },
+    {
+      "epoch": 4.886,
+      "grad_norm": 31.210264205932617,
+      "learning_rate": 2.326e-07,
+      "loss": 0.5656,
+      "step": 48860
+    },
+    {
+      "epoch": 4.8870000000000005,
+      "grad_norm": 89.84651947021484,
+      "learning_rate": 2.3060000000000001e-07,
+      "loss": 0.6617,
+      "step": 48870
+    },
+    {
+      "epoch": 4.888,
+      "grad_norm": 78.52317810058594,
+      "learning_rate": 2.286e-07,
+      "loss": 0.733,
+      "step": 48880
+    },
+    {
+      "epoch": 4.889,
+      "grad_norm": 16.838603973388672,
+      "learning_rate": 2.266e-07,
+      "loss": 0.5299,
+      "step": 48890
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.49068504571914673,
+      "learning_rate": 2.2460000000000004e-07,
+      "loss": 0.3942,
+      "step": 48900
+    },
+    {
+      "epoch": 4.891,
+      "grad_norm": 28.578716278076172,
+      "learning_rate": 2.226e-07,
+      "loss": 0.6955,
+      "step": 48910
+    },
+    {
+      "epoch": 4.892,
+      "grad_norm": 1.8998624086380005,
+      "learning_rate": 2.206e-07,
+      "loss": 0.5737,
+      "step": 48920
+    },
+    {
+      "epoch": 4.893,
+      "grad_norm": 50.37765884399414,
+      "learning_rate": 2.1860000000000004e-07,
+      "loss": 0.6957,
+      "step": 48930
+    },
+    {
+      "epoch": 4.894,
+      "grad_norm": 21.070785522460938,
+      "learning_rate": 2.166e-07,
+      "loss": 0.6923,
+      "step": 48940
+    },
+    {
+      "epoch": 4.895,
+      "grad_norm": 44.579280853271484,
+      "learning_rate": 2.1460000000000003e-07,
+      "loss": 0.5985,
+      "step": 48950
+    },
+    {
+      "epoch": 4.896,
+      "grad_norm": 15.574660301208496,
+      "learning_rate": 2.1260000000000004e-07,
+      "loss": 0.6324,
+      "step": 48960
+    },
+    {
+      "epoch": 4.897,
+      "grad_norm": 54.42012405395508,
+      "learning_rate": 2.106e-07,
+      "loss": 0.746,
+      "step": 48970
+    },
+    {
+      "epoch": 4.898,
+      "grad_norm": 15.344828605651855,
+      "learning_rate": 2.0860000000000003e-07,
+      "loss": 0.5326,
+      "step": 48980
+    },
+    {
+      "epoch": 4.899,
+      "grad_norm": 37.128597259521484,
+      "learning_rate": 2.0660000000000004e-07,
+      "loss": 0.5567,
+      "step": 48990
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 22.833782196044922,
+      "learning_rate": 2.0460000000000002e-07,
+      "loss": 0.6224,
+      "step": 49000
+    },
+    {
+      "epoch": 4.901,
+      "grad_norm": 43.60334396362305,
+      "learning_rate": 2.0260000000000003e-07,
+      "loss": 0.8438,
+      "step": 49010
+    },
+    {
+      "epoch": 4.902,
+      "grad_norm": 17.917469024658203,
+      "learning_rate": 2.0060000000000004e-07,
+      "loss": 0.2035,
+      "step": 49020
+    },
+    {
+      "epoch": 4.9030000000000005,
+      "grad_norm": 22.8342227935791,
+      "learning_rate": 1.9860000000000002e-07,
+      "loss": 0.5491,
+      "step": 49030
+    },
+    {
+      "epoch": 4.904,
+      "grad_norm": 19.834346771240234,
+      "learning_rate": 1.9660000000000003e-07,
+      "loss": 0.411,
+      "step": 49040
+    },
+    {
+      "epoch": 4.905,
+      "grad_norm": 23.630340576171875,
+      "learning_rate": 1.9460000000000004e-07,
+      "loss": 0.6375,
+      "step": 49050
+    },
+    {
+      "epoch": 4.906,
+      "grad_norm": 54.59148025512695,
+      "learning_rate": 1.9260000000000002e-07,
+      "loss": 0.8771,
+      "step": 49060
+    },
+    {
+      "epoch": 4.907,
+      "grad_norm": 30.65956687927246,
+      "learning_rate": 1.9060000000000003e-07,
+      "loss": 0.6741,
+      "step": 49070
+    },
+    {
+      "epoch": 4.908,
+      "grad_norm": 61.472137451171875,
+      "learning_rate": 1.886e-07,
+      "loss": 0.6089,
+      "step": 49080
+    },
+    {
+      "epoch": 4.909,
+      "grad_norm": 38.48383331298828,
+      "learning_rate": 1.8660000000000002e-07,
+      "loss": 0.3453,
+      "step": 49090
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 25.503416061401367,
+      "learning_rate": 1.8460000000000003e-07,
+      "loss": 0.3371,
+      "step": 49100
+    },
+    {
+      "epoch": 4.911,
+      "grad_norm": 26.52053451538086,
+      "learning_rate": 1.826e-07,
+      "loss": 1.0525,
+      "step": 49110
+    },
+    {
+      "epoch": 4.912,
+      "grad_norm": 53.508087158203125,
+      "learning_rate": 1.8060000000000002e-07,
+      "loss": 0.5461,
+      "step": 49120
+    },
+    {
+      "epoch": 4.913,
+      "grad_norm": 36.59590530395508,
+      "learning_rate": 1.7860000000000002e-07,
+      "loss": 0.4117,
+      "step": 49130
+    },
+    {
+      "epoch": 4.914,
+      "grad_norm": 31.170352935791016,
+      "learning_rate": 1.766e-07,
+      "loss": 0.4214,
+      "step": 49140
+    },
+    {
+      "epoch": 4.915,
+      "grad_norm": 72.6932601928711,
+      "learning_rate": 1.7460000000000001e-07,
+      "loss": 0.719,
+      "step": 49150
+    },
+    {
+      "epoch": 4.916,
+      "grad_norm": 34.64444351196289,
+      "learning_rate": 1.7260000000000002e-07,
+      "loss": 0.4933,
+      "step": 49160
+    },
+    {
+      "epoch": 4.917,
+      "grad_norm": 54.596031188964844,
+      "learning_rate": 1.706e-07,
+      "loss": 0.6415,
+      "step": 49170
+    },
+    {
+      "epoch": 4.918,
+      "grad_norm": 34.222679138183594,
+      "learning_rate": 1.6860000000000001e-07,
+      "loss": 0.4704,
+      "step": 49180
+    },
+    {
+      "epoch": 4.9190000000000005,
+      "grad_norm": 3.0823545455932617,
+      "learning_rate": 1.6660000000000002e-07,
+      "loss": 0.6312,
+      "step": 49190
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 61.71739196777344,
+      "learning_rate": 1.646e-07,
+      "loss": 0.6899,
+      "step": 49200
+    },
+    {
+      "epoch": 4.921,
+      "grad_norm": 62.166526794433594,
+      "learning_rate": 1.6260000000000001e-07,
+      "loss": 0.5632,
+      "step": 49210
+    },
+    {
+      "epoch": 4.922,
+      "grad_norm": 75.72866821289062,
+      "learning_rate": 1.6060000000000002e-07,
+      "loss": 0.6717,
+      "step": 49220
+    },
+    {
+      "epoch": 4.923,
+      "grad_norm": 19.29648208618164,
+      "learning_rate": 1.586e-07,
+      "loss": 0.412,
+      "step": 49230
+    },
+    {
+      "epoch": 4.924,
+      "grad_norm": 57.19310760498047,
+      "learning_rate": 1.566e-07,
+      "loss": 0.7854,
+      "step": 49240
+    },
+    {
+      "epoch": 4.925,
+      "grad_norm": 16.37799835205078,
+      "learning_rate": 1.5460000000000002e-07,
+      "loss": 0.4818,
+      "step": 49250
+    },
+    {
+      "epoch": 4.926,
+      "grad_norm": 3.643502712249756,
+      "learning_rate": 1.526e-07,
+      "loss": 0.3695,
+      "step": 49260
+    },
+    {
+      "epoch": 4.927,
+      "grad_norm": 57.638790130615234,
+      "learning_rate": 1.506e-07,
+      "loss": 0.5841,
+      "step": 49270
+    },
+    {
+      "epoch": 4.928,
+      "grad_norm": 38.98382568359375,
+      "learning_rate": 1.4860000000000002e-07,
+      "loss": 0.311,
+      "step": 49280
+    },
+    {
+      "epoch": 4.929,
+      "grad_norm": 31.647018432617188,
+      "learning_rate": 1.466e-07,
+      "loss": 0.5787,
+      "step": 49290
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 5.648628234863281,
+      "learning_rate": 1.446e-07,
+      "loss": 0.5783,
+      "step": 49300
+    },
+    {
+      "epoch": 4.931,
+      "grad_norm": 9.384773254394531,
+      "learning_rate": 1.4260000000000002e-07,
+      "loss": 0.5762,
+      "step": 49310
+    },
+    {
+      "epoch": 4.932,
+      "grad_norm": 55.00576400756836,
+      "learning_rate": 1.406e-07,
+      "loss": 0.4893,
+      "step": 49320
+    },
+    {
+      "epoch": 4.933,
+      "grad_norm": 23.570556640625,
+      "learning_rate": 1.386e-07,
+      "loss": 0.5582,
+      "step": 49330
+    },
+    {
+      "epoch": 4.934,
+      "grad_norm": 43.51008224487305,
+      "learning_rate": 1.3660000000000002e-07,
+      "loss": 0.5739,
+      "step": 49340
+    },
+    {
+      "epoch": 4.9350000000000005,
+      "grad_norm": 58.253780364990234,
+      "learning_rate": 1.346e-07,
+      "loss": 0.5964,
+      "step": 49350
+    },
+    {
+      "epoch": 4.936,
+      "grad_norm": 46.08802795410156,
+      "learning_rate": 1.326e-07,
+      "loss": 0.6438,
+      "step": 49360
+    },
+    {
+      "epoch": 4.937,
+      "grad_norm": 8.2721586227417,
+      "learning_rate": 1.3060000000000002e-07,
+      "loss": 0.372,
+      "step": 49370
+    },
+    {
+      "epoch": 4.938,
+      "grad_norm": 24.139907836914062,
+      "learning_rate": 1.286e-07,
+      "loss": 0.5713,
+      "step": 49380
+    },
+    {
+      "epoch": 4.939,
+      "grad_norm": 16.5051212310791,
+      "learning_rate": 1.266e-07,
+      "loss": 0.7615,
+      "step": 49390
+    },
+    {
+      "epoch": 4.9399999999999995,
+      "grad_norm": 3.0267720222473145,
+      "learning_rate": 1.2460000000000002e-07,
+      "loss": 0.3022,
+      "step": 49400
+    },
+    {
+      "epoch": 4.941,
+      "grad_norm": 53.91283416748047,
+      "learning_rate": 1.226e-07,
+      "loss": 0.7291,
+      "step": 49410
+    },
+    {
+      "epoch": 4.942,
+      "grad_norm": 23.457191467285156,
+      "learning_rate": 1.206e-07,
+      "loss": 0.5664,
+      "step": 49420
+    },
+    {
+      "epoch": 4.943,
+      "grad_norm": 14.853793144226074,
+      "learning_rate": 1.1860000000000002e-07,
+      "loss": 0.4499,
+      "step": 49430
+    },
+    {
+      "epoch": 4.944,
+      "grad_norm": 27.185049057006836,
+      "learning_rate": 1.1660000000000001e-07,
+      "loss": 0.4925,
+      "step": 49440
+    },
+    {
+      "epoch": 4.945,
+      "grad_norm": 69.38607788085938,
+      "learning_rate": 1.1460000000000001e-07,
+      "loss": 0.393,
+      "step": 49450
+    },
+    {
+      "epoch": 4.946,
+      "grad_norm": 35.48990249633789,
+      "learning_rate": 1.126e-07,
+      "loss": 0.4629,
+      "step": 49460
+    },
+    {
+      "epoch": 4.947,
+      "grad_norm": 35.915348052978516,
+      "learning_rate": 1.1060000000000001e-07,
+      "loss": 0.4844,
+      "step": 49470
+    },
+    {
+      "epoch": 4.948,
+      "grad_norm": 47.767757415771484,
+      "learning_rate": 1.086e-07,
+      "loss": 0.6391,
+      "step": 49480
+    },
+    {
+      "epoch": 4.949,
+      "grad_norm": 46.730716705322266,
+      "learning_rate": 1.066e-07,
+      "loss": 0.5344,
+      "step": 49490
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 9.486944198608398,
+      "learning_rate": 1.0460000000000001e-07,
+      "loss": 0.5001,
+      "step": 49500
+    },
+    {
+      "epoch": 4.951,
+      "grad_norm": 49.90391540527344,
+      "learning_rate": 1.026e-07,
+      "loss": 0.4198,
+      "step": 49510
+    },
+    {
+      "epoch": 4.952,
+      "grad_norm": 31.162935256958008,
+      "learning_rate": 1.006e-07,
+      "loss": 0.6015,
+      "step": 49520
+    },
+    {
+      "epoch": 4.953,
+      "grad_norm": 14.631115913391113,
+      "learning_rate": 9.860000000000001e-08,
+      "loss": 0.5281,
+      "step": 49530
+    },
+    {
+      "epoch": 4.954,
+      "grad_norm": 89.07907104492188,
+      "learning_rate": 9.66e-08,
+      "loss": 0.7108,
+      "step": 49540
+    },
+    {
+      "epoch": 4.955,
+      "grad_norm": 29.647022247314453,
+      "learning_rate": 9.46e-08,
+      "loss": 0.4537,
+      "step": 49550
+    },
+    {
+      "epoch": 4.9559999999999995,
+      "grad_norm": 57.06145095825195,
+      "learning_rate": 9.26e-08,
+      "loss": 0.5961,
+      "step": 49560
+    },
+    {
+      "epoch": 4.957,
+      "grad_norm": 28.40804672241211,
+      "learning_rate": 9.060000000000002e-08,
+      "loss": 0.5376,
+      "step": 49570
+    },
+    {
+      "epoch": 4.958,
+      "grad_norm": 69.8585205078125,
+      "learning_rate": 8.86e-08,
+      "loss": 0.5019,
+      "step": 49580
+    },
+    {
+      "epoch": 4.959,
+      "grad_norm": 42.624027252197266,
+      "learning_rate": 8.66e-08,
+      "loss": 0.3875,
+      "step": 49590
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 2.737194538116455,
+      "learning_rate": 8.460000000000002e-08,
+      "loss": 0.563,
+      "step": 49600
+    },
+    {
+      "epoch": 4.961,
+      "grad_norm": 48.68153381347656,
+      "learning_rate": 8.260000000000001e-08,
+      "loss": 0.6256,
+      "step": 49610
+    },
+    {
+      "epoch": 4.962,
+      "grad_norm": 10.589071273803711,
+      "learning_rate": 8.060000000000001e-08,
+      "loss": 0.4583,
+      "step": 49620
+    },
+    {
+      "epoch": 4.963,
+      "grad_norm": 49.633148193359375,
+      "learning_rate": 7.860000000000002e-08,
+      "loss": 0.8238,
+      "step": 49630
+    },
+    {
+      "epoch": 4.964,
+      "grad_norm": 47.60674285888672,
+      "learning_rate": 7.660000000000001e-08,
+      "loss": 0.4204,
+      "step": 49640
+    },
+    {
+      "epoch": 4.965,
+      "grad_norm": 68.43152618408203,
+      "learning_rate": 7.46e-08,
+      "loss": 0.5481,
+      "step": 49650
+    },
+    {
+      "epoch": 4.966,
+      "grad_norm": 62.700435638427734,
+      "learning_rate": 7.26e-08,
+      "loss": 0.5413,
+      "step": 49660
+    },
+    {
+      "epoch": 4.967,
+      "grad_norm": 43.648651123046875,
+      "learning_rate": 7.060000000000001e-08,
+      "loss": 0.4658,
+      "step": 49670
+    },
+    {
+      "epoch": 4.968,
+      "grad_norm": 53.35328674316406,
+      "learning_rate": 6.86e-08,
+      "loss": 0.4889,
+      "step": 49680
+    },
+    {
+      "epoch": 4.969,
+      "grad_norm": 69.61092376708984,
+      "learning_rate": 6.66e-08,
+      "loss": 0.5914,
+      "step": 49690
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 34.3068962097168,
+      "learning_rate": 6.46e-08,
+      "loss": 0.4936,
+      "step": 49700
+    },
+    {
+      "epoch": 4.971,
+      "grad_norm": 70.36450958251953,
+      "learning_rate": 6.26e-08,
+      "loss": 0.4934,
+      "step": 49710
+    },
+    {
+      "epoch": 4.9719999999999995,
+      "grad_norm": 29.908767700195312,
+      "learning_rate": 6.060000000000001e-08,
+      "loss": 0.6727,
+      "step": 49720
+    },
+    {
+      "epoch": 4.973,
+      "grad_norm": 74.44947814941406,
+      "learning_rate": 5.86e-08,
+      "loss": 0.5676,
+      "step": 49730
+    },
+    {
+      "epoch": 4.974,
+      "grad_norm": 36.034976959228516,
+      "learning_rate": 5.6600000000000004e-08,
+      "loss": 0.63,
+      "step": 49740
+    },
+    {
+      "epoch": 4.975,
+      "grad_norm": 2.7080929279327393,
+      "learning_rate": 5.46e-08,
+      "loss": 0.6098,
+      "step": 49750
+    },
+    {
+      "epoch": 4.976,
+      "grad_norm": 56.25693893432617,
+      "learning_rate": 5.26e-08,
+      "loss": 0.3895,
+      "step": 49760
+    },
+    {
+      "epoch": 4.977,
+      "grad_norm": 18.81259536743164,
+      "learning_rate": 5.060000000000001e-08,
+      "loss": 0.5584,
+      "step": 49770
+    },
+    {
+      "epoch": 4.978,
+      "grad_norm": 4.564599990844727,
+      "learning_rate": 4.86e-08,
+      "loss": 0.583,
+      "step": 49780
+    },
+    {
+      "epoch": 4.979,
+      "grad_norm": 35.461177825927734,
+      "learning_rate": 4.660000000000001e-08,
+      "loss": 0.7756,
+      "step": 49790
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 43.247154235839844,
+      "learning_rate": 4.460000000000001e-08,
+      "loss": 0.5009,
+      "step": 49800
+    },
+    {
+      "epoch": 4.981,
+      "grad_norm": 61.074039459228516,
+      "learning_rate": 4.2600000000000004e-08,
+      "loss": 0.6001,
+      "step": 49810
+    },
+    {
+      "epoch": 4.982,
+      "grad_norm": 78.60579681396484,
+      "learning_rate": 4.0600000000000006e-08,
+      "loss": 0.6014,
+      "step": 49820
+    },
+    {
+      "epoch": 4.983,
+      "grad_norm": 32.76850891113281,
+      "learning_rate": 3.86e-08,
+      "loss": 0.4978,
+      "step": 49830
+    },
+    {
+      "epoch": 4.984,
+      "grad_norm": 1.1829404830932617,
+      "learning_rate": 3.6600000000000003e-08,
+      "loss": 0.2477,
+      "step": 49840
+    },
+    {
+      "epoch": 4.985,
+      "grad_norm": 36.923988342285156,
+      "learning_rate": 3.4600000000000005e-08,
+      "loss": 0.5418,
+      "step": 49850
+    },
+    {
+      "epoch": 4.986,
+      "grad_norm": 27.69318199157715,
+      "learning_rate": 3.26e-08,
+      "loss": 0.5853,
+      "step": 49860
+    },
+    {
+      "epoch": 4.987,
+      "grad_norm": 42.778106689453125,
+      "learning_rate": 3.06e-08,
+      "loss": 0.6131,
+      "step": 49870
+    },
+    {
+      "epoch": 4.9879999999999995,
+      "grad_norm": 11.142000198364258,
+      "learning_rate": 2.8600000000000005e-08,
+      "loss": 0.3882,
+      "step": 49880
+    },
+    {
+      "epoch": 4.989,
+      "grad_norm": 37.72666931152344,
+      "learning_rate": 2.6600000000000003e-08,
+      "loss": 0.6759,
+      "step": 49890
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 36.423744201660156,
+      "learning_rate": 2.4600000000000002e-08,
+      "loss": 0.4503,
+      "step": 49900
+    },
+    {
+      "epoch": 4.991,
+      "grad_norm": 37.07157897949219,
+      "learning_rate": 2.26e-08,
+      "loss": 0.7043,
+      "step": 49910
+    },
+    {
+      "epoch": 4.992,
+      "grad_norm": 8.626225471496582,
+      "learning_rate": 2.0600000000000002e-08,
+      "loss": 0.4042,
+      "step": 49920
+    },
+    {
+      "epoch": 4.993,
+      "grad_norm": 66.21869659423828,
+      "learning_rate": 1.86e-08,
+      "loss": 0.9308,
+      "step": 49930
+    },
+    {
+      "epoch": 4.994,
+      "grad_norm": 42.68752670288086,
+      "learning_rate": 1.6600000000000003e-08,
+      "loss": 0.2277,
+      "step": 49940
+    },
+    {
+      "epoch": 4.995,
+      "grad_norm": 23.218891143798828,
+      "learning_rate": 1.46e-08,
+      "loss": 0.6241,
+      "step": 49950
+    },
+    {
+      "epoch": 4.996,
+      "grad_norm": 44.54765319824219,
+      "learning_rate": 1.2600000000000002e-08,
+      "loss": 0.6953,
+      "step": 49960
+    },
+    {
+      "epoch": 4.997,
+      "grad_norm": 19.68933868408203,
+      "learning_rate": 1.06e-08,
+      "loss": 0.4644,
+      "step": 49970
+    },
+    {
+      "epoch": 4.998,
+      "grad_norm": 6.50738525390625,
+      "learning_rate": 8.600000000000001e-09,
+      "loss": 0.8382,
+      "step": 49980
+    },
+    {
+      "epoch": 4.999,
+      "grad_norm": 44.634403228759766,
+      "learning_rate": 6.6e-09,
+      "loss": 0.561,
+      "step": 49990
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 44.17949676513672,
+      "learning_rate": 4.600000000000001e-09,
+      "loss": 0.4599,
+      "step": 50000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 50000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}