diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,36381 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 5193,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.216445284675434,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 1.9775,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.911595589639173,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 2.1666,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.589768673301773,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 1.9854,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.568648529053654,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 2.2656,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.549288858256404,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 1.9927,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.110487979033767,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 2.1646,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.9015150661795,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 2.129,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.501364965637194,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 1.9416,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.453575658521874,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 1.9497,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.607773385934923,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 2.0166,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.464353749675528,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 1.8781,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.97360564518581,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 1.9095,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.992113441163921,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.6395,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.552028290409154,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 1.5434,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.311907698101182,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 1.5281,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.678981025513341,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 1.4648,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0588935799454995,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 0.7232,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.903466856108649,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 1.2737,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.685212001248653,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.3402,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.245435062932411,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 1.3337,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.158382292780615,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 1.3811,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.186622696017208,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 1.2546,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.517891507093355,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 1.18,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1001253598950638,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.3383,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2465679665413227,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 0.832,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0120421636293195,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.3458,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4484948949766845,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 0.8444,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1110997732647743,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 1.2984,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.755277848262445,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.2309,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.860582982001812,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 1.2708,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5897985578473377,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 1.2035,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5495914062962792,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 1.1789,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.389696878446986,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 1.2519,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.076541587027184,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 1.1727,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6797871781763094,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 0.792,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1243426876134257,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 1.1818,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.299174657545576,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 1.2307,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0371458205650415,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 1.1005,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2455222878601098,
+      "learning_rate": 5e-06,
+      "loss": 1.1864,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1266583614163515,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 1.15,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.089620655737586,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 1.1877,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.197147361182619,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 1.1843,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1442985310608946,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 0.8406,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2343649074177634,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 1.2641,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2274293780592904,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 1.2005,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1702020802705553,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 1.2,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2728644356062087,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 1.1639,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.296917779568054,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 1.1447,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1847920179158975,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.1049,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.305602717661454,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 1.1454,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8163614577368297,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 1.1534,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0156958103382907,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.1014,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2061600271142927,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 1.1111,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0514290811875346,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 1.2195,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.009350847624063,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 1.1672,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9383085189275544,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 1.1329,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.231367175822086,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 1.1168,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0008003220100643,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 1.2286,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8980157371763204,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 1.0885,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9742084904113024,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 1.1383,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.130005401042239,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 1.1859,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.986055300283839,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 1.1223,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.901481096038548,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 1.0932,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.125815394626458,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 1.2282,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9414092406206,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.1273,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9746610829599591,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 1.048,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.154415501683974,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 1.0115,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.868695894296507,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 1.1381,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8811818709807824,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 1.1031,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3006046265407827,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 1.1062,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1699669644056137,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 1.1372,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8731108123220086,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.043,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1143191194853337,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 1.2095,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9506052984256566,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 1.1677,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.004755199707491,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 1.1288,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2712419994719104,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 1.1489,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9915382414879983,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 1.1356,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8655506688742343,
+      "learning_rate": 1e-05,
+      "loss": 1.0524,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1482093297445575,
+      "learning_rate": 1.012820512820513e-05,
+      "loss": 0.9386,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9215954323448314,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 1.1801,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9081067065165245,
+      "learning_rate": 1.0384615384615386e-05,
+      "loss": 1.0638,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9407275778581317,
+      "learning_rate": 1.0512820512820514e-05,
+      "loss": 1.1074,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.025638661080253,
+      "learning_rate": 1.0641025641025643e-05,
+      "loss": 1.0776,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.919914713517629,
+      "learning_rate": 1.076923076923077e-05,
+      "loss": 1.0189,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1903909968624977,
+      "learning_rate": 1.0897435897435898e-05,
+      "loss": 1.1085,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.17664798755956,
+      "learning_rate": 1.1025641025641028e-05,
+      "loss": 1.0529,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.113038538305221,
+      "learning_rate": 1.1153846153846154e-05,
+      "loss": 1.1169,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0486185050480894,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 1.1889,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.045888088194733,
+      "learning_rate": 1.1410256410256411e-05,
+      "loss": 1.1538,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7806490104735406,
+      "learning_rate": 1.1538461538461538e-05,
+      "loss": 1.0914,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1971405645771527,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 1.048,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.002750992348389,
+      "learning_rate": 1.1794871794871796e-05,
+      "loss": 1.0975,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.187960188699195,
+      "learning_rate": 1.1923076923076925e-05,
+      "loss": 1.0906,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0174873060151857,
+      "learning_rate": 1.2051282051282051e-05,
+      "loss": 1.0992,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0443319007612852,
+      "learning_rate": 1.217948717948718e-05,
+      "loss": 1.082,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2748628196945109,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.7718,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.896137442216524,
+      "learning_rate": 1.2435897435897436e-05,
+      "loss": 1.1549,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.95843096706086,
+      "learning_rate": 1.2564102564102565e-05,
+      "loss": 1.128,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0686480210411657,
+      "learning_rate": 1.2692307692307693e-05,
+      "loss": 1.1249,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9404669088500452,
+      "learning_rate": 1.2820512820512823e-05,
+      "loss": 1.115,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8653249453155127,
+      "learning_rate": 1.294871794871795e-05,
+      "loss": 1.1051,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.10107543117975,
+      "learning_rate": 1.3076923076923078e-05,
+      "loss": 1.0875,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8747249569118767,
+      "learning_rate": 1.3205128205128207e-05,
+      "loss": 1.0451,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9850084346269088,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.0689,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.049123340018695,
+      "learning_rate": 1.3461538461538463e-05,
+      "loss": 1.0905,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1264307975036925,
+      "learning_rate": 1.3589743589743592e-05,
+      "loss": 1.1224,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0502742046299907,
+      "learning_rate": 1.3717948717948718e-05,
+      "loss": 1.0915,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8303440971792886,
+      "learning_rate": 1.3846153846153847e-05,
+      "loss": 1.0949,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2313836077753413,
+      "learning_rate": 1.3974358974358975e-05,
+      "loss": 1.0429,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8919532863183655,
+      "learning_rate": 1.4102564102564105e-05,
+      "loss": 1.0763,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0704343058304526,
+      "learning_rate": 1.4230769230769232e-05,
+      "loss": 1.1657,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.430724552963001,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 1.0815,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1085720824994245,
+      "learning_rate": 1.4487179487179489e-05,
+      "loss": 0.866,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8279405733949405,
+      "learning_rate": 1.4615384615384615e-05,
+      "loss": 1.0434,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9381964085003,
+      "learning_rate": 1.4743589743589745e-05,
+      "loss": 1.0764,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9487925657079459,
+      "learning_rate": 1.4871794871794874e-05,
+      "loss": 0.7605,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1860059019926847,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.0553,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9458206802695046,
+      "learning_rate": 1.5128205128205129e-05,
+      "loss": 1.048,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.835280245148161,
+      "learning_rate": 1.5256410256410257e-05,
+      "loss": 1.055,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0496294448787333,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 1.1847,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.218142247887536,
+      "learning_rate": 1.5512820512820516e-05,
+      "loss": 1.0944,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4654880331995055,
+      "learning_rate": 1.5641025641025644e-05,
+      "loss": 1.1229,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9292976768956316,
+      "learning_rate": 1.576923076923077e-05,
+      "loss": 1.0633,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9404361588446972,
+      "learning_rate": 1.5897435897435897e-05,
+      "loss": 1.0854,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9362855000410057,
+      "learning_rate": 1.602564102564103e-05,
+      "loss": 1.0061,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.083318365558794,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 1.0612,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.122373700349859,
+      "learning_rate": 1.6282051282051282e-05,
+      "loss": 0.9871,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.359431021405571,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 1.0635,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0429268801216616,
+      "learning_rate": 1.653846153846154e-05,
+      "loss": 0.8406,
+      "step": 129
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.394181875364715,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.0533,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0279839342385624,
+      "learning_rate": 1.6794871794871796e-05,
+      "loss": 1.1405,
+      "step": 131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.133411698576336,
+      "learning_rate": 1.6923076923076924e-05,
+      "loss": 1.1123,
+      "step": 132
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0961335815701654,
+      "learning_rate": 1.7051282051282053e-05,
+      "loss": 1.1525,
+      "step": 133
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0054012186334265,
+      "learning_rate": 1.717948717948718e-05,
+      "loss": 1.0072,
+      "step": 134
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9954089428486153,
+      "learning_rate": 1.730769230769231e-05,
+      "loss": 0.9954,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9266633957752657,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 1.0307,
+      "step": 136
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0235377885233734,
+      "learning_rate": 1.7564102564102566e-05,
+      "loss": 1.0939,
+      "step": 137
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0175192551390526,
+      "learning_rate": 1.7692307692307694e-05,
+      "loss": 1.0717,
+      "step": 138
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0342136013306582,
+      "learning_rate": 1.7820512820512823e-05,
+      "loss": 0.808,
+      "step": 139
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.428830129959296,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 1.082,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9642876987686577,
+      "learning_rate": 1.807692307692308e-05,
+      "loss": 1.1122,
+      "step": 141
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.027656601747293,
+      "learning_rate": 1.8205128205128208e-05,
+      "loss": 1.0719,
+      "step": 142
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.183894708322355,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 1.1253,
+      "step": 143
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.062043365070345,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 1.0173,
+      "step": 144
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.033473969787848,
+      "learning_rate": 1.8589743589743593e-05,
+      "loss": 1.1215,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.062919227268133,
+      "learning_rate": 1.8717948717948718e-05,
+      "loss": 0.9471,
+      "step": 146
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.070479749634218,
+      "learning_rate": 1.8846153846153846e-05,
+      "loss": 1.1518,
+      "step": 147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9754065164309047,
+      "learning_rate": 1.8974358974358975e-05,
+      "loss": 1.0442,
+      "step": 148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.13339324208887,
+      "learning_rate": 1.9102564102564106e-05,
+      "loss": 1.0118,
+      "step": 149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.369137183205933,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 1.092,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1638382502455435,
+      "learning_rate": 1.935897435897436e-05,
+      "loss": 0.9977,
+      "step": 151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1783663043501296,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 1.0939,
+      "step": 152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9497922494089797,
+      "learning_rate": 1.9615384615384617e-05,
+      "loss": 1.0152,
+      "step": 153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0547723801429747,
+      "learning_rate": 1.9743589743589745e-05,
+      "loss": 1.0223,
+      "step": 154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.070242647096933,
+      "learning_rate": 1.9871794871794873e-05,
+      "loss": 0.9549,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9301029883856002,
+      "learning_rate": 2e-05,
+      "loss": 1.0594,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1096159037163598,
+      "learning_rate": 1.9999998054972106e-05,
+      "loss": 1.0369,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1245553998838536,
+      "learning_rate": 1.9999992219889184e-05,
+      "loss": 0.9525,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.131488635617247,
+      "learning_rate": 1.99999824947535e-05,
+      "loss": 1.0323,
+      "step": 159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.363965004377438,
+      "learning_rate": 1.9999968879568835e-05,
+      "loss": 1.0856,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0585329850283522,
+      "learning_rate": 1.9999951374340493e-05,
+      "loss": 1.0892,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9750564938392907,
+      "learning_rate": 1.9999929979075278e-05,
+      "loss": 1.0452,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.141328684995138,
+      "learning_rate": 1.999990469378151e-05,
+      "loss": 1.1101,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.031433061724888,
+      "learning_rate": 1.9999875518469033e-05,
+      "loss": 0.799,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4065812716730095,
+      "learning_rate": 1.9999842453149192e-05,
+      "loss": 1.0711,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.202667832906497,
+      "learning_rate": 1.999980549783485e-05,
+      "loss": 0.994,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.129957736688869,
+      "learning_rate": 1.9999764652540382e-05,
+      "loss": 1.0424,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.32808847121314,
+      "learning_rate": 1.9999719917281682e-05,
+      "loss": 1.1104,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.402682209687328,
+      "learning_rate": 1.9999671292076145e-05,
+      "loss": 1.1413,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1985356868714634,
+      "learning_rate": 1.9999618776942692e-05,
+      "loss": 1.0147,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.174198937460871,
+      "learning_rate": 1.999956237190175e-05,
+      "loss": 1.0795,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5202206051237606,
+      "learning_rate": 1.9999502076975257e-05,
+      "loss": 0.9901,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1957807846109967,
+      "learning_rate": 1.9999437892186673e-05,
+      "loss": 0.9667,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9569728406033189,
+      "learning_rate": 1.9999369817560967e-05,
+      "loss": 1.0522,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.245962863895941,
+      "learning_rate": 1.9999297853124614e-05,
+      "loss": 1.0186,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.413279476262106,
+      "learning_rate": 1.9999221998905613e-05,
+      "loss": 1.0631,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.076972357794391,
+      "learning_rate": 1.9999142254933476e-05,
+      "loss": 1.0691,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.972083747779467,
+      "learning_rate": 1.9999058621239217e-05,
+      "loss": 1.0699,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1755848177736334,
+      "learning_rate": 1.9998971097855372e-05,
+      "loss": 1.1303,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9839915727644544,
+      "learning_rate": 1.999887968481599e-05,
+      "loss": 1.0598,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0373868712914938,
+      "learning_rate": 1.9998784382156622e-05,
+      "loss": 1.0762,
+      "step": 181
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0029728344755298,
+      "learning_rate": 1.9998685189914356e-05,
+      "loss": 1.1297,
+      "step": 182
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9387023208997487,
+      "learning_rate": 1.999858210812777e-05,
+      "loss": 0.9651,
+      "step": 183
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1732843858578414,
+      "learning_rate": 1.9998475136836966e-05,
+      "loss": 1.0479,
+      "step": 184
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.954541614593413,
+      "learning_rate": 1.999836427608355e-05,
+      "loss": 0.9924,
+      "step": 185
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9489826509232497,
+      "learning_rate": 1.9998249525910656e-05,
+      "loss": 1.0869,
+      "step": 186
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.762474252467788,
+      "learning_rate": 1.999813088636292e-05,
+      "loss": 0.9219,
+      "step": 187
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0627539245716164,
+      "learning_rate": 1.999800835748649e-05,
+      "loss": 1.0559,
+      "step": 188
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.137164696299313,
+      "learning_rate": 1.9997881939329034e-05,
+      "loss": 1.0711,
+      "step": 189
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.102050244558165,
+      "learning_rate": 1.9997751631939724e-05,
+      "loss": 1.0547,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9560280671941297,
+      "learning_rate": 1.999761743536926e-05,
+      "loss": 1.0285,
+      "step": 191
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2056335755059853,
+      "learning_rate": 1.9997479349669836e-05,
+      "loss": 1.039,
+      "step": 192
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.857852624351155,
+      "learning_rate": 1.999733737489517e-05,
+      "loss": 1.0525,
+      "step": 193
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1149817844051846,
+      "learning_rate": 1.9997191511100498e-05,
+      "loss": 1.1631,
+      "step": 194
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1742104034580874,
+      "learning_rate": 1.9997041758342554e-05,
+      "loss": 1.0704,
+      "step": 195
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.92170752202666,
+      "learning_rate": 1.9996888116679597e-05,
+      "loss": 1.0977,
+      "step": 196
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8760103416157878,
+      "learning_rate": 1.999673058617139e-05,
+      "loss": 1.0593,
+      "step": 197
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.239018605486002,
+      "learning_rate": 1.9996569166879215e-05,
+      "loss": 1.1451,
+      "step": 198
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9680194040337873,
+      "learning_rate": 1.9996403858865867e-05,
+      "loss": 1.1002,
+      "step": 199
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8952136254504348,
+      "learning_rate": 1.9996234662195653e-05,
+      "loss": 1.13,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7805244417146453,
+      "learning_rate": 1.9996061576934388e-05,
+      "loss": 1.088,
+      "step": 201
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8466727506858036,
+      "learning_rate": 1.9995884603149403e-05,
+      "loss": 1.0727,
+      "step": 202
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8297374299414597,
+      "learning_rate": 1.9995703740909542e-05,
+      "loss": 0.9719,
+      "step": 203
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8430958263907924,
+      "learning_rate": 1.9995518990285166e-05,
+      "loss": 1.0023,
+      "step": 204
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0008787877943846,
+      "learning_rate": 1.9995330351348135e-05,
+      "loss": 1.1435,
+      "step": 205
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.767812451189823,
+      "learning_rate": 1.999513782417184e-05,
+      "loss": 0.988,
+      "step": 206
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9499593819305316,
+      "learning_rate": 1.999494140883117e-05,
+      "loss": 1.0819,
+      "step": 207
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9273733720896404,
+      "learning_rate": 1.9994741105402533e-05,
+      "loss": 0.9283,
+      "step": 208
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.045214086540413,
+      "learning_rate": 1.9994536913963852e-05,
+      "loss": 1.0441,
+      "step": 209
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9323194027489348,
+      "learning_rate": 1.999432883459455e-05,
+      "loss": 1.0739,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.78702758065982,
+      "learning_rate": 1.9994116867375574e-05,
+      "loss": 1.0648,
+      "step": 211
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.999325292321686,
+      "learning_rate": 1.9993901012389386e-05,
+      "loss": 1.0197,
+      "step": 212
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9750895391330754,
+      "learning_rate": 1.999368126971995e-05,
+      "loss": 1.0378,
+      "step": 213
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0645995523227865,
+      "learning_rate": 1.9993457639452748e-05,
+      "loss": 1.0636,
+      "step": 214
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0348647174788024,
+      "learning_rate": 1.999323012167477e-05,
+      "loss": 1.0233,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.955566109346601,
+      "learning_rate": 1.999299871647453e-05,
+      "loss": 1.1114,
+      "step": 216
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0108878112553326,
+      "learning_rate": 1.9992763423942038e-05,
+      "loss": 1.0761,
+      "step": 217
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9661856223780965,
+      "learning_rate": 1.999252424416883e-05,
+      "loss": 1.1253,
+      "step": 218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1696126303119874,
+      "learning_rate": 1.9992281177247942e-05,
+      "loss": 1.0375,
+      "step": 219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2041115439000005,
+      "learning_rate": 1.999203422327393e-05,
+      "loss": 0.8076,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9028998749132995,
+      "learning_rate": 1.9991783382342867e-05,
+      "loss": 1.0468,
+      "step": 221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2722564344114153,
+      "learning_rate": 1.9991528654552326e-05,
+      "loss": 1.066,
+      "step": 222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1410050185723164,
+      "learning_rate": 1.9991270040001396e-05,
+      "loss": 1.0753,
+      "step": 223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0921277876178137,
+      "learning_rate": 1.9991007538790686e-05,
+      "loss": 1.0768,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0882235349073617,
+      "learning_rate": 1.9990741151022302e-05,
+      "loss": 1.0978,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9414030643468547,
+      "learning_rate": 1.9990470876799876e-05,
+      "loss": 1.002,
+      "step": 226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.17093585650916,
+      "learning_rate": 1.9990196716228546e-05,
+      "loss": 1.0542,
+      "step": 227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.955440445323398,
+      "learning_rate": 1.998991866941496e-05,
+      "loss": 1.0849,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0226531291514176,
+      "learning_rate": 1.9989636736467278e-05,
+      "loss": 1.1224,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2336914840149285,
+      "learning_rate": 1.9989350917495177e-05,
+      "loss": 1.0551,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.046392279697416,
+      "learning_rate": 1.9989061212609845e-05,
+      "loss": 1.0738,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.248695264268023,
+      "learning_rate": 1.998876762192397e-05,
+      "loss": 1.019,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3362932243579713,
+      "learning_rate": 1.9988470145551765e-05,
+      "loss": 1.0555,
+      "step": 233
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.149028179411019,
+      "learning_rate": 1.9988168783608955e-05,
+      "loss": 1.0529,
+      "step": 234
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8633459882398478,
+      "learning_rate": 1.9987863536212765e-05,
+      "loss": 1.0074,
+      "step": 235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0594450846907373,
+      "learning_rate": 1.9987554403481938e-05,
+      "loss": 0.966,
+      "step": 236
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9956893021238926,
+      "learning_rate": 1.9987241385536732e-05,
+      "loss": 1.0367,
+      "step": 237
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.025631177166738,
+      "learning_rate": 1.9986924482498907e-05,
+      "loss": 1.0407,
+      "step": 238
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0369291335090365,
+      "learning_rate": 1.998660369449175e-05,
+      "loss": 1.0408,
+      "step": 239
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9777103589259433,
+      "learning_rate": 1.998627902164004e-05,
+      "loss": 1.0964,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9884611294066915,
+      "learning_rate": 1.998595046407008e-05,
+      "loss": 0.7983,
+      "step": 241
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0510845934246746,
+      "learning_rate": 1.998561802190968e-05,
+      "loss": 1.0129,
+      "step": 242
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4092760114986467,
+      "learning_rate": 1.9985281695288165e-05,
+      "loss": 1.076,
+      "step": 243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9933609047274952,
+      "learning_rate": 1.998494148433636e-05,
+      "loss": 1.0225,
+      "step": 244
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0232974321656905,
+      "learning_rate": 1.9984597389186618e-05,
+      "loss": 0.829,
+      "step": 245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0605065022696967,
+      "learning_rate": 1.9984249409972794e-05,
+      "loss": 0.9931,
+      "step": 246
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2454749233665545,
+      "learning_rate": 1.9983897546830243e-05,
+      "loss": 1.0803,
+      "step": 247
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8431535853254453,
+      "learning_rate": 1.998354179989585e-05,
+      "loss": 0.983,
+      "step": 248
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8566618675252857,
+      "learning_rate": 1.9983182169308004e-05,
+      "loss": 0.9921,
+      "step": 249
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9375277651998517,
+      "learning_rate": 1.9982818655206602e-05,
+      "loss": 1.0113,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9572548426006255,
+      "learning_rate": 1.9982451257733047e-05,
+      "loss": 1.1174,
+      "step": 251
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0028626956048434,
+      "learning_rate": 1.9982079977030267e-05,
+      "loss": 1.1028,
+      "step": 252
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.96219145866627,
+      "learning_rate": 1.9981704813242685e-05,
+      "loss": 1.0566,
+      "step": 253
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.072795047778876,
+      "learning_rate": 1.9981325766516248e-05,
+      "loss": 1.0744,
+      "step": 254
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.036098657993234,
+      "learning_rate": 1.9980942836998404e-05,
+      "loss": 1.0763,
+      "step": 255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1388481294252855,
+      "learning_rate": 1.9980556024838116e-05,
+      "loss": 1.1613,
+      "step": 256
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.067974800297108,
+      "learning_rate": 1.9980165330185857e-05,
+      "loss": 0.998,
+      "step": 257
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8790643051942517,
+      "learning_rate": 1.9979770753193605e-05,
+      "loss": 0.996,
+      "step": 258
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9559433697812194,
+      "learning_rate": 1.9979372294014854e-05,
+      "loss": 1.0239,
+      "step": 259
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9308127181761399,
+      "learning_rate": 1.9978969952804613e-05,
+      "loss": 1.1269,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.924957556737311,
+      "learning_rate": 1.9978563729719386e-05,
+      "loss": 1.0322,
+      "step": 261
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.119914670588116,
+      "learning_rate": 1.99781536249172e-05,
+      "loss": 1.111,
+      "step": 262
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2814838916359177,
+      "learning_rate": 1.9977739638557593e-05,
+      "loss": 0.9882,
+      "step": 263
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9764920210183117,
+      "learning_rate": 1.99773217708016e-05,
+      "loss": 1.0747,
+      "step": 264
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8237059936330955,
+      "learning_rate": 1.997690002181178e-05,
+      "loss": 1.0228,
+      "step": 265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7957594475273144,
+      "learning_rate": 1.9976474391752192e-05,
+      "loss": 0.9876,
+      "step": 266
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9980073321283007,
+      "learning_rate": 1.997604488078841e-05,
+      "loss": 1.0911,
+      "step": 267
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0146788024415074,
+      "learning_rate": 1.9975611489087516e-05,
+      "loss": 1.022,
+      "step": 268
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1844582382911173,
+      "learning_rate": 1.9975174216818097e-05,
+      "loss": 1.0328,
+      "step": 269
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.117333899762936,
+      "learning_rate": 1.997473306415026e-05,
+      "loss": 0.9903,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8007107570814573,
+      "learning_rate": 1.997428803125562e-05,
+      "loss": 0.9672,
+      "step": 271
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.094142577887572,
+      "learning_rate": 1.9973839118307286e-05,
+      "loss": 1.0761,
+      "step": 272
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9921256605922988,
+      "learning_rate": 1.9973386325479897e-05,
+      "loss": 1.0929,
+      "step": 273
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7196271013592903,
+      "learning_rate": 1.9972929652949587e-05,
+      "loss": 0.989,
+      "step": 274
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.878734015541386,
+      "learning_rate": 1.9972469100894005e-05,
+      "loss": 1.0751,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9232024443567461,
+      "learning_rate": 1.9972004669492313e-05,
+      "loss": 1.0705,
+      "step": 276
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2481706461115407,
+      "learning_rate": 1.997153635892517e-05,
+      "loss": 0.8429,
+      "step": 277
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8327598842861688,
+      "learning_rate": 1.9971064169374755e-05,
+      "loss": 1.0771,
+      "step": 278
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7554713225318588,
+      "learning_rate": 1.9970588101024754e-05,
+      "loss": 1.0262,
+      "step": 279
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1469666455805716,
+      "learning_rate": 1.9970108154060357e-05,
+      "loss": 1.0043,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.67449508407844,
+      "learning_rate": 1.9969624328668267e-05,
+      "loss": 0.9486,
+      "step": 281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8820097189830565,
+      "learning_rate": 1.9969136625036694e-05,
+      "loss": 1.0906,
+      "step": 282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1303500190062223,
+      "learning_rate": 1.996864504335536e-05,
+      "loss": 1.1237,
+      "step": 283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.780586955227666,
+      "learning_rate": 1.9968149583815494e-05,
+      "loss": 1.003,
+      "step": 284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7489913602886422,
+      "learning_rate": 1.996765024660983e-05,
+      "loss": 0.9842,
+      "step": 285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0155100896265767,
+      "learning_rate": 1.996714703193261e-05,
+      "loss": 1.0557,
+      "step": 286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9438109572777202,
+      "learning_rate": 1.996663993997959e-05,
+      "loss": 0.9971,
+      "step": 287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0728126863662977,
+      "learning_rate": 1.9966128970948038e-05,
+      "loss": 0.8342,
+      "step": 288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9934840280210313,
+      "learning_rate": 1.9965614125036712e-05,
+      "loss": 1.0524,
+      "step": 289
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0936769507538386,
+      "learning_rate": 1.9965095402445898e-05,
+      "loss": 1.1212,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9745427712772865,
+      "learning_rate": 1.996457280337738e-05,
+      "loss": 1.0978,
+      "step": 291
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9179177319145042,
+      "learning_rate": 1.996404632803445e-05,
+      "loss": 1.0146,
+      "step": 292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.052073495470475,
+      "learning_rate": 1.9963515976621914e-05,
+      "loss": 1.0465,
+      "step": 293
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8225286684088682,
+      "learning_rate": 1.996298174934608e-05,
+      "loss": 1.1047,
+      "step": 294
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7580289784696486,
+      "learning_rate": 1.996244364641476e-05,
+      "loss": 0.9965,
+      "step": 295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0161159763669554,
+      "learning_rate": 1.996190166803729e-05,
+      "loss": 1.1507,
+      "step": 296
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8172911318254807,
+      "learning_rate": 1.996135581442449e-05,
+      "loss": 1.0353,
+      "step": 297
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8747317892295494,
+      "learning_rate": 1.996080608578871e-05,
+      "loss": 1.0875,
+      "step": 298
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0573127120786463,
+      "learning_rate": 1.9960252482343797e-05,
+      "loss": 1.0821,
+      "step": 299
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.900633660421174,
+      "learning_rate": 1.9959695004305097e-05,
+      "loss": 1.0045,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7491633697794904,
+      "learning_rate": 1.995913365188948e-05,
+      "loss": 1.0168,
+      "step": 301
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.05599591440065,
+      "learning_rate": 1.9958568425315316e-05,
+      "loss": 0.9449,
+      "step": 302
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8883190997093964,
+      "learning_rate": 1.9957999324802473e-05,
+      "loss": 1.0428,
+      "step": 303
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.720295899770073,
+      "learning_rate": 1.9957426350572345e-05,
+      "loss": 0.9482,
+      "step": 304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9816495220080403,
+      "learning_rate": 1.995684950284781e-05,
+      "loss": 1.0557,
+      "step": 305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.92187930498817,
+      "learning_rate": 1.9956268781853277e-05,
+      "loss": 1.0582,
+      "step": 306
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.795117789160023,
+      "learning_rate": 1.9955684187814644e-05,
+      "loss": 1.0761,
+      "step": 307
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9245112528044024,
+      "learning_rate": 1.9955095720959318e-05,
+      "loss": 1.1684,
+      "step": 308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9632462029062842,
+      "learning_rate": 1.9954503381516225e-05,
+      "loss": 0.7674,
+      "step": 309
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9219486581712448,
+      "learning_rate": 1.995390716971578e-05,
+      "loss": 1.0211,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9147529258042912,
+      "learning_rate": 1.9953307085789916e-05,
+      "loss": 1.0947,
+      "step": 311
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.023656325583728,
+      "learning_rate": 1.995270312997207e-05,
+      "loss": 0.8131,
+      "step": 312
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7903741102627215,
+      "learning_rate": 1.9952095302497185e-05,
+      "loss": 0.9933,
+      "step": 313
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.02294327217036,
+      "learning_rate": 1.9951483603601703e-05,
+      "loss": 0.9262,
+      "step": 314
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7932331946119509,
+      "learning_rate": 1.9950868033523584e-05,
+      "loss": 1.0388,
+      "step": 315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9016906199071297,
+      "learning_rate": 1.9950248592502286e-05,
+      "loss": 1.0161,
+      "step": 316
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9496588477478443,
+      "learning_rate": 1.994962528077878e-05,
+      "loss": 1.0056,
+      "step": 317
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7857218614144716,
+      "learning_rate": 1.9948998098595526e-05,
+      "loss": 1.0717,
+      "step": 318
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2058906093436734,
+      "learning_rate": 1.9948367046196513e-05,
+      "loss": 1.0229,
+      "step": 319
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8833597215221982,
+      "learning_rate": 1.9947732123827218e-05,
+      "loss": 1.0789,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0603717639007564,
+      "learning_rate": 1.994709333173463e-05,
+      "loss": 0.8642,
+      "step": 321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.033922433865109,
+      "learning_rate": 1.994645067016725e-05,
+      "loss": 1.0365,
+      "step": 322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.870516832634325,
+      "learning_rate": 1.9945804139375066e-05,
+      "loss": 0.9902,
+      "step": 323
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9861030358173385,
+      "learning_rate": 1.9945153739609588e-05,
+      "loss": 0.7725,
+      "step": 324
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9496791073979263,
+      "learning_rate": 1.9944499471123824e-05,
+      "loss": 1.0054,
+      "step": 325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0447227654810196,
+      "learning_rate": 1.9943841334172288e-05,
+      "loss": 1.1126,
+      "step": 326
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.24213973693733,
+      "learning_rate": 1.9943179329010997e-05,
+      "loss": 1.0988,
+      "step": 327
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8596491661660886,
+      "learning_rate": 1.994251345589748e-05,
+      "loss": 1.0312,
+      "step": 328
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9170393207754506,
+      "learning_rate": 1.9941843715090758e-05,
+      "loss": 1.0078,
+      "step": 329
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8284329304637323,
+      "learning_rate": 1.994117010685137e-05,
+      "loss": 0.9095,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9552073759311999,
+      "learning_rate": 1.9940492631441352e-05,
+      "loss": 1.0652,
+      "step": 331
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0132724055989812,
+      "learning_rate": 1.9939811289124246e-05,
+      "loss": 1.0849,
+      "step": 332
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.087593794800754,
+      "learning_rate": 1.9939126080165093e-05,
+      "loss": 1.0264,
+      "step": 333
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8617360148956503,
+      "learning_rate": 1.9938437004830453e-05,
+      "loss": 1.0128,
+      "step": 334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6487974768481515,
+      "learning_rate": 1.9937744063388365e-05,
+      "loss": 0.924,
+      "step": 335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8887708831775187,
+      "learning_rate": 1.9937047256108405e-05,
+      "loss": 0.9823,
+      "step": 336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.035534616225895,
+      "learning_rate": 1.9936346583261622e-05,
+      "loss": 1.0182,
+      "step": 337
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.969898654916401,
+      "learning_rate": 1.9935642045120585e-05,
+      "loss": 1.0657,
+      "step": 338
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.009217417503386,
+      "learning_rate": 1.9934933641959367e-05,
+      "loss": 0.9853,
+      "step": 339
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8124399073403066,
+      "learning_rate": 1.9934221374053538e-05,
+      "loss": 0.9696,
+      "step": 340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9066949773049553,
+      "learning_rate": 1.9933505241680176e-05,
+      "loss": 1.1261,
+      "step": 341
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.843104252573575,
+      "learning_rate": 1.9932785245117852e-05,
+      "loss": 1.0277,
+      "step": 342
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0257800926544487,
+      "learning_rate": 1.9932061384646662e-05,
+      "loss": 1.056,
+      "step": 343
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.069660038019677,
+      "learning_rate": 1.993133366054818e-05,
+      "loss": 0.8464,
+      "step": 344
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.047929245583957,
+      "learning_rate": 1.9930602073105503e-05,
+      "loss": 1.0297,
+      "step": 345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7804271272628047,
+      "learning_rate": 1.9929866622603217e-05,
+      "loss": 0.9681,
+      "step": 346
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8036059077290953,
+      "learning_rate": 1.992912730932742e-05,
+      "loss": 1.0982,
+      "step": 347
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1577273526751513,
+      "learning_rate": 1.992838413356571e-05,
+      "loss": 1.0578,
+      "step": 348
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9351592416164614,
+      "learning_rate": 1.992763709560718e-05,
+      "loss": 1.0799,
+      "step": 349
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7613887958883605,
+      "learning_rate": 1.992688619574244e-05,
+      "loss": 1.118,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7710807310053625,
+      "learning_rate": 1.9926131434263587e-05,
+      "loss": 1.0957,
+      "step": 351
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1014264167112024,
+      "learning_rate": 1.9925372811464232e-05,
+      "loss": 1.0916,
+      "step": 352
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.695271742923848,
+      "learning_rate": 1.9924610327639482e-05,
+      "loss": 0.9742,
+      "step": 353
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.946395684960528,
+      "learning_rate": 1.992384398308595e-05,
+      "loss": 1.0277,
+      "step": 354
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.020284309819523,
+      "learning_rate": 1.9923073778101743e-05,
+      "loss": 0.8184,
+      "step": 355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7877368746091395,
+      "learning_rate": 1.9922299712986478e-05,
+      "loss": 1.0512,
+      "step": 356
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9511674507474972,
+      "learning_rate": 1.9921521788041273e-05,
+      "loss": 0.9541,
+      "step": 357
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8548873026796064,
+      "learning_rate": 1.9920740003568743e-05,
+      "loss": 1.0032,
+      "step": 358
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.369642361117581,
+      "learning_rate": 1.9919954359873003e-05,
+      "loss": 1.0463,
+      "step": 359
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6853151801856843,
+      "learning_rate": 1.991916485725968e-05,
+      "loss": 0.9586,
+      "step": 360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.121801973394243,
+      "learning_rate": 1.991837149603589e-05,
+      "loss": 1.007,
+      "step": 361
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8694623317506724,
+      "learning_rate": 1.9917574276510256e-05,
+      "loss": 1.0233,
+      "step": 362
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9936957782176836,
+      "learning_rate": 1.99167731989929e-05,
+      "loss": 1.11,
+      "step": 363
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.044933129387799,
+      "learning_rate": 1.9915968263795448e-05,
+      "loss": 1.0585,
+      "step": 364
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9558810730478424,
+      "learning_rate": 1.9915159471231025e-05,
+      "loss": 0.9883,
+      "step": 365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0169845204190078,
+      "learning_rate": 1.9914346821614247e-05,
+      "loss": 1.0129,
+      "step": 366
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0359412772425673,
+      "learning_rate": 1.9913530315261252e-05,
+      "loss": 1.0086,
+      "step": 367
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.018034818878424,
+      "learning_rate": 1.991270995248966e-05,
+      "loss": 1.0717,
+      "step": 368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9196933985110856,
+      "learning_rate": 1.9911885733618594e-05,
+      "loss": 1.0193,
+      "step": 369
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8924726827316862,
+      "learning_rate": 1.9911057658968677e-05,
+      "loss": 1.0684,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.957954173791168,
+      "learning_rate": 1.9910225728862045e-05,
+      "loss": 1.003,
+      "step": 371
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7327565474786015,
+      "learning_rate": 1.9909389943622316e-05,
+      "loss": 1.0367,
+      "step": 372
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1352232620719276,
+      "learning_rate": 1.9908550303574617e-05,
+      "loss": 1.1118,
+      "step": 373
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0787287987087506,
+      "learning_rate": 1.990770680904557e-05,
+      "loss": 1.0894,
+      "step": 374
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3620095751306678,
+      "learning_rate": 1.9906859460363307e-05,
+      "loss": 0.8552,
+      "step": 375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9670341479798041,
+      "learning_rate": 1.9906008257857447e-05,
+      "loss": 0.9371,
+      "step": 376
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.913167874526329,
+      "learning_rate": 1.990515320185911e-05,
+      "loss": 1.0643,
+      "step": 377
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9217382388951545,
+      "learning_rate": 1.9904294292700917e-05,
+      "loss": 1.0264,
+      "step": 378
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9352416915935249,
+      "learning_rate": 1.9903431530716992e-05,
+      "loss": 1.0726,
+      "step": 379
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.032603191233027,
+      "learning_rate": 1.9902564916242952e-05,
+      "loss": 1.0659,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9618668169033733,
+      "learning_rate": 1.9901694449615917e-05,
+      "loss": 1.0239,
+      "step": 381
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.09195688744226,
+      "learning_rate": 1.9900820131174503e-05,
+      "loss": 1.0984,
+      "step": 382
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9234946021972024,
+      "learning_rate": 1.989994196125882e-05,
+      "loss": 1.0506,
+      "step": 383
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0991909496265084,
+      "learning_rate": 1.989905994021049e-05,
+      "loss": 0.8034,
+      "step": 384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.81014889387723,
+      "learning_rate": 1.989817406837262e-05,
+      "loss": 1.0006,
+      "step": 385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.993529462114099,
+      "learning_rate": 1.989728434608981e-05,
+      "loss": 0.8089,
+      "step": 386
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.882727487143848,
+      "learning_rate": 1.9896390773708182e-05,
+      "loss": 1.0693,
+      "step": 387
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7729876511972384,
+      "learning_rate": 1.9895493351575333e-05,
+      "loss": 1.1044,
+      "step": 388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8930447196671079,
+      "learning_rate": 1.9894592080040364e-05,
+      "loss": 1.0936,
+      "step": 389
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7141797791734572,
+      "learning_rate": 1.9893686959453875e-05,
+      "loss": 0.9307,
+      "step": 390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0206681281369794,
+      "learning_rate": 1.9892777990167967e-05,
+      "loss": 1.0052,
+      "step": 391
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9601360739455025,
+      "learning_rate": 1.9891865172536233e-05,
+      "loss": 1.0426,
+      "step": 392
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6936067659027334,
+      "learning_rate": 1.989094850691376e-05,
+      "loss": 0.9326,
+      "step": 393
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.856166304153269,
+      "learning_rate": 1.989002799365714e-05,
+      "loss": 0.9852,
+      "step": 394
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8367907200949016,
+      "learning_rate": 1.9889103633124457e-05,
+      "loss": 0.9592,
+      "step": 395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.007355585475035,
+      "learning_rate": 1.9888175425675288e-05,
+      "loss": 1.0117,
+      "step": 396
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8516073622878386,
+      "learning_rate": 1.988724337167072e-05,
+      "loss": 1.0603,
+      "step": 397
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.216372170317351,
+      "learning_rate": 1.9886307471473323e-05,
+      "loss": 0.7921,
+      "step": 398
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9320279482943123,
+      "learning_rate": 1.9885367725447162e-05,
+      "loss": 1.0564,
+      "step": 399
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8561275651991433,
+      "learning_rate": 1.9884424133957808e-05,
+      "loss": 1.0022,
+      "step": 400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.912677537686615,
+      "learning_rate": 1.9883476697372327e-05,
+      "loss": 0.9387,
+      "step": 401
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0297626339996624,
+      "learning_rate": 1.9882525416059273e-05,
+      "loss": 0.8481,
+      "step": 402
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0247451338678344,
+      "learning_rate": 1.98815702903887e-05,
+      "loss": 0.9894,
+      "step": 403
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9871374845980427,
+      "learning_rate": 1.9880611320732154e-05,
+      "loss": 1.0576,
+      "step": 404
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9930155207385365,
+      "learning_rate": 1.9879648507462687e-05,
+      "loss": 0.9972,
+      "step": 405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0646801640749652,
+      "learning_rate": 1.9878681850954833e-05,
+      "loss": 1.0854,
+      "step": 406
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8818333915154013,
+      "learning_rate": 1.9877711351584628e-05,
+      "loss": 1.0383,
+      "step": 407
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9722438634458774,
+      "learning_rate": 1.98767370097296e-05,
+      "loss": 1.0116,
+      "step": 408
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7066244506847275,
+      "learning_rate": 1.987575882576878e-05,
+      "loss": 1.0032,
+      "step": 409
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.902672174295704,
+      "learning_rate": 1.987477680008268e-05,
+      "loss": 1.1152,
+      "step": 410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7623694577336437,
+      "learning_rate": 1.9873790933053316e-05,
+      "loss": 0.9807,
+      "step": 411
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9894455403137112,
+      "learning_rate": 1.9872801225064196e-05,
+      "loss": 1.0516,
+      "step": 412
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1739518583820248,
+      "learning_rate": 1.987180767650032e-05,
+      "loss": 1.0914,
+      "step": 413
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7886753698199753,
+      "learning_rate": 1.987081028774819e-05,
+      "loss": 0.9163,
+      "step": 414
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1521609517012579,
+      "learning_rate": 1.9869809059195787e-05,
+      "loss": 0.8685,
+      "step": 415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8376642390611246,
+      "learning_rate": 1.9868803991232602e-05,
+      "loss": 0.9614,
+      "step": 416
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8501436743338675,
+      "learning_rate": 1.986779508424961e-05,
+      "loss": 1.0526,
+      "step": 417
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6869946424617572,
+      "learning_rate": 1.9866782338639278e-05,
+      "loss": 1.0524,
+      "step": 418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8153293231608487,
+      "learning_rate": 1.986576575479557e-05,
+      "loss": 1.0178,
+      "step": 419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7274306454495725,
+      "learning_rate": 1.9864745333113948e-05,
+      "loss": 0.9729,
+      "step": 420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.939912973493156,
+      "learning_rate": 1.986372107399136e-05,
+      "loss": 1.0441,
+      "step": 421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7856564272695379,
+      "learning_rate": 1.9862692977826242e-05,
+      "loss": 1.1091,
+      "step": 422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8341112574380691,
+      "learning_rate": 1.9861661045018537e-05,
+      "loss": 0.8888,
+      "step": 423
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8561948420468832,
+      "learning_rate": 1.9860625275969673e-05,
+      "loss": 1.039,
+      "step": 424
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7934335505219023,
+      "learning_rate": 1.9859585671082562e-05,
+      "loss": 1.0336,
+      "step": 425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8360599761712861,
+      "learning_rate": 1.9858542230761622e-05,
+      "loss": 1.0813,
+      "step": 426
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9867816840398522,
+      "learning_rate": 1.985749495541276e-05,
+      "loss": 1.0322,
+      "step": 427
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9480167546900247,
+      "learning_rate": 1.9856443845443364e-05,
+      "loss": 1.0453,
+      "step": 428
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1140835883562086,
+      "learning_rate": 1.9855388901262324e-05,
+      "loss": 1.0198,
+      "step": 429
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.064005705298389,
+      "learning_rate": 1.9854330123280027e-05,
+      "loss": 1.0869,
+      "step": 430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6915389255562758,
+      "learning_rate": 1.985326751190833e-05,
+      "loss": 1.0063,
+      "step": 431
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8715414211034649,
+      "learning_rate": 1.9852201067560607e-05,
+      "loss": 0.9858,
+      "step": 432
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9636600893192735,
+      "learning_rate": 1.9851130790651706e-05,
+      "loss": 1.0528,
+      "step": 433
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.966849427909118,
+      "learning_rate": 1.9850056681597968e-05,
+      "loss": 1.0915,
+      "step": 434
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9070631155194202,
+      "learning_rate": 1.9848978740817234e-05,
+      "loss": 1.0174,
+      "step": 435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.225217117866621,
+      "learning_rate": 1.984789696872882e-05,
+      "loss": 0.8672,
+      "step": 436
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.038760319545761,
+      "learning_rate": 1.9846811365753548e-05,
+      "loss": 1.0013,
+      "step": 437
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.022940013987618,
+      "learning_rate": 1.9845721932313725e-05,
+      "loss": 0.9708,
+      "step": 438
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7647463910730028,
+      "learning_rate": 1.984462866883314e-05,
+      "loss": 0.9222,
+      "step": 439
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.866488318733101,
+      "learning_rate": 1.9843531575737085e-05,
+      "loss": 1.043,
+      "step": 440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8603090234733615,
+      "learning_rate": 1.9842430653452333e-05,
+      "loss": 0.9491,
+      "step": 441
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8875785862623655,
+      "learning_rate": 1.9841325902407144e-05,
+      "loss": 0.9926,
+      "step": 442
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.942175598283091,
+      "learning_rate": 1.9840217323031278e-05,
+      "loss": 0.9455,
+      "step": 443
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.985476015117722,
+      "learning_rate": 1.983910491575598e-05,
+      "loss": 1.0879,
+      "step": 444
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7275782937334652,
+      "learning_rate": 1.9837988681013977e-05,
+      "loss": 1.0377,
+      "step": 445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0112594657447027,
+      "learning_rate": 1.9836868619239498e-05,
+      "loss": 0.9808,
+      "step": 446
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.839058814041238,
+      "learning_rate": 1.9835744730868244e-05,
+      "loss": 0.8538,
+      "step": 447
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.157369191952226,
+      "learning_rate": 1.9834617016337424e-05,
+      "loss": 1.0736,
+      "step": 448
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.15228050930449,
+      "learning_rate": 1.9833485476085715e-05,
+      "loss": 1.019,
+      "step": 449
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.91857747393063,
+      "learning_rate": 1.98323501105533e-05,
+      "loss": 1.0517,
+      "step": 450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.027868590912365,
+      "learning_rate": 1.983121092018184e-05,
+      "loss": 0.793,
+      "step": 451
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.773990990913634,
+      "learning_rate": 1.9830067905414485e-05,
+      "loss": 1.0535,
+      "step": 452
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7448900958278026,
+      "learning_rate": 1.9828921066695876e-05,
+      "loss": 1.0646,
+      "step": 453
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8842052222344852,
+      "learning_rate": 1.982777040447214e-05,
+      "loss": 0.9916,
+      "step": 454
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.965498295082324,
+      "learning_rate": 1.9826615919190886e-05,
+      "loss": 1.0308,
+      "step": 455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8825201063090378,
+      "learning_rate": 1.9825457611301226e-05,
+      "loss": 1.0029,
+      "step": 456
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0328819583947006,
+      "learning_rate": 1.9824295481253735e-05,
+      "loss": 1.07,
+      "step": 457
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0825105072372898,
+      "learning_rate": 1.98231295295005e-05,
+      "loss": 0.7738,
+      "step": 458
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8165626539270863,
+      "learning_rate": 1.9821959756495075e-05,
+      "loss": 1.0831,
+      "step": 459
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9422401630853525,
+      "learning_rate": 1.982078616269251e-05,
+      "loss": 1.0181,
+      "step": 460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7945631288777881,
+      "learning_rate": 1.981960874854934e-05,
+      "loss": 0.946,
+      "step": 461
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.997337758639312,
+      "learning_rate": 1.9818427514523584e-05,
+      "loss": 1.0037,
+      "step": 462
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9193241141993849,
+      "learning_rate": 1.9817242461074757e-05,
+      "loss": 1.0171,
+      "step": 463
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9978405533297976,
+      "learning_rate": 1.981605358866384e-05,
+      "loss": 1.0468,
+      "step": 464
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8804833385106348,
+      "learning_rate": 1.9814860897753317e-05,
+      "loss": 0.916,
+      "step": 465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8028266192668139,
+      "learning_rate": 1.981366438880715e-05,
+      "loss": 1.0217,
+      "step": 466
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7132667749630526,
+      "learning_rate": 1.9812464062290788e-05,
+      "loss": 1.0445,
+      "step": 467
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9219672750165928,
+      "learning_rate": 1.9811259918671168e-05,
+      "loss": 1.031,
+      "step": 468
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8280416256736887,
+      "learning_rate": 1.9810051958416697e-05,
+      "loss": 1.0184,
+      "step": 469
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6689592882194266,
+      "learning_rate": 1.9808840181997292e-05,
+      "loss": 1.0227,
+      "step": 470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.900220242233606,
+      "learning_rate": 1.9807624589884335e-05,
+      "loss": 1.0608,
+      "step": 471
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6929314568440184,
+      "learning_rate": 1.98064051825507e-05,
+      "loss": 0.9899,
+      "step": 472
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.651258459418155,
+      "learning_rate": 1.980518196047074e-05,
+      "loss": 0.9661,
+      "step": 473
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1874289610603563,
+      "learning_rate": 1.9803954924120297e-05,
+      "loss": 0.9393,
+      "step": 474
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8038819153786403,
+      "learning_rate": 1.9802724073976695e-05,
+      "loss": 0.9559,
+      "step": 475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.770721462261662,
+      "learning_rate": 1.9801489410518736e-05,
+      "loss": 0.9643,
+      "step": 476
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7367654491312072,
+      "learning_rate": 1.9800250934226723e-05,
+      "loss": 0.9443,
+      "step": 477
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9591444421711552,
+      "learning_rate": 1.9799008645582424e-05,
+      "loss": 0.9747,
+      "step": 478
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6828464935006877,
+      "learning_rate": 1.979776254506909e-05,
+      "loss": 0.9754,
+      "step": 479
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.908226452531261,
+      "learning_rate": 1.9796512633171476e-05,
+      "loss": 1.1313,
+      "step": 480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.306422799793475,
+      "learning_rate": 1.9795258910375787e-05,
+      "loss": 0.8817,
+      "step": 481
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9053129529215438,
+      "learning_rate": 1.979400137716974e-05,
+      "loss": 1.0261,
+      "step": 482
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9080635923357674,
+      "learning_rate": 1.979274003404252e-05,
+      "loss": 0.9686,
+      "step": 483
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8037616024087626,
+      "learning_rate": 1.9791474881484793e-05,
+      "loss": 1.0324,
+      "step": 484
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8250327442496437,
+      "learning_rate": 1.9790205919988714e-05,
+      "loss": 1.0082,
+      "step": 485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8511505666088204,
+      "learning_rate": 1.9788933150047916e-05,
+      "loss": 1.1118,
+      "step": 486
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8903056078178093,
+      "learning_rate": 1.9787656572157515e-05,
+      "loss": 1.0646,
+      "step": 487
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9195313614337501,
+      "learning_rate": 1.97863761868141e-05,
+      "loss": 1.0412,
+      "step": 488
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9799264468742026,
+      "learning_rate": 1.9785091994515756e-05,
+      "loss": 1.0861,
+      "step": 489
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6755241928671092,
+      "learning_rate": 1.9783803995762033e-05,
+      "loss": 0.9611,
+      "step": 490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.961898488303003,
+      "learning_rate": 1.9782512191053982e-05,
+      "loss": 1.0589,
+      "step": 491
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.775413845673962,
+      "learning_rate": 1.9781216580894108e-05,
+      "loss": 1.0354,
+      "step": 492
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7928396588919673,
+      "learning_rate": 1.977991716578642e-05,
+      "loss": 1.0508,
+      "step": 493
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9412161829079344,
+      "learning_rate": 1.9778613946236395e-05,
+      "loss": 1.0015,
+      "step": 494
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7096885758847509,
+      "learning_rate": 1.9777306922750995e-05,
+      "loss": 0.9801,
+      "step": 495
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8191213564062976,
+      "learning_rate": 1.9775996095838655e-05,
+      "loss": 1.0899,
+      "step": 496
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.964813155629249,
+      "learning_rate": 1.9774681466009295e-05,
+      "loss": 1.053,
+      "step": 497
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8601508656062495,
+      "learning_rate": 1.9773363033774312e-05,
+      "loss": 1.034,
+      "step": 498
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5995716944146932,
+      "learning_rate": 1.977204079964659e-05,
+      "loss": 0.9518,
+      "step": 499
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9695321697574193,
+      "learning_rate": 1.977071476414048e-05,
+      "loss": 0.9849,
+      "step": 500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9579607025450871,
+      "learning_rate": 1.976938492777182e-05,
+      "loss": 1.1121,
+      "step": 501
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8098631026909815,
+      "learning_rate": 1.976805129105792e-05,
+      "loss": 1.0382,
+      "step": 502
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0532679072516262,
+      "learning_rate": 1.9766713854517575e-05,
+      "loss": 1.0442,
+      "step": 503
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3099052871740984,
+      "learning_rate": 1.9765372618671054e-05,
+      "loss": 0.7801,
+      "step": 504
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.810026265200264,
+      "learning_rate": 1.9764027584040106e-05,
+      "loss": 1.0075,
+      "step": 505
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8637296868640794,
+      "learning_rate": 1.976267875114796e-05,
+      "loss": 0.9448,
+      "step": 506
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9013926216308323,
+      "learning_rate": 1.976132612051931e-05,
+      "loss": 0.9531,
+      "step": 507
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0015119413860785,
+      "learning_rate": 1.975996969268035e-05,
+      "loss": 0.8163,
+      "step": 508
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.896148527285097,
+      "learning_rate": 1.975860946815873e-05,
+      "loss": 1.0849,
+      "step": 509
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7615146405930213,
+      "learning_rate": 1.9757245447483584e-05,
+      "loss": 1.03,
+      "step": 510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.738456291803581,
+      "learning_rate": 1.9755877631185526e-05,
+      "loss": 1.0107,
+      "step": 511
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8183122048094087,
+      "learning_rate": 1.9754506019796646e-05,
+      "loss": 0.9735,
+      "step": 512
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7303141348577542,
+      "learning_rate": 1.9753130613850504e-05,
+      "loss": 1.0331,
+      "step": 513
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7089895406313016,
+      "learning_rate": 1.9751751413882147e-05,
+      "loss": 0.9911,
+      "step": 514
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.758803352802379,
+      "learning_rate": 1.9750368420428082e-05,
+      "loss": 1.0327,
+      "step": 515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0002121182689385,
+      "learning_rate": 1.974898163402631e-05,
+      "loss": 1.0434,
+      "step": 516
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8500757476081868,
+      "learning_rate": 1.9747591055216295e-05,
+      "loss": 1.0831,
+      "step": 517
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6790673482114253,
+      "learning_rate": 1.974619668453898e-05,
+      "loss": 1.016,
+      "step": 518
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0175983416616057,
+      "learning_rate": 1.9744798522536786e-05,
+      "loss": 0.9923,
+      "step": 519
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9402106628437552,
+      "learning_rate": 1.97433965697536e-05,
+      "loss": 1.0542,
+      "step": 520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.731992084995377,
+      "learning_rate": 1.9741990826734793e-05,
+      "loss": 0.9576,
+      "step": 521
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.876259428146914,
+      "learning_rate": 1.9740581294027206e-05,
+      "loss": 0.9504,
+      "step": 522
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7744367158474315,
+      "learning_rate": 1.9739167972179157e-05,
+      "loss": 0.9547,
+      "step": 523
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8912326472653933,
+      "learning_rate": 1.9737750861740434e-05,
+      "loss": 1.0988,
+      "step": 524
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7884555668201632,
+      "learning_rate": 1.97363299632623e-05,
+      "loss": 1.0357,
+      "step": 525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8374367575885093,
+      "learning_rate": 1.9734905277297492e-05,
+      "loss": 0.9488,
+      "step": 526
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.900850598970633,
+      "learning_rate": 1.973347680440023e-05,
+      "loss": 1.0481,
+      "step": 527
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9680722101056642,
+      "learning_rate": 1.9732044545126184e-05,
+      "loss": 0.96,
+      "step": 528
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.777121533767296,
+      "learning_rate": 1.973060850003252e-05,
+      "loss": 0.9076,
+      "step": 529
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7640473754466601,
+      "learning_rate": 1.9729168669677865e-05,
+      "loss": 0.9612,
+      "step": 530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4260387482614596,
+      "learning_rate": 1.972772505462232e-05,
+      "loss": 0.8384,
+      "step": 531
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8390165027204632,
+      "learning_rate": 1.9726277655427462e-05,
+      "loss": 0.9875,
+      "step": 532
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9365708781635196,
+      "learning_rate": 1.9724826472656337e-05,
+      "loss": 1.0735,
+      "step": 533
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8022775523259202,
+      "learning_rate": 1.9723371506873456e-05,
+      "loss": 1.0959,
+      "step": 534
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9427266786884612,
+      "learning_rate": 1.972191275864482e-05,
+      "loss": 1.0465,
+      "step": 535
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.771581543103469,
+      "learning_rate": 1.9720450228537883e-05,
+      "loss": 0.9812,
+      "step": 536
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9122283145634829,
+      "learning_rate": 1.971898391712158e-05,
+      "loss": 1.0029,
+      "step": 537
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.886920368138372,
+      "learning_rate": 1.9717513824966314e-05,
+      "loss": 1.0516,
+      "step": 538
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9513242256729235,
+      "learning_rate": 1.971603995264396e-05,
+      "loss": 1.013,
+      "step": 539
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9856337637621357,
+      "learning_rate": 1.9714562300727856e-05,
+      "loss": 1.0649,
+      "step": 540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2056602684650892,
+      "learning_rate": 1.9713080869792826e-05,
+      "loss": 0.8937,
+      "step": 541
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6906940548878524,
+      "learning_rate": 1.971159566041515e-05,
+      "loss": 0.9744,
+      "step": 542
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7004076601163614,
+      "learning_rate": 1.9710106673172583e-05,
+      "loss": 0.9987,
+      "step": 543
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9129829748887999,
+      "learning_rate": 1.970861390864435e-05,
+      "loss": 1.0169,
+      "step": 544
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8276661627334796,
+      "learning_rate": 1.9707117367411145e-05,
+      "loss": 0.9962,
+      "step": 545
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8558234705727246,
+      "learning_rate": 1.9705617050055134e-05,
+      "loss": 1.0134,
+      "step": 546
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7893354867382198,
+      "learning_rate": 1.970411295715994e-05,
+      "loss": 0.9442,
+      "step": 547
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7169369398061605,
+      "learning_rate": 1.9702605089310673e-05,
+      "loss": 0.9937,
+      "step": 548
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9439725120563627,
+      "learning_rate": 1.9701093447093894e-05,
+      "loss": 0.9936,
+      "step": 549
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.950915283838417,
+      "learning_rate": 1.969957803109765e-05,
+      "loss": 1.0015,
+      "step": 550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0489748530296592,
+      "learning_rate": 1.9698058841911436e-05,
+      "loss": 0.83,
+      "step": 551
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9759823574922812,
+      "learning_rate": 1.9696535880126232e-05,
+      "loss": 1.082,
+      "step": 552
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4093698243508084,
+      "learning_rate": 1.9695009146334474e-05,
+      "loss": 0.9818,
+      "step": 553
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.00492709143234,
+      "learning_rate": 1.9693478641130073e-05,
+      "loss": 1.0321,
+      "step": 554
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8905545329775386,
+      "learning_rate": 1.9691944365108403e-05,
+      "loss": 1.034,
+      "step": 555
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8929788722333032,
+      "learning_rate": 1.969040631886631e-05,
+      "loss": 1.0527,
+      "step": 556
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.041184174508756,
+      "learning_rate": 1.9688864503002095e-05,
+      "loss": 0.9661,
+      "step": 557
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7750117336694395,
+      "learning_rate": 1.9687318918115535e-05,
+      "loss": 0.9632,
+      "step": 558
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7559475624542369,
+      "learning_rate": 1.9685769564807877e-05,
+      "loss": 1.0893,
+      "step": 559
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9236744913981834,
+      "learning_rate": 1.968421644368182e-05,
+      "loss": 1.0122,
+      "step": 560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8612263017089825,
+      "learning_rate": 1.9682659555341544e-05,
+      "loss": 0.9891,
+      "step": 561
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.115880621920905,
+      "learning_rate": 1.9681098900392684e-05,
+      "loss": 0.9876,
+      "step": 562
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6405954526171358,
+      "learning_rate": 1.9679534479442342e-05,
+      "loss": 1.0224,
+      "step": 563
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7178695204028689,
+      "learning_rate": 1.967796629309909e-05,
+      "loss": 1.038,
+      "step": 564
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.976139485228613,
+      "learning_rate": 1.9676394341972954e-05,
+      "loss": 0.9714,
+      "step": 565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1613109754932986,
+      "learning_rate": 1.967481862667544e-05,
+      "loss": 0.8309,
+      "step": 566
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0094545187976824,
+      "learning_rate": 1.967323914781951e-05,
+      "loss": 0.9518,
+      "step": 567
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6446531205954558,
+      "learning_rate": 1.9671655906019583e-05,
+      "loss": 0.9518,
+      "step": 568
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.952934449437181,
+      "learning_rate": 1.9670068901891554e-05,
+      "loss": 0.9326,
+      "step": 569
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7498569470948722,
+      "learning_rate": 1.9668478136052776e-05,
+      "loss": 0.9984,
+      "step": 570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8173618152594502,
+      "learning_rate": 1.9666883609122067e-05,
+      "loss": 0.9551,
+      "step": 571
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.725822447328941,
+      "learning_rate": 1.96652853217197e-05,
+      "loss": 0.9573,
+      "step": 572
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9764695399528318,
+      "learning_rate": 1.966368327446743e-05,
+      "loss": 1.0103,
+      "step": 573
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0407768919579912,
+      "learning_rate": 1.966207746798845e-05,
+      "loss": 0.8198,
+      "step": 574
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8351836881526475,
+      "learning_rate": 1.966046790290743e-05,
+      "loss": 0.9422,
+      "step": 575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.878354477217513,
+      "learning_rate": 1.965885457985051e-05,
+      "loss": 0.9482,
+      "step": 576
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.769439544934077,
+      "learning_rate": 1.965723749944527e-05,
+      "loss": 1.0465,
+      "step": 577
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7711281541405706,
+      "learning_rate": 1.9655616662320766e-05,
+      "loss": 0.9846,
+      "step": 578
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9861940156065996,
+      "learning_rate": 1.965399206910752e-05,
+      "loss": 0.9931,
+      "step": 579
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8599882713004092,
+      "learning_rate": 1.9652363720437497e-05,
+      "loss": 1.0225,
+      "step": 580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7794662919301099,
+      "learning_rate": 1.9650731616944138e-05,
+      "loss": 1.0081,
+      "step": 581
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.946428764498687,
+      "learning_rate": 1.9649095759262344e-05,
+      "loss": 1.0229,
+      "step": 582
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7801234791090594,
+      "learning_rate": 1.964745614802847e-05,
+      "loss": 1.0095,
+      "step": 583
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.903605819566719,
+      "learning_rate": 1.9645812783880333e-05,
+      "loss": 1.0145,
+      "step": 584
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8518982894607283,
+      "learning_rate": 1.964416566745721e-05,
+      "loss": 0.9814,
+      "step": 585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7786004408012115,
+      "learning_rate": 1.9642514799399845e-05,
+      "loss": 0.987,
+      "step": 586
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7574802027403031,
+      "learning_rate": 1.9640860180350425e-05,
+      "loss": 1.0617,
+      "step": 587
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9200605061445495,
+      "learning_rate": 1.9639201810952613e-05,
+      "loss": 0.9635,
+      "step": 588
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0216532403405931,
+      "learning_rate": 1.963753969185152e-05,
+      "loss": 0.8741,
+      "step": 589
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9382755355231525,
+      "learning_rate": 1.9635873823693722e-05,
+      "loss": 0.9884,
+      "step": 590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9101916826560947,
+      "learning_rate": 1.9634204207127254e-05,
+      "loss": 0.9068,
+      "step": 591
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8693955360440746,
+      "learning_rate": 1.9632530842801602e-05,
+      "loss": 1.0419,
+      "step": 592
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9209991430775086,
+      "learning_rate": 1.9630853731367715e-05,
+      "loss": 1.0271,
+      "step": 593
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9257189173495664,
+      "learning_rate": 1.9629172873477995e-05,
+      "loss": 0.9809,
+      "step": 594
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6982518669446218,
+      "learning_rate": 1.962748826978631e-05,
+      "loss": 0.9863,
+      "step": 595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8752391378493762,
+      "learning_rate": 1.9625799920947978e-05,
+      "loss": 1.0739,
+      "step": 596
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8645897959307918,
+      "learning_rate": 1.9624107827619782e-05,
+      "loss": 1.0144,
+      "step": 597
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8999177208030884,
+      "learning_rate": 1.9622411990459946e-05,
+      "loss": 0.9504,
+      "step": 598
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8320111952467946,
+      "learning_rate": 1.9620712410128165e-05,
+      "loss": 1.0647,
+      "step": 599
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7823230268165489,
+      "learning_rate": 1.9619009087285587e-05,
+      "loss": 1.0214,
+      "step": 600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8272929991956897,
+      "learning_rate": 1.9617302022594812e-05,
+      "loss": 1.0105,
+      "step": 601
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7834810203897904,
+      "learning_rate": 1.9615591216719896e-05,
+      "loss": 0.9983,
+      "step": 602
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7827465892442746,
+      "learning_rate": 1.9613876670326353e-05,
+      "loss": 0.9358,
+      "step": 603
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7226171294113795,
+      "learning_rate": 1.9612158384081157e-05,
+      "loss": 0.9363,
+      "step": 604
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8762126714815124,
+      "learning_rate": 1.9610436358652723e-05,
+      "loss": 1.0197,
+      "step": 605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6259762940352678,
+      "learning_rate": 1.9608710594710928e-05,
+      "loss": 0.953,
+      "step": 606
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8644620546361652,
+      "learning_rate": 1.9606981092927112e-05,
+      "loss": 0.977,
+      "step": 607
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.017388984418631,
+      "learning_rate": 1.9605247853974055e-05,
+      "loss": 0.9991,
+      "step": 608
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7816860287872824,
+      "learning_rate": 1.9603510878526e-05,
+      "loss": 0.9822,
+      "step": 609
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8504338459761795,
+      "learning_rate": 1.9601770167258633e-05,
+      "loss": 0.9923,
+      "step": 610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7125258470574116,
+      "learning_rate": 1.9600025720849104e-05,
+      "loss": 0.884,
+      "step": 611
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6259903458318319,
+      "learning_rate": 1.9598277539976015e-05,
+      "loss": 0.9647,
+      "step": 612
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7838540755015266,
+      "learning_rate": 1.9596525625319423e-05,
+      "loss": 0.9658,
+      "step": 613
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8651932885634872,
+      "learning_rate": 1.959476997756082e-05,
+      "loss": 1.0521,
+      "step": 614
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8210726541066404,
+      "learning_rate": 1.959301059738317e-05,
+      "loss": 1.0149,
+      "step": 615
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8803627378278291,
+      "learning_rate": 1.959124748547088e-05,
+      "loss": 1.0156,
+      "step": 616
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.689371961045089,
+      "learning_rate": 1.9589480642509817e-05,
+      "loss": 0.9449,
+      "step": 617
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8578288264731058,
+      "learning_rate": 1.958771006918728e-05,
+      "loss": 1.0559,
+      "step": 618
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8988312781209649,
+      "learning_rate": 1.9585935766192042e-05,
+      "loss": 1.0035,
+      "step": 619
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0135141788781032,
+      "learning_rate": 1.9584157734214312e-05,
+      "loss": 1.0137,
+      "step": 620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7656252811681112,
+      "learning_rate": 1.9582375973945756e-05,
+      "loss": 1.0709,
+      "step": 621
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.909752799816405,
+      "learning_rate": 1.958059048607949e-05,
+      "loss": 0.9888,
+      "step": 622
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8634963942707652,
+      "learning_rate": 1.9578801271310076e-05,
+      "loss": 1.0285,
+      "step": 623
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9494642791144832,
+      "learning_rate": 1.957700833033353e-05,
+      "loss": 0.9605,
+      "step": 624
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9523196873729527,
+      "learning_rate": 1.957521166384732e-05,
+      "loss": 1.0018,
+      "step": 625
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8308691823290184,
+      "learning_rate": 1.957341127255035e-05,
+      "loss": 1.0653,
+      "step": 626
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9240937821093393,
+      "learning_rate": 1.957160715714299e-05,
+      "loss": 0.9381,
+      "step": 627
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.081996503557969,
+      "learning_rate": 1.9569799318327044e-05,
+      "loss": 1.0153,
+      "step": 628
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.893200769767438,
+      "learning_rate": 1.9567987756805782e-05,
+      "loss": 1.075,
+      "step": 629
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8079148765894928,
+      "learning_rate": 1.95661724732839e-05,
+      "loss": 1.0268,
+      "step": 630
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7458820650202476,
+      "learning_rate": 1.9564353468467563e-05,
+      "loss": 0.9522,
+      "step": 631
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0121851152786425,
+      "learning_rate": 1.956253074306437e-05,
+      "loss": 0.819,
+      "step": 632
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6652692877580197,
+      "learning_rate": 1.956070429778337e-05,
+      "loss": 0.9953,
+      "step": 633
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8836370773430482,
+      "learning_rate": 1.955887413333506e-05,
+      "loss": 1.0643,
+      "step": 634
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.144791186424118,
+      "learning_rate": 1.9557040250431385e-05,
+      "loss": 0.9291,
+      "step": 635
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.877148613949905,
+      "learning_rate": 1.9555202649785738e-05,
+      "loss": 1.0046,
+      "step": 636
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6827726503366733,
+      "learning_rate": 1.9553361332112953e-05,
+      "loss": 1.0014,
+      "step": 637
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8603802983446722,
+      "learning_rate": 1.9551516298129317e-05,
+      "loss": 1.0337,
+      "step": 638
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8606967093654574,
+      "learning_rate": 1.9549667548552557e-05,
+      "loss": 1.0276,
+      "step": 639
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6143883313354406,
+      "learning_rate": 1.954781508410184e-05,
+      "loss": 0.9762,
+      "step": 640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.105688140457794,
+      "learning_rate": 1.954595890549779e-05,
+      "loss": 0.8358,
+      "step": 641
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9600008805976168,
+      "learning_rate": 1.9544099013462476e-05,
+      "loss": 0.9342,
+      "step": 642
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6889493431551765,
+      "learning_rate": 1.95422354087194e-05,
+      "loss": 0.9675,
+      "step": 643
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.926930317137505,
+      "learning_rate": 1.9540368091993517e-05,
+      "loss": 1.0889,
+      "step": 644
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8607569160599358,
+      "learning_rate": 1.9538497064011223e-05,
+      "loss": 1.0252,
+      "step": 645
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9817051142790953,
+      "learning_rate": 1.9536622325500357e-05,
+      "loss": 1.0542,
+      "step": 646
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9294283067819495,
+      "learning_rate": 1.9534743877190206e-05,
+      "loss": 1.0189,
+      "step": 647
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9571599164663367,
+      "learning_rate": 1.9532861719811493e-05,
+      "loss": 0.9496,
+      "step": 648
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8989290281200815,
+      "learning_rate": 1.9530975854096394e-05,
+      "loss": 0.9949,
+      "step": 649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.763600721815347,
+      "learning_rate": 1.952908628077851e-05,
+      "loss": 1.0325,
+      "step": 650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.835502817269082,
+      "learning_rate": 1.952719300059291e-05,
+      "loss": 1.0837,
+      "step": 651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7290066139290443,
+      "learning_rate": 1.952529601427608e-05,
+      "loss": 1.0156,
+      "step": 652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0770038756774114,
+      "learning_rate": 1.952339532256596e-05,
+      "loss": 1.04,
+      "step": 653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0099306333394615,
+      "learning_rate": 1.952149092620193e-05,
+      "loss": 1.0622,
+      "step": 654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8540761709508473,
+      "learning_rate": 1.9519582825924814e-05,
+      "loss": 0.9837,
+      "step": 655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.859168333571717,
+      "learning_rate": 1.951767102247687e-05,
+      "loss": 1.0781,
+      "step": 656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7094488005212378,
+      "learning_rate": 1.9515755516601804e-05,
+      "loss": 1.0045,
+      "step": 657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1001276595776517,
+      "learning_rate": 1.9513836309044753e-05,
+      "loss": 0.9457,
+      "step": 658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2615406142825654,
+      "learning_rate": 1.9511913400552303e-05,
+      "loss": 0.9446,
+      "step": 659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6148231321901039,
+      "learning_rate": 1.950998679187248e-05,
+      "loss": 0.9691,
+      "step": 660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8359796293712454,
+      "learning_rate": 1.950805648375474e-05,
+      "loss": 0.9425,
+      "step": 661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9178413815541715,
+      "learning_rate": 1.950612247694998e-05,
+      "loss": 1.0232,
+      "step": 662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8760848422434737,
+      "learning_rate": 1.9504184772210548e-05,
+      "loss": 0.9971,
+      "step": 663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.838490731362393,
+      "learning_rate": 1.9502243370290223e-05,
+      "loss": 0.9216,
+      "step": 664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.104752776709877,
+      "learning_rate": 1.950029827194421e-05,
+      "loss": 1.0305,
+      "step": 665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9417604070097458,
+      "learning_rate": 1.949834947792917e-05,
+      "loss": 1.0304,
+      "step": 666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9359657568543271,
+      "learning_rate": 1.9496396989003195e-05,
+      "loss": 0.9901,
+      "step": 667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7177141910396727,
+      "learning_rate": 1.9494440805925815e-05,
+      "loss": 0.9605,
+      "step": 668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.036379850609099,
+      "learning_rate": 1.949248092945799e-05,
+      "loss": 1.053,
+      "step": 669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.703756011764784,
+      "learning_rate": 1.9490517360362133e-05,
+      "loss": 0.9522,
+      "step": 670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.199581483258029,
+      "learning_rate": 1.9488550099402076e-05,
+      "loss": 0.9585,
+      "step": 671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.032605206926022,
+      "learning_rate": 1.9486579147343096e-05,
+      "loss": 0.9151,
+      "step": 672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.946296864168914,
+      "learning_rate": 1.9484604504951902e-05,
+      "loss": 1.0469,
+      "step": 673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1881602563467488,
+      "learning_rate": 1.9482626172996647e-05,
+      "loss": 0.8594,
+      "step": 674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0554534102703994,
+      "learning_rate": 1.948064415224691e-05,
+      "loss": 1.0739,
+      "step": 675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8898600878003078,
+      "learning_rate": 1.9478658443473704e-05,
+      "loss": 0.9474,
+      "step": 676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0284710118679543,
+      "learning_rate": 1.9476669047449486e-05,
+      "loss": 1.0512,
+      "step": 677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.089426155911072,
+      "learning_rate": 1.9474675964948143e-05,
+      "loss": 0.9444,
+      "step": 678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8092191682970744,
+      "learning_rate": 1.947267919674499e-05,
+      "loss": 1.0226,
+      "step": 679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8643441782112329,
+      "learning_rate": 1.9470678743616783e-05,
+      "loss": 0.9475,
+      "step": 680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9375935666035613,
+      "learning_rate": 1.9468674606341716e-05,
+      "loss": 0.8696,
+      "step": 681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8006139001238568,
+      "learning_rate": 1.9466666785699398e-05,
+      "loss": 0.9545,
+      "step": 682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8348949870956015,
+      "learning_rate": 1.9464655282470892e-05,
+      "loss": 0.9875,
+      "step": 683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8322027782527568,
+      "learning_rate": 1.9462640097438677e-05,
+      "loss": 1.0545,
+      "step": 684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7908258105119117,
+      "learning_rate": 1.946062123138668e-05,
+      "loss": 1.0089,
+      "step": 685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8511485087794988,
+      "learning_rate": 1.945859868510024e-05,
+      "loss": 1.0128,
+      "step": 686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9391361734611727,
+      "learning_rate": 1.9456572459366146e-05,
+      "loss": 1.019,
+      "step": 687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.915612769996821,
+      "learning_rate": 1.9454542554972612e-05,
+      "loss": 1.0427,
+      "step": 688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9243534902148103,
+      "learning_rate": 1.9452508972709278e-05,
+      "loss": 1.1303,
+      "step": 689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.779168811435592,
+      "learning_rate": 1.945047171336722e-05,
+      "loss": 0.9336,
+      "step": 690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.689393608932385,
+      "learning_rate": 1.9448430777738947e-05,
+      "loss": 0.9762,
+      "step": 691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4158815186411915,
+      "learning_rate": 1.9446386166618386e-05,
+      "loss": 0.8338,
+      "step": 692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1984535894354003,
+      "learning_rate": 1.9444337880800913e-05,
+      "loss": 0.8221,
+      "step": 693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.924775577324311,
+      "learning_rate": 1.9442285921083313e-05,
+      "loss": 1.047,
+      "step": 694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9009485214503896,
+      "learning_rate": 1.9440230288263814e-05,
+      "loss": 0.9309,
+      "step": 695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.830574660645173,
+      "learning_rate": 1.943817098314207e-05,
+      "loss": 1.0259,
+      "step": 696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9080730648816597,
+      "learning_rate": 1.9436108006519156e-05,
+      "loss": 1.0596,
+      "step": 697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9040307023402534,
+      "learning_rate": 1.9434041359197592e-05,
+      "loss": 1.078,
+      "step": 698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8942275513868263,
+      "learning_rate": 1.9431971041981305e-05,
+      "loss": 0.9911,
+      "step": 699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.737822272823064,
+      "learning_rate": 1.9429897055675666e-05,
+      "loss": 0.9911,
+      "step": 700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.981786926820268,
+      "learning_rate": 1.9427819401087467e-05,
+      "loss": 1.0077,
+      "step": 701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.712885255090491,
+      "learning_rate": 1.942573807902492e-05,
+      "loss": 1.0302,
+      "step": 702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7904410396803714,
+      "learning_rate": 1.9423653090297683e-05,
+      "loss": 0.9507,
+      "step": 703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7625736409748458,
+      "learning_rate": 1.942156443571682e-05,
+      "loss": 0.8929,
+      "step": 704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8393439702812102,
+      "learning_rate": 1.941947211609483e-05,
+      "loss": 0.9576,
+      "step": 705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7769401837759962,
+      "learning_rate": 1.9417376132245637e-05,
+      "loss": 0.8701,
+      "step": 706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0135945816767893,
+      "learning_rate": 1.9415276484984592e-05,
+      "loss": 1.1089,
+      "step": 707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.855068602331629,
+      "learning_rate": 1.9413173175128472e-05,
+      "loss": 1.0601,
+      "step": 708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8203057295452734,
+      "learning_rate": 1.941106620349547e-05,
+      "loss": 0.9689,
+      "step": 709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7663882308240513,
+      "learning_rate": 1.9408955570905218e-05,
+      "loss": 0.9558,
+      "step": 710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8611601011619379,
+      "learning_rate": 1.9406841278178753e-05,
+      "loss": 0.9238,
+      "step": 711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8807804455185642,
+      "learning_rate": 1.9404723326138554e-05,
+      "loss": 1.0513,
+      "step": 712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.714676720477941,
+      "learning_rate": 1.9402601715608517e-05,
+      "loss": 0.9372,
+      "step": 713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7184448643250585,
+      "learning_rate": 1.9400476447413952e-05,
+      "loss": 0.9251,
+      "step": 714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6550433855691762,
+      "learning_rate": 1.9398347522381613e-05,
+      "loss": 0.9258,
+      "step": 715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8053873915924739,
+      "learning_rate": 1.9396214941339655e-05,
+      "loss": 0.9831,
+      "step": 716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0288974259498747,
+      "learning_rate": 1.9394078705117664e-05,
+      "loss": 0.9911,
+      "step": 717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9191762256475713,
+      "learning_rate": 1.9391938814546652e-05,
+      "loss": 1.0681,
+      "step": 718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.301702597720425,
+      "learning_rate": 1.9389795270459048e-05,
+      "loss": 0.8984,
+      "step": 719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0889296594012627,
+      "learning_rate": 1.9387648073688694e-05,
+      "loss": 0.9859,
+      "step": 720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.001986355960943,
+      "learning_rate": 1.9385497225070872e-05,
+      "loss": 0.9603,
+      "step": 721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1124449652860362,
+      "learning_rate": 1.9383342725442272e-05,
+      "loss": 0.8816,
+      "step": 722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8718120785399868,
+      "learning_rate": 1.9381184575641e-05,
+      "loss": 1.0967,
+      "step": 723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8895059613659613,
+      "learning_rate": 1.9379022776506594e-05,
+      "loss": 0.9189,
+      "step": 724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.777704327938484,
+      "learning_rate": 1.9376857328880006e-05,
+      "loss": 0.9812,
+      "step": 725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.854855708986783,
+      "learning_rate": 1.937468823360361e-05,
+      "loss": 0.8649,
+      "step": 726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.005680578801675,
+      "learning_rate": 1.9372515491521187e-05,
+      "loss": 0.9662,
+      "step": 727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9327359464902558,
+      "learning_rate": 1.937033910347795e-05,
+      "loss": 0.9925,
+      "step": 728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.683507021971863,
+      "learning_rate": 1.9368159070320533e-05,
+      "loss": 0.9434,
+      "step": 729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7822790345322754,
+      "learning_rate": 1.936597539289697e-05,
+      "loss": 0.9535,
+      "step": 730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9786210517133977,
+      "learning_rate": 1.936378807205673e-05,
+      "loss": 1.0063,
+      "step": 731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.809960131490093,
+      "learning_rate": 1.9361597108650694e-05,
+      "loss": 0.9839,
+      "step": 732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7264017379928744,
+      "learning_rate": 1.9359402503531155e-05,
+      "loss": 0.9829,
+      "step": 733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7911745293395993,
+      "learning_rate": 1.935720425755183e-05,
+      "loss": 0.9978,
+      "step": 734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6328333129995931,
+      "learning_rate": 1.9355002371567847e-05,
+      "loss": 1.0279,
+      "step": 735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7316697685940783,
+      "learning_rate": 1.935279684643575e-05,
+      "loss": 1.0043,
+      "step": 736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9938038731345402,
+      "learning_rate": 1.9350587683013503e-05,
+      "loss": 0.8667,
+      "step": 737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9800640945049084,
+      "learning_rate": 1.9348374882160486e-05,
+      "loss": 0.978,
+      "step": 738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7830105639149385,
+      "learning_rate": 1.9346158444737488e-05,
+      "loss": 1.0429,
+      "step": 739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6266067420551416,
+      "learning_rate": 1.9343938371606714e-05,
+      "loss": 0.9442,
+      "step": 740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8841717239719238,
+      "learning_rate": 1.9341714663631784e-05,
+      "loss": 0.9846,
+      "step": 741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8262610561252055,
+      "learning_rate": 1.9339487321677737e-05,
+      "loss": 0.9833,
+      "step": 742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9278221499894135,
+      "learning_rate": 1.933725634661102e-05,
+      "loss": 1.0236,
+      "step": 743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7380423550454758,
+      "learning_rate": 1.9335021739299492e-05,
+      "loss": 0.9503,
+      "step": 744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8316722411184136,
+      "learning_rate": 1.9332783500612432e-05,
+      "loss": 0.956,
+      "step": 745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.825380458769888,
+      "learning_rate": 1.9330541631420524e-05,
+      "loss": 0.9818,
+      "step": 746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7874823816839436,
+      "learning_rate": 1.9328296132595867e-05,
+      "loss": 1.0205,
+      "step": 747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8444671130725447,
+      "learning_rate": 1.9326047005011975e-05,
+      "loss": 1.0453,
+      "step": 748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8659657787057022,
+      "learning_rate": 1.9323794249543773e-05,
+      "loss": 1.0725,
+      "step": 749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8288990003726173,
+      "learning_rate": 1.932153786706759e-05,
+      "loss": 0.9937,
+      "step": 750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7799035925095108,
+      "learning_rate": 1.9319277858461177e-05,
+      "loss": 0.9965,
+      "step": 751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8951526059685713,
+      "learning_rate": 1.9317014224603684e-05,
+      "loss": 0.9969,
+      "step": 752
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5955111075050987,
+      "learning_rate": 1.9314746966375684e-05,
+      "loss": 0.96,
+      "step": 753
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7281513688212402,
+      "learning_rate": 1.931247608465915e-05,
+      "loss": 0.9746,
+      "step": 754
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8365399226236758,
+      "learning_rate": 1.9310201580337464e-05,
+      "loss": 0.9739,
+      "step": 755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9326515863251243,
+      "learning_rate": 1.9307923454295426e-05,
+      "loss": 0.926,
+      "step": 756
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.015056594238777,
+      "learning_rate": 1.9305641707419236e-05,
+      "loss": 1.0162,
+      "step": 757
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.961385512133112,
+      "learning_rate": 1.930335634059651e-05,
+      "loss": 0.9684,
+      "step": 758
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9818196076238774,
+      "learning_rate": 1.9301067354716267e-05,
+      "loss": 0.9548,
+      "step": 759
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8706739745906402,
+      "learning_rate": 1.9298774750668934e-05,
+      "loss": 0.9676,
+      "step": 760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9191774717704708,
+      "learning_rate": 1.9296478529346346e-05,
+      "loss": 0.9316,
+      "step": 761
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8021842972179656,
+      "learning_rate": 1.929417869164175e-05,
+      "loss": 0.9307,
+      "step": 762
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.84774142502998,
+      "learning_rate": 1.9291875238449792e-05,
+      "loss": 0.992,
+      "step": 763
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8788744718102077,
+      "learning_rate": 1.928956817066653e-05,
+      "loss": 1.0372,
+      "step": 764
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7378103364136064,
+      "learning_rate": 1.9287257489189425e-05,
+      "loss": 1.0696,
+      "step": 765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8103327130560956,
+      "learning_rate": 1.928494319491734e-05,
+      "loss": 1.0201,
+      "step": 766
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5930183441602843,
+      "learning_rate": 1.9282625288750566e-05,
+      "loss": 0.979,
+      "step": 767
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6434163897255472,
+      "learning_rate": 1.928030377159076e-05,
+      "loss": 0.9791,
+      "step": 768
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.737384048941076,
+      "learning_rate": 1.9277978644341016e-05,
+      "loss": 0.9885,
+      "step": 769
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8121806941890704,
+      "learning_rate": 1.9275649907905823e-05,
+      "loss": 0.9851,
+      "step": 770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7338799831111298,
+      "learning_rate": 1.927331756319107e-05,
+      "loss": 0.9954,
+      "step": 771
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7818987566131022,
+      "learning_rate": 1.9270981611104046e-05,
+      "loss": 0.9433,
+      "step": 772
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.826309255596021,
+      "learning_rate": 1.926864205255346e-05,
+      "loss": 0.9821,
+      "step": 773
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7278967630268074,
+      "learning_rate": 1.9266298888449406e-05,
+      "loss": 1.0026,
+      "step": 774
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5739618188443811,
+      "learning_rate": 1.926395211970339e-05,
+      "loss": 0.9957,
+      "step": 775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8819173044281157,
+      "learning_rate": 1.9261601747228317e-05,
+      "loss": 0.9541,
+      "step": 776
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8705427772081336,
+      "learning_rate": 1.92592477719385e-05,
+      "loss": 1.0077,
+      "step": 777
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9871131843932932,
+      "learning_rate": 1.9256890194749644e-05,
+      "loss": 0.9522,
+      "step": 778
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8403706596418095,
+      "learning_rate": 1.9254529016578857e-05,
+      "loss": 0.9666,
+      "step": 779
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8088524815463622,
+      "learning_rate": 1.9252164238344656e-05,
+      "loss": 0.9694,
+      "step": 780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.780230596936912,
+      "learning_rate": 1.9249795860966954e-05,
+      "loss": 0.9108,
+      "step": 781
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6808312878957696,
+      "learning_rate": 1.9247423885367057e-05,
+      "loss": 0.9323,
+      "step": 782
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.955058817701851,
+      "learning_rate": 1.924504831246768e-05,
+      "loss": 0.8862,
+      "step": 783
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.91788916334588,
+      "learning_rate": 1.9242669143192932e-05,
+      "loss": 1.0931,
+      "step": 784
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6140692233162057,
+      "learning_rate": 1.924028637846833e-05,
+      "loss": 1.0051,
+      "step": 785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7693300696624477,
+      "learning_rate": 1.923790001922077e-05,
+      "loss": 0.9655,
+      "step": 786
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1615050863585519,
+      "learning_rate": 1.9235510066378574e-05,
+      "loss": 0.846,
+      "step": 787
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.265421298145991,
+      "learning_rate": 1.9233116520871436e-05,
+      "loss": 1.0789,
+      "step": 788
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8266533960387925,
+      "learning_rate": 1.9230719383630462e-05,
+      "loss": 0.9565,
+      "step": 789
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0644204564231006,
+      "learning_rate": 1.9228318655588152e-05,
+      "loss": 0.8786,
+      "step": 790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7707224048541397,
+      "learning_rate": 1.9225914337678405e-05,
+      "loss": 1.0365,
+      "step": 791
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7818146587268162,
+      "learning_rate": 1.922350643083651e-05,
+      "loss": 1.0622,
+      "step": 792
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7170365549960838,
+      "learning_rate": 1.922109493599916e-05,
+      "loss": 0.9927,
+      "step": 793
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7493959465018996,
+      "learning_rate": 1.9218679854104438e-05,
+      "loss": 1.06,
+      "step": 794
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1529617574110014,
+      "learning_rate": 1.921626118609182e-05,
+      "loss": 0.9288,
+      "step": 795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7577955120234203,
+      "learning_rate": 1.921383893290219e-05,
+      "loss": 1.0532,
+      "step": 796
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7323690784969576,
+      "learning_rate": 1.9211413095477807e-05,
+      "loss": 1.0039,
+      "step": 797
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9850326315704374,
+      "learning_rate": 1.920898367476235e-05,
+      "loss": 1.0009,
+      "step": 798
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8682185945447385,
+      "learning_rate": 1.9206550671700864e-05,
+      "loss": 1.0061,
+      "step": 799
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6563113113795207,
+      "learning_rate": 1.9204114087239806e-05,
+      "loss": 1.0119,
+      "step": 800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7729390339817501,
+      "learning_rate": 1.9201673922327023e-05,
+      "loss": 0.9548,
+      "step": 801
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.657182451627104,
+      "learning_rate": 1.9199230177911752e-05,
+      "loss": 0.8926,
+      "step": 802
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6226342644622707,
+      "learning_rate": 1.9196782854944618e-05,
+      "loss": 0.9042,
+      "step": 803
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9230366270285837,
+      "learning_rate": 1.9194331954377648e-05,
+      "loss": 1.0576,
+      "step": 804
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.763821194550213,
+      "learning_rate": 1.9191877477164253e-05,
+      "loss": 1.0191,
+      "step": 805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7987127469409767,
+      "learning_rate": 1.9189419424259245e-05,
+      "loss": 1.0144,
+      "step": 806
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7948736840748094,
+      "learning_rate": 1.9186957796618814e-05,
+      "loss": 0.9722,
+      "step": 807
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.779003732422087,
+      "learning_rate": 1.9184492595200545e-05,
+      "loss": 0.9053,
+      "step": 808
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6516389308973194,
+      "learning_rate": 1.918202382096342e-05,
+      "loss": 0.9249,
+      "step": 809
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8681348870126282,
+      "learning_rate": 1.9179551474867807e-05,
+      "loss": 0.9596,
+      "step": 810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7528583814526333,
+      "learning_rate": 1.917707555787546e-05,
+      "loss": 0.9725,
+      "step": 811
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7048418240162433,
+      "learning_rate": 1.9174596070949523e-05,
+      "loss": 0.9909,
+      "step": 812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8355309142079463,
+      "learning_rate": 1.917211301505453e-05,
+      "loss": 0.943,
+      "step": 813
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8886032938082356,
+      "learning_rate": 1.9169626391156408e-05,
+      "loss": 0.823,
+      "step": 814
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7195575937472773,
+      "learning_rate": 1.9167136200222468e-05,
+      "loss": 0.9589,
+      "step": 815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0615609986307115,
+      "learning_rate": 1.91646424432214e-05,
+      "loss": 0.8354,
+      "step": 816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0408751553473148,
+      "learning_rate": 1.9162145121123293e-05,
+      "loss": 0.7966,
+      "step": 817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.923859833738369,
+      "learning_rate": 1.9159644234899627e-05,
+      "loss": 0.9791,
+      "step": 818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8873924529759574,
+      "learning_rate": 1.9157139785523248e-05,
+      "loss": 0.971,
+      "step": 819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8739154137357001,
+      "learning_rate": 1.9154631773968413e-05,
+      "loss": 1.0441,
+      "step": 820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6808152510916812,
+      "learning_rate": 1.9152120201210743e-05,
+      "loss": 0.9293,
+      "step": 821
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8039346767736815,
+      "learning_rate": 1.9149605068227258e-05,
+      "loss": 0.9581,
+      "step": 822
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5483786799136465,
+      "learning_rate": 1.914708637599636e-05,
+      "loss": 0.9391,
+      "step": 823
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8005297352539191,
+      "learning_rate": 1.914456412549783e-05,
+      "loss": 1.0425,
+      "step": 824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8141615019113242,
+      "learning_rate": 1.914203831771284e-05,
+      "loss": 1.0051,
+      "step": 825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9259089305390866,
+      "learning_rate": 1.9139508953623943e-05,
+      "loss": 0.9461,
+      "step": 826
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7587431572845804,
+      "learning_rate": 1.913697603421508e-05,
+      "loss": 0.9512,
+      "step": 827
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.852943275798576,
+      "learning_rate": 1.9134439560471565e-05,
+      "loss": 1.0068,
+      "step": 828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7545460189572282,
+      "learning_rate": 1.91318995333801e-05,
+      "loss": 0.8822,
+      "step": 829
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6422336075005475,
+      "learning_rate": 1.9129355953928773e-05,
+      "loss": 0.9265,
+      "step": 830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.748777338431876,
+      "learning_rate": 1.912680882310705e-05,
+      "loss": 0.993,
+      "step": 831
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.830935302914684,
+      "learning_rate": 1.9124258141905782e-05,
+      "loss": 1.0304,
+      "step": 832
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9331170987373478,
+      "learning_rate": 1.9121703911317193e-05,
+      "loss": 1.0465,
+      "step": 833
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8736094312043001,
+      "learning_rate": 1.9119146132334895e-05,
+      "loss": 1.0146,
+      "step": 834
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9298501870928926,
+      "learning_rate": 1.9116584805953878e-05,
+      "loss": 1.0229,
+      "step": 835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9194108852613612,
+      "learning_rate": 1.911401993317051e-05,
+      "loss": 0.9494,
+      "step": 836
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8830780653577976,
+      "learning_rate": 1.9111451514982548e-05,
+      "loss": 0.9373,
+      "step": 837
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.985865585368822,
+      "learning_rate": 1.9108879552389113e-05,
+      "loss": 1.0104,
+      "step": 838
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7080896278058746,
+      "learning_rate": 1.9106304046390715e-05,
+      "loss": 0.9323,
+      "step": 839
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8985651012917797,
+      "learning_rate": 1.9103724997989245e-05,
+      "loss": 1.0679,
+      "step": 840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7121570918030733,
+      "learning_rate": 1.910114240818796e-05,
+      "loss": 0.9841,
+      "step": 841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0731678178830184,
+      "learning_rate": 1.909855627799151e-05,
+      "loss": 0.9212,
+      "step": 842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7464822420809258,
+      "learning_rate": 1.9095966608405906e-05,
+      "loss": 0.9496,
+      "step": 843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7599055544830364,
+      "learning_rate": 1.9093373400438545e-05,
+      "loss": 1.0182,
+      "step": 844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7840828264495678,
+      "learning_rate": 1.9090776655098207e-05,
+      "loss": 0.9469,
+      "step": 845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7101535311598335,
+      "learning_rate": 1.908817637339503e-05,
+      "loss": 1.021,
+      "step": 846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0046908152956093,
+      "learning_rate": 1.9085572556340546e-05,
+      "loss": 1.024,
+      "step": 847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9508062771279278,
+      "learning_rate": 1.9082965204947652e-05,
+      "loss": 1.0384,
+      "step": 848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7002131016179325,
+      "learning_rate": 1.9080354320230623e-05,
+      "loss": 0.9058,
+      "step": 849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8986950731114445,
+      "learning_rate": 1.9077739903205105e-05,
+      "loss": 0.9542,
+      "step": 850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9658140939639486,
+      "learning_rate": 1.907512195488812e-05,
+      "loss": 1.0349,
+      "step": 851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7534225234501792,
+      "learning_rate": 1.907250047629807e-05,
+      "loss": 1.049,
+      "step": 852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8776574258337895,
+      "learning_rate": 1.906987546845472e-05,
+      "loss": 1.0251,
+      "step": 853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0141660646876545,
+      "learning_rate": 1.9067246932379216e-05,
+      "loss": 0.8872,
+      "step": 854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9005996511296253,
+      "learning_rate": 1.9064614869094068e-05,
+      "loss": 1.0189,
+      "step": 855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.827293502904554,
+      "learning_rate": 1.906197927962317e-05,
+      "loss": 1.0121,
+      "step": 856
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8512193850780998,
+      "learning_rate": 1.9059340164991778e-05,
+      "loss": 0.9846,
+      "step": 857
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.85530405206384,
+      "learning_rate": 1.905669752622652e-05,
+      "loss": 0.9885,
+      "step": 858
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8023628490389225,
+      "learning_rate": 1.90540513643554e-05,
+      "loss": 0.9288,
+      "step": 859
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8348908046781105,
+      "learning_rate": 1.9051401680407792e-05,
+      "loss": 0.9431,
+      "step": 860
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.813007379240663,
+      "learning_rate": 1.904874847541443e-05,
+      "loss": 0.8987,
+      "step": 861
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7113812607280787,
+      "learning_rate": 1.9046091750407436e-05,
+      "loss": 0.9095,
+      "step": 862
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7241574981423409,
+      "learning_rate": 1.9043431506420284e-05,
+      "loss": 0.9743,
+      "step": 863
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7372927683702382,
+      "learning_rate": 1.904076774448782e-05,
+      "loss": 1.056,
+      "step": 864
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7784693024412865,
+      "learning_rate": 1.9038100465646273e-05,
+      "loss": 0.9043,
+      "step": 865
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7701752149441456,
+      "learning_rate": 1.9035429670933222e-05,
+      "loss": 0.9009,
+      "step": 866
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7809720168773098,
+      "learning_rate": 1.9032755361387622e-05,
+      "loss": 0.883,
+      "step": 867
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8335207452717401,
+      "learning_rate": 1.9030077538049795e-05,
+      "loss": 1.0075,
+      "step": 868
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6749272331917675,
+      "learning_rate": 1.902739620196143e-05,
+      "loss": 0.9947,
+      "step": 869
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8258296233403815,
+      "learning_rate": 1.9024711354165578e-05,
+      "loss": 1.0219,
+      "step": 870
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.790339593669149,
+      "learning_rate": 1.902202299570667e-05,
+      "loss": 1.0081,
+      "step": 871
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.986875707763345,
+      "learning_rate": 1.9019331127630478e-05,
+      "loss": 0.8067,
+      "step": 872
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7072137932045357,
+      "learning_rate": 1.9016635750984164e-05,
+      "loss": 0.978,
+      "step": 873
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.955822837840171,
+      "learning_rate": 1.901393686681624e-05,
+      "loss": 1.0216,
+      "step": 874
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7109791916814112,
+      "learning_rate": 1.901123447617659e-05,
+      "loss": 0.9978,
+      "step": 875
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.942184302781339,
+      "learning_rate": 1.900852858011646e-05,
+      "loss": 0.938,
+      "step": 876
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8123521280916648,
+      "learning_rate": 1.900581917968845e-05,
+      "loss": 0.8991,
+      "step": 877
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.864171902401711,
+      "learning_rate": 1.9003106275946543e-05,
+      "loss": 1.0233,
+      "step": 878
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9540648183740126,
+      "learning_rate": 1.9000389869946064e-05,
+      "loss": 0.9301,
+      "step": 879
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8317037732691612,
+      "learning_rate": 1.899766996274372e-05,
+      "loss": 0.9483,
+      "step": 880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8536064978681959,
+      "learning_rate": 1.899494655539756e-05,
+      "loss": 0.9645,
+      "step": 881
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8261208418580592,
+      "learning_rate": 1.8992219648967014e-05,
+      "loss": 1.0482,
+      "step": 882
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7672399474058402,
+      "learning_rate": 1.8989489244512857e-05,
+      "loss": 1.0267,
+      "step": 883
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8322307949283905,
+      "learning_rate": 1.8986755343097237e-05,
+      "loss": 0.9194,
+      "step": 884
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8153099942855861,
+      "learning_rate": 1.898401794578365e-05,
+      "loss": 1.0861,
+      "step": 885
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7613756418254443,
+      "learning_rate": 1.8981277053636963e-05,
+      "loss": 0.9368,
+      "step": 886
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0752988911654553,
+      "learning_rate": 1.89785326677234e-05,
+      "loss": 0.8455,
+      "step": 887
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8765651804005048,
+      "learning_rate": 1.897578478911054e-05,
+      "loss": 1.0165,
+      "step": 888
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1699584122745206,
+      "learning_rate": 1.897303341886732e-05,
+      "loss": 1.0393,
+      "step": 889
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8626305323030707,
+      "learning_rate": 1.8970278558064046e-05,
+      "loss": 1.0442,
+      "step": 890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7495155379790706,
+      "learning_rate": 1.8967520207772366e-05,
+      "loss": 0.863,
+      "step": 891
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8889493560594923,
+      "learning_rate": 1.8964758369065303e-05,
+      "loss": 0.9569,
+      "step": 892
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8721197928210478,
+      "learning_rate": 1.896199304301722e-05,
+      "loss": 1.0279,
+      "step": 893
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7372893971703123,
+      "learning_rate": 1.8959224230703845e-05,
+      "loss": 0.9051,
+      "step": 894
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.835913448549947,
+      "learning_rate": 1.8956451933202266e-05,
+      "loss": 1.0126,
+      "step": 895
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.94447466625549,
+      "learning_rate": 1.8953676151590917e-05,
+      "loss": 0.9614,
+      "step": 896
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8647931470299668,
+      "learning_rate": 1.8950896886949596e-05,
+      "loss": 0.9441,
+      "step": 897
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7531595622080682,
+      "learning_rate": 1.8948114140359452e-05,
+      "loss": 0.9713,
+      "step": 898
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8481014589856102,
+      "learning_rate": 1.8945327912902986e-05,
+      "loss": 1.0405,
+      "step": 899
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7292997897585691,
+      "learning_rate": 1.894253820566406e-05,
+      "loss": 1.0625,
+      "step": 900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6384765682355549,
+      "learning_rate": 1.8939745019727886e-05,
+      "loss": 1.0307,
+      "step": 901
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8530743024676384,
+      "learning_rate": 1.8936948356181023e-05,
+      "loss": 1.0021,
+      "step": 902
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8293080115673754,
+      "learning_rate": 1.8934148216111393e-05,
+      "loss": 0.9617,
+      "step": 903
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8287085115544068,
+      "learning_rate": 1.8931344600608267e-05,
+      "loss": 0.9547,
+      "step": 904
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8260793779226265,
+      "learning_rate": 1.8928537510762264e-05,
+      "loss": 0.9237,
+      "step": 905
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.804849925465654,
+      "learning_rate": 1.892572694766536e-05,
+      "loss": 0.9346,
+      "step": 906
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0809192650896262,
+      "learning_rate": 1.892291291241088e-05,
+      "loss": 0.8212,
+      "step": 907
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9550464795196383,
+      "learning_rate": 1.89200954060935e-05,
+      "loss": 0.9339,
+      "step": 908
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7146518355369345,
+      "learning_rate": 1.891727442980924e-05,
+      "loss": 0.9011,
+      "step": 909
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7469460997345092,
+      "learning_rate": 1.8914449984655486e-05,
+      "loss": 0.9211,
+      "step": 910
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7022610926551167,
+      "learning_rate": 1.8911622071730952e-05,
+      "loss": 1.027,
+      "step": 911
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.572717626891321,
+      "learning_rate": 1.8908790692135714e-05,
+      "loss": 0.9902,
+      "step": 912
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8092656323972338,
+      "learning_rate": 1.8905955846971204e-05,
+      "loss": 0.9172,
+      "step": 913
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8054865926973673,
+      "learning_rate": 1.890311753734018e-05,
+      "loss": 0.9141,
+      "step": 914
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7668034728958322,
+      "learning_rate": 1.890027576434677e-05,
+      "loss": 0.9733,
+      "step": 915
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8410553011586994,
+      "learning_rate": 1.8897430529096432e-05,
+      "loss": 0.9062,
+      "step": 916
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7619174197954015,
+      "learning_rate": 1.889458183269598e-05,
+      "loss": 1.0699,
+      "step": 917
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9410401199076108,
+      "learning_rate": 1.8891729676253582e-05,
+      "loss": 0.9211,
+      "step": 918
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.75184611259881,
+      "learning_rate": 1.8888874060878728e-05,
+      "loss": 0.9733,
+      "step": 919
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7207475430699342,
+      "learning_rate": 1.888601498768228e-05,
+      "loss": 0.9806,
+      "step": 920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7808756246370172,
+      "learning_rate": 1.8883152457776427e-05,
+      "loss": 0.9573,
+      "step": 921
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9561095184863824,
+      "learning_rate": 1.888028647227471e-05,
+      "loss": 0.9821,
+      "step": 922
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7773230766942696,
+      "learning_rate": 1.887741703229202e-05,
+      "loss": 0.9073,
+      "step": 923
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.129071471662999,
+      "learning_rate": 1.8874544138944575e-05,
+      "loss": 0.9619,
+      "step": 924
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8910190514565346,
+      "learning_rate": 1.887166779334995e-05,
+      "loss": 1.0478,
+      "step": 925
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5677215276011822,
+      "learning_rate": 1.8868787996627062e-05,
+      "loss": 0.9167,
+      "step": 926
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7479651107242147,
+      "learning_rate": 1.886590474989617e-05,
+      "loss": 0.895,
+      "step": 927
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9068821372740623,
+      "learning_rate": 1.886301805427887e-05,
+      "loss": 0.9573,
+      "step": 928
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8835897087197837,
+      "learning_rate": 1.8860127910898096e-05,
+      "loss": 1.0145,
+      "step": 929
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.764510818214623,
+      "learning_rate": 1.8857234320878143e-05,
+      "loss": 1.0005,
+      "step": 930
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6764719832853157,
+      "learning_rate": 1.8854337285344622e-05,
+      "loss": 0.9994,
+      "step": 931
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.680925662948634,
+      "learning_rate": 1.8851436805424502e-05,
+      "loss": 0.8666,
+      "step": 932
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8640272893376575,
+      "learning_rate": 1.8848532882246084e-05,
+      "loss": 0.9816,
+      "step": 933
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6041750250907159,
+      "learning_rate": 1.8845625516939013e-05,
+      "loss": 0.9693,
+      "step": 934
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8318879885690347,
+      "learning_rate": 1.884271471063427e-05,
+      "loss": 1.0006,
+      "step": 935
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7196295921526732,
+      "learning_rate": 1.8839800464464173e-05,
+      "loss": 0.9136,
+      "step": 936
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1161464159670889,
+      "learning_rate": 1.883688277956238e-05,
+      "loss": 0.8585,
+      "step": 937
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6957290545980972,
+      "learning_rate": 1.8833961657063887e-05,
+      "loss": 1.016,
+      "step": 938
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8212003177045848,
+      "learning_rate": 1.8831037098105027e-05,
+      "loss": 1.0086,
+      "step": 939
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7802330188046225,
+      "learning_rate": 1.882810910382347e-05,
+      "loss": 0.9704,
+      "step": 940
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8626496074886685,
+      "learning_rate": 1.8825177675358218e-05,
+      "loss": 1.0282,
+      "step": 941
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7632515912518898,
+      "learning_rate": 1.882224281384962e-05,
+      "loss": 0.9159,
+      "step": 942
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.83003157160647,
+      "learning_rate": 1.8819304520439356e-05,
+      "loss": 1.0748,
+      "step": 943
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7393204750013316,
+      "learning_rate": 1.8816362796270423e-05,
+      "loss": 0.9811,
+      "step": 944
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7236295857260255,
+      "learning_rate": 1.8813417642487188e-05,
+      "loss": 0.9968,
+      "step": 945
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7842741294009632,
+      "learning_rate": 1.8810469060235316e-05,
+      "loss": 0.9955,
+      "step": 946
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8850015100069208,
+      "learning_rate": 1.880751705066183e-05,
+      "loss": 1.0119,
+      "step": 947
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7103680655019537,
+      "learning_rate": 1.8804561614915074e-05,
+      "loss": 0.982,
+      "step": 948
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7457737900802666,
+      "learning_rate": 1.880160275414473e-05,
+      "loss": 1.0117,
+      "step": 949
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7380880327585835,
+      "learning_rate": 1.879864046950182e-05,
+      "loss": 0.9814,
+      "step": 950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6168810096782424,
+      "learning_rate": 1.879567476213868e-05,
+      "loss": 1.0266,
+      "step": 951
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6896180201871307,
+      "learning_rate": 1.8792705633208984e-05,
+      "loss": 1.0973,
+      "step": 952
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9883924393816006,
+      "learning_rate": 1.878973308386775e-05,
+      "loss": 0.9656,
+      "step": 953
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.751124281045345,
+      "learning_rate": 1.8786757115271304e-05,
+      "loss": 1.076,
+      "step": 954
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9397698973675674,
+      "learning_rate": 1.878377772857733e-05,
+      "loss": 0.8821,
+      "step": 955
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.825924450835264,
+      "learning_rate": 1.878079492494481e-05,
+      "loss": 0.918,
+      "step": 956
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.598177775662806,
+      "learning_rate": 1.877780870553408e-05,
+      "loss": 0.9548,
+      "step": 957
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.754444109667171,
+      "learning_rate": 1.87748190715068e-05,
+      "loss": 0.9269,
+      "step": 958
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7185734654649434,
+      "learning_rate": 1.8771826024025944e-05,
+      "loss": 0.9515,
+      "step": 959
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7516480228209133,
+      "learning_rate": 1.876882956425583e-05,
+      "loss": 0.9077,
+      "step": 960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5390233926288543,
+      "learning_rate": 1.8765829693362097e-05,
+      "loss": 0.9383,
+      "step": 961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7511779654205533,
+      "learning_rate": 1.8762826412511707e-05,
+      "loss": 0.8654,
+      "step": 962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8337278369894814,
+      "learning_rate": 1.8759819722872962e-05,
+      "loss": 0.9682,
+      "step": 963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.630463001648313,
+      "learning_rate": 1.8756809625615476e-05,
+      "loss": 0.9301,
+      "step": 964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8259198480757137,
+      "learning_rate": 1.875379612191019e-05,
+      "loss": 1.0628,
+      "step": 965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7225958216426362,
+      "learning_rate": 1.875077921292938e-05,
+      "loss": 0.9861,
+      "step": 966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1997463284412684,
+      "learning_rate": 1.8747758899846636e-05,
+      "loss": 0.857,
+      "step": 967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8198138794011165,
+      "learning_rate": 1.8744735183836883e-05,
+      "loss": 0.9274,
+      "step": 968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8280718208867093,
+      "learning_rate": 1.8741708066076355e-05,
+      "loss": 0.9962,
+      "step": 969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.885386949678351,
+      "learning_rate": 1.873867754774262e-05,
+      "loss": 0.9317,
+      "step": 970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.650817221311481,
+      "learning_rate": 1.8735643630014573e-05,
+      "loss": 0.9479,
+      "step": 971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.030456179189788,
+      "learning_rate": 1.8732606314072418e-05,
+      "loss": 0.8117,
+      "step": 972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7255823855574908,
+      "learning_rate": 1.872956560109769e-05,
+      "loss": 1.0099,
+      "step": 973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7880899405107156,
+      "learning_rate": 1.8726521492273243e-05,
+      "loss": 0.9599,
+      "step": 974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6315086551223301,
+      "learning_rate": 1.872347398878325e-05,
+      "loss": 1.0135,
+      "step": 975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7439493090706077,
+      "learning_rate": 1.8720423091813212e-05,
+      "loss": 0.9534,
+      "step": 976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8029359266871865,
+      "learning_rate": 1.8717368802549942e-05,
+      "loss": 0.9886,
+      "step": 977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.744701375651719,
+      "learning_rate": 1.8714311122181577e-05,
+      "loss": 0.9795,
+      "step": 978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.137523058173478,
+      "learning_rate": 1.8711250051897568e-05,
+      "loss": 0.8049,
+      "step": 979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.899946301921143,
+      "learning_rate": 1.8708185592888694e-05,
+      "loss": 0.9562,
+      "step": 980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.860447508568594,
+      "learning_rate": 1.8705117746347043e-05,
+      "loss": 0.9351,
+      "step": 981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9853952642587809,
+      "learning_rate": 1.8702046513466023e-05,
+      "loss": 1.1098,
+      "step": 982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0232788851356247,
+      "learning_rate": 1.8698971895440364e-05,
+      "loss": 0.9697,
+      "step": 983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6182938883940627,
+      "learning_rate": 1.869589389346611e-05,
+      "loss": 0.8622,
+      "step": 984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8016042106370669,
+      "learning_rate": 1.869281250874062e-05,
+      "loss": 0.9718,
+      "step": 985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.842631970704674,
+      "learning_rate": 1.8689727742462568e-05,
+      "loss": 0.9628,
+      "step": 986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7757501677552117,
+      "learning_rate": 1.8686639595831945e-05,
+      "loss": 1.0369,
+      "step": 987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7862503410057553,
+      "learning_rate": 1.868354807005006e-05,
+      "loss": 1.0037,
+      "step": 988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9713640072750347,
+      "learning_rate": 1.8680453166319534e-05,
+      "loss": 1.0668,
+      "step": 989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5834161475349875,
+      "learning_rate": 1.8677354885844298e-05,
+      "loss": 0.9117,
+      "step": 990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7558629074148957,
+      "learning_rate": 1.8674253229829603e-05,
+      "loss": 0.9565,
+      "step": 991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7897207769634744,
+      "learning_rate": 1.8671148199482008e-05,
+      "loss": 0.8187,
+      "step": 992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6939542401434835,
+      "learning_rate": 1.8668039796009395e-05,
+      "loss": 0.9575,
+      "step": 993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7199831061387953,
+      "learning_rate": 1.866492802062094e-05,
+      "loss": 0.9861,
+      "step": 994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6602703366513138,
+      "learning_rate": 1.8661812874527146e-05,
+      "loss": 0.9197,
+      "step": 995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.870769391335709,
+      "learning_rate": 1.8658694358939825e-05,
+      "loss": 0.9745,
+      "step": 996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0984317950085014,
+      "learning_rate": 1.865557247507209e-05,
+      "loss": 0.9383,
+      "step": 997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1346115720788092,
+      "learning_rate": 1.865244722413838e-05,
+      "loss": 1.014,
+      "step": 998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7820798091137076,
+      "learning_rate": 1.8649318607354425e-05,
+      "loss": 1.0927,
+      "step": 999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8432658848332057,
+      "learning_rate": 1.8646186625937283e-05,
+      "loss": 1.0396,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6606423884661052,
+      "learning_rate": 1.8643051281105306e-05,
+      "loss": 0.9916,
+      "step": 1001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7356514802738832,
+      "learning_rate": 1.8639912574078166e-05,
+      "loss": 1.0397,
+      "step": 1002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9654005366842533,
+      "learning_rate": 1.863677050607683e-05,
+      "loss": 1.0416,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7007752362612942,
+      "learning_rate": 1.863362507832359e-05,
+      "loss": 0.9567,
+      "step": 1004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7140015188950268,
+      "learning_rate": 1.863047629204203e-05,
+      "loss": 0.9889,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8038816957710546,
+      "learning_rate": 1.8627324148457045e-05,
+      "loss": 1.0358,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7918336922714801,
+      "learning_rate": 1.8624168648794833e-05,
+      "loss": 1.0899,
+      "step": 1007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9893741940116425,
+      "learning_rate": 1.8621009794282908e-05,
+      "loss": 1.0072,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.880951803823825,
+      "learning_rate": 1.8617847586150074e-05,
+      "loss": 0.9179,
+      "step": 1009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.773048768472068,
+      "learning_rate": 1.8614682025626457e-05,
+      "loss": 0.9337,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7238139599014435,
+      "learning_rate": 1.8611513113943474e-05,
+      "loss": 0.936,
+      "step": 1011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9274738617630254,
+      "learning_rate": 1.8608340852333848e-05,
+      "loss": 1.0219,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7261211949806228,
+      "learning_rate": 1.86051652420316e-05,
+      "loss": 0.8676,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7339997882443183,
+      "learning_rate": 1.8601986284272074e-05,
+      "loss": 0.9865,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8505045248417213,
+      "learning_rate": 1.8598803980291894e-05,
+      "loss": 0.9859,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1629854089616662,
+      "learning_rate": 1.8595618331328994e-05,
+      "loss": 0.8771,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1001200272274703,
+      "learning_rate": 1.859242933862261e-05,
+      "loss": 0.8365,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7887746000305487,
+      "learning_rate": 1.858923700341328e-05,
+      "loss": 1.044,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9444771408024482,
+      "learning_rate": 1.858604132694284e-05,
+      "loss": 1.0438,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8072937597189,
+      "learning_rate": 1.858284231045442e-05,
+      "loss": 0.9188,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.914526422574316,
+      "learning_rate": 1.8579639955192462e-05,
+      "loss": 0.9582,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8002540275946919,
+      "learning_rate": 1.85764342624027e-05,
+      "loss": 0.9141,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9356616736953243,
+      "learning_rate": 1.8573225233332158e-05,
+      "loss": 1.0024,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.730066184444454,
+      "learning_rate": 1.8570012869229178e-05,
+      "loss": 0.976,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7225255060219573,
+      "learning_rate": 1.856679717134338e-05,
+      "loss": 0.8718,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6393537075143196,
+      "learning_rate": 1.8563578140925692e-05,
+      "loss": 1.0115,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6299298628068148,
+      "learning_rate": 1.8560355779228332e-05,
+      "loss": 0.9343,
+      "step": 1027
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7029097849722552,
+      "learning_rate": 1.855713008750482e-05,
+      "loss": 0.9686,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.969844605788607,
+      "learning_rate": 1.8553901067009964e-05,
+      "loss": 0.9893,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.704086919456949,
+      "learning_rate": 1.8550668718999873e-05,
+      "loss": 0.8566,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.724266119391234,
+      "learning_rate": 1.8547433044731947e-05,
+      "loss": 0.9257,
+      "step": 1031
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5771383258769143,
+      "learning_rate": 1.8544194045464888e-05,
+      "loss": 0.8341,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7096799900684885,
+      "learning_rate": 1.8540951722458677e-05,
+      "loss": 0.8842,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.907328083319732,
+      "learning_rate": 1.8537706076974597e-05,
+      "loss": 0.9625,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8785595681104557,
+      "learning_rate": 1.8534457110275222e-05,
+      "loss": 1.0047,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9149343630950748,
+      "learning_rate": 1.8531204823624423e-05,
+      "loss": 0.9813,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9192916798154525,
+      "learning_rate": 1.852794921828735e-05,
+      "loss": 1.0411,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.792138121037841,
+      "learning_rate": 1.852469029553046e-05,
+      "loss": 0.9539,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7377921391977782,
+      "learning_rate": 1.8521428056621488e-05,
+      "loss": 0.9869,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.836138914488659,
+      "learning_rate": 1.851816250282946e-05,
+      "loss": 0.9277,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.696135903850965,
+      "learning_rate": 1.85148936354247e-05,
+      "loss": 0.8891,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9453997222142316,
+      "learning_rate": 1.8511621455678814e-05,
+      "loss": 0.9673,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6244801367188275,
+      "learning_rate": 1.8508345964864697e-05,
+      "loss": 0.9709,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8987256850412175,
+      "learning_rate": 1.8505067164256534e-05,
+      "loss": 0.9947,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9918725683598992,
+      "learning_rate": 1.8501785055129795e-05,
+      "loss": 0.9426,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3606628898180593,
+      "learning_rate": 1.8498499638761244e-05,
+      "loss": 0.859,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9075468754263132,
+      "learning_rate": 1.849521091642892e-05,
+      "loss": 1.0131,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8017078685401475,
+      "learning_rate": 1.8491918889412155e-05,
+      "loss": 0.9883,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5936255603464167,
+      "learning_rate": 1.848862355899157e-05,
+      "loss": 0.9388,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.713480368859398,
+      "learning_rate": 1.8485324926449065e-05,
+      "loss": 0.9242,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8737085606443047,
+      "learning_rate": 1.8482022993067828e-05,
+      "loss": 1.0481,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7994992554529972,
+      "learning_rate": 1.8478717760132327e-05,
+      "loss": 1.0248,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9273237236049192,
+      "learning_rate": 1.8475409228928314e-05,
+      "loss": 1.1252,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.811256243389598,
+      "learning_rate": 1.847209740074283e-05,
+      "loss": 0.9975,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7665314252777147,
+      "learning_rate": 1.8468782276864194e-05,
+      "loss": 1.0071,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6853950651577243,
+      "learning_rate": 1.8465463858582008e-05,
+      "loss": 0.9103,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0474240570356756,
+      "learning_rate": 1.8462142147187152e-05,
+      "loss": 0.9535,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.762291868979056,
+      "learning_rate": 1.8458817143971796e-05,
+      "loss": 1.0347,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8450923509174453,
+      "learning_rate": 1.845548885022938e-05,
+      "loss": 0.9764,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.737647606721249,
+      "learning_rate": 1.8452157267254628e-05,
+      "loss": 0.9281,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7282435390337312,
+      "learning_rate": 1.8448822396343547e-05,
+      "loss": 1.0776,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6878995204830884,
+      "learning_rate": 1.8445484238793426e-05,
+      "loss": 1.0418,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7672972126509847,
+      "learning_rate": 1.8442142795902815e-05,
+      "loss": 0.9585,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7251599746817046,
+      "learning_rate": 1.8438798068971563e-05,
+      "loss": 0.996,
+      "step": 1064
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7214933444620912,
+      "learning_rate": 1.843545005930078e-05,
+      "loss": 1.0199,
+      "step": 1065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.95885512538602,
+      "learning_rate": 1.843209876819287e-05,
+      "loss": 0.9727,
+      "step": 1066
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8864908531664697,
+      "learning_rate": 1.8428744196951497e-05,
+      "loss": 1.0298,
+      "step": 1067
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6377396633741799,
+      "learning_rate": 1.842538634688161e-05,
+      "loss": 0.9738,
+      "step": 1068
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5770143994554613,
+      "learning_rate": 1.842202521928943e-05,
+      "loss": 0.9777,
+      "step": 1069
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7537617684698388,
+      "learning_rate": 1.8418660815482455e-05,
+      "loss": 0.892,
+      "step": 1070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.749951027765996,
+      "learning_rate": 1.841529313676946e-05,
+      "loss": 1.0151,
+      "step": 1071
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.84947905592699,
+      "learning_rate": 1.8411922184460485e-05,
+      "loss": 0.9415,
+      "step": 1072
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7052366678756565,
+      "learning_rate": 1.8408547959866854e-05,
+      "loss": 1.0444,
+      "step": 1073
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8815412991321163,
+      "learning_rate": 1.8405170464301157e-05,
+      "loss": 0.9091,
+      "step": 1074
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6427529500460285,
+      "learning_rate": 1.840178969907726e-05,
+      "loss": 0.9417,
+      "step": 1075
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8055789848950878,
+      "learning_rate": 1.83984056655103e-05,
+      "loss": 1.0194,
+      "step": 1076
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8021950217700953,
+      "learning_rate": 1.839501836491668e-05,
+      "loss": 1.0476,
+      "step": 1077
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9178254864295985,
+      "learning_rate": 1.8391627798614083e-05,
+      "loss": 0.927,
+      "step": 1078
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6772722800337356,
+      "learning_rate": 1.838823396792146e-05,
+      "loss": 0.9693,
+      "step": 1079
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.826535932649843,
+      "learning_rate": 1.8384836874159027e-05,
+      "loss": 0.9583,
+      "step": 1080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7952870078363883,
+      "learning_rate": 1.8381436518648274e-05,
+      "loss": 0.9533,
+      "step": 1081
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9080114507989396,
+      "learning_rate": 1.8378032902711957e-05,
+      "loss": 1.0318,
+      "step": 1082
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5988402480469883,
+      "learning_rate": 1.8374626027674097e-05,
+      "loss": 0.8866,
+      "step": 1083
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.889505068458654,
+      "learning_rate": 1.837121589486e-05,
+      "loss": 1.005,
+      "step": 1084
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8685037585681314,
+      "learning_rate": 1.8367802505596212e-05,
+      "loss": 0.9165,
+      "step": 1085
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7810033420773723,
+      "learning_rate": 1.8364385861210573e-05,
+      "loss": 0.9902,
+      "step": 1086
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.833315963377804,
+      "learning_rate": 1.836096596303217e-05,
+      "loss": 0.8384,
+      "step": 1087
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7518403623650973,
+      "learning_rate": 1.8357542812391363e-05,
+      "loss": 0.9596,
+      "step": 1088
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9486533221657822,
+      "learning_rate": 1.835411641061978e-05,
+      "loss": 0.9678,
+      "step": 1089
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7593487486740689,
+      "learning_rate": 1.8350686759050306e-05,
+      "loss": 0.8999,
+      "step": 1090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.688500614021737,
+      "learning_rate": 1.8347253859017096e-05,
+      "loss": 1.0182,
+      "step": 1091
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7649273183460408,
+      "learning_rate": 1.8343817711855564e-05,
+      "loss": 0.951,
+      "step": 1092
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.704027407774857,
+      "learning_rate": 1.83403783189024e-05,
+      "loss": 1.0187,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.768447118521444,
+      "learning_rate": 1.833693568149554e-05,
+      "loss": 1.004,
+      "step": 1094
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5561392928343287,
+      "learning_rate": 1.833348980097419e-05,
+      "loss": 0.9605,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.759604941599594,
+      "learning_rate": 1.8330040678678815e-05,
+      "loss": 0.9477,
+      "step": 1096
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6833482858188118,
+      "learning_rate": 1.8326588315951145e-05,
+      "loss": 0.9744,
+      "step": 1097
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8602823837747016,
+      "learning_rate": 1.8323132714134174e-05,
+      "loss": 1.0181,
+      "step": 1098
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.72725277723906,
+      "learning_rate": 1.831967387457214e-05,
+      "loss": 0.9717,
+      "step": 1099
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7488839951788413,
+      "learning_rate": 1.8316211798610553e-05,
+      "loss": 0.9287,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.715600714192395,
+      "learning_rate": 1.8312746487596187e-05,
+      "loss": 0.9275,
+      "step": 1101
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9234276900594538,
+      "learning_rate": 1.830927794287706e-05,
+      "loss": 1.0362,
+      "step": 1102
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.780128882189903,
+      "learning_rate": 1.830580616580246e-05,
+      "loss": 0.8657,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.058785840695658,
+      "learning_rate": 1.8302331157722926e-05,
+      "loss": 0.7972,
+      "step": 1104
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7508411264214212,
+      "learning_rate": 1.8298852919990254e-05,
+      "loss": 0.9381,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7776534965786406,
+      "learning_rate": 1.8295371453957498e-05,
+      "loss": 0.9355,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8407458427528127,
+      "learning_rate": 1.829188676097897e-05,
+      "loss": 0.938,
+      "step": 1107
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9754630784316629,
+      "learning_rate": 1.828839884241023e-05,
+      "loss": 0.7886,
+      "step": 1108
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6729876825438816,
+      "learning_rate": 1.8284907699608107e-05,
+      "loss": 0.9862,
+      "step": 1109
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.70364445386777,
+      "learning_rate": 1.8281413333930667e-05,
+      "loss": 0.9166,
+      "step": 1110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7943773887955248,
+      "learning_rate": 1.8277915746737244e-05,
+      "loss": 0.9559,
+      "step": 1111
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7706311048243588,
+      "learning_rate": 1.827441493938841e-05,
+      "loss": 1.0027,
+      "step": 1112
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6402229998611062,
+      "learning_rate": 1.8270910913246006e-05,
+      "loss": 0.9496,
+      "step": 1113
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7163309338152166,
+      "learning_rate": 1.8267403669673114e-05,
+      "loss": 0.909,
+      "step": 1114
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7633141996715775,
+      "learning_rate": 1.8263893210034075e-05,
+      "loss": 1.016,
+      "step": 1115
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8789837144004635,
+      "learning_rate": 1.8260379535694476e-05,
+      "loss": 1.0426,
+      "step": 1116
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9562642292283585,
+      "learning_rate": 1.825686264802115e-05,
+      "loss": 1.017,
+      "step": 1117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6423320642020387,
+      "learning_rate": 1.8253342548382196e-05,
+      "loss": 0.9763,
+      "step": 1118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0178675553150265,
+      "learning_rate": 1.8249819238146947e-05,
+      "loss": 0.8347,
+      "step": 1119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0005200585395415,
+      "learning_rate": 1.824629271868599e-05,
+      "loss": 1.0056,
+      "step": 1120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9043558006375518,
+      "learning_rate": 1.824276299137116e-05,
+      "loss": 1.0329,
+      "step": 1121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6465605497101876,
+      "learning_rate": 1.8239230057575542e-05,
+      "loss": 0.9662,
+      "step": 1122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7575700422042988,
+      "learning_rate": 1.8235693918673472e-05,
+      "loss": 1.0516,
+      "step": 1123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6179029539225438,
+      "learning_rate": 1.823215457604052e-05,
+      "loss": 0.8443,
+      "step": 1124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.659719987126641,
+      "learning_rate": 1.822861203105351e-05,
+      "loss": 0.9977,
+      "step": 1125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7202277656336742,
+      "learning_rate": 1.8225066285090517e-05,
+      "loss": 0.945,
+      "step": 1126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7788664620864658,
+      "learning_rate": 1.8221517339530854e-05,
+      "loss": 0.9998,
+      "step": 1127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7827706610732965,
+      "learning_rate": 1.8217965195755082e-05,
+      "loss": 0.9792,
+      "step": 1128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7869417460616888,
+      "learning_rate": 1.8214409855145e-05,
+      "loss": 0.8796,
+      "step": 1129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.604801121348349,
+      "learning_rate": 1.821085131908366e-05,
+      "loss": 0.8961,
+      "step": 1130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.789825055915408,
+      "learning_rate": 1.820728958895535e-05,
+      "loss": 1.0058,
+      "step": 1131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5874110655595879,
+      "learning_rate": 1.8203724666145604e-05,
+      "loss": 0.8884,
+      "step": 1132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9852213381493118,
+      "learning_rate": 1.8200156552041196e-05,
+      "loss": 0.9552,
+      "step": 1133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6948650585221623,
+      "learning_rate": 1.819658524803014e-05,
+      "loss": 0.9663,
+      "step": 1134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.708131778641045,
+      "learning_rate": 1.81930107555017e-05,
+      "loss": 1.0013,
+      "step": 1135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7240336688010882,
+      "learning_rate": 1.818943307584637e-05,
+      "loss": 0.9397,
+      "step": 1136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.600038727635933,
+      "learning_rate": 1.8185852210455883e-05,
+      "loss": 0.963,
+      "step": 1137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5714940339757,
+      "learning_rate": 1.818226816072322e-05,
+      "loss": 0.9257,
+      "step": 1138
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.884892347702815,
+      "learning_rate": 1.8178680928042597e-05,
+      "loss": 0.8838,
+      "step": 1139
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9168753175003004,
+      "learning_rate": 1.8175090513809472e-05,
+      "loss": 0.9836,
+      "step": 1140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6732740841507494,
+      "learning_rate": 1.817149691942052e-05,
+      "loss": 0.9369,
+      "step": 1141
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5371119006052867,
+      "learning_rate": 1.816790014627369e-05,
+      "loss": 0.9285,
+      "step": 1142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8055078013755588,
+      "learning_rate": 1.8164300195768127e-05,
+      "loss": 0.9205,
+      "step": 1143
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.969601309755434,
+      "learning_rate": 1.8160697069304248e-05,
+      "loss": 0.9895,
+      "step": 1144
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.703227974778778,
+      "learning_rate": 1.815709076828368e-05,
+      "loss": 0.9485,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7476636511720447,
+      "learning_rate": 1.8153481294109297e-05,
+      "loss": 0.9276,
+      "step": 1146
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8518174020875764,
+      "learning_rate": 1.8149868648185207e-05,
+      "loss": 0.8492,
+      "step": 1147
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7772716744902848,
+      "learning_rate": 1.814625283191674e-05,
+      "loss": 0.9507,
+      "step": 1148
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7614630138684018,
+      "learning_rate": 1.8142633846710484e-05,
+      "loss": 0.9293,
+      "step": 1149
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.797671924779979,
+      "learning_rate": 1.813901169397423e-05,
+      "loss": 1.0318,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.582075581088096,
+      "learning_rate": 1.8135386375117025e-05,
+      "loss": 0.9799,
+      "step": 1151
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6874904371674906,
+      "learning_rate": 1.8131757891549134e-05,
+      "loss": 0.8737,
+      "step": 1152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.786427729474046,
+      "learning_rate": 1.812812624468206e-05,
+      "loss": 0.8877,
+      "step": 1153
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8026853355595733,
+      "learning_rate": 1.812449143592853e-05,
+      "loss": 0.983,
+      "step": 1154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.645285885074254,
+      "learning_rate": 1.8120853466702505e-05,
+      "loss": 0.853,
+      "step": 1155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7225585582736722,
+      "learning_rate": 1.811721233841918e-05,
+      "loss": 1.0056,
+      "step": 1156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6531037428657749,
+      "learning_rate": 1.8113568052494973e-05,
+      "loss": 0.946,
+      "step": 1157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6477077633924566,
+      "learning_rate": 1.8109920610347526e-05,
+      "loss": 0.9772,
+      "step": 1158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.759635383594622,
+      "learning_rate": 1.8106270013395722e-05,
+      "loss": 0.9663,
+      "step": 1159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8755922461636803,
+      "learning_rate": 1.810261626305966e-05,
+      "loss": 0.9845,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.785581379644735,
+      "learning_rate": 1.8098959360760665e-05,
+      "loss": 0.9824,
+      "step": 1161
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9949668309791637,
+      "learning_rate": 1.80952993079213e-05,
+      "loss": 1.0,
+      "step": 1162
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7405863384045483,
+      "learning_rate": 1.8091636105965343e-05,
+      "loss": 0.9767,
+      "step": 1163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7391281951276767,
+      "learning_rate": 1.80879697563178e-05,
+      "loss": 0.8337,
+      "step": 1164
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6795034455390772,
+      "learning_rate": 1.80843002604049e-05,
+      "loss": 0.9712,
+      "step": 1165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.63723530693541,
+      "learning_rate": 1.80806276196541e-05,
+      "loss": 0.9864,
+      "step": 1166
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8333977460629227,
+      "learning_rate": 1.8076951835494077e-05,
+      "loss": 1.0439,
+      "step": 1167
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.710751816346255,
+      "learning_rate": 1.8073272909354727e-05,
+      "loss": 1.0028,
+      "step": 1168
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6962919023621972,
+      "learning_rate": 1.806959084266718e-05,
+      "loss": 0.9167,
+      "step": 1169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7069224647946217,
+      "learning_rate": 1.8065905636863778e-05,
+      "loss": 0.9273,
+      "step": 1170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7938161497267153,
+      "learning_rate": 1.8062217293378082e-05,
+      "loss": 0.9723,
+      "step": 1171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1089072790922694,
+      "learning_rate": 1.8058525813644882e-05,
+      "loss": 0.8665,
+      "step": 1172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.69892710179464,
+      "learning_rate": 1.8054831199100187e-05,
+      "loss": 1.0218,
+      "step": 1173
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0451491099423602,
+      "learning_rate": 1.8051133451181218e-05,
+      "loss": 0.8929,
+      "step": 1174
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8796359371454332,
+      "learning_rate": 1.8047432571326424e-05,
+      "loss": 0.9927,
+      "step": 1175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7233987627495029,
+      "learning_rate": 1.804372856097546e-05,
+      "loss": 1.0978,
+      "step": 1176
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9292585998511684,
+      "learning_rate": 1.8040021421569217e-05,
+      "loss": 0.9651,
+      "step": 1177
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.694444802207448,
+      "learning_rate": 1.8036311154549783e-05,
+      "loss": 1.0246,
+      "step": 1178
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7171806882910932,
+      "learning_rate": 1.8032597761360483e-05,
+      "loss": 0.9737,
+      "step": 1179
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6837885479019155,
+      "learning_rate": 1.802888124344584e-05,
+      "loss": 0.9178,
+      "step": 1180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8826883675534296,
+      "learning_rate": 1.8025161602251598e-05,
+      "loss": 0.9846,
+      "step": 1181
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0322980918296496,
+      "learning_rate": 1.8021438839224727e-05,
+      "loss": 0.8489,
+      "step": 1182
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8079028925157001,
+      "learning_rate": 1.8017712955813396e-05,
+      "loss": 0.9028,
+      "step": 1183
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7122036600160135,
+      "learning_rate": 1.8013983953467e-05,
+      "loss": 0.9006,
+      "step": 1184
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4596731921645267,
+      "learning_rate": 1.8010251833636134e-05,
+      "loss": 0.9003,
+      "step": 1185
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8961229586110253,
+      "learning_rate": 1.800651659777262e-05,
+      "loss": 1.0811,
+      "step": 1186
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.613416292292287,
+      "learning_rate": 1.8002778247329483e-05,
+      "loss": 0.8975,
+      "step": 1187
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6739171767192602,
+      "learning_rate": 1.7999036783760962e-05,
+      "loss": 0.9107,
+      "step": 1188
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.650906616060577,
+      "learning_rate": 1.7995292208522505e-05,
+      "loss": 0.9537,
+      "step": 1189
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9625538582641633,
+      "learning_rate": 1.799154452307078e-05,
+      "loss": 0.9207,
+      "step": 1190
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6753887830581213,
+      "learning_rate": 1.798779372886365e-05,
+      "loss": 0.9792,
+      "step": 1191
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7758901785218542,
+      "learning_rate": 1.7984039827360196e-05,
+      "loss": 0.9522,
+      "step": 1192
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6125505526071824,
+      "learning_rate": 1.7980282820020716e-05,
+      "loss": 0.983,
+      "step": 1193
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7384287720116136,
+      "learning_rate": 1.797652270830669e-05,
+      "loss": 0.9703,
+      "step": 1194
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.861555723738114,
+      "learning_rate": 1.7972759493680837e-05,
+      "loss": 0.9137,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.041259419376844,
+      "learning_rate": 1.7968993177607062e-05,
+      "loss": 0.9454,
+      "step": 1196
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8739241610760922,
+      "learning_rate": 1.7965223761550482e-05,
+      "loss": 0.9806,
+      "step": 1197
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.674145922056197,
+      "learning_rate": 1.7961451246977426e-05,
+      "loss": 0.9276,
+      "step": 1198
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7127329957842803,
+      "learning_rate": 1.7957675635355417e-05,
+      "loss": 0.9682,
+      "step": 1199
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8149078483496093,
+      "learning_rate": 1.7953896928153194e-05,
+      "loss": 0.998,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9318185389017108,
+      "learning_rate": 1.7950115126840694e-05,
+      "loss": 1.0629,
+      "step": 1201
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0548626243338954,
+      "learning_rate": 1.7946330232889056e-05,
+      "loss": 0.8551,
+      "step": 1202
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8128007476904608,
+      "learning_rate": 1.794254224777063e-05,
+      "loss": 0.9968,
+      "step": 1203
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8426640315270553,
+      "learning_rate": 1.793875117295896e-05,
+      "loss": 1.0071,
+      "step": 1204
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8622523541744536,
+      "learning_rate": 1.793495700992879e-05,
+      "loss": 1.0226,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8592975553732978,
+      "learning_rate": 1.793115976015608e-05,
+      "loss": 0.9464,
+      "step": 1206
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7145817947231954,
+      "learning_rate": 1.7927359425117977e-05,
+      "loss": 0.9987,
+      "step": 1207
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7928354696423001,
+      "learning_rate": 1.7923556006292833e-05,
+      "loss": 0.9877,
+      "step": 1208
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5668009311186468,
+      "learning_rate": 1.79197495051602e-05,
+      "loss": 0.8424,
+      "step": 1209
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7574464067051558,
+      "learning_rate": 1.7915939923200826e-05,
+      "loss": 0.9927,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.664921587813246,
+      "learning_rate": 1.7912127261896663e-05,
+      "loss": 0.9248,
+      "step": 1211
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7752558765064403,
+      "learning_rate": 1.790831152273085e-05,
+      "loss": 0.9306,
+      "step": 1212
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8892846210070513,
+      "learning_rate": 1.7904492707187742e-05,
+      "loss": 0.8761,
+      "step": 1213
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7332662965522532,
+      "learning_rate": 1.7900670816752875e-05,
+      "loss": 0.9938,
+      "step": 1214
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7419628732267034,
+      "learning_rate": 1.789684585291298e-05,
+      "loss": 0.9207,
+      "step": 1215
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7692266699476438,
+      "learning_rate": 1.7893017817155994e-05,
+      "loss": 0.9323,
+      "step": 1216
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.686099900584496,
+      "learning_rate": 1.788918671097105e-05,
+      "loss": 1.0403,
+      "step": 1217
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7049271897856655,
+      "learning_rate": 1.7885352535848457e-05,
+      "loss": 0.8763,
+      "step": 1218
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8842867623470574,
+      "learning_rate": 1.788151529327974e-05,
+      "loss": 0.8992,
+      "step": 1219
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5101257193691588,
+      "learning_rate": 1.7877674984757606e-05,
+      "loss": 0.8578,
+      "step": 1220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7458704500708562,
+      "learning_rate": 1.7873831611775953e-05,
+      "loss": 1.0039,
+      "step": 1221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9588037137122554,
+      "learning_rate": 1.786998517582988e-05,
+      "loss": 1.0242,
+      "step": 1222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6960489702196067,
+      "learning_rate": 1.7866135678415664e-05,
+      "loss": 0.9264,
+      "step": 1223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6446035957311669,
+      "learning_rate": 1.786228312103079e-05,
+      "loss": 0.9222,
+      "step": 1224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6385096331956297,
+      "learning_rate": 1.785842750517392e-05,
+      "loss": 0.9651,
+      "step": 1225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8288884624971344,
+      "learning_rate": 1.785456883234491e-05,
+      "loss": 0.884,
+      "step": 1226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7612917137043658,
+      "learning_rate": 1.7850707104044805e-05,
+      "loss": 0.9367,
+      "step": 1227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.693272299332214,
+      "learning_rate": 1.784684232177584e-05,
+      "loss": 1.0184,
+      "step": 1228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7787276387045592,
+      "learning_rate": 1.7842974487041438e-05,
+      "loss": 0.8999,
+      "step": 1229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8491883688334283,
+      "learning_rate": 1.7839103601346208e-05,
+      "loss": 1.0087,
+      "step": 1230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8445884971669637,
+      "learning_rate": 1.783522966619594e-05,
+      "loss": 0.9743,
+      "step": 1231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6844175504540648,
+      "learning_rate": 1.7831352683097624e-05,
+      "loss": 0.8907,
+      "step": 1232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9564826874811936,
+      "learning_rate": 1.7827472653559423e-05,
+      "loss": 0.8386,
+      "step": 1233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6730253156245942,
+      "learning_rate": 1.7823589579090693e-05,
+      "loss": 0.9593,
+      "step": 1234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6714737183653596,
+      "learning_rate": 1.7819703461201972e-05,
+      "loss": 0.9599,
+      "step": 1235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6822494430380985,
+      "learning_rate": 1.781581430140498e-05,
+      "loss": 0.9372,
+      "step": 1236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6314341135474497,
+      "learning_rate": 1.7811922101212622e-05,
+      "loss": 0.8846,
+      "step": 1237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.682802339581403,
+      "learning_rate": 1.7808026862138988e-05,
+      "loss": 0.9077,
+      "step": 1238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.770602061803176,
+      "learning_rate": 1.7804128585699344e-05,
+      "loss": 0.9694,
+      "step": 1239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8401386638335337,
+      "learning_rate": 1.7800227273410143e-05,
+      "loss": 0.8913,
+      "step": 1240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7827670056297955,
+      "learning_rate": 1.7796322926789015e-05,
+      "loss": 0.91,
+      "step": 1241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8302380251708938,
+      "learning_rate": 1.779241554735478e-05,
+      "loss": 1.0125,
+      "step": 1242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.657268435720547,
+      "learning_rate": 1.7788505136627422e-05,
+      "loss": 0.914,
+      "step": 1243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6904335179764767,
+      "learning_rate": 1.7784591696128117e-05,
+      "loss": 0.8872,
+      "step": 1244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6113331535786208,
+      "learning_rate": 1.778067522737921e-05,
+      "loss": 0.891,
+      "step": 1245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7436649439724654,
+      "learning_rate": 1.7776755731904242e-05,
+      "loss": 1.0426,
+      "step": 1246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7145594894019331,
+      "learning_rate": 1.7772833211227903e-05,
+      "loss": 0.9528,
+      "step": 1247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8811027877747375,
+      "learning_rate": 1.7768907666876084e-05,
+      "loss": 0.9386,
+      "step": 1248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4322069299752822,
+      "learning_rate": 1.776497910037584e-05,
+      "loss": 0.8494,
+      "step": 1249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8580943241155263,
+      "learning_rate": 1.7761047513255413e-05,
+      "loss": 1.0434,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6560980953182558,
+      "learning_rate": 1.77571129070442e-05,
+      "loss": 0.8866,
+      "step": 1251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4946413349691632,
+      "learning_rate": 1.775317528327279e-05,
+      "loss": 0.8984,
+      "step": 1252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7735826158427899,
+      "learning_rate": 1.7749234643472948e-05,
+      "loss": 0.9594,
+      "step": 1253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6888493777188174,
+      "learning_rate": 1.77452909891776e-05,
+      "loss": 0.933,
+      "step": 1254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7622039365048128,
+      "learning_rate": 1.774134432192084e-05,
+      "loss": 1.0269,
+      "step": 1255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7546436993651273,
+      "learning_rate": 1.773739464323796e-05,
+      "loss": 0.9723,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7373925506250298,
+      "learning_rate": 1.7733441954665394e-05,
+      "loss": 0.8289,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7500153315759612,
+      "learning_rate": 1.7729486257740767e-05,
+      "loss": 0.9889,
+      "step": 1258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6897884730229005,
+      "learning_rate": 1.7725527554002864e-05,
+      "loss": 1.0051,
+      "step": 1259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7690009483496882,
+      "learning_rate": 1.7721565844991643e-05,
+      "loss": 0.9932,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6744024759552112,
+      "learning_rate": 1.7717601132248234e-05,
+      "loss": 0.9504,
+      "step": 1261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7751298206078183,
+      "learning_rate": 1.771363341731493e-05,
+      "loss": 0.949,
+      "step": 1262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.833265667937307,
+      "learning_rate": 1.7709662701735193e-05,
+      "loss": 1.0175,
+      "step": 1263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7270681819406895,
+      "learning_rate": 1.7705688987053652e-05,
+      "loss": 0.975,
+      "step": 1264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6798932485466451,
+      "learning_rate": 1.7701712274816113e-05,
+      "loss": 0.9937,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7095578408020877,
+      "learning_rate": 1.769773256656953e-05,
+      "loss": 0.9702,
+      "step": 1266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8207148781154288,
+      "learning_rate": 1.7693749863862035e-05,
+      "loss": 0.9651,
+      "step": 1267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.845259681594431,
+      "learning_rate": 1.768976416824292e-05,
+      "loss": 1.0077,
+      "step": 1268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7485953502759874,
+      "learning_rate": 1.768577548126265e-05,
+      "loss": 0.9662,
+      "step": 1269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8264179475383608,
+      "learning_rate": 1.7681783804472836e-05,
+      "loss": 0.95,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7067415453888903,
+      "learning_rate": 1.767778913942627e-05,
+      "loss": 0.9203,
+      "step": 1271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7252838914659816,
+      "learning_rate": 1.7673791487676894e-05,
+      "loss": 0.9391,
+      "step": 1272
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0610984788222169,
+      "learning_rate": 1.7669790850779822e-05,
+      "loss": 0.7934,
+      "step": 1273
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7095293063932833,
+      "learning_rate": 1.766578723029132e-05,
+      "loss": 0.9266,
+      "step": 1274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6715441384185652,
+      "learning_rate": 1.766178062776882e-05,
+      "loss": 0.9304,
+      "step": 1275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.639120335091084,
+      "learning_rate": 1.7657771044770912e-05,
+      "loss": 0.9935,
+      "step": 1276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7966880501531552,
+      "learning_rate": 1.7653758482857348e-05,
+      "loss": 1.0562,
+      "step": 1277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7164304291943553,
+      "learning_rate": 1.764974294358903e-05,
+      "loss": 1.0884,
+      "step": 1278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8441469317600085,
+      "learning_rate": 1.764572442852804e-05,
+      "loss": 0.959,
+      "step": 1279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6335311063786688,
+      "learning_rate": 1.764170293923759e-05,
+      "loss": 0.887,
+      "step": 1280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7677392819497102,
+      "learning_rate": 1.7637678477282065e-05,
+      "loss": 0.9491,
+      "step": 1281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8041857447383878,
+      "learning_rate": 1.7633651044227005e-05,
+      "loss": 0.9792,
+      "step": 1282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6570719994532657,
+      "learning_rate": 1.7629620641639102e-05,
+      "loss": 1.0132,
+      "step": 1283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.721058407235359,
+      "learning_rate": 1.7625587271086207e-05,
+      "loss": 1.0047,
+      "step": 1284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6142108646088418,
+      "learning_rate": 1.762155093413732e-05,
+      "loss": 0.9252,
+      "step": 1285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.748832552144545,
+      "learning_rate": 1.7617511632362603e-05,
+      "loss": 0.9342,
+      "step": 1286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8505294991033425,
+      "learning_rate": 1.7613469367333363e-05,
+      "loss": 0.9765,
+      "step": 1287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7453633452022101,
+      "learning_rate": 1.760942414062207e-05,
+      "loss": 0.9851,
+      "step": 1288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5833112618634952,
+      "learning_rate": 1.7605375953802332e-05,
+      "loss": 0.9169,
+      "step": 1289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5834514685871437,
+      "learning_rate": 1.7601324808448915e-05,
+      "loss": 0.9231,
+      "step": 1290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6817000805206421,
+      "learning_rate": 1.7597270706137747e-05,
+      "loss": 0.9839,
+      "step": 1291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6772971684665223,
+      "learning_rate": 1.759321364844589e-05,
+      "loss": 0.8571,
+      "step": 1292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5621440437562841,
+      "learning_rate": 1.7589153636951567e-05,
+      "loss": 0.9332,
+      "step": 1293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5962926453698725,
+      "learning_rate": 1.7585090673234138e-05,
+      "loss": 0.8703,
+      "step": 1294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5723614016371874,
+      "learning_rate": 1.7581024758874123e-05,
+      "loss": 0.9494,
+      "step": 1295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6347725666167945,
+      "learning_rate": 1.7576955895453184e-05,
+      "loss": 0.8609,
+      "step": 1296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7010657367474895,
+      "learning_rate": 1.757288408455413e-05,
+      "loss": 0.9558,
+      "step": 1297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6446945731877154,
+      "learning_rate": 1.7568809327760922e-05,
+      "loss": 0.8834,
+      "step": 1298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.792591676958605,
+      "learning_rate": 1.756473162665866e-05,
+      "loss": 0.798,
+      "step": 1299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.151768387268869,
+      "learning_rate": 1.7560650982833592e-05,
+      "loss": 0.8523,
+      "step": 1300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7460125410714868,
+      "learning_rate": 1.7556567397873117e-05,
+      "loss": 0.9024,
+      "step": 1301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6342852768637341,
+      "learning_rate": 1.7552480873365765e-05,
+      "loss": 0.9765,
+      "step": 1302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.684615258994111,
+      "learning_rate": 1.754839141090122e-05,
+      "loss": 1.0184,
+      "step": 1303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6818895691958813,
+      "learning_rate": 1.754429901207031e-05,
+      "loss": 0.9283,
+      "step": 1304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7676616619144674,
+      "learning_rate": 1.754020367846499e-05,
+      "loss": 0.9291,
+      "step": 1305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9134734047690318,
+      "learning_rate": 1.753610541167838e-05,
+      "loss": 0.9758,
+      "step": 1306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.946346940186412,
+      "learning_rate": 1.753200421330472e-05,
+      "loss": 0.8958,
+      "step": 1307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6484980905877256,
+      "learning_rate": 1.7527900084939402e-05,
+      "loss": 0.906,
+      "step": 1308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6261114619558765,
+      "learning_rate": 1.7523793028178954e-05,
+      "loss": 0.8721,
+      "step": 1309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7217602467893347,
+      "learning_rate": 1.7519683044621046e-05,
+      "loss": 0.9437,
+      "step": 1310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5590892914165577,
+      "learning_rate": 1.7515570135864485e-05,
+      "loss": 0.9221,
+      "step": 1311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5607965152921113,
+      "learning_rate": 1.7511454303509212e-05,
+      "loss": 0.8696,
+      "step": 1312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8381051267610902,
+      "learning_rate": 1.750733554915631e-05,
+      "loss": 0.9752,
+      "step": 1313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7922865110360966,
+      "learning_rate": 1.7503213874408e-05,
+      "loss": 0.8665,
+      "step": 1314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7390196593419154,
+      "learning_rate": 1.749908928086763e-05,
+      "loss": 1.0602,
+      "step": 1315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7023500109759495,
+      "learning_rate": 1.74949617701397e-05,
+      "loss": 0.9,
+      "step": 1316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.658163097939603,
+      "learning_rate": 1.7490831343829823e-05,
+      "loss": 0.7851,
+      "step": 1317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5807997153264632,
+      "learning_rate": 1.748669800354477e-05,
+      "loss": 0.9329,
+      "step": 1318
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8035234778546025,
+      "learning_rate": 1.7482561750892425e-05,
+      "loss": 0.9496,
+      "step": 1319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.450714734188658,
+      "learning_rate": 1.7478422587481817e-05,
+      "loss": 0.9077,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6983712855319606,
+      "learning_rate": 1.7474280514923097e-05,
+      "loss": 0.8783,
+      "step": 1321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.66200770375001,
+      "learning_rate": 1.7470135534827567e-05,
+      "loss": 0.9625,
+      "step": 1322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6519131679458783,
+      "learning_rate": 1.7465987648807633e-05,
+      "loss": 0.924,
+      "step": 1323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5783591428910106,
+      "learning_rate": 1.7461836858476858e-05,
+      "loss": 0.9293,
+      "step": 1324
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5165609739884254,
+      "learning_rate": 1.7457683165449913e-05,
+      "loss": 0.9601,
+      "step": 1325
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.694066880629218,
+      "learning_rate": 1.7453526571342614e-05,
+      "loss": 1.0174,
+      "step": 1326
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.687023177732187,
+      "learning_rate": 1.7449367077771898e-05,
+      "loss": 0.8966,
+      "step": 1327
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.565604076180921,
+      "learning_rate": 1.7445204686355833e-05,
+      "loss": 0.8819,
+      "step": 1328
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.939897064602898,
+      "learning_rate": 1.744103939871361e-05,
+      "loss": 0.9717,
+      "step": 1329
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6171385859970593,
+      "learning_rate": 1.7436871216465548e-05,
+      "loss": 0.909,
+      "step": 1330
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7086249573623178,
+      "learning_rate": 1.743270014123309e-05,
+      "loss": 0.9276,
+      "step": 1331
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8202933633315734,
+      "learning_rate": 1.742852617463882e-05,
+      "loss": 1.0013,
+      "step": 1332
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.687572656962771,
+      "learning_rate": 1.742434931830642e-05,
+      "loss": 0.889,
+      "step": 1333
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7025577492387172,
+      "learning_rate": 1.742016957386072e-05,
+      "loss": 0.9424,
+      "step": 1334
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7547053048428878,
+      "learning_rate": 1.7415986942927657e-05,
+      "loss": 0.9419,
+      "step": 1335
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7573434554913565,
+      "learning_rate": 1.7411801427134305e-05,
+      "loss": 0.9505,
+      "step": 1336
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7465607882057324,
+      "learning_rate": 1.740761302810885e-05,
+      "loss": 1.0197,
+      "step": 1337
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.065296651052321,
+      "learning_rate": 1.7403421747480597e-05,
+      "loss": 0.7856,
+      "step": 1338
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6740797612291074,
+      "learning_rate": 1.7399227586879986e-05,
+      "loss": 0.9429,
+      "step": 1339
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.681997547710722,
+      "learning_rate": 1.7395030547938567e-05,
+      "loss": 0.8764,
+      "step": 1340
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6536689845879644,
+      "learning_rate": 1.739083063228901e-05,
+      "loss": 0.9271,
+      "step": 1341
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.701903849737158,
+      "learning_rate": 1.7386627841565103e-05,
+      "loss": 1.0073,
+      "step": 1342
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5844710414761023,
+      "learning_rate": 1.7382422177401755e-05,
+      "loss": 0.9231,
+      "step": 1343
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6245159505899607,
+      "learning_rate": 1.7378213641435e-05,
+      "loss": 0.921,
+      "step": 1344
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5884932716931588,
+      "learning_rate": 1.7374002235301974e-05,
+      "loss": 0.9007,
+      "step": 1345
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.755168760894522,
+      "learning_rate": 1.7369787960640945e-05,
+      "loss": 0.9059,
+      "step": 1346
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6355104165072563,
+      "learning_rate": 1.7365570819091282e-05,
+      "loss": 0.9211,
+      "step": 1347
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7186630398671447,
+      "learning_rate": 1.736135081229348e-05,
+      "loss": 0.9482,
+      "step": 1348
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6620935728043695,
+      "learning_rate": 1.7357127941889146e-05,
+      "loss": 0.9321,
+      "step": 1349
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7216215714547267,
+      "learning_rate": 1.7352902209521e-05,
+      "loss": 0.9848,
+      "step": 1350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7153227854125548,
+      "learning_rate": 1.7348673616832874e-05,
+      "loss": 0.9762,
+      "step": 1351
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6550354928018525,
+      "learning_rate": 1.7344442165469714e-05,
+      "loss": 0.98,
+      "step": 1352
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7388866050542833,
+      "learning_rate": 1.734020785707758e-05,
+      "loss": 0.9413,
+      "step": 1353
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6626600488325456,
+      "learning_rate": 1.733597069330364e-05,
+      "loss": 0.9432,
+      "step": 1354
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6121940553974758,
+      "learning_rate": 1.7331730675796175e-05,
+      "loss": 1.0248,
+      "step": 1355
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8863193811308028,
+      "learning_rate": 1.7327487806204574e-05,
+      "loss": 0.9194,
+      "step": 1356
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.143010751486766,
+      "learning_rate": 1.732324208617934e-05,
+      "loss": 0.8708,
+      "step": 1357
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7316138622944446,
+      "learning_rate": 1.7318993517372077e-05,
+      "loss": 0.9101,
+      "step": 1358
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8332797866104589,
+      "learning_rate": 1.7314742101435506e-05,
+      "loss": 0.9385,
+      "step": 1359
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0857696738555198,
+      "learning_rate": 1.7310487840023452e-05,
+      "loss": 0.8742,
+      "step": 1360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7486026869609543,
+      "learning_rate": 1.730623073479085e-05,
+      "loss": 0.9603,
+      "step": 1361
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.712016710790856,
+      "learning_rate": 1.7301970787393726e-05,
+      "loss": 1.0098,
+      "step": 1362
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6509980947692182,
+      "learning_rate": 1.729770799948923e-05,
+      "loss": 0.9048,
+      "step": 1363
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6824944974993126,
+      "learning_rate": 1.7293442372735613e-05,
+      "loss": 0.9626,
+      "step": 1364
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7717063468555012,
+      "learning_rate": 1.728917390879222e-05,
+      "loss": 0.9248,
+      "step": 1365
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6438753412438467,
+      "learning_rate": 1.7284902609319517e-05,
+      "loss": 0.9318,
+      "step": 1366
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8029043347662361,
+      "learning_rate": 1.7280628475979057e-05,
+      "loss": 0.9231,
+      "step": 1367
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7474733470229593,
+      "learning_rate": 1.72763515104335e-05,
+      "loss": 1.0219,
+      "step": 1368
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6665957363511321,
+      "learning_rate": 1.727207171434662e-05,
+      "loss": 0.9412,
+      "step": 1369
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8529390584294096,
+      "learning_rate": 1.7267789089383268e-05,
+      "loss": 0.9895,
+      "step": 1370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8004099592342182,
+      "learning_rate": 1.7263503637209418e-05,
+      "loss": 0.8964,
+      "step": 1371
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7417275319940497,
+      "learning_rate": 1.725921535949213e-05,
+      "loss": 0.9362,
+      "step": 1372
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7305279162936646,
+      "learning_rate": 1.7254924257899573e-05,
+      "loss": 1.0102,
+      "step": 1373
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6404416120622276,
+      "learning_rate": 1.7250630334101e-05,
+      "loss": 0.9562,
+      "step": 1374
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6250543320203037,
+      "learning_rate": 1.7246333589766786e-05,
+      "loss": 0.9175,
+      "step": 1375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7347982919779557,
+      "learning_rate": 1.7242034026568376e-05,
+      "loss": 0.9374,
+      "step": 1376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.827171957637176,
+      "learning_rate": 1.723773164617833e-05,
+      "loss": 0.9697,
+      "step": 1377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6453217695646947,
+      "learning_rate": 1.723342645027029e-05,
+      "loss": 0.9124,
+      "step": 1378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.602806326269613,
+      "learning_rate": 1.7229118440519013e-05,
+      "loss": 1.0038,
+      "step": 1379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.675520425919604,
+      "learning_rate": 1.7224807618600336e-05,
+      "loss": 0.9897,
+      "step": 1380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6734809530308237,
+      "learning_rate": 1.7220493986191184e-05,
+      "loss": 0.9806,
+      "step": 1381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6239506093823162,
+      "learning_rate": 1.7216177544969594e-05,
+      "loss": 0.9851,
+      "step": 1382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7430290905964447,
+      "learning_rate": 1.7211858296614677e-05,
+      "loss": 0.8954,
+      "step": 1383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5960957909978695,
+      "learning_rate": 1.7207536242806656e-05,
+      "loss": 0.8413,
+      "step": 1384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.676634084437962,
+      "learning_rate": 1.7203211385226822e-05,
+      "loss": 0.9327,
+      "step": 1385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5473433223671649,
+      "learning_rate": 1.719888372555758e-05,
+      "loss": 0.891,
+      "step": 1386
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8279633573160905,
+      "learning_rate": 1.7194553265482404e-05,
+      "loss": 0.946,
+      "step": 1387
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7191594302423903,
+      "learning_rate": 1.7190220006685873e-05,
+      "loss": 0.9382,
+      "step": 1388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6916636711725745,
+      "learning_rate": 1.7185883950853645e-05,
+      "loss": 0.9764,
+      "step": 1389
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8953712601914787,
+      "learning_rate": 1.7181545099672476e-05,
+      "loss": 0.9107,
+      "step": 1390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5318001331404214,
+      "learning_rate": 1.7177203454830194e-05,
+      "loss": 0.9538,
+      "step": 1391
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8397516689655071,
+      "learning_rate": 1.7172859018015734e-05,
+      "loss": 0.9645,
+      "step": 1392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8048769235572197,
+      "learning_rate": 1.7168511790919095e-05,
+      "loss": 0.959,
+      "step": 1393
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.853430204630511,
+      "learning_rate": 1.716416177523138e-05,
+      "loss": 1.0354,
+      "step": 1394
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.572479369413313,
+      "learning_rate": 1.7159808972644772e-05,
+      "loss": 1.0001,
+      "step": 1395
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6790760149276092,
+      "learning_rate": 1.7155453384852527e-05,
+      "loss": 0.9809,
+      "step": 1396
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5973042140718492,
+      "learning_rate": 1.7151095013548996e-05,
+      "loss": 1.0299,
+      "step": 1397
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8204067496897727,
+      "learning_rate": 1.7146733860429614e-05,
+      "loss": 0.9435,
+      "step": 1398
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7383247220232247,
+      "learning_rate": 1.7142369927190886e-05,
+      "loss": 1.0098,
+      "step": 1399
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9876749431741878,
+      "learning_rate": 1.7138003215530412e-05,
+      "loss": 0.9639,
+      "step": 1400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6565100277618772,
+      "learning_rate": 1.7133633727146867e-05,
+      "loss": 0.9656,
+      "step": 1401
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6116819193736143,
+      "learning_rate": 1.7129261463740007e-05,
+      "loss": 0.9413,
+      "step": 1402
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.733088891276575,
+      "learning_rate": 1.7124886427010662e-05,
+      "loss": 0.9465,
+      "step": 1403
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.558229014092365,
+      "learning_rate": 1.7120508618660754e-05,
+      "loss": 0.9383,
+      "step": 1404
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7664209425448871,
+      "learning_rate": 1.7116128040393263e-05,
+      "loss": 0.9144,
+      "step": 1405
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7277333482293653,
+      "learning_rate": 1.711174469391227e-05,
+      "loss": 0.916,
+      "step": 1406
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8548716126654878,
+      "learning_rate": 1.710735858092292e-05,
+      "loss": 0.8979,
+      "step": 1407
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6377900041554916,
+      "learning_rate": 1.710296970313143e-05,
+      "loss": 0.9391,
+      "step": 1408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8923707296002565,
+      "learning_rate": 1.7098578062245096e-05,
+      "loss": 0.9882,
+      "step": 1409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7995462053009967,
+      "learning_rate": 1.7094183659972294e-05,
+      "loss": 0.9337,
+      "step": 1410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7599104526369203,
+      "learning_rate": 1.7089786498022476e-05,
+      "loss": 0.9249,
+      "step": 1411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9179389974150218,
+      "learning_rate": 1.7085386578106155e-05,
+      "loss": 0.9443,
+      "step": 1412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6740814222059806,
+      "learning_rate": 1.7080983901934927e-05,
+      "loss": 0.9556,
+      "step": 1413
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6887203864671438,
+      "learning_rate": 1.7076578471221462e-05,
+      "loss": 1.0092,
+      "step": 1414
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5965915080160404,
+      "learning_rate": 1.7072170287679494e-05,
+      "loss": 0.9281,
+      "step": 1415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6272320582586894,
+      "learning_rate": 1.7067759353023828e-05,
+      "loss": 0.9668,
+      "step": 1416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6966463326459986,
+      "learning_rate": 1.7063345668970345e-05,
+      "loss": 0.9126,
+      "step": 1417
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6530271448543825,
+      "learning_rate": 1.705892923723599e-05,
+      "loss": 1.0023,
+      "step": 1418
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.719365409215874,
+      "learning_rate": 1.7054510059538784e-05,
+      "loss": 0.8989,
+      "step": 1419
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1579710316999434,
+      "learning_rate": 1.705008813759781e-05,
+      "loss": 0.8738,
+      "step": 1420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.746637666051279,
+      "learning_rate": 1.7045663473133215e-05,
+      "loss": 0.9863,
+      "step": 1421
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7465100295846994,
+      "learning_rate": 1.704123606786623e-05,
+      "loss": 0.9276,
+      "step": 1422
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.67679519909722,
+      "learning_rate": 1.7036805923519128e-05,
+      "loss": 0.9904,
+      "step": 1423
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.750902856914541,
+      "learning_rate": 1.7032373041815268e-05,
+      "loss": 0.927,
+      "step": 1424
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7418168282265958,
+      "learning_rate": 1.7027937424479063e-05,
+      "loss": 1.0384,
+      "step": 1425
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7730685759688836,
+      "learning_rate": 1.7023499073235994e-05,
+      "loss": 0.9529,
+      "step": 1426
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9178929793678605,
+      "learning_rate": 1.7019057989812602e-05,
+      "loss": 0.9699,
+      "step": 1427
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7072662633791194,
+      "learning_rate": 1.7014614175936495e-05,
+      "loss": 1.0113,
+      "step": 1428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.075764479888501,
+      "learning_rate": 1.7010167633336338e-05,
+      "loss": 0.8318,
+      "step": 1429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7507116856720615,
+      "learning_rate": 1.7005718363741868e-05,
+      "loss": 0.9248,
+      "step": 1430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6901265037861908,
+      "learning_rate": 1.7001266368883868e-05,
+      "loss": 1.0013,
+      "step": 1431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9208989037547692,
+      "learning_rate": 1.6996811650494195e-05,
+      "loss": 0.9004,
+      "step": 1432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6682815884456546,
+      "learning_rate": 1.6992354210305756e-05,
+      "loss": 0.9552,
+      "step": 1433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.632633628473611,
+      "learning_rate": 1.698789405005252e-05,
+      "loss": 0.9425,
+      "step": 1434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.627018650418955,
+      "learning_rate": 1.6983431171469513e-05,
+      "loss": 0.9362,
+      "step": 1435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8415309323980804,
+      "learning_rate": 1.6978965576292824e-05,
+      "loss": 0.8749,
+      "step": 1436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7071204350397233,
+      "learning_rate": 1.6974497266259593e-05,
+      "loss": 0.8779,
+      "step": 1437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.603496790030763,
+      "learning_rate": 1.6970026243108013e-05,
+      "loss": 0.9853,
+      "step": 1438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5768847618227466,
+      "learning_rate": 1.6965552508577346e-05,
+      "loss": 0.9544,
+      "step": 1439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6351071668169876,
+      "learning_rate": 1.6961076064407888e-05,
+      "loss": 0.9107,
+      "step": 1440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7743138659324722,
+      "learning_rate": 1.6956596912341012e-05,
+      "loss": 0.967,
+      "step": 1441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6846171001587962,
+      "learning_rate": 1.6952115054119125e-05,
+      "loss": 1.0468,
+      "step": 1442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7701257031191913,
+      "learning_rate": 1.69476304914857e-05,
+      "loss": 0.9121,
+      "step": 1443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9287482561634184,
+      "learning_rate": 1.6943143226185252e-05,
+      "loss": 0.8934,
+      "step": 1444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6466217949670046,
+      "learning_rate": 1.6938653259963355e-05,
+      "loss": 0.8066,
+      "step": 1445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7313894900577378,
+      "learning_rate": 1.6934160594566634e-05,
+      "loss": 0.9395,
+      "step": 1446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.63399803200492,
+      "learning_rate": 1.692966523174276e-05,
+      "loss": 0.9348,
+      "step": 1447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7930326300631263,
+      "learning_rate": 1.6925167173240445e-05,
+      "loss": 0.9104,
+      "step": 1448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5792669544776021,
+      "learning_rate": 1.692066642080947e-05,
+      "loss": 0.8501,
+      "step": 1449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9512073611154526,
+      "learning_rate": 1.691616297620065e-05,
+      "loss": 0.9133,
+      "step": 1450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7158077127223021,
+      "learning_rate": 1.6911656841165847e-05,
+      "loss": 0.9274,
+      "step": 1451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.603079523782172,
+      "learning_rate": 1.6907148017457973e-05,
+      "loss": 0.9322,
+      "step": 1452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.644797874679676,
+      "learning_rate": 1.6902636506830988e-05,
+      "loss": 0.9688,
+      "step": 1453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7552465905302959,
+      "learning_rate": 1.6898122311039894e-05,
+      "loss": 0.9403,
+      "step": 1454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7662234716249925,
+      "learning_rate": 1.689360543184074e-05,
+      "loss": 0.9915,
+      "step": 1455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.761969423247699,
+      "learning_rate": 1.6889085870990616e-05,
+      "loss": 0.9474,
+      "step": 1456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.780084975185576,
+      "learning_rate": 1.688456363024765e-05,
+      "loss": 0.9822,
+      "step": 1457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7356405868921618,
+      "learning_rate": 1.688003871137103e-05,
+      "loss": 0.9207,
+      "step": 1458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6730033078462045,
+      "learning_rate": 1.6875511116120968e-05,
+      "loss": 0.9898,
+      "step": 1459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5786435504868848,
+      "learning_rate": 1.6870980846258725e-05,
+      "loss": 0.8656,
+      "step": 1460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.605084194891479,
+      "learning_rate": 1.6866447903546603e-05,
+      "loss": 0.9143,
+      "step": 1461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7108913485920274,
+      "learning_rate": 1.686191228974794e-05,
+      "loss": 0.9279,
+      "step": 1462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7572971289549633,
+      "learning_rate": 1.6857374006627116e-05,
+      "loss": 1.0326,
+      "step": 1463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7526278615174398,
+      "learning_rate": 1.6852833055949544e-05,
+      "loss": 0.9208,
+      "step": 1464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7875742092692188,
+      "learning_rate": 1.6848289439481685e-05,
+      "loss": 0.9456,
+      "step": 1465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7499551461132643,
+      "learning_rate": 1.6843743158991032e-05,
+      "loss": 0.9426,
+      "step": 1466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1504249541055573,
+      "learning_rate": 1.683919421624611e-05,
+      "loss": 0.8286,
+      "step": 1467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7151282325157646,
+      "learning_rate": 1.6834642613016478e-05,
+      "loss": 0.9629,
+      "step": 1468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7771040491261743,
+      "learning_rate": 1.6830088351072748e-05,
+      "loss": 0.9117,
+      "step": 1469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6993238899693832,
+      "learning_rate": 1.6825531432186545e-05,
+      "loss": 0.9368,
+      "step": 1470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7879837139280792,
+      "learning_rate": 1.6820971858130533e-05,
+      "loss": 0.9607,
+      "step": 1471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5401983560500536,
+      "learning_rate": 1.681640963067842e-05,
+      "loss": 0.9501,
+      "step": 1472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5319811709794429,
+      "learning_rate": 1.681184475160493e-05,
+      "loss": 0.9347,
+      "step": 1473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5630213342068766,
+      "learning_rate": 1.680727722268583e-05,
+      "loss": 0.9501,
+      "step": 1474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.683734473819289,
+      "learning_rate": 1.680270704569792e-05,
+      "loss": 0.7793,
+      "step": 1475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8266202087418464,
+      "learning_rate": 1.6798134222419014e-05,
+      "loss": 0.9966,
+      "step": 1476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5528051015205264,
+      "learning_rate": 1.679355875462797e-05,
+      "loss": 0.8275,
+      "step": 1477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8914243341908228,
+      "learning_rate": 1.6788980644104677e-05,
+      "loss": 1.0082,
+      "step": 1478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.732799484679875,
+      "learning_rate": 1.6784399892630034e-05,
+      "loss": 1.0337,
+      "step": 1479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5783816942440898,
+      "learning_rate": 1.6779816501985984e-05,
+      "loss": 0.9443,
+      "step": 1480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.804847271214007,
+      "learning_rate": 1.6775230473955494e-05,
+      "loss": 0.9471,
+      "step": 1481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6401944178324206,
+      "learning_rate": 1.6770641810322553e-05,
+      "loss": 0.9046,
+      "step": 1482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.696567947177858,
+      "learning_rate": 1.6766050512872173e-05,
+      "loss": 0.8929,
+      "step": 1483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1565684846121331,
+      "learning_rate": 1.67614565833904e-05,
+      "loss": 0.8652,
+      "step": 1484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.826885768493143,
+      "learning_rate": 1.6756860023664297e-05,
+      "loss": 0.9831,
+      "step": 1485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7627104026742375,
+      "learning_rate": 1.6752260835481948e-05,
+      "loss": 0.9316,
+      "step": 1486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7641332755112662,
+      "learning_rate": 1.6747659020632466e-05,
+      "loss": 0.8461,
+      "step": 1487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7888026098430971,
+      "learning_rate": 1.674305458090598e-05,
+      "loss": 0.9491,
+      "step": 1488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6437330131265935,
+      "learning_rate": 1.6738447518093646e-05,
+      "loss": 0.9629,
+      "step": 1489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6524282201125482,
+      "learning_rate": 1.6733837833987634e-05,
+      "loss": 0.9001,
+      "step": 1490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4956735363656297,
+      "learning_rate": 1.672922553038114e-05,
+      "loss": 0.9098,
+      "step": 1491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8519590080597437,
+      "learning_rate": 1.672461060906837e-05,
+      "loss": 0.876,
+      "step": 1492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7066132774811174,
+      "learning_rate": 1.6719993071844562e-05,
+      "loss": 0.9608,
+      "step": 1493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7105613565543796,
+      "learning_rate": 1.671537292050596e-05,
+      "loss": 0.9411,
+      "step": 1494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6513874782453577,
+      "learning_rate": 1.671075015684983e-05,
+      "loss": 0.9777,
+      "step": 1495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7441156581397124,
+      "learning_rate": 1.6706124782674447e-05,
+      "loss": 0.8624,
+      "step": 1496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8573690986372053,
+      "learning_rate": 1.6701496799779112e-05,
+      "loss": 0.9529,
+      "step": 1497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6183013728401463,
+      "learning_rate": 1.669686620996414e-05,
+      "loss": 0.9973,
+      "step": 1498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5347797872577595,
+      "learning_rate": 1.669223301503085e-05,
+      "loss": 0.9486,
+      "step": 1499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6028668855972816,
+      "learning_rate": 1.6687597216781584e-05,
+      "loss": 0.8145,
+      "step": 1500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6814585945564913,
+      "learning_rate": 1.668295881701969e-05,
+      "loss": 0.9505,
+      "step": 1501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8977009968762184,
+      "learning_rate": 1.667831781754954e-05,
+      "loss": 0.9042,
+      "step": 1502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8711464591936275,
+      "learning_rate": 1.6673674220176496e-05,
+      "loss": 0.9478,
+      "step": 1503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7180539423192296,
+      "learning_rate": 1.666902802670695e-05,
+      "loss": 0.962,
+      "step": 1504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6666626104248825,
+      "learning_rate": 1.66643792389483e-05,
+      "loss": 0.9123,
+      "step": 1505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1973527998514766,
+      "learning_rate": 1.6659727858708947e-05,
+      "loss": 0.856,
+      "step": 1506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5673259814240763,
+      "learning_rate": 1.66550738877983e-05,
+      "loss": 0.9118,
+      "step": 1507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5728628446079507,
+      "learning_rate": 1.6650417328026793e-05,
+      "loss": 0.9139,
+      "step": 1508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5857396432980118,
+      "learning_rate": 1.6645758181205834e-05,
+      "loss": 0.8822,
+      "step": 1509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8890331201978154,
+      "learning_rate": 1.6641096449147875e-05,
+      "loss": 0.9863,
+      "step": 1510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6118515343399076,
+      "learning_rate": 1.6636432133666342e-05,
+      "loss": 0.9455,
+      "step": 1511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6742412673222384,
+      "learning_rate": 1.663176523657569e-05,
+      "loss": 0.9777,
+      "step": 1512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8259098282089554,
+      "learning_rate": 1.6627095759691364e-05,
+      "loss": 1.0421,
+      "step": 1513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7330916535087395,
+      "learning_rate": 1.6622423704829812e-05,
+      "loss": 0.8842,
+      "step": 1514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6424763863303073,
+      "learning_rate": 1.66177490738085e-05,
+      "loss": 0.8998,
+      "step": 1515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7477206896837032,
+      "learning_rate": 1.661307186844588e-05,
+      "loss": 0.9033,
+      "step": 1516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.557691444398473,
+      "learning_rate": 1.660839209056141e-05,
+      "loss": 0.8975,
+      "step": 1517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.689261636764078,
+      "learning_rate": 1.6603709741975545e-05,
+      "loss": 0.9635,
+      "step": 1518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7840501705145897,
+      "learning_rate": 1.6599024824509752e-05,
+      "loss": 0.9215,
+      "step": 1519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9083255044750755,
+      "learning_rate": 1.659433733998649e-05,
+      "loss": 0.9896,
+      "step": 1520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6931027780438765,
+      "learning_rate": 1.6589647290229215e-05,
+      "loss": 0.858,
+      "step": 1521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7778446951035027,
+      "learning_rate": 1.6584954677062382e-05,
+      "loss": 0.9867,
+      "step": 1522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6874904324233353,
+      "learning_rate": 1.6580259502311445e-05,
+      "loss": 1.0151,
+      "step": 1523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8272429728687376,
+      "learning_rate": 1.657556176780285e-05,
+      "loss": 0.9711,
+      "step": 1524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8411467883969832,
+      "learning_rate": 1.6570861475364046e-05,
+      "loss": 0.9206,
+      "step": 1525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8710930737452116,
+      "learning_rate": 1.6566158626823465e-05,
+      "loss": 0.9872,
+      "step": 1526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.720506195988624,
+      "learning_rate": 1.6561453224010554e-05,
+      "loss": 0.9681,
+      "step": 1527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6647183073763208,
+      "learning_rate": 1.655674526875573e-05,
+      "loss": 0.9251,
+      "step": 1528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.577609447670818,
+      "learning_rate": 1.6552034762890422e-05,
+      "loss": 0.9072,
+      "step": 1529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.066399167027944,
+      "learning_rate": 1.6547321708247037e-05,
+      "loss": 0.9379,
+      "step": 1530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.840506676774067,
+      "learning_rate": 1.654260610665898e-05,
+      "loss": 0.9352,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7208292126684335,
+      "learning_rate": 1.653788795996065e-05,
+      "loss": 0.899,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8414566876249059,
+      "learning_rate": 1.6533167269987435e-05,
+      "loss": 0.8995,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8163661936250057,
+      "learning_rate": 1.65284440385757e-05,
+      "loss": 1.0081,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.705322353027036,
+      "learning_rate": 1.6523718267562815e-05,
+      "loss": 0.9674,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7366560844179042,
+      "learning_rate": 1.6518989958787126e-05,
+      "loss": 0.9463,
+      "step": 1536
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1520789836347844,
+      "learning_rate": 1.651425911408798e-05,
+      "loss": 0.8954,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8026482343246513,
+      "learning_rate": 1.6509525735305697e-05,
+      "loss": 0.9674,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8502578015677411,
+      "learning_rate": 1.6504789824281586e-05,
+      "loss": 0.9929,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7705139842133477,
+      "learning_rate": 1.6500051382857944e-05,
+      "loss": 0.8824,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8066747761487192,
+      "learning_rate": 1.6495310412878053e-05,
+      "loss": 0.8852,
+      "step": 1541
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.673141236636469,
+      "learning_rate": 1.649056691618618e-05,
+      "loss": 0.9782,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8277718379488843,
+      "learning_rate": 1.648582089462756e-05,
+      "loss": 0.9895,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.62439404205175,
+      "learning_rate": 1.648107235004843e-05,
+      "loss": 0.9448,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6323693740805412,
+      "learning_rate": 1.6476321284296005e-05,
+      "loss": 0.9419,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8284109095149372,
+      "learning_rate": 1.647156769921847e-05,
+      "loss": 0.8998,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6500951414316578,
+      "learning_rate": 1.6466811596664997e-05,
+      "loss": 0.897,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6636394238778047,
+      "learning_rate": 1.6462052978485733e-05,
+      "loss": 0.991,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6471253017947654,
+      "learning_rate": 1.645729184653181e-05,
+      "loss": 0.9912,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.900480841118098,
+      "learning_rate": 1.645252820265534e-05,
+      "loss": 0.9342,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7274929771579584,
+      "learning_rate": 1.6447762048709397e-05,
+      "loss": 1.0001,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8024044277346902,
+      "learning_rate": 1.644299338654805e-05,
+      "loss": 0.8785,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.650686902566554,
+      "learning_rate": 1.643822221802633e-05,
+      "loss": 0.9584,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5474371202081496,
+      "learning_rate": 1.6433448545000252e-05,
+      "loss": 0.9775,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5597079451380407,
+      "learning_rate": 1.64286723693268e-05,
+      "loss": 0.9947,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6385235106949365,
+      "learning_rate": 1.642389369286393e-05,
+      "loss": 0.8831,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6742026888304702,
+      "learning_rate": 1.6419112517470577e-05,
+      "loss": 0.9836,
+      "step": 1557
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7075133360117316,
+      "learning_rate": 1.6414328845006644e-05,
+      "loss": 0.9422,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5892520663236174,
+      "learning_rate": 1.6409542677333007e-05,
+      "loss": 0.8709,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6384876281252254,
+      "learning_rate": 1.640475401631151e-05,
+      "loss": 1.0045,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.693979144639973,
+      "learning_rate": 1.6399962863804974e-05,
+      "loss": 0.9034,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8320477072069328,
+      "learning_rate": 1.6395169221677176e-05,
+      "loss": 1.0343,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7645628809213245,
+      "learning_rate": 1.6390373091792873e-05,
+      "loss": 1.0095,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6613618239883878,
+      "learning_rate": 1.6385574476017795e-05,
+      "loss": 0.9025,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.870626964940079,
+      "learning_rate": 1.638077337621861e-05,
+      "loss": 0.9294,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6781919217989782,
+      "learning_rate": 1.6375969794262997e-05,
+      "loss": 0.9265,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6732809661254076,
+      "learning_rate": 1.637116373201956e-05,
+      "loss": 0.9116,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0888776005788916,
+      "learning_rate": 1.6366355191357885e-05,
+      "loss": 1.0467,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8650433291396247,
+      "learning_rate": 1.6361544174148527e-05,
+      "loss": 0.9516,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7447193484240076,
+      "learning_rate": 1.6356730682262997e-05,
+      "loss": 0.9328,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.544513865520117,
+      "learning_rate": 1.6351914717573767e-05,
+      "loss": 0.8765,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7101828627501428,
+      "learning_rate": 1.6347096281954277e-05,
+      "loss": 0.9101,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7209199573338974,
+      "learning_rate": 1.6342275377278923e-05,
+      "loss": 0.8928,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.730187436877261,
+      "learning_rate": 1.6337452005423067e-05,
+      "loss": 0.9241,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6976839779004649,
+      "learning_rate": 1.6332626168263025e-05,
+      "loss": 0.8881,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6303874238073626,
+      "learning_rate": 1.6327797867676077e-05,
+      "loss": 0.943,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.670673560161254,
+      "learning_rate": 1.6322967105540457e-05,
+      "loss": 0.9151,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7163060222751576,
+      "learning_rate": 1.6318133883735358e-05,
+      "loss": 0.9639,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.687703952131334,
+      "learning_rate": 1.6313298204140934e-05,
+      "loss": 0.9525,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9942262110876094,
+      "learning_rate": 1.6308460068638287e-05,
+      "loss": 0.955,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8919831342466207,
+      "learning_rate": 1.6303619479109475e-05,
+      "loss": 0.9573,
+      "step": 1581
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.778747910790833,
+      "learning_rate": 1.6298776437437526e-05,
+      "loss": 0.9421,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.745824812138257,
+      "learning_rate": 1.62939309455064e-05,
+      "loss": 0.9841,
+      "step": 1583
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.005077387084947,
+      "learning_rate": 1.6289083005201026e-05,
+      "loss": 0.9857,
+      "step": 1584
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.898514749339766,
+      "learning_rate": 1.6284232618407276e-05,
+      "loss": 0.9036,
+      "step": 1585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5823702179562669,
+      "learning_rate": 1.627937978701198e-05,
+      "loss": 0.9336,
+      "step": 1586
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6580159968617827,
+      "learning_rate": 1.6274524512902912e-05,
+      "loss": 0.9015,
+      "step": 1587
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6328945384243476,
+      "learning_rate": 1.626966679796881e-05,
+      "loss": 0.9527,
+      "step": 1588
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7353326953684673,
+      "learning_rate": 1.6264806644099347e-05,
+      "loss": 0.9191,
+      "step": 1589
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.731152621721883,
+      "learning_rate": 1.6259944053185148e-05,
+      "loss": 0.9689,
+      "step": 1590
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8267882683217032,
+      "learning_rate": 1.625507902711779e-05,
+      "loss": 0.9006,
+      "step": 1591
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.660673153228956,
+      "learning_rate": 1.6250211567789796e-05,
+      "loss": 0.9531,
+      "step": 1592
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7726958038187408,
+      "learning_rate": 1.6245341677094632e-05,
+      "loss": 0.9605,
+      "step": 1593
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5638525823490939,
+      "learning_rate": 1.6240469356926714e-05,
+      "loss": 0.9121,
+      "step": 1594
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.615819492487161,
+      "learning_rate": 1.6235594609181404e-05,
+      "loss": 0.9112,
+      "step": 1595
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1452214964318819,
+      "learning_rate": 1.6230717435755002e-05,
+      "loss": 0.8227,
+      "step": 1596
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1150563856276265,
+      "learning_rate": 1.6225837838544763e-05,
+      "loss": 0.8186,
+      "step": 1597
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8136187860164037,
+      "learning_rate": 1.622095581944887e-05,
+      "loss": 0.8627,
+      "step": 1598
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9577379459272641,
+      "learning_rate": 1.6216071380366453e-05,
+      "loss": 0.9854,
+      "step": 1599
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4458602998328087,
+      "learning_rate": 1.62111845231976e-05,
+      "loss": 0.9476,
+      "step": 1600
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6626009143381728,
+      "learning_rate": 1.6206295249843312e-05,
+      "loss": 0.9418,
+      "step": 1601
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8417180161902174,
+      "learning_rate": 1.6201403562205547e-05,
+      "loss": 0.928,
+      "step": 1602
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6342966301969772,
+      "learning_rate": 1.6196509462187205e-05,
+      "loss": 0.8499,
+      "step": 1603
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4967835565015328,
+      "learning_rate": 1.619161295169211e-05,
+      "loss": 0.8553,
+      "step": 1604
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7224864536848172,
+      "learning_rate": 1.6186714032625036e-05,
+      "loss": 0.925,
+      "step": 1605
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7763197537104958,
+      "learning_rate": 1.618181270689169e-05,
+      "loss": 0.9975,
+      "step": 1606
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.831926739868879,
+      "learning_rate": 1.6176908976398713e-05,
+      "loss": 1.0196,
+      "step": 1607
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4575672638416406,
+      "learning_rate": 1.6172002843053687e-05,
+      "loss": 0.8872,
+      "step": 1608
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8265500625497886,
+      "learning_rate": 1.6167094308765122e-05,
+      "loss": 0.8805,
+      "step": 1609
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6430601567156466,
+      "learning_rate": 1.6162183375442464e-05,
+      "loss": 0.9035,
+      "step": 1610
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4940280767553107,
+      "learning_rate": 1.6157270044996098e-05,
+      "loss": 0.8646,
+      "step": 1611
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7627770422295945,
+      "learning_rate": 1.6152354319337336e-05,
+      "loss": 0.8689,
+      "step": 1612
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6788474884705797,
+      "learning_rate": 1.614743620037842e-05,
+      "loss": 0.8535,
+      "step": 1613
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8638924023543273,
+      "learning_rate": 1.6142515690032525e-05,
+      "loss": 0.9121,
+      "step": 1614
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0612905331097275,
+      "learning_rate": 1.6137592790213765e-05,
+      "loss": 0.9246,
+      "step": 1615
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7216050549850093,
+      "learning_rate": 1.6132667502837164e-05,
+      "loss": 0.9596,
+      "step": 1616
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5634428918355618,
+      "learning_rate": 1.6127739829818695e-05,
+      "loss": 0.8957,
+      "step": 1617
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6959549456380303,
+      "learning_rate": 1.6122809773075247e-05,
+      "loss": 0.752,
+      "step": 1618
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8052961117745467,
+      "learning_rate": 1.6117877334524635e-05,
+      "loss": 0.9041,
+      "step": 1619
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.82138238416882,
+      "learning_rate": 1.6112942516085616e-05,
+      "loss": 0.9431,
+      "step": 1620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8821211929280084,
+      "learning_rate": 1.6108005319677853e-05,
+      "loss": 0.9189,
+      "step": 1621
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6797384699671483,
+      "learning_rate": 1.6103065747221945e-05,
+      "loss": 0.9559,
+      "step": 1622
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6335614363214426,
+      "learning_rate": 1.6098123800639415e-05,
+      "loss": 1.0283,
+      "step": 1623
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.663077182850145,
+      "learning_rate": 1.6093179481852706e-05,
+      "loss": 0.8679,
+      "step": 1624
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.736097411032046,
+      "learning_rate": 1.6088232792785187e-05,
+      "loss": 0.9939,
+      "step": 1625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9331864705161093,
+      "learning_rate": 1.608328373536115e-05,
+      "loss": 0.8924,
+      "step": 1626
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7154518294143721,
+      "learning_rate": 1.60783323115058e-05,
+      "loss": 0.9131,
+      "step": 1627
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6772390452717216,
+      "learning_rate": 1.6073378523145272e-05,
+      "loss": 1.0287,
+      "step": 1628
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7439611127964232,
+      "learning_rate": 1.6068422372206613e-05,
+      "loss": 0.9183,
+      "step": 1629
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.691458239632585,
+      "learning_rate": 1.6063463860617802e-05,
+      "loss": 0.9311,
+      "step": 1630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.658304169586574,
+      "learning_rate": 1.605850299030772e-05,
+      "loss": 0.9134,
+      "step": 1631
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7083337750889018,
+      "learning_rate": 1.6053539763206177e-05,
+      "loss": 0.9304,
+      "step": 1632
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7508484136879259,
+      "learning_rate": 1.6048574181243895e-05,
+      "loss": 0.939,
+      "step": 1633
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7852876914422533,
+      "learning_rate": 1.6043606246352516e-05,
+      "loss": 0.9663,
+      "step": 1634
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6659108647906151,
+      "learning_rate": 1.6038635960464586e-05,
+      "loss": 0.9219,
+      "step": 1635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8060772275287673,
+      "learning_rate": 1.6033663325513583e-05,
+      "loss": 0.9479,
+      "step": 1636
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.894086011906853,
+      "learning_rate": 1.602868834343388e-05,
+      "loss": 0.9187,
+      "step": 1637
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7600387051701272,
+      "learning_rate": 1.602371101616079e-05,
+      "loss": 0.9938,
+      "step": 1638
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6732938933125896,
+      "learning_rate": 1.60187313456305e-05,
+      "loss": 0.941,
+      "step": 1639
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.543451364693145,
+      "learning_rate": 1.6013749333780143e-05,
+      "loss": 0.9488,
+      "step": 1640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6555084246553577,
+      "learning_rate": 1.600876498254775e-05,
+      "loss": 0.96,
+      "step": 1641
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8317018556108886,
+      "learning_rate": 1.600377829387225e-05,
+      "loss": 0.9523,
+      "step": 1642
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.635513167409773,
+      "learning_rate": 1.5998789269693508e-05,
+      "loss": 0.9135,
+      "step": 1643
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6171341568108941,
+      "learning_rate": 1.599379791195227e-05,
+      "loss": 0.8896,
+      "step": 1644
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8946636753822537,
+      "learning_rate": 1.598880422259021e-05,
+      "loss": 1.0003,
+      "step": 1645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.76766590571786,
+      "learning_rate": 1.598380820354989e-05,
+      "loss": 0.9844,
+      "step": 1646
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6685097327074356,
+      "learning_rate": 1.5978809856774807e-05,
+      "loss": 0.9786,
+      "step": 1647
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5918070005843183,
+      "learning_rate": 1.5973809184209328e-05,
+      "loss": 0.8771,
+      "step": 1648
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7023742031839773,
+      "learning_rate": 1.596880618779875e-05,
+      "loss": 0.8801,
+      "step": 1649
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.724381937302016,
+      "learning_rate": 1.5963800869489273e-05,
+      "loss": 0.9386,
+      "step": 1650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8494707991079804,
+      "learning_rate": 1.595879323122798e-05,
+      "loss": 1.0036,
+      "step": 1651
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5226561189767724,
+      "learning_rate": 1.5953783274962883e-05,
+      "loss": 0.8788,
+      "step": 1652
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6318911919928796,
+      "learning_rate": 1.5948771002642874e-05,
+      "loss": 0.9095,
+      "step": 1653
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.912279971229655,
+      "learning_rate": 1.5943756416217758e-05,
+      "loss": 1.0128,
+      "step": 1654
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1650261536863196,
+      "learning_rate": 1.5938739517638236e-05,
+      "loss": 0.8428,
+      "step": 1655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7415643450863434,
+      "learning_rate": 1.5933720308855915e-05,
+      "loss": 0.9211,
+      "step": 1656
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8820348107062908,
+      "learning_rate": 1.5928698791823292e-05,
+      "loss": 0.955,
+      "step": 1657
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6280225538899704,
+      "learning_rate": 1.5923674968493762e-05,
+      "loss": 1.0581,
+      "step": 1658
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6107594870919844,
+      "learning_rate": 1.5918648840821622e-05,
+      "loss": 0.9198,
+      "step": 1659
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9027448285073223,
+      "learning_rate": 1.5913620410762066e-05,
+      "loss": 0.9414,
+      "step": 1660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1187156575317108,
+      "learning_rate": 1.590858968027118e-05,
+      "loss": 0.8779,
+      "step": 1661
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.632013326547582,
+      "learning_rate": 1.5903556651305945e-05,
+      "loss": 0.8975,
+      "step": 1662
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6182938839187448,
+      "learning_rate": 1.5898521325824238e-05,
+      "loss": 0.855,
+      "step": 1663
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7066599421699202,
+      "learning_rate": 1.589348370578483e-05,
+      "loss": 0.9101,
+      "step": 1664
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5337852130580103,
+      "learning_rate": 1.588844379314738e-05,
+      "loss": 0.9258,
+      "step": 1665
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5552796689485406,
+      "learning_rate": 1.5883401589872444e-05,
+      "loss": 0.8547,
+      "step": 1666
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.637891557924487,
+      "learning_rate": 1.587835709792147e-05,
+      "loss": 0.9402,
+      "step": 1667
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6360176706001976,
+      "learning_rate": 1.587331031925679e-05,
+      "loss": 0.8344,
+      "step": 1668
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7108867430309864,
+      "learning_rate": 1.586826125584163e-05,
+      "loss": 0.8892,
+      "step": 1669
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7038609470120825,
+      "learning_rate": 1.5863209909640106e-05,
+      "loss": 0.9929,
+      "step": 1670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5206042599249054,
+      "learning_rate": 1.5858156282617215e-05,
+      "loss": 0.8909,
+      "step": 1671
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6711511627920845,
+      "learning_rate": 1.585310037673885e-05,
+      "loss": 0.8614,
+      "step": 1672
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8427061122981878,
+      "learning_rate": 1.5848042193971788e-05,
+      "loss": 1.0136,
+      "step": 1673
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6340322224497956,
+      "learning_rate": 1.5842981736283686e-05,
+      "loss": 0.977,
+      "step": 1674
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6997279400264118,
+      "learning_rate": 1.583791900564309e-05,
+      "loss": 0.9009,
+      "step": 1675
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7262098742750869,
+      "learning_rate": 1.583285400401944e-05,
+      "loss": 0.9738,
+      "step": 1676
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5265533453619249,
+      "learning_rate": 1.5827786733383033e-05,
+      "loss": 0.855,
+      "step": 1677
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5822635288749958,
+      "learning_rate": 1.5822717195705082e-05,
+      "loss": 0.8655,
+      "step": 1678
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4231646639501143,
+      "learning_rate": 1.581764539295765e-05,
+      "loss": 0.9177,
+      "step": 1679
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6751968342336907,
+      "learning_rate": 1.5812571327113715e-05,
+      "loss": 0.8745,
+      "step": 1680
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6747266634292668,
+      "learning_rate": 1.5807495000147103e-05,
+      "loss": 0.867,
+      "step": 1681
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8696931576923919,
+      "learning_rate": 1.5802416414032537e-05,
+      "loss": 0.9896,
+      "step": 1682
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6926163574834707,
+      "learning_rate": 1.5797335570745614e-05,
+      "loss": 0.868,
+      "step": 1683
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7188418247492503,
+      "learning_rate": 1.5792252472262813e-05,
+      "loss": 0.9944,
+      "step": 1684
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.566218028695576,
+      "learning_rate": 1.578716712056149e-05,
+      "loss": 0.9525,
+      "step": 1685
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.699566762060431,
+      "learning_rate": 1.578207951761987e-05,
+      "loss": 0.9356,
+      "step": 1686
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6877601287806214,
+      "learning_rate": 1.577698966541706e-05,
+      "loss": 1.0246,
+      "step": 1687
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.514033754520559,
+      "learning_rate": 1.577189756593304e-05,
+      "loss": 0.9412,
+      "step": 1688
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8053392098687253,
+      "learning_rate": 1.5766803221148676e-05,
+      "loss": 0.9257,
+      "step": 1689
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.637881254327796,
+      "learning_rate": 1.5761706633045677e-05,
+      "loss": 0.9698,
+      "step": 1690
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4978510106761729,
+      "learning_rate": 1.575660780360666e-05,
+      "loss": 0.8801,
+      "step": 1691
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6878496657903825,
+      "learning_rate": 1.575150673481509e-05,
+      "loss": 0.8644,
+      "step": 1692
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6593397363940259,
+      "learning_rate": 1.5746403428655318e-05,
+      "loss": 1.023,
+      "step": 1693
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6235841003825928,
+      "learning_rate": 1.5741297887112554e-05,
+      "loss": 0.9081,
+      "step": 1694
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8109036039046256,
+      "learning_rate": 1.573619011217288e-05,
+      "loss": 1.008,
+      "step": 1695
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7366906202563615,
+      "learning_rate": 1.5731080105823253e-05,
+      "loss": 1.0069,
+      "step": 1696
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7910543152195322,
+      "learning_rate": 1.572596787005149e-05,
+      "loss": 1.009,
+      "step": 1697
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7770513140362456,
+      "learning_rate": 1.5720853406846284e-05,
+      "loss": 0.9124,
+      "step": 1698
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.663563980279973,
+      "learning_rate": 1.571573671819719e-05,
+      "loss": 0.9501,
+      "step": 1699
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.798302917814401,
+      "learning_rate": 1.5710617806094623e-05,
+      "loss": 1.0329,
+      "step": 1700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.591105191580427,
+      "learning_rate": 1.5705496672529875e-05,
+      "loss": 0.9011,
+      "step": 1701
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6919167077809922,
+      "learning_rate": 1.5700373319495087e-05,
+      "loss": 0.954,
+      "step": 1702
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6001732503077335,
+      "learning_rate": 1.5695247748983277e-05,
+      "loss": 0.7931,
+      "step": 1703
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7339939929632062,
+      "learning_rate": 1.5690119962988325e-05,
+      "loss": 0.915,
+      "step": 1704
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.548451551637994,
+      "learning_rate": 1.568498996350496e-05,
+      "loss": 0.9463,
+      "step": 1705
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.748437218428189,
+      "learning_rate": 1.5679857752528788e-05,
+      "loss": 0.9676,
+      "step": 1706
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6471556654707846,
+      "learning_rate": 1.567472333205626e-05,
+      "loss": 0.926,
+      "step": 1707
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6877746907602753,
+      "learning_rate": 1.5669586704084697e-05,
+      "loss": 1.0535,
+      "step": 1708
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7356876446324196,
+      "learning_rate": 1.566444787061228e-05,
+      "loss": 0.9476,
+      "step": 1709
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6872419025030534,
+      "learning_rate": 1.565930683363804e-05,
+      "loss": 0.9989,
+      "step": 1710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0437985973600723,
+      "learning_rate": 1.565416359516187e-05,
+      "loss": 0.8111,
+      "step": 1711
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0088547902055691,
+      "learning_rate": 1.564901815718452e-05,
+      "loss": 0.8493,
+      "step": 1712
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8308929603793904,
+      "learning_rate": 1.5643870521707588e-05,
+      "loss": 0.9714,
+      "step": 1713
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6678011598396532,
+      "learning_rate": 1.5638720690733543e-05,
+      "loss": 0.8796,
+      "step": 1714
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6676069738231614,
+      "learning_rate": 1.563356866626569e-05,
+      "loss": 0.9325,
+      "step": 1715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.687942802391812,
+      "learning_rate": 1.56284144503082e-05,
+      "loss": 0.879,
+      "step": 1716
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5751332093640165,
+      "learning_rate": 1.5623258044866087e-05,
+      "loss": 0.9392,
+      "step": 1717
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5575620694150407,
+      "learning_rate": 1.561809945194522e-05,
+      "loss": 0.9539,
+      "step": 1718
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6785024493599945,
+      "learning_rate": 1.561293867355233e-05,
+      "loss": 0.8994,
+      "step": 1719
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5229589448071077,
+      "learning_rate": 1.560777571169498e-05,
+      "loss": 0.9325,
+      "step": 1720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.70296765640194,
+      "learning_rate": 1.5602610568381594e-05,
+      "loss": 0.9706,
+      "step": 1721
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6492987597197222,
+      "learning_rate": 1.5597443245621444e-05,
+      "loss": 0.9846,
+      "step": 1722
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5539259406249721,
+      "learning_rate": 1.5592273745424643e-05,
+      "loss": 0.8847,
+      "step": 1723
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9574877365730032,
+      "learning_rate": 1.5587102069802156e-05,
+      "loss": 0.9819,
+      "step": 1724
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.059334241300379,
+      "learning_rate": 1.5581928220765792e-05,
+      "loss": 0.8633,
+      "step": 1725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8777654599053444,
+      "learning_rate": 1.5576752200328217e-05,
+      "loss": 0.9534,
+      "step": 1726
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6369275243616173,
+      "learning_rate": 1.5571574010502925e-05,
+      "loss": 0.9465,
+      "step": 1727
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.956314276556218,
+      "learning_rate": 1.5566393653304255e-05,
+      "loss": 0.9679,
+      "step": 1728
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6159927205594145,
+      "learning_rate": 1.5561211130747405e-05,
+      "loss": 0.892,
+      "step": 1729
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7028870865116463,
+      "learning_rate": 1.5556026444848398e-05,
+      "loss": 0.9115,
+      "step": 1730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8208461469027895,
+      "learning_rate": 1.5550839597624106e-05,
+      "loss": 0.8889,
+      "step": 1731
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.91661243257076,
+      "learning_rate": 1.5545650591092248e-05,
+      "loss": 0.97,
+      "step": 1732
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.60671095140956,
+      "learning_rate": 1.5540459427271373e-05,
+      "loss": 0.9894,
+      "step": 1733
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6111175098454973,
+      "learning_rate": 1.553526610818087e-05,
+      "loss": 0.9202,
+      "step": 1734
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7534583643676716,
+      "learning_rate": 1.553007063584097e-05,
+      "loss": 0.8502,
+      "step": 1735
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6062441383768804,
+      "learning_rate": 1.5524873012272742e-05,
+      "loss": 1.0042,
+      "step": 1736
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5342773548854218,
+      "learning_rate": 1.551967323949809e-05,
+      "loss": 0.867,
+      "step": 1737
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6482540145045306,
+      "learning_rate": 1.551447131953976e-05,
+      "loss": 0.9283,
+      "step": 1738
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.616014640944003,
+      "learning_rate": 1.550926725442132e-05,
+      "loss": 0.8996,
+      "step": 1739
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.514494552942554,
+      "learning_rate": 1.550406104616718e-05,
+      "loss": 0.9248,
+      "step": 1740
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5363529361022905,
+      "learning_rate": 1.5498852696802595e-05,
+      "loss": 0.948,
+      "step": 1741
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6602545790223517,
+      "learning_rate": 1.5493642208353627e-05,
+      "loss": 0.9086,
+      "step": 1742
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.585488148395561,
+      "learning_rate": 1.5488429582847194e-05,
+      "loss": 0.9484,
+      "step": 1743
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7637723283346054,
+      "learning_rate": 1.5483214822311036e-05,
+      "loss": 0.9477,
+      "step": 1744
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7728810975772227,
+      "learning_rate": 1.5477997928773722e-05,
+      "loss": 0.8948,
+      "step": 1745
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.627418598739184,
+      "learning_rate": 1.5472778904264655e-05,
+      "loss": 0.9482,
+      "step": 1746
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.737617924919354,
+      "learning_rate": 1.546755775081406e-05,
+      "loss": 0.9492,
+      "step": 1747
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6996510871147033,
+      "learning_rate": 1.5462334470453e-05,
+      "loss": 0.806,
+      "step": 1748
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.749003030553493,
+      "learning_rate": 1.5457109065213357e-05,
+      "loss": 0.8445,
+      "step": 1749
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6613913869959902,
+      "learning_rate": 1.5451881537127846e-05,
+      "loss": 0.8682,
+      "step": 1750
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8209928107994644,
+      "learning_rate": 1.5446651888230002e-05,
+      "loss": 1.0161,
+      "step": 1751
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7742834895232118,
+      "learning_rate": 1.5441420120554183e-05,
+      "loss": 0.9762,
+      "step": 1752
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4903916353227185,
+      "learning_rate": 1.5436186236135586e-05,
+      "loss": 0.8358,
+      "step": 1753
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6523896850698925,
+      "learning_rate": 1.5430950237010215e-05,
+      "loss": 0.918,
+      "step": 1754
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.827097756747899,
+      "learning_rate": 1.5425712125214905e-05,
+      "loss": 0.9376,
+      "step": 1755
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6366236729315444,
+      "learning_rate": 1.542047190278731e-05,
+      "loss": 1.0006,
+      "step": 1756
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0524696605763608,
+      "learning_rate": 1.54152295717659e-05,
+      "loss": 0.8466,
+      "step": 1757
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.592256955879349,
+      "learning_rate": 1.5409985134189984e-05,
+      "loss": 0.8885,
+      "step": 1758
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5958006272826617,
+      "learning_rate": 1.5404738592099665e-05,
+      "loss": 0.914,
+      "step": 1759
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7383735027873934,
+      "learning_rate": 1.5399489947535884e-05,
+      "loss": 0.9507,
+      "step": 1760
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7829696954430776,
+      "learning_rate": 1.539423920254039e-05,
+      "loss": 0.9411,
+      "step": 1761
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.591157229474544,
+      "learning_rate": 1.538898635915576e-05,
+      "loss": 0.9009,
+      "step": 1762
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8961964721472366,
+      "learning_rate": 1.5383731419425363e-05,
+      "loss": 0.987,
+      "step": 1763
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7139184167666317,
+      "learning_rate": 1.5378474385393412e-05,
+      "loss": 0.9263,
+      "step": 1764
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5793588095105051,
+      "learning_rate": 1.537321525910492e-05,
+      "loss": 0.8779,
+      "step": 1765
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.806016217319507,
+      "learning_rate": 1.536795404260572e-05,
+      "loss": 0.9229,
+      "step": 1766
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.756089856756367,
+      "learning_rate": 1.5362690737942446e-05,
+      "loss": 0.8605,
+      "step": 1767
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8572118885860662,
+      "learning_rate": 1.5357425347162564e-05,
+      "loss": 0.9234,
+      "step": 1768
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7856241806716384,
+      "learning_rate": 1.5352157872314327e-05,
+      "loss": 1.0146,
+      "step": 1769
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.610512365395507,
+      "learning_rate": 1.5346888315446824e-05,
+      "loss": 0.873,
+      "step": 1770
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5782211017109056,
+      "learning_rate": 1.5341616678609935e-05,
+      "loss": 0.8567,
+      "step": 1771
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.654570207794557,
+      "learning_rate": 1.533634296385436e-05,
+      "loss": 0.8263,
+      "step": 1772
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.393170469294069,
+      "learning_rate": 1.5331067173231603e-05,
+      "loss": 0.882,
+      "step": 1773
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5878170235295708,
+      "learning_rate": 1.5325789308793974e-05,
+      "loss": 0.9394,
+      "step": 1774
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7372761113657915,
+      "learning_rate": 1.5320509372594596e-05,
+      "loss": 0.9523,
+      "step": 1775
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7300471659324526,
+      "learning_rate": 1.5315227366687386e-05,
+      "loss": 1.0319,
+      "step": 1776
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9409527939786382,
+      "learning_rate": 1.5309943293127076e-05,
+      "loss": 1.0208,
+      "step": 1777
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5078314012936311,
+      "learning_rate": 1.5304657153969206e-05,
+      "loss": 0.9007,
+      "step": 1778
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6940177485783197,
+      "learning_rate": 1.529936895127011e-05,
+      "loss": 0.9291,
+      "step": 1779
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6538468073075518,
+      "learning_rate": 1.5294078687086926e-05,
+      "loss": 0.8828,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6245630899167927,
+      "learning_rate": 1.5288786363477603e-05,
+      "loss": 0.9154,
+      "step": 1781
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.644068790077352,
+      "learning_rate": 1.5283491982500874e-05,
+      "loss": 0.9155,
+      "step": 1782
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7621859790247505,
+      "learning_rate": 1.5278195546216293e-05,
+      "loss": 0.9127,
+      "step": 1783
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6412691064740812,
+      "learning_rate": 1.5272897056684196e-05,
+      "loss": 0.9642,
+      "step": 1784
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7284551662658276,
+      "learning_rate": 1.526759651596573e-05,
+      "loss": 0.9682,
+      "step": 1785
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7534777399859998,
+      "learning_rate": 1.526229392612283e-05,
+      "loss": 0.9304,
+      "step": 1786
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.610391858816259,
+      "learning_rate": 1.5256989289218237e-05,
+      "loss": 0.9148,
+      "step": 1787
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8330198235703115,
+      "learning_rate": 1.5251682607315485e-05,
+      "loss": 0.9917,
+      "step": 1788
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6130949199706817,
+      "learning_rate": 1.5246373882478899e-05,
+      "loss": 0.8627,
+      "step": 1789
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6917099405793583,
+      "learning_rate": 1.5241063116773606e-05,
+      "loss": 0.8093,
+      "step": 1790
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.611514855787027,
+      "learning_rate": 1.5235750312265522e-05,
+      "loss": 0.8823,
+      "step": 1791
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.508488536223312,
+      "learning_rate": 1.5230435471021356e-05,
+      "loss": 0.8525,
+      "step": 1792
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8488504173801166,
+      "learning_rate": 1.5225118595108615e-05,
+      "loss": 0.9723,
+      "step": 1793
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.81787454238434,
+      "learning_rate": 1.5219799686595588e-05,
+      "loss": 0.9186,
+      "step": 1794
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.835940653590753,
+      "learning_rate": 1.5214478747551367e-05,
+      "loss": 1.0315,
+      "step": 1795
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7183650877031766,
+      "learning_rate": 1.520915578004582e-05,
+      "loss": 0.9409,
+      "step": 1796
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6105105683049763,
+      "learning_rate": 1.5203830786149615e-05,
+      "loss": 0.9783,
+      "step": 1797
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6828073709950533,
+      "learning_rate": 1.51985037679342e-05,
+      "loss": 0.95,
+      "step": 1798
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7659646765890635,
+      "learning_rate": 1.5193174727471822e-05,
+      "loss": 0.9881,
+      "step": 1799
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7186747858397031,
+      "learning_rate": 1.5187843666835502e-05,
+      "loss": 0.9142,
+      "step": 1800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5025090660678702,
+      "learning_rate": 1.5182510588099058e-05,
+      "loss": 0.8743,
+      "step": 1801
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7934959998223838,
+      "learning_rate": 1.5177175493337077e-05,
+      "loss": 0.9239,
+      "step": 1802
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5578282419512874,
+      "learning_rate": 1.5171838384624952e-05,
+      "loss": 0.869,
+      "step": 1803
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7377002208663477,
+      "learning_rate": 1.516649926403884e-05,
+      "loss": 0.968,
+      "step": 1804
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.775042883902276,
+      "learning_rate": 1.5161158133655694e-05,
+      "loss": 0.9978,
+      "step": 1805
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5911271575560177,
+      "learning_rate": 1.5155814995553239e-05,
+      "loss": 0.8461,
+      "step": 1806
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9003801818751707,
+      "learning_rate": 1.515046985180999e-05,
+      "loss": 0.9565,
+      "step": 1807
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7235091223002925,
+      "learning_rate": 1.5145122704505238e-05,
+      "loss": 0.9244,
+      "step": 1808
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5836993320305877,
+      "learning_rate": 1.5139773555719046e-05,
+      "loss": 0.9256,
+      "step": 1809
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7698468539495165,
+      "learning_rate": 1.5134422407532268e-05,
+      "loss": 0.9067,
+      "step": 1810
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6257859529986807,
+      "learning_rate": 1.512906926202653e-05,
+      "loss": 0.879,
+      "step": 1811
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6883738442738565,
+      "learning_rate": 1.512371412128424e-05,
+      "loss": 0.9165,
+      "step": 1812
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.641267657042082,
+      "learning_rate": 1.5118356987388567e-05,
+      "loss": 0.9208,
+      "step": 1813
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5234773704787912,
+      "learning_rate": 1.5112997862423472e-05,
+      "loss": 0.939,
+      "step": 1814
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.699284171054198,
+      "learning_rate": 1.5107636748473687e-05,
+      "loss": 0.927,
+      "step": 1815
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7766671571230526,
+      "learning_rate": 1.5102273647624714e-05,
+      "loss": 0.9559,
+      "step": 1816
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8071128001622048,
+      "learning_rate": 1.5096908561962824e-05,
+      "loss": 0.9329,
+      "step": 1817
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5602195899860873,
+      "learning_rate": 1.509154149357507e-05,
+      "loss": 0.8408,
+      "step": 1818
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6681745264243264,
+      "learning_rate": 1.5086172444549274e-05,
+      "loss": 0.9715,
+      "step": 1819
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8299212257238782,
+      "learning_rate": 1.508080141697402e-05,
+      "loss": 0.9271,
+      "step": 1820
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6199015194976085,
+      "learning_rate": 1.5075428412938672e-05,
+      "loss": 0.8942,
+      "step": 1821
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.637504821627329,
+      "learning_rate": 1.5070053434533351e-05,
+      "loss": 0.9063,
+      "step": 1822
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.676609757359368,
+      "learning_rate": 1.5064676483848968e-05,
+      "loss": 0.8435,
+      "step": 1823
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.759464544373338,
+      "learning_rate": 1.5059297562977174e-05,
+      "loss": 0.9497,
+      "step": 1824
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8044880904814558,
+      "learning_rate": 1.5053916674010403e-05,
+      "loss": 0.921,
+      "step": 1825
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6629588750577138,
+      "learning_rate": 1.5048533819041853e-05,
+      "loss": 0.9421,
+      "step": 1826
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0351132912201115,
+      "learning_rate": 1.5043149000165482e-05,
+      "loss": 0.9275,
+      "step": 1827
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7221826608550477,
+      "learning_rate": 1.5037762219476016e-05,
+      "loss": 0.9359,
+      "step": 1828
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8863983035535,
+      "learning_rate": 1.503237347906894e-05,
+      "loss": 0.9219,
+      "step": 1829
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7877241947739209,
+      "learning_rate": 1.5026982781040511e-05,
+      "loss": 0.9239,
+      "step": 1830
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7559006781624251,
+      "learning_rate": 1.5021590127487731e-05,
+      "loss": 0.9135,
+      "step": 1831
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7628850458243945,
+      "learning_rate": 1.5016195520508383e-05,
+      "loss": 1.0021,
+      "step": 1832
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.712262152372984,
+      "learning_rate": 1.501079896220099e-05,
+      "loss": 0.9462,
+      "step": 1833
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5312995616872846,
+      "learning_rate": 1.5005400454664847e-05,
+      "loss": 0.8234,
+      "step": 1834
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4310758854139343,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.8858,
+      "step": 1835
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.546597318526974,
+      "learning_rate": 1.4994597600307263e-05,
+      "loss": 0.9568,
+      "step": 1836
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.582830465670841,
+      "learning_rate": 1.4989193257688195e-05,
+      "loss": 0.9045,
+      "step": 1837
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6151541196243508,
+      "learning_rate": 1.4983786974245118e-05,
+      "loss": 0.9019,
+      "step": 1838
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6357268555554538,
+      "learning_rate": 1.4978378752081105e-05,
+      "loss": 0.9621,
+      "step": 1839
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7828133609360786,
+      "learning_rate": 1.497296859329998e-05,
+      "loss": 0.8937,
+      "step": 1840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6209326547083096,
+      "learning_rate": 1.4967556500006336e-05,
+      "loss": 0.9472,
+      "step": 1841
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6715487795705135,
+      "learning_rate": 1.4962142474305496e-05,
+      "loss": 0.8342,
+      "step": 1842
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.644054514212963,
+      "learning_rate": 1.4956726518303553e-05,
+      "loss": 0.8793,
+      "step": 1843
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6778795735018404,
+      "learning_rate": 1.495130863410734e-05,
+      "loss": 0.9961,
+      "step": 1844
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7530107949989808,
+      "learning_rate": 1.4945888823824449e-05,
+      "loss": 0.8523,
+      "step": 1845
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.600308338620861,
+      "learning_rate": 1.494046708956321e-05,
+      "loss": 0.8763,
+      "step": 1846
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7748182051211745,
+      "learning_rate": 1.4935043433432715e-05,
+      "loss": 0.9168,
+      "step": 1847
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6428997560126952,
+      "learning_rate": 1.4929617857542793e-05,
+      "loss": 0.8927,
+      "step": 1848
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7377574670461893,
+      "learning_rate": 1.4924190364004023e-05,
+      "loss": 0.869,
+      "step": 1849
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7021319792833238,
+      "learning_rate": 1.4918760954927729e-05,
+      "loss": 0.9108,
+      "step": 1850
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.591116905628271,
+      "learning_rate": 1.4913329632425984e-05,
+      "loss": 0.949,
+      "step": 1851
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6276173765055855,
+      "learning_rate": 1.4907896398611603e-05,
+      "loss": 0.89,
+      "step": 1852
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8225115626463677,
+      "learning_rate": 1.4902461255598141e-05,
+      "loss": 0.9123,
+      "step": 1853
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5498562183799023,
+      "learning_rate": 1.4897024205499902e-05,
+      "loss": 0.8801,
+      "step": 1854
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7255906804323513,
+      "learning_rate": 1.4891585250431929e-05,
+      "loss": 0.88,
+      "step": 1855
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7392433916238563,
+      "learning_rate": 1.4886144392510004e-05,
+      "loss": 0.9908,
+      "step": 1856
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.819892755574117,
+      "learning_rate": 1.4880701633850652e-05,
+      "loss": 0.9699,
+      "step": 1857
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6453268799227307,
+      "learning_rate": 1.4875256976571135e-05,
+      "loss": 0.9985,
+      "step": 1858
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7570555784318829,
+      "learning_rate": 1.4869810422789459e-05,
+      "loss": 0.9425,
+      "step": 1859
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6103885798432909,
+      "learning_rate": 1.4864361974624357e-05,
+      "loss": 0.9111,
+      "step": 1860
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6392043539397223,
+      "learning_rate": 1.4858911634195312e-05,
+      "loss": 0.9116,
+      "step": 1861
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6241801480289666,
+      "learning_rate": 1.4853459403622535e-05,
+      "loss": 0.9424,
+      "step": 1862
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.634986046998513,
+      "learning_rate": 1.4848005285026976e-05,
+      "loss": 0.9494,
+      "step": 1863
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5937895894975387,
+      "learning_rate": 1.4842549280530312e-05,
+      "loss": 0.877,
+      "step": 1864
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.572921397651084,
+      "learning_rate": 1.4837091392254964e-05,
+      "loss": 0.8937,
+      "step": 1865
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6988453094864369,
+      "learning_rate": 1.4831631622324078e-05,
+      "loss": 0.8695,
+      "step": 1866
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5134644774277317,
+      "learning_rate": 1.4826169972861539e-05,
+      "loss": 0.8927,
+      "step": 1867
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.627315805696269,
+      "learning_rate": 1.4820706445991954e-05,
+      "loss": 0.9352,
+      "step": 1868
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6393371999448805,
+      "learning_rate": 1.4815241043840669e-05,
+      "loss": 0.8528,
+      "step": 1869
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8952895153405638,
+      "learning_rate": 1.4809773768533757e-05,
+      "loss": 0.9755,
+      "step": 1870
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6091945516084094,
+      "learning_rate": 1.4804304622198013e-05,
+      "loss": 0.8705,
+      "step": 1871
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.624495437307758,
+      "learning_rate": 1.4798833606960972e-05,
+      "loss": 0.8524,
+      "step": 1872
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8744172065380642,
+      "learning_rate": 1.4793360724950887e-05,
+      "loss": 1.0719,
+      "step": 1873
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5575380698894332,
+      "learning_rate": 1.4787885978296737e-05,
+      "loss": 0.9028,
+      "step": 1874
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7892688635911647,
+      "learning_rate": 1.478240936912823e-05,
+      "loss": 0.9931,
+      "step": 1875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6951433638470188,
+      "learning_rate": 1.4776930899575801e-05,
+      "loss": 0.9601,
+      "step": 1876
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8511925621033531,
+      "learning_rate": 1.4771450571770603e-05,
+      "loss": 1.0311,
+      "step": 1877
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1442810135716395,
+      "learning_rate": 1.4765968387844516e-05,
+      "loss": 0.884,
+      "step": 1878
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9225596331812123,
+      "learning_rate": 1.4760484349930134e-05,
+      "loss": 0.8523,
+      "step": 1879
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.808853808526827,
+      "learning_rate": 1.4754998460160784e-05,
+      "loss": 0.9429,
+      "step": 1880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5600398733677634,
+      "learning_rate": 1.4749510720670506e-05,
+      "loss": 0.9736,
+      "step": 1881
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5320412874387632,
+      "learning_rate": 1.4744021133594059e-05,
+      "loss": 0.8701,
+      "step": 1882
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.523014076301379,
+      "learning_rate": 1.4738529701066928e-05,
+      "loss": 0.8883,
+      "step": 1883
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7505097977570019,
+      "learning_rate": 1.4733036425225306e-05,
+      "loss": 0.9172,
+      "step": 1884
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5554959069631156,
+      "learning_rate": 1.4727541308206114e-05,
+      "loss": 0.8644,
+      "step": 1885
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7958310637324932,
+      "learning_rate": 1.4722044352146978e-05,
+      "loss": 0.866,
+      "step": 1886
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.855213005785535,
+      "learning_rate": 1.4716545559186244e-05,
+      "loss": 0.8861,
+      "step": 1887
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.660788717810155,
+      "learning_rate": 1.4711044931462976e-05,
+      "loss": 0.888,
+      "step": 1888
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5748794402819912,
+      "learning_rate": 1.4705542471116949e-05,
+      "loss": 0.9492,
+      "step": 1889
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.671940289814532,
+      "learning_rate": 1.470003818028865e-05,
+      "loss": 0.9392,
+      "step": 1890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6682541425119886,
+      "learning_rate": 1.4694532061119277e-05,
+      "loss": 0.9954,
+      "step": 1891
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6270272723817845,
+      "learning_rate": 1.4689024115750745e-05,
+      "loss": 0.9257,
+      "step": 1892
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6944012589743433,
+      "learning_rate": 1.4683514346325671e-05,
+      "loss": 0.8054,
+      "step": 1893
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7534054491921223,
+      "learning_rate": 1.4678002754987388e-05,
+      "loss": 0.9034,
+      "step": 1894
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.658585836806205,
+      "learning_rate": 1.4672489343879939e-05,
+      "loss": 0.9457,
+      "step": 1895
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.640884368651283,
+      "learning_rate": 1.4666974115148065e-05,
+      "loss": 0.8148,
+      "step": 1896
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5939722795602012,
+      "learning_rate": 1.4661457070937226e-05,
+      "loss": 0.8452,
+      "step": 1897
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6167631945289578,
+      "learning_rate": 1.4655938213393582e-05,
+      "loss": 0.9033,
+      "step": 1898
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9454139532441512,
+      "learning_rate": 1.4650417544663997e-05,
+      "loss": 0.9275,
+      "step": 1899
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5858648706624496,
+      "learning_rate": 1.4644895066896046e-05,
+      "loss": 0.9054,
+      "step": 1900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6936034257279007,
+      "learning_rate": 1.4639370782238e-05,
+      "loss": 0.9825,
+      "step": 1901
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6703890968967032,
+      "learning_rate": 1.4633844692838837e-05,
+      "loss": 0.9524,
+      "step": 1902
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5551617347389055,
+      "learning_rate": 1.4628316800848241e-05,
+      "loss": 0.9375,
+      "step": 1903
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8783960822922252,
+      "learning_rate": 1.4622787108416585e-05,
+      "loss": 1.0336,
+      "step": 1904
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7771071816180781,
+      "learning_rate": 1.4617255617694957e-05,
+      "loss": 0.8875,
+      "step": 1905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6400552487600897,
+      "learning_rate": 1.4611722330835133e-05,
+      "loss": 0.9945,
+      "step": 1906
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7424809111938788,
+      "learning_rate": 1.4606187249989598e-05,
+      "loss": 0.947,
+      "step": 1907
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6042784586765482,
+      "learning_rate": 1.4600650377311523e-05,
+      "loss": 0.9005,
+      "step": 1908
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6758776492466088,
+      "learning_rate": 1.4595111714954787e-05,
+      "loss": 0.9794,
+      "step": 1909
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5913304752712094,
+      "learning_rate": 1.4589571265073959e-05,
+      "loss": 0.7905,
+      "step": 1910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7798933897483598,
+      "learning_rate": 1.4584029029824305e-05,
+      "loss": 0.9255,
+      "step": 1911
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6437473812141559,
+      "learning_rate": 1.4578485011361783e-05,
+      "loss": 0.8905,
+      "step": 1912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.701664671213589,
+      "learning_rate": 1.4572939211843051e-05,
+      "loss": 0.9686,
+      "step": 1913
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.611101839884725,
+      "learning_rate": 1.4567391633425455e-05,
+      "loss": 0.9226,
+      "step": 1914
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4477358561753013,
+      "learning_rate": 1.4561842278267033e-05,
+      "loss": 0.956,
+      "step": 1915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5313991293553062,
+      "learning_rate": 1.4556291148526516e-05,
+      "loss": 0.8799,
+      "step": 1916
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.570437900284019,
+      "learning_rate": 1.4550738246363322e-05,
+      "loss": 0.8546,
+      "step": 1917
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6393386291966954,
+      "learning_rate": 1.4545183573937566e-05,
+      "loss": 0.8923,
+      "step": 1918
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7235853278804611,
+      "learning_rate": 1.4539627133410042e-05,
+      "loss": 0.8886,
+      "step": 1919
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6947056773766112,
+      "learning_rate": 1.4534068926942235e-05,
+      "loss": 0.8453,
+      "step": 1920
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5835750935087545,
+      "learning_rate": 1.4528508956696326e-05,
+      "loss": 0.9339,
+      "step": 1921
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.742088241485645,
+      "learning_rate": 1.4522947224835165e-05,
+      "loss": 0.9946,
+      "step": 1922
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4949963929651728,
+      "learning_rate": 1.4517383733522304e-05,
+      "loss": 0.927,
+      "step": 1923
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5253164901288647,
+      "learning_rate": 1.451181848492197e-05,
+      "loss": 0.8769,
+      "step": 1924
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4889989022162844,
+      "learning_rate": 1.4506251481199071e-05,
+      "loss": 0.872,
+      "step": 1925
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8096496107868196,
+      "learning_rate": 1.450068272451921e-05,
+      "loss": 0.9849,
+      "step": 1926
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6165599569378286,
+      "learning_rate": 1.449511221704866e-05,
+      "loss": 0.8824,
+      "step": 1927
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.696888487069437,
+      "learning_rate": 1.4489539960954382e-05,
+      "loss": 0.9493,
+      "step": 1928
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5617807092050424,
+      "learning_rate": 1.4483965958404012e-05,
+      "loss": 0.9603,
+      "step": 1929
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7772163500967342,
+      "learning_rate": 1.447839021156587e-05,
+      "loss": 0.8918,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.060365842099134,
+      "learning_rate": 1.4472812722608948e-05,
+      "loss": 0.8678,
+      "step": 1931
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6120849117340765,
+      "learning_rate": 1.446723349370293e-05,
+      "loss": 0.9337,
+      "step": 1932
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0949366371155416,
+      "learning_rate": 1.4461652527018158e-05,
+      "loss": 0.8726,
+      "step": 1933
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6705665592860641,
+      "learning_rate": 1.4456069824725663e-05,
+      "loss": 0.9106,
+      "step": 1934
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.842606010652722,
+      "learning_rate": 1.4450485388997149e-05,
+      "loss": 0.9954,
+      "step": 1935
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.057321791400095,
+      "learning_rate": 1.444489922200499e-05,
+      "loss": 0.8959,
+      "step": 1936
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8140608348248906,
+      "learning_rate": 1.4439311325922235e-05,
+      "loss": 0.9943,
+      "step": 1937
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7098410462471028,
+      "learning_rate": 1.4433721702922607e-05,
+      "loss": 0.8261,
+      "step": 1938
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0270872514099714,
+      "learning_rate": 1.4428130355180505e-05,
+      "loss": 0.8597,
+      "step": 1939
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.596577887827628,
+      "learning_rate": 1.4422537284870993e-05,
+      "loss": 0.8603,
+      "step": 1940
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0437894211858543,
+      "learning_rate": 1.4416942494169801e-05,
+      "loss": 0.8863,
+      "step": 1941
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.022257618024313,
+      "learning_rate": 1.441134598525334e-05,
+      "loss": 0.8304,
+      "step": 1942
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.630469813224106,
+      "learning_rate": 1.4405747760298681e-05,
+      "loss": 0.8952,
+      "step": 1943
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7663508160729156,
+      "learning_rate": 1.4400147821483564e-05,
+      "loss": 0.9825,
+      "step": 1944
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6935303282233554,
+      "learning_rate": 1.4394546170986394e-05,
+      "loss": 0.9867,
+      "step": 1945
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.811332894249063,
+      "learning_rate": 1.4388942810986253e-05,
+      "loss": 0.9215,
+      "step": 1946
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5936262053219343,
+      "learning_rate": 1.4383337743662872e-05,
+      "loss": 1.0064,
+      "step": 1947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2221389486212109,
+      "learning_rate": 1.4377730971196658e-05,
+      "loss": 0.8219,
+      "step": 1948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.692758782000757,
+      "learning_rate": 1.437212249576867e-05,
+      "loss": 0.8998,
+      "step": 1949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6050066128005005,
+      "learning_rate": 1.4366512319560642e-05,
+      "loss": 0.9318,
+      "step": 1950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.820927381354136,
+      "learning_rate": 1.4360900444754959e-05,
+      "loss": 0.9633,
+      "step": 1951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.759938786543528,
+      "learning_rate": 1.4355286873534677e-05,
+      "loss": 0.9468,
+      "step": 1952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6752343970748882,
+      "learning_rate": 1.4349671608083504e-05,
+      "loss": 0.8669,
+      "step": 1953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7184980534359486,
+      "learning_rate": 1.4344054650585813e-05,
+      "loss": 0.9065,
+      "step": 1954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.672450031095198,
+      "learning_rate": 1.4338436003226623e-05,
+      "loss": 1.0158,
+      "step": 1955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7822544858188532,
+      "learning_rate": 1.433281566819163e-05,
+      "loss": 0.8884,
+      "step": 1956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2323855892676456,
+      "learning_rate": 1.4327193647667167e-05,
+      "loss": 0.857,
+      "step": 1957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6749892119624912,
+      "learning_rate": 1.4321569943840235e-05,
+      "loss": 0.9273,
+      "step": 1958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.562532316981109,
+      "learning_rate": 1.4315944558898486e-05,
+      "loss": 0.9543,
+      "step": 1959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.733048227533296,
+      "learning_rate": 1.4310317495030226e-05,
+      "loss": 0.9416,
+      "step": 1960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6187249948968137,
+      "learning_rate": 1.4304688754424417e-05,
+      "loss": 0.9877,
+      "step": 1961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.544184212908568,
+      "learning_rate": 1.4299058339270663e-05,
+      "loss": 0.8356,
+      "step": 1962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5597864305215874,
+      "learning_rate": 1.4293426251759234e-05,
+      "loss": 0.8891,
+      "step": 1963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6764875535211585,
+      "learning_rate": 1.4287792494081042e-05,
+      "loss": 0.9978,
+      "step": 1964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.628972353736311,
+      "learning_rate": 1.428215706842765e-05,
+      "loss": 0.8752,
+      "step": 1965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6175343059783494,
+      "learning_rate": 1.4276519976991267e-05,
+      "loss": 0.9888,
+      "step": 1966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6569498876168287,
+      "learning_rate": 1.4270881221964756e-05,
+      "loss": 0.9243,
+      "step": 1967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5486333423207015,
+      "learning_rate": 1.4265240805541628e-05,
+      "loss": 0.875,
+      "step": 1968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6050782117840383,
+      "learning_rate": 1.4259598729916027e-05,
+      "loss": 0.8632,
+      "step": 1969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.613167697354811,
+      "learning_rate": 1.425395499728276e-05,
+      "loss": 0.8139,
+      "step": 1970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0995807826218689,
+      "learning_rate": 1.4248309609837262e-05,
+      "loss": 0.8318,
+      "step": 1971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7339742844852268,
+      "learning_rate": 1.4242662569775632e-05,
+      "loss": 1.0169,
+      "step": 1972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.775570046531735,
+      "learning_rate": 1.423701387929459e-05,
+      "loss": 0.8689,
+      "step": 1973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.712780324624818,
+      "learning_rate": 1.4231363540591512e-05,
+      "loss": 0.8534,
+      "step": 1974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9214141386062493,
+      "learning_rate": 1.4225711555864413e-05,
+      "loss": 0.9083,
+      "step": 1975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0979553405135083,
+      "learning_rate": 1.4220057927311944e-05,
+      "loss": 0.8881,
+      "step": 1976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6731728263110943,
+      "learning_rate": 1.4214402657133396e-05,
+      "loss": 0.968,
+      "step": 1977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5442188820300313,
+      "learning_rate": 1.4208745747528705e-05,
+      "loss": 0.8429,
+      "step": 1978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.560545156571736,
+      "learning_rate": 1.420308720069844e-05,
+      "loss": 0.8704,
+      "step": 1979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5465513905600146,
+      "learning_rate": 1.4197427018843804e-05,
+      "loss": 0.8807,
+      "step": 1980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9032763143445839,
+      "learning_rate": 1.4191765204166643e-05,
+      "loss": 0.8731,
+      "step": 1981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5074974422961198,
+      "learning_rate": 1.418610175886943e-05,
+      "loss": 0.8785,
+      "step": 1982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7085486460822743,
+      "learning_rate": 1.4180436685155284e-05,
+      "loss": 0.9437,
+      "step": 1983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0614399683461917,
+      "learning_rate": 1.4174769985227943e-05,
+      "loss": 0.869,
+      "step": 1984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8102639056663725,
+      "learning_rate": 1.4169101661291789e-05,
+      "loss": 0.9407,
+      "step": 1985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.677696426013255,
+      "learning_rate": 1.416343171555183e-05,
+      "loss": 0.9164,
+      "step": 1986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5983097724232058,
+      "learning_rate": 1.4157760150213706e-05,
+      "loss": 0.8976,
+      "step": 1987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5633357942983659,
+      "learning_rate": 1.415208696748369e-05,
+      "loss": 0.9583,
+      "step": 1988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8757667991615432,
+      "learning_rate": 1.414641216956868e-05,
+      "loss": 0.8841,
+      "step": 1989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7104011632366787,
+      "learning_rate": 1.4140735758676205e-05,
+      "loss": 0.9663,
+      "step": 1990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6278015170837083,
+      "learning_rate": 1.4135057737014416e-05,
+      "loss": 0.9017,
+      "step": 1991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6085948855074665,
+      "learning_rate": 1.4129378106792103e-05,
+      "loss": 0.9748,
+      "step": 1992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7212729021378976,
+      "learning_rate": 1.412369687021867e-05,
+      "loss": 0.8887,
+      "step": 1993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4847775674762957,
+      "learning_rate": 1.411801402950415e-05,
+      "loss": 0.8589,
+      "step": 1994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7254985997235885,
+      "learning_rate": 1.4112329586859196e-05,
+      "loss": 0.9851,
+      "step": 1995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6497964331392245,
+      "learning_rate": 1.4106643544495092e-05,
+      "loss": 0.9671,
+      "step": 1996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7500475617926596,
+      "learning_rate": 1.4100955904623742e-05,
+      "loss": 0.9816,
+      "step": 1997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6055781534173013,
+      "learning_rate": 1.4095266669457664e-05,
+      "loss": 0.7968,
+      "step": 1998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7201656182922893,
+      "learning_rate": 1.4089575841210004e-05,
+      "loss": 0.9469,
+      "step": 1999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7198154929517466,
+      "learning_rate": 1.4083883422094528e-05,
+      "loss": 0.9505,
+      "step": 2000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7677187878568104,
+      "learning_rate": 1.407818941432562e-05,
+      "loss": 0.9059,
+      "step": 2001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5424839722311516,
+      "learning_rate": 1.4072493820118278e-05,
+      "loss": 0.8941,
+      "step": 2002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.648842223017535,
+      "learning_rate": 1.4066796641688118e-05,
+      "loss": 0.9272,
+      "step": 2003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5963846420986898,
+      "learning_rate": 1.406109788125138e-05,
+      "loss": 0.9239,
+      "step": 2004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1024301374398795,
+      "learning_rate": 1.4055397541024908e-05,
+      "loss": 0.82,
+      "step": 2005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7369095392393668,
+      "learning_rate": 1.4049695623226172e-05,
+      "loss": 0.9127,
+      "step": 2006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6698097280492465,
+      "learning_rate": 1.4043992130073244e-05,
+      "loss": 0.8411,
+      "step": 2007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.618638834647394,
+      "learning_rate": 1.4038287063784817e-05,
+      "loss": 0.9056,
+      "step": 2008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7683780144261347,
+      "learning_rate": 1.4032580426580194e-05,
+      "loss": 0.9605,
+      "step": 2009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6496070307271165,
+      "learning_rate": 1.4026872220679289e-05,
+      "loss": 0.8955,
+      "step": 2010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6288855500157886,
+      "learning_rate": 1.4021162448302623e-05,
+      "loss": 0.8678,
+      "step": 2011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6294433513649282,
+      "learning_rate": 1.4015451111671334e-05,
+      "loss": 0.9101,
+      "step": 2012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7232305944530801,
+      "learning_rate": 1.400973821300716e-05,
+      "loss": 0.9553,
+      "step": 2013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7979358873403866,
+      "learning_rate": 1.4004023754532451e-05,
+      "loss": 0.9593,
+      "step": 2014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7148781900261514,
+      "learning_rate": 1.3998307738470168e-05,
+      "loss": 0.9923,
+      "step": 2015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.685929158664252,
+      "learning_rate": 1.3992590167043867e-05,
+      "loss": 0.8531,
+      "step": 2016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5364133228307006,
+      "learning_rate": 1.3986871042477717e-05,
+      "loss": 0.9131,
+      "step": 2017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8923374990592905,
+      "learning_rate": 1.398115036699649e-05,
+      "loss": 0.961,
+      "step": 2018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9213485352679245,
+      "learning_rate": 1.3975428142825562e-05,
+      "loss": 0.973,
+      "step": 2019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6898456454219928,
+      "learning_rate": 1.3969704372190907e-05,
+      "loss": 0.9556,
+      "step": 2020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.639789866546958,
+      "learning_rate": 1.3963979057319104e-05,
+      "loss": 0.951,
+      "step": 2021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.778718491999352,
+      "learning_rate": 1.3958252200437335e-05,
+      "loss": 0.9443,
+      "step": 2022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4470740232227306,
+      "learning_rate": 1.3952523803773379e-05,
+      "loss": 0.8811,
+      "step": 2023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.573422266308468,
+      "learning_rate": 1.394679386955561e-05,
+      "loss": 0.8615,
+      "step": 2024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6111185628370515,
+      "learning_rate": 1.394106240001301e-05,
+      "loss": 0.9061,
+      "step": 2025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.712324474317859,
+      "learning_rate": 1.3935329397375153e-05,
+      "loss": 0.8465,
+      "step": 2026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.611475452328952,
+      "learning_rate": 1.3929594863872203e-05,
+      "loss": 0.9116,
+      "step": 2027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6674866516301947,
+      "learning_rate": 1.3923858801734929e-05,
+      "loss": 0.9775,
+      "step": 2028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6947441578154065,
+      "learning_rate": 1.391812121319469e-05,
+      "loss": 0.875,
+      "step": 2029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6569751725313584,
+      "learning_rate": 1.3912382100483444e-05,
+      "loss": 0.9456,
+      "step": 2030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7052166387386258,
+      "learning_rate": 1.3906641465833732e-05,
+      "loss": 0.9772,
+      "step": 2031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.77203005430193,
+      "learning_rate": 1.3900899311478695e-05,
+      "loss": 0.9253,
+      "step": 2032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.674871514983555,
+      "learning_rate": 1.3895155639652067e-05,
+      "loss": 0.8341,
+      "step": 2033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7947589153855095,
+      "learning_rate": 1.3889410452588164e-05,
+      "loss": 0.9746,
+      "step": 2034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.233363205539803,
+      "learning_rate": 1.3883663752521895e-05,
+      "loss": 0.8633,
+      "step": 2035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7405978697481788,
+      "learning_rate": 1.387791554168876e-05,
+      "loss": 1.0023,
+      "step": 2036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7252602631323557,
+      "learning_rate": 1.387216582232485e-05,
+      "loss": 0.9743,
+      "step": 2037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6973224466147088,
+      "learning_rate": 1.3866414596666828e-05,
+      "loss": 0.954,
+      "step": 2038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.50826708052,
+      "learning_rate": 1.3860661866951961e-05,
+      "loss": 0.8543,
+      "step": 2039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0496205349562528,
+      "learning_rate": 1.385490763541809e-05,
+      "loss": 0.8579,
+      "step": 2040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6004874934135802,
+      "learning_rate": 1.3849151904303646e-05,
+      "loss": 0.9211,
+      "step": 2041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7077977894436696,
+      "learning_rate": 1.3843394675847635e-05,
+      "loss": 0.9311,
+      "step": 2042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9729241867776557,
+      "learning_rate": 1.3837635952289653e-05,
+      "loss": 0.8186,
+      "step": 2043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.591122849853647,
+      "learning_rate": 1.3831875735869875e-05,
+      "loss": 0.8782,
+      "step": 2044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6694338698420805,
+      "learning_rate": 1.3826114028829062e-05,
+      "loss": 0.8907,
+      "step": 2045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8035349756233845,
+      "learning_rate": 1.3820350833408547e-05,
+      "loss": 0.9971,
+      "step": 2046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7115337511488642,
+      "learning_rate": 1.3814586151850242e-05,
+      "loss": 0.9591,
+      "step": 2047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.661407758649582,
+      "learning_rate": 1.3808819986396646e-05,
+      "loss": 0.9572,
+      "step": 2048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6266972953361736,
+      "learning_rate": 1.3803052339290828e-05,
+      "loss": 0.9373,
+      "step": 2049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.585764866694993,
+      "learning_rate": 1.3797283212776429e-05,
+      "loss": 0.9437,
+      "step": 2050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6417729840191095,
+      "learning_rate": 1.3791512609097678e-05,
+      "loss": 0.9525,
+      "step": 2051
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6620631943947641,
+      "learning_rate": 1.3785740530499374e-05,
+      "loss": 0.9463,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6146509478141164,
+      "learning_rate": 1.3779966979226883e-05,
+      "loss": 0.9176,
+      "step": 2053
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6995311583630783,
+      "learning_rate": 1.3774191957526144e-05,
+      "loss": 0.9522,
+      "step": 2054
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.559150610076637,
+      "learning_rate": 1.3768415467643686e-05,
+      "loss": 0.883,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6554811245841967,
+      "learning_rate": 1.3762637511826583e-05,
+      "loss": 0.9502,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5175284639127558,
+      "learning_rate": 1.3756858092322499e-05,
+      "loss": 0.9335,
+      "step": 2057
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7878594480964296,
+      "learning_rate": 1.3751077211379655e-05,
+      "loss": 0.9755,
+      "step": 2058
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6251856215670673,
+      "learning_rate": 1.3745294871246855e-05,
+      "loss": 0.8954,
+      "step": 2059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.607142038643464,
+      "learning_rate": 1.3739511074173453e-05,
+      "loss": 0.8548,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7208876744054031,
+      "learning_rate": 1.3733725822409381e-05,
+      "loss": 0.935,
+      "step": 2061
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7148019456000763,
+      "learning_rate": 1.3727939118205138e-05,
+      "loss": 0.8617,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7717928228905133,
+      "learning_rate": 1.372215096381178e-05,
+      "loss": 0.9462,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.608567233231734,
+      "learning_rate": 1.371636136148093e-05,
+      "loss": 0.8995,
+      "step": 2064
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.654669774638764,
+      "learning_rate": 1.3710570313464778e-05,
+      "loss": 0.8675,
+      "step": 2065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5784192802996106,
+      "learning_rate": 1.3704777822016073e-05,
+      "loss": 0.8348,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5752177782109855,
+      "learning_rate": 1.369898388938813e-05,
+      "loss": 0.9154,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6057503297690165,
+      "learning_rate": 1.3693188517834817e-05,
+      "loss": 0.8867,
+      "step": 2068
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.580357790732273,
+      "learning_rate": 1.3687391709610567e-05,
+      "loss": 0.9331,
+      "step": 2069
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5956666067919434,
+      "learning_rate": 1.368159346697037e-05,
+      "loss": 0.915,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8065683319014694,
+      "learning_rate": 1.3675793792169778e-05,
+      "loss": 0.9793,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.767908804361483,
+      "learning_rate": 1.3669992687464892e-05,
+      "loss": 0.8071,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.830273395372375,
+      "learning_rate": 1.3664190155112378e-05,
+      "loss": 0.8879,
+      "step": 2073
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7181577530862,
+      "learning_rate": 1.3658386197369454e-05,
+      "loss": 0.9701,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6819476835531595,
+      "learning_rate": 1.3652580816493888e-05,
+      "loss": 0.9009,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7995504374088598,
+      "learning_rate": 1.3646774014744006e-05,
+      "loss": 0.9607,
+      "step": 2076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8940395913914327,
+      "learning_rate": 1.364096579437869e-05,
+      "loss": 0.9216,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0633131945514072,
+      "learning_rate": 1.3635156157657367e-05,
+      "loss": 0.8286,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6433065037689483,
+      "learning_rate": 1.3629345106840021e-05,
+      "loss": 0.9918,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6715856898091073,
+      "learning_rate": 1.3623532644187178e-05,
+      "loss": 0.8932,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7379870777670272,
+      "learning_rate": 1.3617718771959927e-05,
+      "loss": 0.9138,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9913975678897632,
+      "learning_rate": 1.3611903492419887e-05,
+      "loss": 0.8786,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8222590608862976,
+      "learning_rate": 1.360608680782924e-05,
+      "loss": 0.9796,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6060510642771517,
+      "learning_rate": 1.3600268720450709e-05,
+      "loss": 0.9731,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6124188670023307,
+      "learning_rate": 1.3594449232547558e-05,
+      "loss": 0.8643,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.020617297116168,
+      "learning_rate": 1.3588628346383606e-05,
+      "loss": 0.873,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.585782449241958,
+      "learning_rate": 1.3582806064223205e-05,
+      "loss": 0.9332,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6714192697193047,
+      "learning_rate": 1.3576982388331258e-05,
+      "loss": 0.847,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7295472680055475,
+      "learning_rate": 1.3571157320973205e-05,
+      "loss": 0.7776,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.732269731832295,
+      "learning_rate": 1.3565330864415033e-05,
+      "loss": 0.8969,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5583853586606025,
+      "learning_rate": 1.3559503020923263e-05,
+      "loss": 0.9501,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6768535366687451,
+      "learning_rate": 1.3553673792764966e-05,
+      "loss": 0.9482,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8569229269314258,
+      "learning_rate": 1.3547843182207734e-05,
+      "loss": 0.8831,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0464434904024547,
+      "learning_rate": 1.3542011191519713e-05,
+      "loss": 0.8096,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.630394864250798,
+      "learning_rate": 1.3536177822969578e-05,
+      "loss": 0.8882,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7200281264322186,
+      "learning_rate": 1.3530343078826542e-05,
+      "loss": 0.89,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6023929624185314,
+      "learning_rate": 1.3524506961360352e-05,
+      "loss": 0.8554,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7246353702642154,
+      "learning_rate": 1.3518669472841294e-05,
+      "loss": 0.909,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5612325542636551,
+      "learning_rate": 1.3512830615540185e-05,
+      "loss": 0.9564,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6661129481013794,
+      "learning_rate": 1.3506990391728363e-05,
+      "loss": 1.0003,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6051601661556567,
+      "learning_rate": 1.3501148803677716e-05,
+      "loss": 0.8952,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7025732828565276,
+      "learning_rate": 1.3495305853660655e-05,
+      "loss": 0.8573,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6973326205783885,
+      "learning_rate": 1.3489461543950115e-05,
+      "loss": 0.9265,
+      "step": 2103
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8285882993365843,
+      "learning_rate": 1.3483615876819567e-05,
+      "loss": 0.9949,
+      "step": 2104
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6550660129849564,
+      "learning_rate": 1.347776885454301e-05,
+      "loss": 0.8315,
+      "step": 2105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.017610199117295,
+      "learning_rate": 1.3471920479394968e-05,
+      "loss": 0.8346,
+      "step": 2106
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7493753959956675,
+      "learning_rate": 1.3466070753650488e-05,
+      "loss": 0.9144,
+      "step": 2107
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8105696401658662,
+      "learning_rate": 1.3460219679585151e-05,
+      "loss": 0.864,
+      "step": 2108
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0341798381431317,
+      "learning_rate": 1.3454367259475054e-05,
+      "loss": 0.8413,
+      "step": 2109
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7241112987242089,
+      "learning_rate": 1.3448513495596825e-05,
+      "loss": 0.9258,
+      "step": 2110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7343487102172213,
+      "learning_rate": 1.3442658390227604e-05,
+      "loss": 0.9995,
+      "step": 2111
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.005184510279151,
+      "learning_rate": 1.3436801945645064e-05,
+      "loss": 0.8499,
+      "step": 2112
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6737959446984936,
+      "learning_rate": 1.3430944164127393e-05,
+      "loss": 0.9313,
+      "step": 2113
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.60355195723054,
+      "learning_rate": 1.3425085047953305e-05,
+      "loss": 0.8879,
+      "step": 2114
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9581229890711807,
+      "learning_rate": 1.3419224599402022e-05,
+      "loss": 0.8998,
+      "step": 2115
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7350906426964268,
+      "learning_rate": 1.34133628207533e-05,
+      "loss": 0.9496,
+      "step": 2116
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5649967802054536,
+      "learning_rate": 1.3407499714287394e-05,
+      "loss": 0.9279,
+      "step": 2117
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6807796590794255,
+      "learning_rate": 1.3401635282285093e-05,
+      "loss": 0.9326,
+      "step": 2118
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.980590924594308,
+      "learning_rate": 1.3395769527027686e-05,
+      "loss": 0.8208,
+      "step": 2119
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.724955608361257,
+      "learning_rate": 1.3389902450796992e-05,
+      "loss": 0.9574,
+      "step": 2120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8312009983843984,
+      "learning_rate": 1.3384034055875335e-05,
+      "loss": 0.9464,
+      "step": 2121
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5650853859886844,
+      "learning_rate": 1.3378164344545547e-05,
+      "loss": 0.9365,
+      "step": 2122
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.807221266940543,
+      "learning_rate": 1.3372293319090986e-05,
+      "loss": 0.918,
+      "step": 2123
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7290038198400746,
+      "learning_rate": 1.336642098179551e-05,
+      "loss": 0.9254,
+      "step": 2124
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5328454429910365,
+      "learning_rate": 1.3360547334943492e-05,
+      "loss": 0.9176,
+      "step": 2125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7538645237566617,
+      "learning_rate": 1.3354672380819812e-05,
+      "loss": 0.9798,
+      "step": 2126
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7618174910360094,
+      "learning_rate": 1.3348796121709862e-05,
+      "loss": 0.9204,
+      "step": 2127
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6894788948068933,
+      "learning_rate": 1.334291855989954e-05,
+      "loss": 0.9981,
+      "step": 2128
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5644951693738753,
+      "learning_rate": 1.3337039697675248e-05,
+      "loss": 0.9003,
+      "step": 2129
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8106486528731447,
+      "learning_rate": 1.3331159537323895e-05,
+      "loss": 0.9581,
+      "step": 2130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.701541614506914,
+      "learning_rate": 1.3325278081132903e-05,
+      "loss": 0.9828,
+      "step": 2131
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7188890375486765,
+      "learning_rate": 1.3319395331390183e-05,
+      "loss": 0.8649,
+      "step": 2132
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.551975461490393,
+      "learning_rate": 1.3313511290384162e-05,
+      "loss": 0.9076,
+      "step": 2133
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.623627593156588,
+      "learning_rate": 1.3307625960403763e-05,
+      "loss": 0.934,
+      "step": 2134
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.562255111835152,
+      "learning_rate": 1.3301739343738411e-05,
+      "loss": 0.893,
+      "step": 2135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.062152753605498,
+      "learning_rate": 1.3295851442678041e-05,
+      "loss": 0.8802,
+      "step": 2136
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6294603562027585,
+      "learning_rate": 1.3289962259513068e-05,
+      "loss": 0.9409,
+      "step": 2137
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0305824729818394,
+      "learning_rate": 1.3284071796534423e-05,
+      "loss": 0.856,
+      "step": 2138
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8795952307768637,
+      "learning_rate": 1.327818005603353e-05,
+      "loss": 0.9633,
+      "step": 2139
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7920550562875857,
+      "learning_rate": 1.3272287040302305e-05,
+      "loss": 0.9558,
+      "step": 2140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8401686895677056,
+      "learning_rate": 1.3266392751633167e-05,
+      "loss": 0.9857,
+      "step": 2141
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7050562583924986,
+      "learning_rate": 1.3260497192319023e-05,
+      "loss": 0.8695,
+      "step": 2142
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.626224124134655,
+      "learning_rate": 1.325460036465329e-05,
+      "loss": 0.9499,
+      "step": 2143
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.768167052796858,
+      "learning_rate": 1.3248702270929852e-05,
+      "loss": 0.8659,
+      "step": 2144
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4464909722775954,
+      "learning_rate": 1.3242802913443107e-05,
+      "loss": 0.8654,
+      "step": 2145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6593594259957878,
+      "learning_rate": 1.323690229448794e-05,
+      "loss": 0.9109,
+      "step": 2146
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6258466682289654,
+      "learning_rate": 1.3231000416359722e-05,
+      "loss": 0.9012,
+      "step": 2147
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7124571183734703,
+      "learning_rate": 1.3225097281354318e-05,
+      "loss": 0.9403,
+      "step": 2148
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.730141741177968,
+      "learning_rate": 1.3219192891768078e-05,
+      "loss": 0.8308,
+      "step": 2149
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5433957845062611,
+      "learning_rate": 1.3213287249897846e-05,
+      "loss": 0.9221,
+      "step": 2150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6438867568282847,
+      "learning_rate": 1.3207380358040947e-05,
+      "loss": 0.9838,
+      "step": 2151
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7556351915446156,
+      "learning_rate": 1.3201472218495192e-05,
+      "loss": 0.9837,
+      "step": 2152
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6453055727975792,
+      "learning_rate": 1.319556283355889e-05,
+      "loss": 0.955,
+      "step": 2153
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6318175278955305,
+      "learning_rate": 1.3189652205530815e-05,
+      "loss": 0.846,
+      "step": 2154
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.574637859261439,
+      "learning_rate": 1.318374033671024e-05,
+      "loss": 0.8639,
+      "step": 2155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8077737747977336,
+      "learning_rate": 1.3177827229396912e-05,
+      "loss": 0.8721,
+      "step": 2156
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8078152012544024,
+      "learning_rate": 1.3171912885891063e-05,
+      "loss": 0.8708,
+      "step": 2157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2103081584599216,
+      "learning_rate": 1.3165997308493405e-05,
+      "loss": 0.8289,
+      "step": 2158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6617691728326804,
+      "learning_rate": 1.3160080499505132e-05,
+      "loss": 0.9497,
+      "step": 2159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6410481882492853,
+      "learning_rate": 1.3154162461227915e-05,
+      "loss": 0.9713,
+      "step": 2160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6486208082257108,
+      "learning_rate": 1.3148243195963908e-05,
+      "loss": 0.8473,
+      "step": 2161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7802577278092968,
+      "learning_rate": 1.314232270601573e-05,
+      "loss": 0.9142,
+      "step": 2162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5514367407880878,
+      "learning_rate": 1.313640099368649e-05,
+      "loss": 0.9225,
+      "step": 2163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0018914189904142,
+      "learning_rate": 1.313047806127977e-05,
+      "loss": 0.8552,
+      "step": 2164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6484350227909816,
+      "learning_rate": 1.3124553911099619e-05,
+      "loss": 0.9388,
+      "step": 2165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6031495161599543,
+      "learning_rate": 1.3118628545450563e-05,
+      "loss": 0.8759,
+      "step": 2166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5933471072536454,
+      "learning_rate": 1.3112701966637604e-05,
+      "loss": 0.9293,
+      "step": 2167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8729990217388943,
+      "learning_rate": 1.310677417696622e-05,
+      "loss": 0.8789,
+      "step": 2168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0028615267541232,
+      "learning_rate": 1.3100845178742346e-05,
+      "loss": 0.8229,
+      "step": 2169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6568317959437235,
+      "learning_rate": 1.3094914974272397e-05,
+      "loss": 0.9349,
+      "step": 2170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6054963140057126,
+      "learning_rate": 1.3088983565863259e-05,
+      "loss": 0.8916,
+      "step": 2171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0006483785770817,
+      "learning_rate": 1.3083050955822283e-05,
+      "loss": 0.8644,
+      "step": 2172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7302213264267754,
+      "learning_rate": 1.3077117146457282e-05,
+      "loss": 0.9034,
+      "step": 2173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.72650261320599,
+      "learning_rate": 1.3071182140076546e-05,
+      "loss": 1.0115,
+      "step": 2174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7268068432061778,
+      "learning_rate": 1.3065245938988826e-05,
+      "loss": 0.8801,
+      "step": 2175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6958264686976343,
+      "learning_rate": 1.3059308545503334e-05,
+      "loss": 0.8276,
+      "step": 2176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.576051262342631,
+      "learning_rate": 1.3053369961929749e-05,
+      "loss": 0.9285,
+      "step": 2177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8194664679222294,
+      "learning_rate": 1.304743019057822e-05,
+      "loss": 0.8408,
+      "step": 2178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5860808533998068,
+      "learning_rate": 1.3041489233759345e-05,
+      "loss": 0.9085,
+      "step": 2179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6844259906206056,
+      "learning_rate": 1.3035547093784187e-05,
+      "loss": 0.8861,
+      "step": 2180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6024806798103937,
+      "learning_rate": 1.3029603772964276e-05,
+      "loss": 0.9385,
+      "step": 2181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6215205652746287,
+      "learning_rate": 1.3023659273611595e-05,
+      "loss": 0.9247,
+      "step": 2182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7550326649533516,
+      "learning_rate": 1.3017713598038593e-05,
+      "loss": 0.9246,
+      "step": 2183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5696001515161895,
+      "learning_rate": 1.301176674855816e-05,
+      "loss": 0.9022,
+      "step": 2184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8141654615147336,
+      "learning_rate": 1.3005818727483663e-05,
+      "loss": 0.9301,
+      "step": 2185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.746264821965584,
+      "learning_rate": 1.2999869537128916e-05,
+      "loss": 0.9236,
+      "step": 2186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.700661838664337,
+      "learning_rate": 1.2993919179808179e-05,
+      "loss": 0.7285,
+      "step": 2187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.749034432876835,
+      "learning_rate": 1.2987967657836178e-05,
+      "loss": 0.939,
+      "step": 2188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6517553740816686,
+      "learning_rate": 1.2982014973528089e-05,
+      "loss": 0.9995,
+      "step": 2189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6588664777345277,
+      "learning_rate": 1.2976061129199544e-05,
+      "loss": 0.807,
+      "step": 2190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5995544700750486,
+      "learning_rate": 1.2970106127166613e-05,
+      "loss": 0.9226,
+      "step": 2191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4921003783830158,
+      "learning_rate": 1.2964149969745832e-05,
+      "loss": 0.8215,
+      "step": 2192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5682006495868175,
+      "learning_rate": 1.2958192659254176e-05,
+      "loss": 0.8373,
+      "step": 2193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8121344651724471,
+      "learning_rate": 1.295223419800907e-05,
+      "loss": 0.8835,
+      "step": 2194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8579962574660704,
+      "learning_rate": 1.2946274588328391e-05,
+      "loss": 0.8364,
+      "step": 2195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7335016298545531,
+      "learning_rate": 1.2940313832530462e-05,
+      "loss": 0.9407,
+      "step": 2196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6863232204553487,
+      "learning_rate": 1.2934351932934054e-05,
+      "loss": 0.8441,
+      "step": 2197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6245080302961667,
+      "learning_rate": 1.2928388891858368e-05,
+      "loss": 0.9209,
+      "step": 2198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6236456124399417,
+      "learning_rate": 1.2922424711623064e-05,
+      "loss": 0.9917,
+      "step": 2199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7689775512188415,
+      "learning_rate": 1.291645939454825e-05,
+      "loss": 0.8672,
+      "step": 2200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.674815471137174,
+      "learning_rate": 1.2910492942954462e-05,
+      "loss": 0.8703,
+      "step": 2201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7285291212581586,
+      "learning_rate": 1.2904525359162678e-05,
+      "loss": 0.8252,
+      "step": 2202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1125595515996292,
+      "learning_rate": 1.2898556645494327e-05,
+      "loss": 0.8613,
+      "step": 2203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7194621253684532,
+      "learning_rate": 1.2892586804271269e-05,
+      "loss": 0.9301,
+      "step": 2204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.613244846699818,
+      "learning_rate": 1.2886615837815813e-05,
+      "loss": 0.8608,
+      "step": 2205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6017740000898104,
+      "learning_rate": 1.2880643748450685e-05,
+      "loss": 0.9411,
+      "step": 2206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7134987639829544,
+      "learning_rate": 1.2874670538499072e-05,
+      "loss": 0.8782,
+      "step": 2207
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.65427332668514,
+      "learning_rate": 1.2868696210284584e-05,
+      "loss": 0.9187,
+      "step": 2208
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.66535080513408,
+      "learning_rate": 1.2862720766131265e-05,
+      "loss": 0.9116,
+      "step": 2209
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4882330222915572,
+      "learning_rate": 1.2856744208363593e-05,
+      "loss": 0.9322,
+      "step": 2210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6079386941052378,
+      "learning_rate": 1.2850766539306493e-05,
+      "loss": 0.857,
+      "step": 2211
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6437647431889684,
+      "learning_rate": 1.2844787761285304e-05,
+      "loss": 0.8721,
+      "step": 2212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8521788343574743,
+      "learning_rate": 1.2838807876625806e-05,
+      "loss": 0.9379,
+      "step": 2213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7981006740472716,
+      "learning_rate": 1.2832826887654206e-05,
+      "loss": 0.9974,
+      "step": 2214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.127192375994006,
+      "learning_rate": 1.2826844796697145e-05,
+      "loss": 0.8825,
+      "step": 2215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7045053953528582,
+      "learning_rate": 1.2820861606081686e-05,
+      "loss": 0.8435,
+      "step": 2216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.640423832277713,
+      "learning_rate": 1.2814877318135323e-05,
+      "loss": 0.9068,
+      "step": 2217
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5459230507089665,
+      "learning_rate": 1.2808891935185983e-05,
+      "loss": 0.883,
+      "step": 2218
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5878013894112704,
+      "learning_rate": 1.2802905459562011e-05,
+      "loss": 0.8999,
+      "step": 2219
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6329005532631673,
+      "learning_rate": 1.2796917893592178e-05,
+      "loss": 0.9404,
+      "step": 2220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6296586405116358,
+      "learning_rate": 1.2790929239605679e-05,
+      "loss": 0.9259,
+      "step": 2221
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6675947493054515,
+      "learning_rate": 1.2784939499932139e-05,
+      "loss": 0.8121,
+      "step": 2222
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.662349414865135,
+      "learning_rate": 1.2778948676901594e-05,
+      "loss": 0.8929,
+      "step": 2223
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.734165064222773,
+      "learning_rate": 1.2772956772844514e-05,
+      "loss": 0.9075,
+      "step": 2224
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5038776683085109,
+      "learning_rate": 1.2766963790091778e-05,
+      "loss": 0.9186,
+      "step": 2225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5247340237807017,
+      "learning_rate": 1.2760969730974692e-05,
+      "loss": 0.9173,
+      "step": 2226
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8069677268604598,
+      "learning_rate": 1.2754974597824976e-05,
+      "loss": 0.9038,
+      "step": 2227
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6199303166844012,
+      "learning_rate": 1.2748978392974774e-05,
+      "loss": 0.8373,
+      "step": 2228
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7279692893731582,
+      "learning_rate": 1.274298111875664e-05,
+      "loss": 0.9137,
+      "step": 2229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7320981883039976,
+      "learning_rate": 1.273698277750355e-05,
+      "loss": 0.8817,
+      "step": 2230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.728074798057771,
+      "learning_rate": 1.2730983371548889e-05,
+      "loss": 0.8619,
+      "step": 2231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5570522244638392,
+      "learning_rate": 1.2724982903226462e-05,
+      "loss": 0.8921,
+      "step": 2232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7353716032674855,
+      "learning_rate": 1.2718981374870483e-05,
+      "loss": 0.9845,
+      "step": 2233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8364316722298129,
+      "learning_rate": 1.2712978788815582e-05,
+      "loss": 0.8488,
+      "step": 2234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6048953222179199,
+      "learning_rate": 1.2706975147396796e-05,
+      "loss": 0.9246,
+      "step": 2235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.61648716407882,
+      "learning_rate": 1.2700970452949575e-05,
+      "loss": 0.926,
+      "step": 2236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8028802433566433,
+      "learning_rate": 1.2694964707809785e-05,
+      "loss": 0.9069,
+      "step": 2237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4184900807377758,
+      "learning_rate": 1.2688957914313685e-05,
+      "loss": 0.8225,
+      "step": 2238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5305327020064978,
+      "learning_rate": 1.2682950074797953e-05,
+      "loss": 0.907,
+      "step": 2239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7132477628343878,
+      "learning_rate": 1.2676941191599679e-05,
+      "loss": 0.9243,
+      "step": 2240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5788775045201433,
+      "learning_rate": 1.2670931267056345e-05,
+      "loss": 0.878,
+      "step": 2241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7411515709825545,
+      "learning_rate": 1.2664920303505848e-05,
+      "loss": 1.0053,
+      "step": 2242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1063496392164596,
+      "learning_rate": 1.2658908303286488e-05,
+      "loss": 0.8664,
+      "step": 2243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6567897764327646,
+      "learning_rate": 1.2652895268736964e-05,
+      "loss": 0.8687,
+      "step": 2244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9406180230539718,
+      "learning_rate": 1.2646881202196379e-05,
+      "loss": 0.9703,
+      "step": 2245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.621820127024766,
+      "learning_rate": 1.2640866106004239e-05,
+      "loss": 0.8967,
+      "step": 2246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7354894987788128,
+      "learning_rate": 1.2634849982500455e-05,
+      "loss": 0.9914,
+      "step": 2247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.633748396653854,
+      "learning_rate": 1.262883283402533e-05,
+      "loss": 0.9158,
+      "step": 2248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7013231982938417,
+      "learning_rate": 1.2622814662919562e-05,
+      "loss": 0.9733,
+      "step": 2249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7829135678934118,
+      "learning_rate": 1.2616795471524259e-05,
+      "loss": 0.9022,
+      "step": 2250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7325470649885222,
+      "learning_rate": 1.261077526218092e-05,
+      "loss": 0.9507,
+      "step": 2251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6048025828772114,
+      "learning_rate": 1.260475403723144e-05,
+      "loss": 0.9148,
+      "step": 2252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.780596287732145,
+      "learning_rate": 1.2598731799018107e-05,
+      "loss": 0.8589,
+      "step": 2253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6779552236055686,
+      "learning_rate": 1.2592708549883607e-05,
+      "loss": 0.8587,
+      "step": 2254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7644297333465717,
+      "learning_rate": 1.2586684292171019e-05,
+      "loss": 0.9134,
+      "step": 2255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.465663545629934,
+      "learning_rate": 1.2580659028223808e-05,
+      "loss": 0.8563,
+      "step": 2256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6018615680731225,
+      "learning_rate": 1.2574632760385838e-05,
+      "loss": 0.8088,
+      "step": 2257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6080384665846545,
+      "learning_rate": 1.2568605491001363e-05,
+      "loss": 0.8747,
+      "step": 2258
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.782831198650503,
+      "learning_rate": 1.2562577222415022e-05,
+      "loss": 0.9122,
+      "step": 2259
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5293840639936092,
+      "learning_rate": 1.2556547956971845e-05,
+      "loss": 0.8949,
+      "step": 2260
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7924663165454586,
+      "learning_rate": 1.255051769701725e-05,
+      "loss": 0.9104,
+      "step": 2261
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5676440120579376,
+      "learning_rate": 1.2544486444897044e-05,
+      "loss": 0.8705,
+      "step": 2262
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7781024345656955,
+      "learning_rate": 1.2538454202957415e-05,
+      "loss": 0.8851,
+      "step": 2263
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6789120064479233,
+      "learning_rate": 1.2532420973544938e-05,
+      "loss": 0.9167,
+      "step": 2264
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8134846817706436,
+      "learning_rate": 1.2526386759006577e-05,
+      "loss": 0.9897,
+      "step": 2265
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7020622467786453,
+      "learning_rate": 1.2520351561689674e-05,
+      "loss": 0.9584,
+      "step": 2266
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6800125867312588,
+      "learning_rate": 1.2514315383941948e-05,
+      "loss": 0.9096,
+      "step": 2267
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6135339771505852,
+      "learning_rate": 1.2508278228111514e-05,
+      "loss": 0.8535,
+      "step": 2268
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.690543928956837,
+      "learning_rate": 1.2502240096546858e-05,
+      "loss": 0.909,
+      "step": 2269
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.823302563255051,
+      "learning_rate": 1.2496200991596844e-05,
+      "loss": 0.958,
+      "step": 2270
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.555411436419574,
+      "learning_rate": 1.2490160915610718e-05,
+      "loss": 0.8705,
+      "step": 2271
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6502037654859913,
+      "learning_rate": 1.2484119870938102e-05,
+      "loss": 0.9156,
+      "step": 2272
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6652870157828632,
+      "learning_rate": 1.2478077859929e-05,
+      "loss": 0.9056,
+      "step": 2273
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8492068390440737,
+      "learning_rate": 1.2472034884933789e-05,
+      "loss": 0.8911,
+      "step": 2274
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6642878857744101,
+      "learning_rate": 1.2465990948303212e-05,
+      "loss": 0.9521,
+      "step": 2275
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.680749007902511,
+      "learning_rate": 1.2459946052388398e-05,
+      "loss": 0.9656,
+      "step": 2276
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.560089131970013,
+      "learning_rate": 1.2453900199540852e-05,
+      "loss": 0.9098,
+      "step": 2277
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7466988829477026,
+      "learning_rate": 1.2447853392112434e-05,
+      "loss": 0.9297,
+      "step": 2278
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5116354152618778,
+      "learning_rate": 1.2441805632455392e-05,
+      "loss": 0.8241,
+      "step": 2279
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7161078251844877,
+      "learning_rate": 1.2435756922922337e-05,
+      "loss": 0.8956,
+      "step": 2280
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1819975014057913,
+      "learning_rate": 1.242970726586625e-05,
+      "loss": 0.9319,
+      "step": 2281
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7311335263758756,
+      "learning_rate": 1.2423656663640482e-05,
+      "loss": 0.9073,
+      "step": 2282
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.743868725006009,
+      "learning_rate": 1.2417605118598753e-05,
+      "loss": 0.9334,
+      "step": 2283
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.777151030781481,
+      "learning_rate": 1.2411552633095142e-05,
+      "loss": 0.927,
+      "step": 2284
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7194671949721274,
+      "learning_rate": 1.2405499209484106e-05,
+      "loss": 0.9191,
+      "step": 2285
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.018930896191571,
+      "learning_rate": 1.2399444850120457e-05,
+      "loss": 0.8408,
+      "step": 2286
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7842311368856647,
+      "learning_rate": 1.2393389557359374e-05,
+      "loss": 0.965,
+      "step": 2287
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5456058038168958,
+      "learning_rate": 1.2387333333556404e-05,
+      "loss": 0.8451,
+      "step": 2288
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.657035277059945,
+      "learning_rate": 1.2381276181067447e-05,
+      "loss": 0.9866,
+      "step": 2289
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4790603932491677,
+      "learning_rate": 1.2375218102248771e-05,
+      "loss": 0.8309,
+      "step": 2290
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6958909773004571,
+      "learning_rate": 1.2369159099457004e-05,
+      "loss": 0.9544,
+      "step": 2291
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6604832765013011,
+      "learning_rate": 1.2363099175049129e-05,
+      "loss": 0.9005,
+      "step": 2292
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7741994169945483,
+      "learning_rate": 1.2357038331382494e-05,
+      "loss": 0.8186,
+      "step": 2293
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6886518796300005,
+      "learning_rate": 1.2350976570814797e-05,
+      "loss": 0.8124,
+      "step": 2294
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6071131672202916,
+      "learning_rate": 1.2344913895704099e-05,
+      "loss": 0.8536,
+      "step": 2295
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7065589961634942,
+      "learning_rate": 1.2338850308408814e-05,
+      "loss": 0.8726,
+      "step": 2296
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6731293256167514,
+      "learning_rate": 1.233278581128771e-05,
+      "loss": 0.8854,
+      "step": 2297
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7318675518850375,
+      "learning_rate": 1.2326720406699913e-05,
+      "loss": 0.9208,
+      "step": 2298
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5244847991863824,
+      "learning_rate": 1.2320654097004898e-05,
+      "loss": 0.914,
+      "step": 2299
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.599215140424623,
+      "learning_rate": 1.231458688456249e-05,
+      "loss": 0.9427,
+      "step": 2300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7470299127086406,
+      "learning_rate": 1.2308518771732874e-05,
+      "loss": 0.9115,
+      "step": 2301
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7717492806662127,
+      "learning_rate": 1.2302449760876574e-05,
+      "loss": 0.8745,
+      "step": 2302
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.720737263267829,
+      "learning_rate": 1.2296379854354477e-05,
+      "loss": 0.9952,
+      "step": 2303
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.597836460637408,
+      "learning_rate": 1.2290309054527803e-05,
+      "loss": 0.9984,
+      "step": 2304
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4952013412601537,
+      "learning_rate": 1.228423736375813e-05,
+      "loss": 0.8173,
+      "step": 2305
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.600898558500933,
+      "learning_rate": 1.227816478440738e-05,
+      "loss": 0.8645,
+      "step": 2306
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6303508576087478,
+      "learning_rate": 1.2272091318837817e-05,
+      "loss": 0.8196,
+      "step": 2307
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6768714104416356,
+      "learning_rate": 1.2266016969412059e-05,
+      "loss": 0.8531,
+      "step": 2308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5709323607292336,
+      "learning_rate": 1.2259941738493056e-05,
+      "loss": 0.944,
+      "step": 2309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4680770737737074,
+      "learning_rate": 1.225386562844411e-05,
+      "loss": 0.9263,
+      "step": 2310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6718306566887422,
+      "learning_rate": 1.2247788641628862e-05,
+      "loss": 0.937,
+      "step": 2311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6625544513702304,
+      "learning_rate": 1.2241710780411291e-05,
+      "loss": 0.9315,
+      "step": 2312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5852664391006877,
+      "learning_rate": 1.2235632047155723e-05,
+      "loss": 0.9473,
+      "step": 2313
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6381393387799825,
+      "learning_rate": 1.2229552444226813e-05,
+      "loss": 0.8948,
+      "step": 2314
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6365418078185097,
+      "learning_rate": 1.2223471973989564e-05,
+      "loss": 0.8681,
+      "step": 2315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.570472559133025,
+      "learning_rate": 1.2217390638809317e-05,
+      "loss": 0.9631,
+      "step": 2316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.569978179660728,
+      "learning_rate": 1.2211308441051741e-05,
+      "loss": 0.834,
+      "step": 2317
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.081582386198703,
+      "learning_rate": 1.2205225383082844e-05,
+      "loss": 0.8543,
+      "step": 2318
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6849035750148287,
+      "learning_rate": 1.219914146726897e-05,
+      "loss": 0.9299,
+      "step": 2319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6973317971362039,
+      "learning_rate": 1.2193056695976795e-05,
+      "loss": 0.8657,
+      "step": 2320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.707185823048885,
+      "learning_rate": 1.2186971071573333e-05,
+      "loss": 0.9916,
+      "step": 2321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6368448163606089,
+      "learning_rate": 1.2180884596425924e-05,
+      "loss": 0.9059,
+      "step": 2322
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7449469231031471,
+      "learning_rate": 1.2174797272902238e-05,
+      "loss": 0.9292,
+      "step": 2323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6883743583640474,
+      "learning_rate": 1.2168709103370282e-05,
+      "loss": 1.0231,
+      "step": 2324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7102200609053733,
+      "learning_rate": 1.2162620090198384e-05,
+      "loss": 0.9768,
+      "step": 2325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5029872590217672,
+      "learning_rate": 1.2156530235755209e-05,
+      "loss": 0.7996,
+      "step": 2326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.515762515861755,
+      "learning_rate": 1.215043954240974e-05,
+      "loss": 0.8744,
+      "step": 2327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4870573157571088,
+      "learning_rate": 1.2144348012531295e-05,
+      "loss": 0.9023,
+      "step": 2328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5722520774377307,
+      "learning_rate": 1.2138255648489507e-05,
+      "loss": 0.9547,
+      "step": 2329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6054220739476521,
+      "learning_rate": 1.2132162452654344e-05,
+      "loss": 0.9617,
+      "step": 2330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5581735658279259,
+      "learning_rate": 1.2126068427396094e-05,
+      "loss": 0.9257,
+      "step": 2331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0726213556349586,
+      "learning_rate": 1.2119973575085363e-05,
+      "loss": 0.8889,
+      "step": 2332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9812318871869911,
+      "learning_rate": 1.2113877898093082e-05,
+      "loss": 0.8586,
+      "step": 2333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7528300101695584,
+      "learning_rate": 1.210778139879051e-05,
+      "loss": 0.8612,
+      "step": 2334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6447946149297836,
+      "learning_rate": 1.2101684079549214e-05,
+      "loss": 0.9005,
+      "step": 2335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.614926812137702,
+      "learning_rate": 1.2095585942741086e-05,
+      "loss": 0.9346,
+      "step": 2336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.636018399184569,
+      "learning_rate": 1.2089486990738335e-05,
+      "loss": 0.9194,
+      "step": 2337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7820242310159078,
+      "learning_rate": 1.208338722591349e-05,
+      "loss": 0.8558,
+      "step": 2338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.641815954357134,
+      "learning_rate": 1.207728665063939e-05,
+      "loss": 0.9079,
+      "step": 2339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9713948751022217,
+      "learning_rate": 1.2071185267289196e-05,
+      "loss": 0.9082,
+      "step": 2340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0881037116257228,
+      "learning_rate": 1.2065083078236375e-05,
+      "loss": 0.8544,
+      "step": 2341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7218735663972184,
+      "learning_rate": 1.2058980085854719e-05,
+      "loss": 0.8386,
+      "step": 2342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7207476838500912,
+      "learning_rate": 1.2052876292518323e-05,
+      "loss": 0.9535,
+      "step": 2343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6781793466035655,
+      "learning_rate": 1.2046771700601593e-05,
+      "loss": 0.9131,
+      "step": 2344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0496520830386282,
+      "learning_rate": 1.2040666312479257e-05,
+      "loss": 0.9015,
+      "step": 2345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6204501477922486,
+      "learning_rate": 1.2034560130526341e-05,
+      "loss": 0.8013,
+      "step": 2346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6213860619078162,
+      "learning_rate": 1.2028453157118181e-05,
+      "loss": 0.8374,
+      "step": 2347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.748124267870794,
+      "learning_rate": 1.2022345394630427e-05,
+      "loss": 0.9355,
+      "step": 2348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6642755852057176,
+      "learning_rate": 1.2016236845439035e-05,
+      "loss": 0.9601,
+      "step": 2349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5834804252170924,
+      "learning_rate": 1.2010127511920261e-05,
+      "loss": 0.8813,
+      "step": 2350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8231193886211687,
+      "learning_rate": 1.2004017396450671e-05,
+      "loss": 0.9881,
+      "step": 2351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.629151284252244,
+      "learning_rate": 1.1997906501407132e-05,
+      "loss": 0.9539,
+      "step": 2352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.588829809654717,
+      "learning_rate": 1.1991794829166823e-05,
+      "loss": 0.8471,
+      "step": 2353
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.58915263636295,
+      "learning_rate": 1.1985682382107208e-05,
+      "loss": 0.9533,
+      "step": 2354
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5463505817070597,
+      "learning_rate": 1.1979569162606073e-05,
+      "loss": 0.8634,
+      "step": 2355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1705507882738844,
+      "learning_rate": 1.1973455173041489e-05,
+      "loss": 0.9334,
+      "step": 2356
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6698798026371962,
+      "learning_rate": 1.1967340415791832e-05,
+      "loss": 0.8404,
+      "step": 2357
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8084057266815006,
+      "learning_rate": 1.196122489323578e-05,
+      "loss": 1.0155,
+      "step": 2358
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8154745631021973,
+      "learning_rate": 1.1955108607752301e-05,
+      "loss": 0.9918,
+      "step": 2359
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5894215322465934,
+      "learning_rate": 1.1948991561720669e-05,
+      "loss": 0.9316,
+      "step": 2360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6897606305080461,
+      "learning_rate": 1.1942873757520445e-05,
+      "loss": 0.9895,
+      "step": 2361
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.718742150115532,
+      "learning_rate": 1.1936755197531491e-05,
+      "loss": 0.881,
+      "step": 2362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4838695183397146,
+      "learning_rate": 1.193063588413396e-05,
+      "loss": 0.8451,
+      "step": 2363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5546745507809492,
+      "learning_rate": 1.19245158197083e-05,
+      "loss": 0.8679,
+      "step": 2364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6565188851066053,
+      "learning_rate": 1.1918395006635254e-05,
+      "loss": 0.9966,
+      "step": 2365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7487486513340875,
+      "learning_rate": 1.1912273447295842e-05,
+      "loss": 0.9032,
+      "step": 2366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5551194900957086,
+      "learning_rate": 1.1906151144071392e-05,
+      "loss": 0.8731,
+      "step": 2367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5572547842836877,
+      "learning_rate": 1.1900028099343515e-05,
+      "loss": 0.7964,
+      "step": 2368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.751955626922658,
+      "learning_rate": 1.1893904315494107e-05,
+      "loss": 0.9325,
+      "step": 2369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6290817447037917,
+      "learning_rate": 1.1887779794905351e-05,
+      "loss": 0.9115,
+      "step": 2370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.684528024725099,
+      "learning_rate": 1.1881654539959729e-05,
+      "loss": 0.9037,
+      "step": 2371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8034315441220967,
+      "learning_rate": 1.1875528553039992e-05,
+      "loss": 0.9,
+      "step": 2372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6066181936136708,
+      "learning_rate": 1.1869401836529184e-05,
+      "loss": 0.8758,
+      "step": 2373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9278723102991506,
+      "learning_rate": 1.186327439281063e-05,
+      "loss": 1.023,
+      "step": 2374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5698552200541538,
+      "learning_rate": 1.1857146224267947e-05,
+      "loss": 0.9205,
+      "step": 2375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.653755503771635,
+      "learning_rate": 1.1851017333285022e-05,
+      "loss": 0.8546,
+      "step": 2376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.650343141166476,
+      "learning_rate": 1.1844887722246026e-05,
+      "loss": 0.9252,
+      "step": 2377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.758919208429365,
+      "learning_rate": 1.1838757393535412e-05,
+      "loss": 0.916,
+      "step": 2378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6626598322690156,
+      "learning_rate": 1.1832626349537919e-05,
+      "loss": 0.9084,
+      "step": 2379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6283236310131255,
+      "learning_rate": 1.182649459263855e-05,
+      "loss": 0.9248,
+      "step": 2380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8994396789042338,
+      "learning_rate": 1.1820362125222594e-05,
+      "loss": 0.8982,
+      "step": 2381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8803804290683304,
+      "learning_rate": 1.1814228949675617e-05,
+      "loss": 0.8694,
+      "step": 2382
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9651204603942862,
+      "learning_rate": 1.1808095068383456e-05,
+      "loss": 0.9695,
+      "step": 2383
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7784943781767095,
+      "learning_rate": 1.180196048373223e-05,
+      "loss": 0.9357,
+      "step": 2384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1645318256103658,
+      "learning_rate": 1.1795825198108318e-05,
+      "loss": 0.8955,
+      "step": 2385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.705681392834632,
+      "learning_rate": 1.1789689213898392e-05,
+      "loss": 0.9168,
+      "step": 2386
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7306300017911898,
+      "learning_rate": 1.1783552533489372e-05,
+      "loss": 0.8582,
+      "step": 2387
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6092561732001907,
+      "learning_rate": 1.1777415159268468e-05,
+      "loss": 0.8716,
+      "step": 2388
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6121961454177964,
+      "learning_rate": 1.1771277093623152e-05,
+      "loss": 0.9345,
+      "step": 2389
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0235940071709773,
+      "learning_rate": 1.1765138338941166e-05,
+      "loss": 0.8743,
+      "step": 2390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6650725542166134,
+      "learning_rate": 1.1758998897610516e-05,
+      "loss": 0.9128,
+      "step": 2391
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6838240849652575,
+      "learning_rate": 1.1752858772019485e-05,
+      "loss": 0.8248,
+      "step": 2392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5405143785069657,
+      "learning_rate": 1.174671796455661e-05,
+      "loss": 0.8519,
+      "step": 2393
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7075192846554887,
+      "learning_rate": 1.1740576477610705e-05,
+      "loss": 0.9705,
+      "step": 2394
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5915559469139586,
+      "learning_rate": 1.173443431357084e-05,
+      "loss": 0.8119,
+      "step": 2395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5341539604443246,
+      "learning_rate": 1.1728291474826348e-05,
+      "loss": 0.8359,
+      "step": 2396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7111959044728358,
+      "learning_rate": 1.1722147963766835e-05,
+      "loss": 0.9043,
+      "step": 2397
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.662496762446925,
+      "learning_rate": 1.1716003782782153e-05,
+      "loss": 0.8818,
+      "step": 2398
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6296461909582525,
+      "learning_rate": 1.1709858934262426e-05,
+      "loss": 0.8846,
+      "step": 2399
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5766469513574761,
+      "learning_rate": 1.1703713420598035e-05,
+      "loss": 0.8518,
+      "step": 2400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6211540601985241,
+      "learning_rate": 1.1697567244179622e-05,
+      "loss": 0.904,
+      "step": 2401
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5114466052898647,
+      "learning_rate": 1.169142040739808e-05,
+      "loss": 0.9241,
+      "step": 2402
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5822019774644347,
+      "learning_rate": 1.1685272912644561e-05,
+      "loss": 0.8509,
+      "step": 2403
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0018430388719812,
+      "learning_rate": 1.167912476231048e-05,
+      "loss": 0.8854,
+      "step": 2404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6631887409329853,
+      "learning_rate": 1.1672975958787496e-05,
+      "loss": 0.9394,
+      "step": 2405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.444186862995282,
+      "learning_rate": 1.1666826504467532e-05,
+      "loss": 0.7888,
+      "step": 2406
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7944503005469847,
+      "learning_rate": 1.166067640174276e-05,
+      "loss": 0.8832,
+      "step": 2407
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7834728453717437,
+      "learning_rate": 1.1654525653005604e-05,
+      "loss": 0.9259,
+      "step": 2408
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7332128164682012,
+      "learning_rate": 1.164837426064874e-05,
+      "loss": 0.9517,
+      "step": 2409
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.659247438604464,
+      "learning_rate": 1.164222222706509e-05,
+      "loss": 0.7546,
+      "step": 2410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5563453090100108,
+      "learning_rate": 1.163606955464783e-05,
+      "loss": 0.8656,
+      "step": 2411
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6293861354370793,
+      "learning_rate": 1.1629916245790392e-05,
+      "loss": 0.8636,
+      "step": 2412
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0863983546542098,
+      "learning_rate": 1.1623762302886437e-05,
+      "loss": 0.831,
+      "step": 2413
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.828209954059361,
+      "learning_rate": 1.161760772832989e-05,
+      "loss": 0.9173,
+      "step": 2414
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0137121324801834,
+      "learning_rate": 1.161145252451491e-05,
+      "loss": 0.8057,
+      "step": 2415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9857713425480262,
+      "learning_rate": 1.160529669383591e-05,
+      "loss": 0.8389,
+      "step": 2416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6352355688813127,
+      "learning_rate": 1.1599140238687538e-05,
+      "loss": 0.8737,
+      "step": 2417
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6840348552573543,
+      "learning_rate": 1.1592983161464694e-05,
+      "loss": 0.9751,
+      "step": 2418
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0528803637679105,
+      "learning_rate": 1.1586825464562515e-05,
+      "loss": 0.941,
+      "step": 2419
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.726980622135025,
+      "learning_rate": 1.1580667150376375e-05,
+      "loss": 0.8909,
+      "step": 2420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.585472067063811,
+      "learning_rate": 1.1574508221301897e-05,
+      "loss": 0.9406,
+      "step": 2421
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7269981417087819,
+      "learning_rate": 1.1568348679734938e-05,
+      "loss": 0.8186,
+      "step": 2422
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5885606087580384,
+      "learning_rate": 1.1562188528071591e-05,
+      "loss": 0.9174,
+      "step": 2423
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.535029440925942,
+      "learning_rate": 1.1556027768708191e-05,
+      "loss": 0.824,
+      "step": 2424
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.56283264762811,
+      "learning_rate": 1.154986640404131e-05,
+      "loss": 0.8747,
+      "step": 2425
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5490847865812722,
+      "learning_rate": 1.1543704436467753e-05,
+      "loss": 0.8951,
+      "step": 2426
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8494513289263999,
+      "learning_rate": 1.1537541868384557e-05,
+      "loss": 0.9657,
+      "step": 2427
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.584004582857036,
+      "learning_rate": 1.1531378702188993e-05,
+      "loss": 0.919,
+      "step": 2428
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6687200664721955,
+      "learning_rate": 1.1525214940278575e-05,
+      "loss": 0.9514,
+      "step": 2429
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.600534689188985,
+      "learning_rate": 1.1519050585051036e-05,
+      "loss": 0.8146,
+      "step": 2430
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6736830048680211,
+      "learning_rate": 1.1512885638904343e-05,
+      "loss": 0.8767,
+      "step": 2431
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.579972992425518,
+      "learning_rate": 1.1506720104236699e-05,
+      "loss": 0.8742,
+      "step": 2432
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5549897458293858,
+      "learning_rate": 1.1500553983446527e-05,
+      "loss": 0.8583,
+      "step": 2433
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6716704698557523,
+      "learning_rate": 1.1494387278932489e-05,
+      "loss": 0.8663,
+      "step": 2434
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7083774250110153,
+      "learning_rate": 1.1488219993093457e-05,
+      "loss": 0.8671,
+      "step": 2435
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6067018846964083,
+      "learning_rate": 1.1482052128328544e-05,
+      "loss": 0.9127,
+      "step": 2436
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4769938342211295,
+      "learning_rate": 1.147588368703709e-05,
+      "loss": 0.8525,
+      "step": 2437
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6923231003889867,
+      "learning_rate": 1.1469714671618646e-05,
+      "loss": 0.8387,
+      "step": 2438
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6461448341840523,
+      "learning_rate": 1.1463545084472996e-05,
+      "loss": 0.9499,
+      "step": 2439
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7360314461905522,
+      "learning_rate": 1.1457374928000143e-05,
+      "loss": 0.9681,
+      "step": 2440
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5457334882850975,
+      "learning_rate": 1.1451204204600314e-05,
+      "loss": 0.8696,
+      "step": 2441
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.561625586656231,
+      "learning_rate": 1.1445032916673952e-05,
+      "loss": 0.8564,
+      "step": 2442
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.747483387309643,
+      "learning_rate": 1.1438861066621723e-05,
+      "loss": 0.8932,
+      "step": 2443
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3094491636523606,
+      "learning_rate": 1.1432688656844515e-05,
+      "loss": 0.803,
+      "step": 2444
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7379661376040465,
+      "learning_rate": 1.1426515689743424e-05,
+      "loss": 0.8556,
+      "step": 2445
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8655534910329177,
+      "learning_rate": 1.1420342167719774e-05,
+      "loss": 0.9865,
+      "step": 2446
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5825039469820188,
+      "learning_rate": 1.1414168093175095e-05,
+      "loss": 0.8872,
+      "step": 2447
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5783432114883513,
+      "learning_rate": 1.140799346851114e-05,
+      "loss": 1.0184,
+      "step": 2448
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6838048949133646,
+      "learning_rate": 1.140181829612987e-05,
+      "loss": 0.824,
+      "step": 2449
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7059529383130307,
+      "learning_rate": 1.1395642578433463e-05,
+      "loss": 0.934,
+      "step": 2450
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.532459597194217,
+      "learning_rate": 1.1389466317824308e-05,
+      "loss": 0.8124,
+      "step": 2451
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6707682104800727,
+      "learning_rate": 1.1383289516705002e-05,
+      "loss": 0.8355,
+      "step": 2452
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5905150836191113,
+      "learning_rate": 1.1377112177478357e-05,
+      "loss": 0.9399,
+      "step": 2453
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.476541345700363,
+      "learning_rate": 1.1370934302547392e-05,
+      "loss": 0.91,
+      "step": 2454
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8163916004232388,
+      "learning_rate": 1.1364755894315338e-05,
+      "loss": 0.9637,
+      "step": 2455
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7278870188469235,
+      "learning_rate": 1.135857695518563e-05,
+      "loss": 0.8965,
+      "step": 2456
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6552499182917635,
+      "learning_rate": 1.1352397487561902e-05,
+      "loss": 0.8417,
+      "step": 2457
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.598933278901883,
+      "learning_rate": 1.1346217493848006e-05,
+      "loss": 0.8921,
+      "step": 2458
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.687926256672301,
+      "learning_rate": 1.1340036976448e-05,
+      "loss": 0.8655,
+      "step": 2459
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6035334090769586,
+      "learning_rate": 1.133385593776613e-05,
+      "loss": 0.9147,
+      "step": 2460
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5264152477990032,
+      "learning_rate": 1.132767438020686e-05,
+      "loss": 0.9431,
+      "step": 2461
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5920847675311918,
+      "learning_rate": 1.1321492306174846e-05,
+      "loss": 0.8807,
+      "step": 2462
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.112832842734782,
+      "learning_rate": 1.1315309718074959e-05,
+      "loss": 0.8337,
+      "step": 2463
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5825485343461176,
+      "learning_rate": 1.130912661831225e-05,
+      "loss": 0.8071,
+      "step": 2464
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.688405918216043,
+      "learning_rate": 1.1302943009291983e-05,
+      "loss": 0.8733,
+      "step": 2465
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.692881378968125,
+      "learning_rate": 1.1296758893419619e-05,
+      "loss": 0.9034,
+      "step": 2466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7686242260062677,
+      "learning_rate": 1.1290574273100809e-05,
+      "loss": 0.9479,
+      "step": 2467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6238557173785837,
+      "learning_rate": 1.1284389150741409e-05,
+      "loss": 0.8756,
+      "step": 2468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6007895176406324,
+      "learning_rate": 1.1278203528747462e-05,
+      "loss": 0.8185,
+      "step": 2469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6593428719736762,
+      "learning_rate": 1.1272017409525214e-05,
+      "loss": 0.8083,
+      "step": 2470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7002917279111165,
+      "learning_rate": 1.1265830795481094e-05,
+      "loss": 0.923,
+      "step": 2471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7304355587774407,
+      "learning_rate": 1.1259643689021735e-05,
+      "loss": 0.9441,
+      "step": 2472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6679592855665224,
+      "learning_rate": 1.1253456092553953e-05,
+      "loss": 0.9144,
+      "step": 2473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0243338142166214,
+      "learning_rate": 1.124726800848476e-05,
+      "loss": 0.83,
+      "step": 2474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.604542125189678,
+      "learning_rate": 1.1241079439221351e-05,
+      "loss": 0.9074,
+      "step": 2475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6043307471060633,
+      "learning_rate": 1.1234890387171117e-05,
+      "loss": 0.9517,
+      "step": 2476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5780944028408468,
+      "learning_rate": 1.1228700854741636e-05,
+      "loss": 0.9303,
+      "step": 2477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.775283469539117,
+      "learning_rate": 1.1222510844340665e-05,
+      "loss": 0.8424,
+      "step": 2478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.676019589554559,
+      "learning_rate": 1.1216320358376158e-05,
+      "loss": 0.965,
+      "step": 2479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.638811723020345,
+      "learning_rate": 1.1210129399256244e-05,
+      "loss": 0.8638,
+      "step": 2480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6328412829747403,
+      "learning_rate": 1.1203937969389248e-05,
+      "loss": 0.9258,
+      "step": 2481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7741009577725613,
+      "learning_rate": 1.1197746071183657e-05,
+      "loss": 0.9279,
+      "step": 2482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6400466618549117,
+      "learning_rate": 1.1191553707048165e-05,
+      "loss": 0.9113,
+      "step": 2483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5192636829539932,
+      "learning_rate": 1.1185360879391637e-05,
+      "loss": 0.8468,
+      "step": 2484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0318138922746853,
+      "learning_rate": 1.1179167590623112e-05,
+      "loss": 0.8137,
+      "step": 2485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.732616877086671,
+      "learning_rate": 1.1172973843151816e-05,
+      "loss": 0.9841,
+      "step": 2486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.602059082255045,
+      "learning_rate": 1.116677963938715e-05,
+      "loss": 0.9456,
+      "step": 2487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6484754975073685,
+      "learning_rate": 1.1160584981738698e-05,
+      "loss": 0.9802,
+      "step": 2488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5816914312978256,
+      "learning_rate": 1.1154389872616211e-05,
+      "loss": 0.951,
+      "step": 2489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7201399996596591,
+      "learning_rate": 1.1148194314429624e-05,
+      "loss": 0.9038,
+      "step": 2490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6381211720035134,
+      "learning_rate": 1.1141998309589044e-05,
+      "loss": 0.9237,
+      "step": 2491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6996608044968606,
+      "learning_rate": 1.113580186050475e-05,
+      "loss": 0.8609,
+      "step": 2492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6024660122379049,
+      "learning_rate": 1.1129604969587195e-05,
+      "loss": 0.9167,
+      "step": 2493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6257769816518974,
+      "learning_rate": 1.1123407639247005e-05,
+      "loss": 0.9076,
+      "step": 2494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.704993220468897,
+      "learning_rate": 1.1117209871894981e-05,
+      "loss": 0.8974,
+      "step": 2495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6435068974581712,
+      "learning_rate": 1.1111011669942078e-05,
+      "loss": 0.9255,
+      "step": 2496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.749630028646259,
+      "learning_rate": 1.110481303579944e-05,
+      "loss": 0.9103,
+      "step": 2497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6028774232788787,
+      "learning_rate": 1.109861397187837e-05,
+      "loss": 0.9591,
+      "step": 2498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7460273196123488,
+      "learning_rate": 1.1092414480590332e-05,
+      "loss": 0.9395,
+      "step": 2499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6342526877007284,
+      "learning_rate": 1.1086214564346967e-05,
+      "loss": 0.8496,
+      "step": 2500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6626376416304331,
+      "learning_rate": 1.1080014225560079e-05,
+      "loss": 0.8866,
+      "step": 2501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7229097443862398,
+      "learning_rate": 1.1073813466641633e-05,
+      "loss": 0.9501,
+      "step": 2502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6160182537779568,
+      "learning_rate": 1.1067612290003757e-05,
+      "loss": 0.862,
+      "step": 2503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6000573133865095,
+      "learning_rate": 1.1061410698058744e-05,
+      "loss": 0.968,
+      "step": 2504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5828066722218466,
+      "learning_rate": 1.1055208693219046e-05,
+      "loss": 0.8768,
+      "step": 2505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.706476808576579,
+      "learning_rate": 1.1049006277897283e-05,
+      "loss": 0.8708,
+      "step": 2506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.683797466125745,
+      "learning_rate": 1.1042803454506221e-05,
+      "loss": 0.9767,
+      "step": 2507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0517005900779952,
+      "learning_rate": 1.10366002254588e-05,
+      "loss": 0.8866,
+      "step": 2508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.55527652863224,
+      "learning_rate": 1.1030396593168109e-05,
+      "loss": 0.8991,
+      "step": 2509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.632483914614969,
+      "learning_rate": 1.1024192560047397e-05,
+      "loss": 0.8591,
+      "step": 2510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7280086086399848,
+      "learning_rate": 1.1017988128510062e-05,
+      "loss": 0.8552,
+      "step": 2511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8449699204080128,
+      "learning_rate": 1.1011783300969667e-05,
+      "loss": 0.9839,
+      "step": 2512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1051453107783602,
+      "learning_rate": 1.1005578079839925e-05,
+      "loss": 0.8716,
+      "step": 2513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.066439513175387,
+      "learning_rate": 1.0999372467534699e-05,
+      "loss": 0.8639,
+      "step": 2514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7121892816761888,
+      "learning_rate": 1.0993166466468005e-05,
+      "loss": 0.8558,
+      "step": 2515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6694838802243908,
+      "learning_rate": 1.098696007905402e-05,
+      "loss": 0.9295,
+      "step": 2516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5842819484955362,
+      "learning_rate": 1.0980753307707057e-05,
+      "loss": 0.8784,
+      "step": 2517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6604557498417662,
+      "learning_rate": 1.0974546154841584e-05,
+      "loss": 0.8892,
+      "step": 2518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7836070439679585,
+      "learning_rate": 1.0968338622872219e-05,
+      "loss": 0.9087,
+      "step": 2519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.895327252689359,
+      "learning_rate": 1.0962130714213733e-05,
+      "loss": 0.8696,
+      "step": 2520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6633919360494855,
+      "learning_rate": 1.0955922431281028e-05,
+      "loss": 0.8676,
+      "step": 2521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6419736231200284,
+      "learning_rate": 1.0949713776489164e-05,
+      "loss": 0.9691,
+      "step": 2522
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.561296384767736,
+      "learning_rate": 1.0943504752253342e-05,
+      "loss": 0.7954,
+      "step": 2523
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0538384040555189,
+      "learning_rate": 1.0937295360988909e-05,
+      "loss": 0.8068,
+      "step": 2524
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0548922682154,
+      "learning_rate": 1.0931085605111354e-05,
+      "loss": 0.8274,
+      "step": 2525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5934174901568194,
+      "learning_rate": 1.0924875487036302e-05,
+      "loss": 0.899,
+      "step": 2526
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.704362272374043,
+      "learning_rate": 1.0918665009179525e-05,
+      "loss": 0.859,
+      "step": 2527
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8225572438991193,
+      "learning_rate": 1.0912454173956937e-05,
+      "loss": 0.8878,
+      "step": 2528
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.489382296989872,
+      "learning_rate": 1.0906242983784583e-05,
+      "loss": 0.8924,
+      "step": 2529
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6215320488880864,
+      "learning_rate": 1.0900031441078654e-05,
+      "loss": 0.882,
+      "step": 2530
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6538062323393943,
+      "learning_rate": 1.0893819548255471e-05,
+      "loss": 0.8929,
+      "step": 2531
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7543484617420304,
+      "learning_rate": 1.0887607307731501e-05,
+      "loss": 0.9544,
+      "step": 2532
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5574059064623862,
+      "learning_rate": 1.0881394721923332e-05,
+      "loss": 0.9129,
+      "step": 2533
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8188816245701696,
+      "learning_rate": 1.0875181793247702e-05,
+      "loss": 0.9458,
+      "step": 2534
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.708446147466221,
+      "learning_rate": 1.0868968524121473e-05,
+      "loss": 0.91,
+      "step": 2535
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5085938102934138,
+      "learning_rate": 1.0862754916961637e-05,
+      "loss": 0.8415,
+      "step": 2536
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9704063271563848,
+      "learning_rate": 1.0856540974185328e-05,
+      "loss": 0.7696,
+      "step": 2537
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7056566077912285,
+      "learning_rate": 1.0850326698209798e-05,
+      "loss": 0.8952,
+      "step": 2538
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5772566696700812,
+      "learning_rate": 1.0844112091452443e-05,
+      "loss": 0.8908,
+      "step": 2539
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6313038628318326,
+      "learning_rate": 1.0837897156330774e-05,
+      "loss": 0.8325,
+      "step": 2540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.63719309335538,
+      "learning_rate": 1.0831681895262435e-05,
+      "loss": 0.86,
+      "step": 2541
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0020153565871255,
+      "learning_rate": 1.0825466310665202e-05,
+      "loss": 0.8439,
+      "step": 2542
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5400550721122903,
+      "learning_rate": 1.0819250404956967e-05,
+      "loss": 0.8367,
+      "step": 2543
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6940483805250788,
+      "learning_rate": 1.0813034180555755e-05,
+      "loss": 0.9796,
+      "step": 2544
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5723257187622544,
+      "learning_rate": 1.080681763987971e-05,
+      "loss": 0.9397,
+      "step": 2545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8321927963341342,
+      "learning_rate": 1.0800600785347103e-05,
+      "loss": 0.8792,
+      "step": 2546
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9839156891038674,
+      "learning_rate": 1.0794383619376325e-05,
+      "loss": 0.8371,
+      "step": 2547
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8998655020364166,
+      "learning_rate": 1.0788166144385888e-05,
+      "loss": 0.9333,
+      "step": 2548
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.980212742905011,
+      "learning_rate": 1.078194836279442e-05,
+      "loss": 0.8316,
+      "step": 2549
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6263547141122499,
+      "learning_rate": 1.0775730277020682e-05,
+      "loss": 0.8021,
+      "step": 2550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7378580074208123,
+      "learning_rate": 1.0769511889483534e-05,
+      "loss": 0.92,
+      "step": 2551
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6601559370094976,
+      "learning_rate": 1.076329320260197e-05,
+      "loss": 0.9671,
+      "step": 2552
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7234333502524333,
+      "learning_rate": 1.0757074218795091e-05,
+      "loss": 0.8602,
+      "step": 2553
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5449905353717643,
+      "learning_rate": 1.0750854940482118e-05,
+      "loss": 0.8734,
+      "step": 2554
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6058871051008872,
+      "learning_rate": 1.0744635370082384e-05,
+      "loss": 0.8336,
+      "step": 2555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.735680484791797,
+      "learning_rate": 1.0738415510015336e-05,
+      "loss": 0.9391,
+      "step": 2556
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6228456121338202,
+      "learning_rate": 1.073219536270054e-05,
+      "loss": 0.903,
+      "step": 2557
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8983600406186754,
+      "learning_rate": 1.072597493055766e-05,
+      "loss": 0.9682,
+      "step": 2558
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6693268848764007,
+      "learning_rate": 1.071975421600648e-05,
+      "loss": 0.9151,
+      "step": 2559
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7551807879851804,
+      "learning_rate": 1.0713533221466897e-05,
+      "loss": 0.8728,
+      "step": 2560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6700286455405353,
+      "learning_rate": 1.0707311949358906e-05,
+      "loss": 0.946,
+      "step": 2561
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.660246567369808,
+      "learning_rate": 1.0701090402102624e-05,
+      "loss": 0.9628,
+      "step": 2562
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4993621366836158,
+      "learning_rate": 1.0694868582118266e-05,
+      "loss": 0.8558,
+      "step": 2563
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.79913245437129,
+      "learning_rate": 1.0688646491826154e-05,
+      "loss": 0.9495,
+      "step": 2564
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5871582655142278,
+      "learning_rate": 1.0682424133646712e-05,
+      "loss": 0.8386,
+      "step": 2565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6374166057551383,
+      "learning_rate": 1.0676201510000474e-05,
+      "loss": 0.8815,
+      "step": 2566
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7580364721790434,
+      "learning_rate": 1.0669978623308077e-05,
+      "loss": 0.902,
+      "step": 2567
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0617083751057919,
+      "learning_rate": 1.0663755475990261e-05,
+      "loss": 0.8752,
+      "step": 2568
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6778028355635968,
+      "learning_rate": 1.0657532070467857e-05,
+      "loss": 0.905,
+      "step": 2569
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7262810567136453,
+      "learning_rate": 1.065130840916181e-05,
+      "loss": 0.9073,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5880254713941317,
+      "learning_rate": 1.0645084494493166e-05,
+      "loss": 0.9341,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7076889903354375,
+      "learning_rate": 1.0638860328883049e-05,
+      "loss": 0.8659,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.530980372166329,
+      "learning_rate": 1.06326359147527e-05,
+      "loss": 0.8848,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7293964084217277,
+      "learning_rate": 1.062641125452345e-05,
+      "loss": 0.8931,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6328744954290797,
+      "learning_rate": 1.062018635061673e-05,
+      "loss": 0.8845,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5920135010535654,
+      "learning_rate": 1.0613961205454057e-05,
+      "loss": 0.9125,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6689150965013047,
+      "learning_rate": 1.0607735821457053e-05,
+      "loss": 0.8153,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5580108963116983,
+      "learning_rate": 1.060151020104742e-05,
+      "loss": 0.9602,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5384216662323622,
+      "learning_rate": 1.0595284346646968e-05,
+      "loss": 0.8817,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5694727827641175,
+      "learning_rate": 1.0589058260677581e-05,
+      "loss": 0.9036,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6509380161701281,
+      "learning_rate": 1.0582831945561242e-05,
+      "loss": 0.9603,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6022754231614071,
+      "learning_rate": 1.0576605403720031e-05,
+      "loss": 0.9359,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6312373889447316,
+      "learning_rate": 1.0570378637576097e-05,
+      "loss": 0.9703,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6026495273894386,
+      "learning_rate": 1.056415164955169e-05,
+      "loss": 0.8243,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5141648134637216,
+      "learning_rate": 1.0557924442069148e-05,
+      "loss": 0.8867,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7748722238784276,
+      "learning_rate": 1.0551697017550883e-05,
+      "loss": 0.8468,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7905639173609451,
+      "learning_rate": 1.0545469378419402e-05,
+      "loss": 0.9255,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7953906769802488,
+      "learning_rate": 1.0539241527097288e-05,
+      "loss": 0.8867,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7697074206170684,
+      "learning_rate": 1.0533013466007213e-05,
+      "loss": 0.921,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.604101851664955,
+      "learning_rate": 1.0526785197571925e-05,
+      "loss": 0.8313,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5548373928984829,
+      "learning_rate": 1.0520556724214258e-05,
+      "loss": 0.8476,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5483436837571907,
+      "learning_rate": 1.0514328048357118e-05,
+      "loss": 0.8595,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6865799306159428,
+      "learning_rate": 1.0508099172423503e-05,
+      "loss": 0.958,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5223350419556696,
+      "learning_rate": 1.0501870098836473e-05,
+      "loss": 0.8857,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.652060183344791,
+      "learning_rate": 1.0495640830019173e-05,
+      "loss": 0.93,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8777297945381404,
+      "learning_rate": 1.0489411368394825e-05,
+      "loss": 0.9434,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.618584309310693,
+      "learning_rate": 1.0483181716386727e-05,
+      "loss": 0.8668,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.788852906124346,
+      "learning_rate": 1.0476951876418244e-05,
+      "loss": 1.0191,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.036832185244328,
+      "learning_rate": 1.0470721850912821e-05,
+      "loss": 0.8631,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.804085698473126,
+      "learning_rate": 1.046449164229397e-05,
+      "loss": 0.9201,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.577800079219274,
+      "learning_rate": 1.0458261252985283e-05,
+      "loss": 0.9055,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.701003594870017,
+      "learning_rate": 1.0452030685410409e-05,
+      "loss": 0.8759,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6669343906105938,
+      "learning_rate": 1.0445799941993077e-05,
+      "loss": 0.9175,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6455853941570702,
+      "learning_rate": 1.0439569025157081e-05,
+      "loss": 0.8811,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6448332243656791,
+      "learning_rate": 1.0433337937326282e-05,
+      "loss": 0.9224,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6590686518032083,
+      "learning_rate": 1.0427106680924608e-05,
+      "loss": 0.8032,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.724091098480085,
+      "learning_rate": 1.042087525837605e-05,
+      "loss": 0.9284,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6894900411041467,
+      "learning_rate": 1.0414643672104673e-05,
+      "loss": 0.9109,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.944866261038826,
+      "learning_rate": 1.0408411924534592e-05,
+      "loss": 0.7628,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0015248418941,
+      "learning_rate": 1.0402180018089993e-05,
+      "loss": 0.7692,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6503490381223511,
+      "learning_rate": 1.0395947955195124e-05,
+      "loss": 0.8492,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9934133514431954,
+      "learning_rate": 1.0389715738274294e-05,
+      "loss": 0.7984,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5913575149467811,
+      "learning_rate": 1.0383483369751864e-05,
+      "loss": 0.916,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6442960948374028,
+      "learning_rate": 1.0377250852052263e-05,
+      "loss": 0.8679,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6205029098205084,
+      "learning_rate": 1.0371018187599982e-05,
+      "loss": 0.9516,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7795649536483642,
+      "learning_rate": 1.0364785378819551e-05,
+      "loss": 0.8446,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6327253275166451,
+      "learning_rate": 1.0358552428135576e-05,
+      "loss": 0.9582,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.969972007671843,
+      "learning_rate": 1.0352319337972703e-05,
+      "loss": 0.9707,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.70479406272249,
+      "learning_rate": 1.0346086110755644e-05,
+      "loss": 0.8643,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5431323092797076,
+      "learning_rate": 1.0339852748909155e-05,
+      "loss": 0.8064,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.759324447728648,
+      "learning_rate": 1.0333619254858054e-05,
+      "loss": 0.8488,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3668765862921222,
+      "learning_rate": 1.0327385631027199e-05,
+      "loss": 0.8383,
+      "step": 2622
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7748499689520665,
+      "learning_rate": 1.032115187984151e-05,
+      "loss": 0.925,
+      "step": 2623
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7707471979501763,
+      "learning_rate": 1.0314918003725945e-05,
+      "loss": 0.9398,
+      "step": 2624
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5663114145470098,
+      "learning_rate": 1.030868400510552e-05,
+      "loss": 0.8632,
+      "step": 2625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6057451746915887,
+      "learning_rate": 1.0302449886405297e-05,
+      "loss": 0.8129,
+      "step": 2626
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6631617752176138,
+      "learning_rate": 1.0296215650050378e-05,
+      "loss": 0.8626,
+      "step": 2627
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.749909447767286,
+      "learning_rate": 1.0289981298465921e-05,
+      "loss": 0.8608,
+      "step": 2628
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5997919024504486,
+      "learning_rate": 1.028374683407712e-05,
+      "loss": 0.8067,
+      "step": 2629
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7545790986314025,
+      "learning_rate": 1.0277512259309219e-05,
+      "loss": 0.7941,
+      "step": 2630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6582118363557157,
+      "learning_rate": 1.0271277576587499e-05,
+      "loss": 0.7819,
+      "step": 2631
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7928696081410378,
+      "learning_rate": 1.026504278833729e-05,
+      "loss": 1.0137,
+      "step": 2632
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7651589172161002,
+      "learning_rate": 1.0258807896983957e-05,
+      "loss": 0.9539,
+      "step": 2633
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.733500098449855,
+      "learning_rate": 1.0252572904952909e-05,
+      "loss": 0.9811,
+      "step": 2634
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6157288899218702,
+      "learning_rate": 1.0246337814669589e-05,
+      "loss": 0.9352,
+      "step": 2635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6820995408578228,
+      "learning_rate": 1.0240102628559487e-05,
+      "loss": 0.9355,
+      "step": 2636
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7392672004338872,
+      "learning_rate": 1.0233867349048125e-05,
+      "loss": 0.8796,
+      "step": 2637
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5766758483730425,
+      "learning_rate": 1.0227631978561057e-05,
+      "loss": 0.8922,
+      "step": 2638
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.667170082694874,
+      "learning_rate": 1.0221396519523881e-05,
+      "loss": 0.8648,
+      "step": 2639
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7055214693990044,
+      "learning_rate": 1.0215160974362224e-05,
+      "loss": 0.8865,
+      "step": 2640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5642308389440769,
+      "learning_rate": 1.0208925345501747e-05,
+      "loss": 0.8733,
+      "step": 2641
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7446340667025126,
+      "learning_rate": 1.0202689635368143e-05,
+      "loss": 0.8817,
+      "step": 2642
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6081738516053856,
+      "learning_rate": 1.0196453846387142e-05,
+      "loss": 1.0279,
+      "step": 2643
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5564230881542247,
+      "learning_rate": 1.01902179809845e-05,
+      "loss": 0.9169,
+      "step": 2644
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6271563208949373,
+      "learning_rate": 1.0183982041586001e-05,
+      "loss": 0.9401,
+      "step": 2645
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7240913517326828,
+      "learning_rate": 1.017774603061746e-05,
+      "loss": 1.0292,
+      "step": 2646
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1045348553568979,
+      "learning_rate": 1.0171509950504724e-05,
+      "loss": 0.8517,
+      "step": 2647
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7194730100037476,
+      "learning_rate": 1.0165273803673658e-05,
+      "loss": 0.8299,
+      "step": 2648
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6152857454031748,
+      "learning_rate": 1.0159037592550162e-05,
+      "loss": 0.9132,
+      "step": 2649
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7503180469098836,
+      "learning_rate": 1.0152801319560152e-05,
+      "loss": 0.8828,
+      "step": 2650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7123915855665384,
+      "learning_rate": 1.0146564987129581e-05,
+      "loss": 0.92,
+      "step": 2651
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.730123034388108,
+      "learning_rate": 1.014032859768441e-05,
+      "loss": 0.9958,
+      "step": 2652
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5950184960317233,
+      "learning_rate": 1.0134092153650631e-05,
+      "loss": 0.827,
+      "step": 2653
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6636339087834497,
+      "learning_rate": 1.0127855657454258e-05,
+      "loss": 0.9717,
+      "step": 2654
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6689770277511726,
+      "learning_rate": 1.0121619111521322e-05,
+      "loss": 0.8787,
+      "step": 2655
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6433761233345823,
+      "learning_rate": 1.011538251827787e-05,
+      "loss": 0.9031,
+      "step": 2656
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.843680533395224,
+      "learning_rate": 1.0109145880149975e-05,
+      "loss": 0.9406,
+      "step": 2657
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4700620403103788,
+      "learning_rate": 1.0102909199563726e-05,
+      "loss": 0.8803,
+      "step": 2658
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6998878550112213,
+      "learning_rate": 1.0096672478945222e-05,
+      "loss": 0.8799,
+      "step": 2659
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.704048201924426,
+      "learning_rate": 1.0090435720720585e-05,
+      "loss": 0.8514,
+      "step": 2660
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6532241710042284,
+      "learning_rate": 1.0084198927315948e-05,
+      "loss": 0.8854,
+      "step": 2661
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7039978276310852,
+      "learning_rate": 1.0077962101157462e-05,
+      "loss": 0.9448,
+      "step": 2662
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7357466691093242,
+      "learning_rate": 1.0071725244671281e-05,
+      "loss": 0.9426,
+      "step": 2663
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5965968442553984,
+      "learning_rate": 1.006548836028358e-05,
+      "loss": 0.8188,
+      "step": 2664
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6832792572274011,
+      "learning_rate": 1.005925145042054e-05,
+      "loss": 0.886,
+      "step": 2665
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7185159518002575,
+      "learning_rate": 1.0053014517508357e-05,
+      "loss": 0.9258,
+      "step": 2666
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6676893789715022,
+      "learning_rate": 1.0046777563973229e-05,
+      "loss": 0.9161,
+      "step": 2667
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6215091649752167,
+      "learning_rate": 1.0040540592241369e-05,
+      "loss": 0.8537,
+      "step": 2668
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7942476610526024,
+      "learning_rate": 1.0034303604738991e-05,
+      "loss": 0.8486,
+      "step": 2669
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6828762346113042,
+      "learning_rate": 1.0028066603892321e-05,
+      "loss": 0.994,
+      "step": 2670
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6073183478788597,
+      "learning_rate": 1.0021829592127586e-05,
+      "loss": 0.8439,
+      "step": 2671
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8453040215795065,
+      "learning_rate": 1.0015592571871016e-05,
+      "loss": 0.8188,
+      "step": 2672
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7595581370364481,
+      "learning_rate": 1.000935554554885e-05,
+      "loss": 1.0148,
+      "step": 2673
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7266685362864227,
+      "learning_rate": 1.0003118515587324e-05,
+      "loss": 0.8457,
+      "step": 2674
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5411105545827308,
+      "learning_rate": 9.99688148441268e-06,
+      "loss": 0.902,
+      "step": 2675
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7193590789198092,
+      "learning_rate": 9.990644454451155e-06,
+      "loss": 0.9516,
+      "step": 2676
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.804369296890046,
+      "learning_rate": 9.984407428128988e-06,
+      "loss": 0.8689,
+      "step": 2677
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8373927163972414,
+      "learning_rate": 9.978170407872419e-06,
+      "loss": 0.9253,
+      "step": 2678
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5566264846944136,
+      "learning_rate": 9.971933396107682e-06,
+      "loss": 0.8905,
+      "step": 2679
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7790722258658354,
+      "learning_rate": 9.965696395261012e-06,
+      "loss": 0.9341,
+      "step": 2680
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6083660993105502,
+      "learning_rate": 9.959459407758637e-06,
+      "loss": 0.8901,
+      "step": 2681
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.656324819410977,
+      "learning_rate": 9.953222436026776e-06,
+      "loss": 0.8254,
+      "step": 2682
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6222188172335699,
+      "learning_rate": 9.946985482491648e-06,
+      "loss": 0.8604,
+      "step": 2683
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8065269391988577,
+      "learning_rate": 9.940748549579463e-06,
+      "loss": 0.8606,
+      "step": 2684
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6029232804771094,
+      "learning_rate": 9.934511639716427e-06,
+      "loss": 0.8341,
+      "step": 2685
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5038291032142035,
+      "learning_rate": 9.928274755328724e-06,
+      "loss": 0.8392,
+      "step": 2686
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5999197440458548,
+      "learning_rate": 9.922037898842541e-06,
+      "loss": 0.8171,
+      "step": 2687
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9590512411669851,
+      "learning_rate": 9.915801072684055e-06,
+      "loss": 0.8816,
+      "step": 2688
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.767352321903527,
+      "learning_rate": 9.909564279279416e-06,
+      "loss": 0.9132,
+      "step": 2689
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6048755713165277,
+      "learning_rate": 9.90332752105478e-06,
+      "loss": 0.9553,
+      "step": 2690
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5714941041644446,
+      "learning_rate": 9.897090800436277e-06,
+      "loss": 0.9023,
+      "step": 2691
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6031902743614117,
+      "learning_rate": 9.890854119850027e-06,
+      "loss": 0.7784,
+      "step": 2692
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5367992451846328,
+      "learning_rate": 9.884617481722133e-06,
+      "loss": 0.8808,
+      "step": 2693
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8292338595882096,
+      "learning_rate": 9.878380888478681e-06,
+      "loss": 0.8974,
+      "step": 2694
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8808832153869115,
+      "learning_rate": 9.872144342545745e-06,
+      "loss": 0.8886,
+      "step": 2695
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6647137580503564,
+      "learning_rate": 9.86590784634937e-06,
+      "loss": 0.8641,
+      "step": 2696
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5415878707172042,
+      "learning_rate": 9.859671402315592e-06,
+      "loss": 0.9605,
+      "step": 2697
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6556483095855636,
+      "learning_rate": 9.85343501287042e-06,
+      "loss": 0.8947,
+      "step": 2698
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5290110275061344,
+      "learning_rate": 9.84719868043985e-06,
+      "loss": 0.8059,
+      "step": 2699
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6441651461039437,
+      "learning_rate": 9.840962407449841e-06,
+      "loss": 0.8939,
+      "step": 2700
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5588573803770556,
+      "learning_rate": 9.834726196326344e-06,
+      "loss": 0.8649,
+      "step": 2701
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6195868579049366,
+      "learning_rate": 9.828490049495281e-06,
+      "loss": 0.9323,
+      "step": 2702
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.214387222290087,
+      "learning_rate": 9.822253969382542e-06,
+      "loss": 0.8674,
+      "step": 2703
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.661137771115788,
+      "learning_rate": 9.816017958414002e-06,
+      "loss": 0.8358,
+      "step": 2704
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9057113313335587,
+      "learning_rate": 9.809782019015503e-06,
+      "loss": 0.9369,
+      "step": 2705
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5710964496995494,
+      "learning_rate": 9.80354615361286e-06,
+      "loss": 0.8651,
+      "step": 2706
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5178464795948843,
+      "learning_rate": 9.797310364631859e-06,
+      "loss": 0.9048,
+      "step": 2707
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6501629321488704,
+      "learning_rate": 9.791074654498254e-06,
+      "loss": 0.8935,
+      "step": 2708
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0116546529607562,
+      "learning_rate": 9.78483902563778e-06,
+      "loss": 0.7828,
+      "step": 2709
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5951035485104152,
+      "learning_rate": 9.77860348047612e-06,
+      "loss": 0.9576,
+      "step": 2710
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7229865236771669,
+      "learning_rate": 9.772368021438943e-06,
+      "loss": 0.8432,
+      "step": 2711
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6306035716808758,
+      "learning_rate": 9.766132650951874e-06,
+      "loss": 0.8785,
+      "step": 2712
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6705622889489669,
+      "learning_rate": 9.759897371440513e-06,
+      "loss": 0.8872,
+      "step": 2713
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6313171382399358,
+      "learning_rate": 9.75366218533041e-06,
+      "loss": 0.8913,
+      "step": 2714
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5668687057252857,
+      "learning_rate": 9.747427095047091e-06,
+      "loss": 0.8911,
+      "step": 2715
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0201958055128055,
+      "learning_rate": 9.741192103016044e-06,
+      "loss": 0.8422,
+      "step": 2716
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4812933421290355,
+      "learning_rate": 9.734957211662712e-06,
+      "loss": 0.8336,
+      "step": 2717
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4552819620391966,
+      "learning_rate": 9.728722423412501e-06,
+      "loss": 0.9294,
+      "step": 2718
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6243711305628117,
+      "learning_rate": 9.722487740690781e-06,
+      "loss": 0.8332,
+      "step": 2719
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5539162337791288,
+      "learning_rate": 9.71625316592288e-06,
+      "loss": 0.8171,
+      "step": 2720
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8147539453558093,
+      "learning_rate": 9.710018701534084e-06,
+      "loss": 0.9224,
+      "step": 2721
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8366109133899138,
+      "learning_rate": 9.703784349949627e-06,
+      "loss": 0.9049,
+      "step": 2722
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4517796704452655,
+      "learning_rate": 9.697550113594708e-06,
+      "loss": 0.8616,
+      "step": 2723
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8868446840350614,
+      "learning_rate": 9.691315994894484e-06,
+      "loss": 0.8284,
+      "step": 2724
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7855944868884226,
+      "learning_rate": 9.68508199627406e-06,
+      "loss": 0.8641,
+      "step": 2725
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0425103460772933,
+      "learning_rate": 9.678848120158496e-06,
+      "loss": 0.8758,
+      "step": 2726
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7518379659639014,
+      "learning_rate": 9.672614368972805e-06,
+      "loss": 0.8636,
+      "step": 2727
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7520032016096,
+      "learning_rate": 9.666380745141953e-06,
+      "loss": 0.9324,
+      "step": 2728
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6114363159725842,
+      "learning_rate": 9.660147251090848e-06,
+      "loss": 0.887,
+      "step": 2729
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6150956410057926,
+      "learning_rate": 9.653913889244361e-06,
+      "loss": 0.8463,
+      "step": 2730
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1122149070850331,
+      "learning_rate": 9.6476806620273e-06,
+      "loss": 0.8311,
+      "step": 2731
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7809864619675158,
+      "learning_rate": 9.641447571864429e-06,
+      "loss": 0.8596,
+      "step": 2732
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5506462870826228,
+      "learning_rate": 9.635214621180452e-06,
+      "loss": 0.8777,
+      "step": 2733
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0420118383686607,
+      "learning_rate": 9.628981812400022e-06,
+      "loss": 0.8338,
+      "step": 2734
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6281409989400513,
+      "learning_rate": 9.622749147947739e-06,
+      "loss": 0.933,
+      "step": 2735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7488269065905813,
+      "learning_rate": 9.61651663024814e-06,
+      "loss": 0.8688,
+      "step": 2736
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.892099085704269,
+      "learning_rate": 9.610284261725711e-06,
+      "loss": 0.9789,
+      "step": 2737
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5666296874974874,
+      "learning_rate": 9.604052044804877e-06,
+      "loss": 0.9279,
+      "step": 2738
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6307432073103498,
+      "learning_rate": 9.59781998191001e-06,
+      "loss": 0.9003,
+      "step": 2739
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5652833226501377,
+      "learning_rate": 9.591588075465411e-06,
+      "loss": 0.8563,
+      "step": 2740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.857327510902887,
+      "learning_rate": 9.58535632789533e-06,
+      "loss": 0.9486,
+      "step": 2741
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.55435370797295,
+      "learning_rate": 9.579124741623953e-06,
+      "loss": 0.8031,
+      "step": 2742
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6698556356369156,
+      "learning_rate": 9.572893319075396e-06,
+      "loss": 0.8579,
+      "step": 2743
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7445207416915745,
+      "learning_rate": 9.56666206267372e-06,
+      "loss": 0.9423,
+      "step": 2744
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6680905933702188,
+      "learning_rate": 9.56043097484292e-06,
+      "loss": 0.9176,
+      "step": 2745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.795640370198453,
+      "learning_rate": 9.554200058006926e-06,
+      "loss": 0.9395,
+      "step": 2746
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6996323033281253,
+      "learning_rate": 9.547969314589593e-06,
+      "loss": 0.8588,
+      "step": 2747
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6997459546236668,
+      "learning_rate": 9.541738747014719e-06,
+      "loss": 0.9418,
+      "step": 2748
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7748432851364173,
+      "learning_rate": 9.535508357706031e-06,
+      "loss": 0.8796,
+      "step": 2749
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7854786737289723,
+      "learning_rate": 9.529278149087182e-06,
+      "loss": 0.8415,
+      "step": 2750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.770385366206982,
+      "learning_rate": 9.523048123581758e-06,
+      "loss": 0.7899,
+      "step": 2751
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6366548207955924,
+      "learning_rate": 9.516818283613274e-06,
+      "loss": 0.9371,
+      "step": 2752
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.839373020842207,
+      "learning_rate": 9.510588631605176e-06,
+      "loss": 0.9645,
+      "step": 2753
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9008678155055228,
+      "learning_rate": 9.50435916998083e-06,
+      "loss": 0.9297,
+      "step": 2754
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5696009779369189,
+      "learning_rate": 9.49812990116353e-06,
+      "loss": 0.7761,
+      "step": 2755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7728095139118927,
+      "learning_rate": 9.491900827576502e-06,
+      "loss": 0.8757,
+      "step": 2756
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6743173632051649,
+      "learning_rate": 9.485671951642884e-06,
+      "loss": 0.9276,
+      "step": 2757
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8619025924233175,
+      "learning_rate": 9.479443275785744e-06,
+      "loss": 0.9839,
+      "step": 2758
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6563261443174515,
+      "learning_rate": 9.473214802428076e-06,
+      "loss": 0.9335,
+      "step": 2759
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6365247584476026,
+      "learning_rate": 9.46698653399279e-06,
+      "loss": 0.8441,
+      "step": 2760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6826758567766007,
+      "learning_rate": 9.460758472902714e-06,
+      "loss": 0.8519,
+      "step": 2761
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7773363340148975,
+      "learning_rate": 9.4545306215806e-06,
+      "loss": 0.8723,
+      "step": 2762
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6373183570799716,
+      "learning_rate": 9.448302982449119e-06,
+      "loss": 0.8511,
+      "step": 2763
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.87536734059661,
+      "learning_rate": 9.442075557930854e-06,
+      "loss": 0.9205,
+      "step": 2764
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.687793104717551,
+      "learning_rate": 9.43584835044831e-06,
+      "loss": 0.8572,
+      "step": 2765
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6710465700967605,
+      "learning_rate": 9.429621362423903e-06,
+      "loss": 0.9558,
+      "step": 2766
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5365386871994817,
+      "learning_rate": 9.423394596279974e-06,
+      "loss": 0.8584,
+      "step": 2767
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.596781391833034,
+      "learning_rate": 9.417168054438761e-06,
+      "loss": 0.886,
+      "step": 2768
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6868043609833412,
+      "learning_rate": 9.410941739322424e-06,
+      "loss": 0.8879,
+      "step": 2769
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6804530289583974,
+      "learning_rate": 9.404715653353039e-06,
+      "loss": 0.9526,
+      "step": 2770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5902494438971557,
+      "learning_rate": 9.398489798952582e-06,
+      "loss": 0.9107,
+      "step": 2771
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6784503668340927,
+      "learning_rate": 9.392264178542954e-06,
+      "loss": 0.9301,
+      "step": 2772
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7735091600671817,
+      "learning_rate": 9.386038794545948e-06,
+      "loss": 0.7587,
+      "step": 2773
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.120800593265598,
+      "learning_rate": 9.379813649383274e-06,
+      "loss": 0.9265,
+      "step": 2774
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8926406169244583,
+      "learning_rate": 9.373588745476555e-06,
+      "loss": 0.9112,
+      "step": 2775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6599883578978085,
+      "learning_rate": 9.367364085247306e-06,
+      "loss": 0.8855,
+      "step": 2776
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5775486781209005,
+      "learning_rate": 9.361139671116956e-06,
+      "loss": 0.8847,
+      "step": 2777
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7403389277744925,
+      "learning_rate": 9.354915505506839e-06,
+      "loss": 0.882,
+      "step": 2778
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.630862033418489,
+      "learning_rate": 9.348691590838191e-06,
+      "loss": 0.8179,
+      "step": 2779
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6902613120507355,
+      "learning_rate": 9.342467929532146e-06,
+      "loss": 0.9301,
+      "step": 2780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6224706144049812,
+      "learning_rate": 9.336244524009744e-06,
+      "loss": 0.8695,
+      "step": 2781
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7837960166866742,
+      "learning_rate": 9.330021376691926e-06,
+      "loss": 0.9399,
+      "step": 2782
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7225231730601518,
+      "learning_rate": 9.323798489999529e-06,
+      "loss": 0.8059,
+      "step": 2783
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4866751441193808,
+      "learning_rate": 9.317575866353293e-06,
+      "loss": 0.8574,
+      "step": 2784
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6662702227359207,
+      "learning_rate": 9.31135350817385e-06,
+      "loss": 0.9331,
+      "step": 2785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.524876444930354,
+      "learning_rate": 9.305131417881736e-06,
+      "loss": 0.9129,
+      "step": 2786
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5495361096031077,
+      "learning_rate": 9.298909597897377e-06,
+      "loss": 0.8761,
+      "step": 2787
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6778036930775928,
+      "learning_rate": 9.292688050641095e-06,
+      "loss": 0.913,
+      "step": 2788
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5866709431365775,
+      "learning_rate": 9.286466778533107e-06,
+      "loss": 0.9243,
+      "step": 2789
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6470173264746972,
+      "learning_rate": 9.280245783993524e-06,
+      "loss": 0.8999,
+      "step": 2790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5557053792433595,
+      "learning_rate": 9.274025069442346e-06,
+      "loss": 0.8846,
+      "step": 2791
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0936770191818879,
+      "learning_rate": 9.267804637299464e-06,
+      "loss": 0.8567,
+      "step": 2792
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.793148210921839,
+      "learning_rate": 9.261584489984666e-06,
+      "loss": 0.876,
+      "step": 2793
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6488187210330383,
+      "learning_rate": 9.255364629917618e-06,
+      "loss": 0.9208,
+      "step": 2794
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6600249165863183,
+      "learning_rate": 9.249145059517884e-06,
+      "loss": 0.8407,
+      "step": 2795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9794255140887218,
+      "learning_rate": 9.24292578120491e-06,
+      "loss": 0.7815,
+      "step": 2796
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6245396295614274,
+      "learning_rate": 9.236706797398033e-06,
+      "loss": 0.8743,
+      "step": 2797
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0751690579715993,
+      "learning_rate": 9.230488110516468e-06,
+      "loss": 0.836,
+      "step": 2798
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7327803047337416,
+      "learning_rate": 9.224269722979322e-06,
+      "loss": 0.8167,
+      "step": 2799
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6874444093703491,
+      "learning_rate": 9.218051637205581e-06,
+      "loss": 0.9271,
+      "step": 2800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.583593130189749,
+      "learning_rate": 9.211833855614115e-06,
+      "loss": 0.939,
+      "step": 2801
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4600652268053038,
+      "learning_rate": 9.205616380623676e-06,
+      "loss": 0.8482,
+      "step": 2802
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5548180033483459,
+      "learning_rate": 9.199399214652897e-06,
+      "loss": 0.8916,
+      "step": 2803
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6196516714079978,
+      "learning_rate": 9.19318236012029e-06,
+      "loss": 0.9157,
+      "step": 2804
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6975192007158488,
+      "learning_rate": 9.186965819444245e-06,
+      "loss": 0.8882,
+      "step": 2805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5975043424529383,
+      "learning_rate": 9.180749595043033e-06,
+      "loss": 0.8964,
+      "step": 2806
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0550851702735273,
+      "learning_rate": 9.1745336893348e-06,
+      "loss": 0.7974,
+      "step": 2807
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7791421366646047,
+      "learning_rate": 9.168318104737565e-06,
+      "loss": 0.9304,
+      "step": 2808
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6233042747700792,
+      "learning_rate": 9.162102843669227e-06,
+      "loss": 0.8334,
+      "step": 2809
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8712068900948575,
+      "learning_rate": 9.155887908547557e-06,
+      "loss": 0.8762,
+      "step": 2810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8419421935171567,
+      "learning_rate": 9.149673301790202e-06,
+      "loss": 0.9353,
+      "step": 2811
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.995362902803064,
+      "learning_rate": 9.143459025814673e-06,
+      "loss": 0.8242,
+      "step": 2812
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5932851905915137,
+      "learning_rate": 9.137245083038363e-06,
+      "loss": 0.9527,
+      "step": 2813
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6317124893687645,
+      "learning_rate": 9.131031475878532e-06,
+      "loss": 0.9472,
+      "step": 2814
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6198318639967977,
+      "learning_rate": 9.124818206752301e-06,
+      "loss": 0.9524,
+      "step": 2815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6695432421633027,
+      "learning_rate": 9.118605278076673e-06,
+      "loss": 0.8814,
+      "step": 2816
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7116812936283503,
+      "learning_rate": 9.112392692268506e-06,
+      "loss": 0.9454,
+      "step": 2817
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7949429452740104,
+      "learning_rate": 9.106180451744532e-06,
+      "loss": 0.9934,
+      "step": 2818
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8199922468280891,
+      "learning_rate": 9.099968558921353e-06,
+      "loss": 0.8584,
+      "step": 2819
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6252991588982315,
+      "learning_rate": 9.093757016215423e-06,
+      "loss": 0.951,
+      "step": 2820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9962893615050404,
+      "learning_rate": 9.087545826043068e-06,
+      "loss": 0.8235,
+      "step": 2821
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6744432082676144,
+      "learning_rate": 9.081334990820477e-06,
+      "loss": 0.8505,
+      "step": 2822
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7451521444706546,
+      "learning_rate": 9.075124512963703e-06,
+      "loss": 0.8831,
+      "step": 2823
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4767059332590342,
+      "learning_rate": 9.068914394888651e-06,
+      "loss": 0.8509,
+      "step": 2824
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6915767429754536,
+      "learning_rate": 9.062704639011093e-06,
+      "loss": 0.9327,
+      "step": 2825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5994343971229656,
+      "learning_rate": 9.056495247746661e-06,
+      "loss": 0.8994,
+      "step": 2826
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.693871789768774,
+      "learning_rate": 9.05028622351084e-06,
+      "loss": 0.9057,
+      "step": 2827
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6783338474545895,
+      "learning_rate": 9.044077568718977e-06,
+      "loss": 0.8302,
+      "step": 2828
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6136950502037228,
+      "learning_rate": 9.03786928578627e-06,
+      "loss": 0.9243,
+      "step": 2829
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6793609278826633,
+      "learning_rate": 9.031661377127783e-06,
+      "loss": 0.8468,
+      "step": 2830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5797414314504952,
+      "learning_rate": 9.02545384515842e-06,
+      "loss": 0.9046,
+      "step": 2831
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4854448353409146,
+      "learning_rate": 9.019246692292946e-06,
+      "loss": 0.8769,
+      "step": 2832
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4373392447600624,
+      "learning_rate": 9.013039920945984e-06,
+      "loss": 0.7729,
+      "step": 2833
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7640857731480182,
+      "learning_rate": 9.006833533531996e-06,
+      "loss": 0.8615,
+      "step": 2834
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7879420990647688,
+      "learning_rate": 9.000627532465304e-06,
+      "loss": 0.8834,
+      "step": 2835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6535091251144516,
+      "learning_rate": 8.994421920160077e-06,
+      "loss": 0.9078,
+      "step": 2836
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6806745467472761,
+      "learning_rate": 8.988216699030335e-06,
+      "loss": 0.8312,
+      "step": 2837
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7044192090995676,
+      "learning_rate": 8.98201187148994e-06,
+      "loss": 0.8219,
+      "step": 2838
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6251835695817942,
+      "learning_rate": 8.975807439952606e-06,
+      "loss": 0.8901,
+      "step": 2839
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6068931889314269,
+      "learning_rate": 8.969603406831894e-06,
+      "loss": 0.8993,
+      "step": 2840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6906146199936813,
+      "learning_rate": 8.963399774541201e-06,
+      "loss": 0.8244,
+      "step": 2841
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9902891595935833,
+      "learning_rate": 8.95719654549378e-06,
+      "loss": 0.8107,
+      "step": 2842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.545724591299349,
+      "learning_rate": 8.950993722102722e-06,
+      "loss": 0.8879,
+      "step": 2843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.754980698057507,
+      "learning_rate": 8.944791306780958e-06,
+      "loss": 0.9186,
+      "step": 2844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0385907656715172,
+      "learning_rate": 8.938589301941261e-06,
+      "loss": 0.8078,
+      "step": 2845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.655903117828014,
+      "learning_rate": 8.932387709996246e-06,
+      "loss": 0.9066,
+      "step": 2846
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.692318861781103,
+      "learning_rate": 8.92618653335837e-06,
+      "loss": 0.8861,
+      "step": 2847
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.776409106768995,
+      "learning_rate": 8.919985774439923e-06,
+      "loss": 0.9176,
+      "step": 2848
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9907055916429031,
+      "learning_rate": 8.913785435653033e-06,
+      "loss": 0.8483,
+      "step": 2849
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7841981616101148,
+      "learning_rate": 8.907585519409668e-06,
+      "loss": 0.9033,
+      "step": 2850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7007461741551388,
+      "learning_rate": 8.901386028121634e-06,
+      "loss": 0.9044,
+      "step": 2851
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.712908262114012,
+      "learning_rate": 8.895186964200561e-06,
+      "loss": 0.8364,
+      "step": 2852
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6251011494370893,
+      "learning_rate": 8.888988330057922e-06,
+      "loss": 0.8449,
+      "step": 2853
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.686231118447992,
+      "learning_rate": 8.882790128105024e-06,
+      "loss": 0.8375,
+      "step": 2854
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8175092241741957,
+      "learning_rate": 8.876592360752995e-06,
+      "loss": 0.9047,
+      "step": 2855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.641559518116367,
+      "learning_rate": 8.870395030412805e-06,
+      "loss": 0.8614,
+      "step": 2856
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.787906785858516,
+      "learning_rate": 8.86419813949525e-06,
+      "loss": 0.8802,
+      "step": 2857
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5524803931883946,
+      "learning_rate": 8.858001690410958e-06,
+      "loss": 0.8387,
+      "step": 2858
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7419593060382064,
+      "learning_rate": 8.851805685570377e-06,
+      "loss": 0.9802,
+      "step": 2859
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6113765550732522,
+      "learning_rate": 8.845610127383795e-06,
+      "loss": 0.8941,
+      "step": 2860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.584927432302916,
+      "learning_rate": 8.839415018261307e-06,
+      "loss": 0.8219,
+      "step": 2861
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5566071440891167,
+      "learning_rate": 8.833220360612853e-06,
+      "loss": 0.8292,
+      "step": 2862
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.012004094892325,
+      "learning_rate": 8.827026156848189e-06,
+      "loss": 0.7744,
+      "step": 2863
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6693909564775262,
+      "learning_rate": 8.820832409376893e-06,
+      "loss": 0.9393,
+      "step": 2864
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.659875884926293,
+      "learning_rate": 8.814639120608368e-06,
+      "loss": 0.8823,
+      "step": 2865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6702199726930549,
+      "learning_rate": 8.808446292951838e-06,
+      "loss": 0.8842,
+      "step": 2866
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7558535447594004,
+      "learning_rate": 8.802253928816347e-06,
+      "loss": 0.8638,
+      "step": 2867
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6692529859197347,
+      "learning_rate": 8.796062030610759e-06,
+      "loss": 0.8638,
+      "step": 2868
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6954796102011807,
+      "learning_rate": 8.789870600743757e-06,
+      "loss": 0.9407,
+      "step": 2869
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7395733630036525,
+      "learning_rate": 8.783679641623845e-06,
+      "loss": 0.8989,
+      "step": 2870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5388947683696628,
+      "learning_rate": 8.777489155659336e-06,
+      "loss": 0.8187,
+      "step": 2871
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6366034567277383,
+      "learning_rate": 8.771299145258367e-06,
+      "loss": 0.7989,
+      "step": 2872
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5843796545323887,
+      "learning_rate": 8.765109612828884e-06,
+      "loss": 0.871,
+      "step": 2873
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5109392892565023,
+      "learning_rate": 8.75892056077865e-06,
+      "loss": 0.881,
+      "step": 2874
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7802710665532078,
+      "learning_rate": 8.752731991515243e-06,
+      "loss": 0.8493,
+      "step": 2875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.538895263494408,
+      "learning_rate": 8.746543907446049e-06,
+      "loss": 0.7993,
+      "step": 2876
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5256376074153322,
+      "learning_rate": 8.740356310978268e-06,
+      "loss": 0.9027,
+      "step": 2877
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7159958084705211,
+      "learning_rate": 8.73416920451891e-06,
+      "loss": 0.9155,
+      "step": 2878
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8230005457481084,
+      "learning_rate": 8.72798259047479e-06,
+      "loss": 0.9255,
+      "step": 2879
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6563365197323283,
+      "learning_rate": 8.72179647125254e-06,
+      "loss": 0.8647,
+      "step": 2880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.696362881222672,
+      "learning_rate": 8.715610849258595e-06,
+      "loss": 0.8174,
+      "step": 2881
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.709962094377892,
+      "learning_rate": 8.709425726899193e-06,
+      "loss": 0.8785,
+      "step": 2882
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.669087288148626,
+      "learning_rate": 8.703241106580384e-06,
+      "loss": 0.725,
+      "step": 2883
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7052853436274893,
+      "learning_rate": 8.697056990708019e-06,
+      "loss": 0.8743,
+      "step": 2884
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6038738634301302,
+      "learning_rate": 8.690873381687753e-06,
+      "loss": 0.9203,
+      "step": 2885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7786125804659687,
+      "learning_rate": 8.684690281925045e-06,
+      "loss": 0.9231,
+      "step": 2886
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6532355107926053,
+      "learning_rate": 8.678507693825156e-06,
+      "loss": 0.8883,
+      "step": 2887
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8446625621145372,
+      "learning_rate": 8.672325619793145e-06,
+      "loss": 0.8459,
+      "step": 2888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6678338071101526,
+      "learning_rate": 8.666144062233874e-06,
+      "loss": 0.8686,
+      "step": 2889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8178687869874979,
+      "learning_rate": 8.659963023552004e-06,
+      "loss": 0.903,
+      "step": 2890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7418292455237305,
+      "learning_rate": 8.653782506151996e-06,
+      "loss": 0.9018,
+      "step": 2891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5687708693670412,
+      "learning_rate": 8.647602512438103e-06,
+      "loss": 0.9524,
+      "step": 2892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5704017836401136,
+      "learning_rate": 8.641423044814375e-06,
+      "loss": 0.8398,
+      "step": 2893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7723233356594223,
+      "learning_rate": 8.635244105684664e-06,
+      "loss": 0.943,
+      "step": 2894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6441339938451631,
+      "learning_rate": 8.629065697452608e-06,
+      "loss": 0.836,
+      "step": 2895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.690628046950744,
+      "learning_rate": 8.622887822521645e-06,
+      "loss": 0.8661,
+      "step": 2896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5033387497647257,
+      "learning_rate": 8.616710483295e-06,
+      "loss": 0.8449,
+      "step": 2897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5713782081158334,
+      "learning_rate": 8.610533682175694e-06,
+      "loss": 0.8009,
+      "step": 2898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7389977390787457,
+      "learning_rate": 8.604357421566539e-06,
+      "loss": 0.9398,
+      "step": 2899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7360022202247412,
+      "learning_rate": 8.59818170387013e-06,
+      "loss": 0.9454,
+      "step": 2900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6802001098960284,
+      "learning_rate": 8.59200653148886e-06,
+      "loss": 0.8774,
+      "step": 2901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7998873802078137,
+      "learning_rate": 8.585831906824906e-06,
+      "loss": 0.8907,
+      "step": 2902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5981185161698253,
+      "learning_rate": 8.579657832280228e-06,
+      "loss": 0.8089,
+      "step": 2903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6730053839367558,
+      "learning_rate": 8.573484310256574e-06,
+      "loss": 0.9235,
+      "step": 2904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7250994162159647,
+      "learning_rate": 8.567311343155487e-06,
+      "loss": 0.9042,
+      "step": 2905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6306492223273898,
+      "learning_rate": 8.561138933378279e-06,
+      "loss": 0.8445,
+      "step": 2906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.488009193606293,
+      "learning_rate": 8.554967083326053e-06,
+      "loss": 0.8976,
+      "step": 2907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.752994511076105,
+      "learning_rate": 8.54879579539969e-06,
+      "loss": 0.9311,
+      "step": 2908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4676944262924772,
+      "learning_rate": 8.54262507199986e-06,
+      "loss": 0.9131,
+      "step": 2909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7003174552669684,
+      "learning_rate": 8.536454915527009e-06,
+      "loss": 0.8125,
+      "step": 2910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5656437117943105,
+      "learning_rate": 8.530285328381358e-06,
+      "loss": 0.8694,
+      "step": 2911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7461658655934071,
+      "learning_rate": 8.524116312962913e-06,
+      "loss": 0.9595,
+      "step": 2912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6564554434428689,
+      "learning_rate": 8.517947871671457e-06,
+      "loss": 0.8962,
+      "step": 2913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.601575335590067,
+      "learning_rate": 8.51178000690655e-06,
+      "loss": 0.8929,
+      "step": 2914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6141187717395675,
+      "learning_rate": 8.505612721067518e-06,
+      "loss": 0.9809,
+      "step": 2915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7465279264548101,
+      "learning_rate": 8.499446016553475e-06,
+      "loss": 0.9291,
+      "step": 2916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4809387447217077,
+      "learning_rate": 8.493279895763305e-06,
+      "loss": 0.9089,
+      "step": 2917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8149365102907096,
+      "learning_rate": 8.48711436109566e-06,
+      "loss": 0.8288,
+      "step": 2918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8088453692938176,
+      "learning_rate": 8.480949414948967e-06,
+      "loss": 0.9195,
+      "step": 2919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6322738087707473,
+      "learning_rate": 8.474785059721426e-06,
+      "loss": 0.8814,
+      "step": 2920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.602232490429101,
+      "learning_rate": 8.468621297811009e-06,
+      "loss": 0.9343,
+      "step": 2921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6673432830265518,
+      "learning_rate": 8.462458131615448e-06,
+      "loss": 0.8832,
+      "step": 2922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5993487747504578,
+      "learning_rate": 8.45629556353225e-06,
+      "loss": 0.9265,
+      "step": 2923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5466354109187563,
+      "learning_rate": 8.450133595958693e-06,
+      "loss": 0.8085,
+      "step": 2924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5982304665275466,
+      "learning_rate": 8.443972231291812e-06,
+      "loss": 0.8856,
+      "step": 2925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7991918100565665,
+      "learning_rate": 8.437811471928412e-06,
+      "loss": 0.8978,
+      "step": 2926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7103454115278325,
+      "learning_rate": 8.431651320265065e-06,
+      "loss": 0.8484,
+      "step": 2927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0863158287943084,
+      "learning_rate": 8.425491778698106e-06,
+      "loss": 0.8008,
+      "step": 2928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5579773941335044,
+      "learning_rate": 8.419332849623626e-06,
+      "loss": 0.8905,
+      "step": 2929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9258823332724813,
+      "learning_rate": 8.413174535437486e-06,
+      "loss": 0.9649,
+      "step": 2930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.70002143255522,
+      "learning_rate": 8.40701683853531e-06,
+      "loss": 0.9819,
+      "step": 2931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.551880101331,
+      "learning_rate": 8.400859761312463e-06,
+      "loss": 0.9635,
+      "step": 2932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7064274120032745,
+      "learning_rate": 8.394703306164093e-06,
+      "loss": 0.8715,
+      "step": 2933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6119039315115984,
+      "learning_rate": 8.388547475485092e-06,
+      "loss": 0.8743,
+      "step": 2934
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5340381395377936,
+      "learning_rate": 8.382392271670115e-06,
+      "loss": 0.843,
+      "step": 2935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6497003007828541,
+      "learning_rate": 8.376237697113566e-06,
+      "loss": 0.8174,
+      "step": 2936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0782446090514706,
+      "learning_rate": 8.370083754209612e-06,
+      "loss": 0.8565,
+      "step": 2937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7801266016354835,
+      "learning_rate": 8.363930445352172e-06,
+      "loss": 0.8494,
+      "step": 2938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7858596456451363,
+      "learning_rate": 8.357777772934914e-06,
+      "loss": 0.8467,
+      "step": 2939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.766105246125054,
+      "learning_rate": 8.351625739351263e-06,
+      "loss": 0.9009,
+      "step": 2940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5434631259134224,
+      "learning_rate": 8.345474346994396e-06,
+      "loss": 0.9481,
+      "step": 2941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.667008399866263,
+      "learning_rate": 8.33932359825724e-06,
+      "loss": 0.8181,
+      "step": 2942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6243544723712247,
+      "learning_rate": 8.333173495532468e-06,
+      "loss": 0.9295,
+      "step": 2943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5076598615547359,
+      "learning_rate": 8.327024041212504e-06,
+      "loss": 0.8418,
+      "step": 2944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6733058279880815,
+      "learning_rate": 8.320875237689522e-06,
+      "loss": 0.827,
+      "step": 2945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4574407668743208,
+      "learning_rate": 8.31472708735544e-06,
+      "loss": 0.855,
+      "step": 2946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.497053962089164,
+      "learning_rate": 8.308579592601922e-06,
+      "loss": 0.8743,
+      "step": 2947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5973812773928333,
+      "learning_rate": 8.302432755820378e-06,
+      "loss": 0.8357,
+      "step": 2948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6774408051842553,
+      "learning_rate": 8.296286579401965e-06,
+      "loss": 0.9271,
+      "step": 2949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.757701269527345,
+      "learning_rate": 8.290141065737574e-06,
+      "loss": 0.8914,
+      "step": 2950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6560923888517387,
+      "learning_rate": 8.283996217217848e-06,
+      "loss": 0.9063,
+      "step": 2951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6190321689050888,
+      "learning_rate": 8.277852036233172e-06,
+      "loss": 0.91,
+      "step": 2952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5575728517113605,
+      "learning_rate": 8.271708525173656e-06,
+      "loss": 0.9043,
+      "step": 2953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6359204976480444,
+      "learning_rate": 8.265565686429167e-06,
+      "loss": 0.8539,
+      "step": 2954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8358625292793525,
+      "learning_rate": 8.259423522389299e-06,
+      "loss": 0.8722,
+      "step": 2955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.578442491332812,
+      "learning_rate": 8.253282035443393e-06,
+      "loss": 0.8313,
+      "step": 2956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6655981195788856,
+      "learning_rate": 8.24714122798052e-06,
+      "loss": 0.808,
+      "step": 2957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.664261149670698,
+      "learning_rate": 8.241001102389487e-06,
+      "loss": 0.9248,
+      "step": 2958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7463186250647078,
+      "learning_rate": 8.23486166105884e-06,
+      "loss": 0.8536,
+      "step": 2959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7886515312291449,
+      "learning_rate": 8.228722906376851e-06,
+      "loss": 0.9303,
+      "step": 2960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6862531813774804,
+      "learning_rate": 8.222584840731535e-06,
+      "loss": 0.9625,
+      "step": 2961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6517024568409715,
+      "learning_rate": 8.216447466510633e-06,
+      "loss": 0.8769,
+      "step": 2962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6337298078387912,
+      "learning_rate": 8.210310786101613e-06,
+      "loss": 0.9752,
+      "step": 2963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6446051639534425,
+      "learning_rate": 8.204174801891684e-06,
+      "loss": 0.9023,
+      "step": 2964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.622045835850986,
+      "learning_rate": 8.198039516267776e-06,
+      "loss": 0.8356,
+      "step": 2965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.689962489273412,
+      "learning_rate": 8.191904931616547e-06,
+      "loss": 0.8511,
+      "step": 2966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6759672895512594,
+      "learning_rate": 8.185771050324385e-06,
+      "loss": 0.9159,
+      "step": 2967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5309928773860466,
+      "learning_rate": 8.17963787477741e-06,
+      "loss": 0.7929,
+      "step": 2968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7620299819122374,
+      "learning_rate": 8.173505407361454e-06,
+      "loss": 0.8198,
+      "step": 2969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.620371108996678,
+      "learning_rate": 8.167373650462084e-06,
+      "loss": 0.8951,
+      "step": 2970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5939168226469527,
+      "learning_rate": 8.16124260646459e-06,
+      "loss": 0.8199,
+      "step": 2971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7091746128202194,
+      "learning_rate": 8.155112277753978e-06,
+      "loss": 0.8879,
+      "step": 2972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9705874199836579,
+      "learning_rate": 8.148982666714981e-06,
+      "loss": 0.8033,
+      "step": 2973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8399974547894165,
+      "learning_rate": 8.142853775732054e-06,
+      "loss": 0.9543,
+      "step": 2974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4516340922194422,
+      "learning_rate": 8.136725607189371e-06,
+      "loss": 0.7915,
+      "step": 2975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5930593260326569,
+      "learning_rate": 8.130598163470819e-06,
+      "loss": 0.9091,
+      "step": 2976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.735883734767861,
+      "learning_rate": 8.124471446960012e-06,
+      "loss": 0.9062,
+      "step": 2977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5512646476671164,
+      "learning_rate": 8.118345460040276e-06,
+      "loss": 0.8497,
+      "step": 2978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7996595576685017,
+      "learning_rate": 8.11222020509465e-06,
+      "loss": 0.8556,
+      "step": 2979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.752686012985231,
+      "learning_rate": 8.106095684505897e-06,
+      "loss": 0.9199,
+      "step": 2980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6206337850966999,
+      "learning_rate": 8.099971900656487e-06,
+      "loss": 0.8581,
+      "step": 2981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.539073395815719,
+      "learning_rate": 8.093848855928611e-06,
+      "loss": 0.9163,
+      "step": 2982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.592610583545328,
+      "learning_rate": 8.087726552704162e-06,
+      "loss": 0.9052,
+      "step": 2983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.654853413631213,
+      "learning_rate": 8.081604993364752e-06,
+      "loss": 0.8843,
+      "step": 2984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6777775910945665,
+      "learning_rate": 8.075484180291702e-06,
+      "loss": 0.9052,
+      "step": 2985
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6454712717660718,
+      "learning_rate": 8.069364115866043e-06,
+      "loss": 0.8979,
+      "step": 2986
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.651920726174267,
+      "learning_rate": 8.06324480246851e-06,
+      "loss": 0.8572,
+      "step": 2987
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9460339819899883,
+      "learning_rate": 8.057126242479557e-06,
+      "loss": 0.7833,
+      "step": 2988
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6947391350710808,
+      "learning_rate": 8.051008438279333e-06,
+      "loss": 0.8792,
+      "step": 2989
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0040821303484626,
+      "learning_rate": 8.0448913922477e-06,
+      "loss": 0.9691,
+      "step": 2990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6313978411682766,
+      "learning_rate": 8.038775106764222e-06,
+      "loss": 0.8861,
+      "step": 2991
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5505460946019247,
+      "learning_rate": 8.03265958420817e-06,
+      "loss": 0.8515,
+      "step": 2992
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0086630847539766,
+      "learning_rate": 8.026544826958513e-06,
+      "loss": 0.7491,
+      "step": 2993
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8636040184232463,
+      "learning_rate": 8.020430837393928e-06,
+      "loss": 0.9098,
+      "step": 2994
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6364637808567868,
+      "learning_rate": 8.01431761789279e-06,
+      "loss": 0.8639,
+      "step": 2995
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7232748969328247,
+      "learning_rate": 8.00820517083318e-06,
+      "loss": 0.8198,
+      "step": 2996
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7090523904573154,
+      "learning_rate": 8.002093498592866e-06,
+      "loss": 0.8965,
+      "step": 2997
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4758153159321423,
+      "learning_rate": 7.995982603549334e-06,
+      "loss": 0.8313,
+      "step": 2998
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5946021194615747,
+      "learning_rate": 7.989872488079744e-06,
+      "loss": 0.8755,
+      "step": 2999
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9830289083269651,
+      "learning_rate": 7.983763154560968e-06,
+      "loss": 0.7821,
+      "step": 3000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6693992817984953,
+      "learning_rate": 7.977654605369576e-06,
+      "loss": 0.8728,
+      "step": 3001
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.688302764746823,
+      "learning_rate": 7.971546842881822e-06,
+      "loss": 0.8816,
+      "step": 3002
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4856811366533669,
+      "learning_rate": 7.965439869473664e-06,
+      "loss": 0.8838,
+      "step": 3003
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7359187904314755,
+      "learning_rate": 7.959333687520748e-06,
+      "loss": 0.9064,
+      "step": 3004
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5965192319047086,
+      "learning_rate": 7.95322829939841e-06,
+      "loss": 0.9237,
+      "step": 3005
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6965191121343643,
+      "learning_rate": 7.947123707481682e-06,
+      "loss": 0.9566,
+      "step": 3006
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6386128658642458,
+      "learning_rate": 7.941019914145285e-06,
+      "loss": 0.8635,
+      "step": 3007
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5964849875603513,
+      "learning_rate": 7.93491692176363e-06,
+      "loss": 0.8486,
+      "step": 3008
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6203889819501085,
+      "learning_rate": 7.92881473271081e-06,
+      "loss": 0.8362,
+      "step": 3009
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7067103859299149,
+      "learning_rate": 7.922713349360613e-06,
+      "loss": 0.8485,
+      "step": 3010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7487187258694665,
+      "learning_rate": 7.916612774086513e-06,
+      "loss": 0.9068,
+      "step": 3011
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4974596749567788,
+      "learning_rate": 7.910513009261667e-06,
+      "loss": 0.8174,
+      "step": 3012
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.661945406174596,
+      "learning_rate": 7.904414057258917e-06,
+      "loss": 0.8815,
+      "step": 3013
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.810429850369526,
+      "learning_rate": 7.898315920450788e-06,
+      "loss": 0.8243,
+      "step": 3014
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7623817534980704,
+      "learning_rate": 7.892218601209493e-06,
+      "loss": 0.9242,
+      "step": 3015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6481374777747804,
+      "learning_rate": 7.88612210190692e-06,
+      "loss": 0.8241,
+      "step": 3016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6201104897254557,
+      "learning_rate": 7.88002642491464e-06,
+      "loss": 0.8975,
+      "step": 3017
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.588041167931714,
+      "learning_rate": 7.87393157260391e-06,
+      "loss": 0.9271,
+      "step": 3018
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6714686830797574,
+      "learning_rate": 7.867837547345659e-06,
+      "loss": 0.8826,
+      "step": 3019
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7184864559313786,
+      "learning_rate": 7.861744351510496e-06,
+      "loss": 0.8318,
+      "step": 3020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5702066757978494,
+      "learning_rate": 7.855651987468709e-06,
+      "loss": 0.9101,
+      "step": 3021
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9583830006628503,
+      "learning_rate": 7.849560457590264e-06,
+      "loss": 0.8164,
+      "step": 3022
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5438961164815266,
+      "learning_rate": 7.843469764244795e-06,
+      "loss": 0.8414,
+      "step": 3023
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7718665299326102,
+      "learning_rate": 7.837379909801617e-06,
+      "loss": 0.9856,
+      "step": 3024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5862156100567921,
+      "learning_rate": 7.831290896629721e-06,
+      "loss": 0.838,
+      "step": 3025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6569244657120084,
+      "learning_rate": 7.825202727097766e-06,
+      "loss": 0.8013,
+      "step": 3026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5028164968572124,
+      "learning_rate": 7.81911540357408e-06,
+      "loss": 0.8764,
+      "step": 3027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.674058157022971,
+      "learning_rate": 7.813028928426669e-06,
+      "loss": 0.8489,
+      "step": 3028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6228536525531558,
+      "learning_rate": 7.806943304023209e-06,
+      "loss": 0.8933,
+      "step": 3029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.63310464093078,
+      "learning_rate": 7.800858532731035e-06,
+      "loss": 0.8993,
+      "step": 3030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7082070819696427,
+      "learning_rate": 7.79477461691716e-06,
+      "loss": 0.9155,
+      "step": 3031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6158697871122063,
+      "learning_rate": 7.78869155894826e-06,
+      "loss": 0.9677,
+      "step": 3032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5147683383703177,
+      "learning_rate": 7.782609361190685e-06,
+      "loss": 0.8046,
+      "step": 3033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.607117862003437,
+      "learning_rate": 7.776528026010434e-06,
+      "loss": 0.9289,
+      "step": 3034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6283957041178678,
+      "learning_rate": 7.770447555773187e-06,
+      "loss": 0.9199,
+      "step": 3035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6387061501872977,
+      "learning_rate": 7.76436795284428e-06,
+      "loss": 0.9032,
+      "step": 3036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.80351925533845,
+      "learning_rate": 7.758289219588709e-06,
+      "loss": 0.8993,
+      "step": 3037
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6967569449724593,
+      "learning_rate": 7.75221135837114e-06,
+      "loss": 0.9058,
+      "step": 3038
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6217497675044128,
+      "learning_rate": 7.74613437155589e-06,
+      "loss": 0.9123,
+      "step": 3039
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8011448566689687,
+      "learning_rate": 7.740058261506944e-06,
+      "loss": 0.9078,
+      "step": 3040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7854625493428522,
+      "learning_rate": 7.733983030587941e-06,
+      "loss": 0.9957,
+      "step": 3041
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9734309108874523,
+      "learning_rate": 7.727908681162181e-06,
+      "loss": 0.7836,
+      "step": 3042
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6583174570761303,
+      "learning_rate": 7.721835215592623e-06,
+      "loss": 0.8473,
+      "step": 3043
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8564572225147289,
+      "learning_rate": 7.71576263624187e-06,
+      "loss": 0.9714,
+      "step": 3044
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.600059396266203,
+      "learning_rate": 7.709690945472202e-06,
+      "loss": 0.9074,
+      "step": 3045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.623146121702975,
+      "learning_rate": 7.703620145645528e-06,
+      "loss": 0.8928,
+      "step": 3046
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.011861175957159,
+      "learning_rate": 7.697550239123427e-06,
+      "loss": 0.7937,
+      "step": 3047
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6123345964824651,
+      "learning_rate": 7.691481228267133e-06,
+      "loss": 0.8457,
+      "step": 3048
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6017527820543433,
+      "learning_rate": 7.685413115437514e-06,
+      "loss": 0.8948,
+      "step": 3049
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5544931005993492,
+      "learning_rate": 7.679345902995107e-06,
+      "loss": 0.8029,
+      "step": 3050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5920991525015225,
+      "learning_rate": 7.67327959330009e-06,
+      "loss": 0.8231,
+      "step": 3051
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4828586589957515,
+      "learning_rate": 7.667214188712295e-06,
+      "loss": 0.8514,
+      "step": 3052
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6821353317874144,
+      "learning_rate": 7.66114969159119e-06,
+      "loss": 0.8833,
+      "step": 3053
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5212183317349368,
+      "learning_rate": 7.655086104295904e-06,
+      "loss": 0.8562,
+      "step": 3054
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5852215243322778,
+      "learning_rate": 7.649023429185206e-06,
+      "loss": 0.9152,
+      "step": 3055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.884711011533379,
+      "learning_rate": 7.64296166861751e-06,
+      "loss": 0.8934,
+      "step": 3056
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.609768788971651,
+      "learning_rate": 7.636900824950873e-06,
+      "loss": 0.8729,
+      "step": 3057
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7009317039211018,
+      "learning_rate": 7.630840900542998e-06,
+      "loss": 0.8748,
+      "step": 3058
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.512674601953858,
+      "learning_rate": 7.624781897751231e-06,
+      "loss": 0.8444,
+      "step": 3059
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.665991190009202,
+      "learning_rate": 7.6187238189325555e-06,
+      "loss": 0.895,
+      "step": 3060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.506428511216225,
+      "learning_rate": 7.612666666443598e-06,
+      "loss": 0.8677,
+      "step": 3061
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5187920888001651,
+      "learning_rate": 7.6066104426406285e-06,
+      "loss": 0.9158,
+      "step": 3062
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.481907655926501,
+      "learning_rate": 7.600555149879546e-06,
+      "loss": 0.8562,
+      "step": 3063
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.578518747949265,
+      "learning_rate": 7.594500790515896e-06,
+      "loss": 0.932,
+      "step": 3064
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.877626627253111,
+      "learning_rate": 7.588447366904858e-06,
+      "loss": 0.9065,
+      "step": 3065
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5256217005031816,
+      "learning_rate": 7.582394881401251e-06,
+      "loss": 0.97,
+      "step": 3066
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6225989666570764,
+      "learning_rate": 7.576343336359519e-06,
+      "loss": 0.82,
+      "step": 3067
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7265807624217753,
+      "learning_rate": 7.570292734133752e-06,
+      "loss": 0.8256,
+      "step": 3068
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6171882133157245,
+      "learning_rate": 7.564243077077667e-06,
+      "loss": 0.9291,
+      "step": 3069
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6905364936582288,
+      "learning_rate": 7.558194367544611e-06,
+      "loss": 0.8791,
+      "step": 3070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7127951240944437,
+      "learning_rate": 7.552146607887568e-06,
+      "loss": 0.8972,
+      "step": 3071
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6899821947410933,
+      "learning_rate": 7.546099800459151e-06,
+      "loss": 0.9272,
+      "step": 3072
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5306507663508566,
+      "learning_rate": 7.540053947611603e-06,
+      "loss": 0.8818,
+      "step": 3073
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5062576814001796,
+      "learning_rate": 7.5340090516967914e-06,
+      "loss": 0.8726,
+      "step": 3074
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6193364540814856,
+      "learning_rate": 7.5279651150662155e-06,
+      "loss": 0.9092,
+      "step": 3075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6687877072273696,
+      "learning_rate": 7.521922140071003e-06,
+      "loss": 0.9002,
+      "step": 3076
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7134031591503625,
+      "learning_rate": 7.5158801290619e-06,
+      "loss": 0.8974,
+      "step": 3077
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7961049255470638,
+      "learning_rate": 7.509839084389284e-06,
+      "loss": 0.9479,
+      "step": 3078
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.548976159521323,
+      "learning_rate": 7.503799008403158e-06,
+      "loss": 0.8005,
+      "step": 3079
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7587980049174923,
+      "learning_rate": 7.497759903453144e-06,
+      "loss": 0.856,
+      "step": 3080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6847935419837459,
+      "learning_rate": 7.491721771888486e-06,
+      "loss": 0.8361,
+      "step": 3081
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4986156062298495,
+      "learning_rate": 7.485684616058051e-06,
+      "loss": 0.7818,
+      "step": 3082
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6595369417005381,
+      "learning_rate": 7.479648438310329e-06,
+      "loss": 0.9151,
+      "step": 3083
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7030478144226884,
+      "learning_rate": 7.4736132409934245e-06,
+      "loss": 0.8069,
+      "step": 3084
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.853490590663634,
+      "learning_rate": 7.467579026455061e-06,
+      "loss": 0.8431,
+      "step": 3085
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.660284431224386,
+      "learning_rate": 7.461545797042585e-06,
+      "loss": 0.8686,
+      "step": 3086
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.690879146766381,
+      "learning_rate": 7.455513555102958e-06,
+      "loss": 0.9707,
+      "step": 3087
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.659064507883572,
+      "learning_rate": 7.4494823029827505e-06,
+      "loss": 0.9232,
+      "step": 3088
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.647579546986959,
+      "learning_rate": 7.443452043028155e-06,
+      "loss": 0.8334,
+      "step": 3089
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.612274773309948,
+      "learning_rate": 7.43742277758498e-06,
+      "loss": 0.7824,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.631179585108783,
+      "learning_rate": 7.43139450899864e-06,
+      "loss": 0.9762,
+      "step": 3091
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6681692534240398,
+      "learning_rate": 7.425367239614166e-06,
+      "loss": 0.8348,
+      "step": 3092
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.612907994977427,
+      "learning_rate": 7.419340971776196e-06,
+      "loss": 0.9304,
+      "step": 3093
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6377669597592006,
+      "learning_rate": 7.413315707828986e-06,
+      "loss": 0.8725,
+      "step": 3094
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5225526785299495,
+      "learning_rate": 7.407291450116399e-06,
+      "loss": 0.8552,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.02202257756661,
+      "learning_rate": 7.401268200981897e-06,
+      "loss": 0.8136,
+      "step": 3096
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5282166365527594,
+      "learning_rate": 7.395245962768564e-06,
+      "loss": 0.8392,
+      "step": 3097
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6397628073422055,
+      "learning_rate": 7.389224737819083e-06,
+      "loss": 0.883,
+      "step": 3098
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6254418717859656,
+      "learning_rate": 7.383204528475745e-06,
+      "loss": 0.9128,
+      "step": 3099
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5437027031087878,
+      "learning_rate": 7.377185337080443e-06,
+      "loss": 0.8846,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6689977285251483,
+      "learning_rate": 7.371167165974675e-06,
+      "loss": 0.9319,
+      "step": 3101
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7801250629576035,
+      "learning_rate": 7.365150017499549e-06,
+      "loss": 0.9235,
+      "step": 3102
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6284905335891178,
+      "learning_rate": 7.359133893995762e-06,
+      "loss": 0.8862,
+      "step": 3103
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7574940249045918,
+      "learning_rate": 7.3531187978036245e-06,
+      "loss": 0.9386,
+      "step": 3104
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.822737429641499,
+      "learning_rate": 7.347104731263039e-06,
+      "loss": 0.8347,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6889157283124279,
+      "learning_rate": 7.341091696713516e-06,
+      "loss": 0.8928,
+      "step": 3106
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6813044155079595,
+      "learning_rate": 7.335079696494154e-06,
+      "loss": 0.9376,
+      "step": 3107
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4815235612993205,
+      "learning_rate": 7.329068732943657e-06,
+      "loss": 0.868,
+      "step": 3108
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.565166557131615,
+      "learning_rate": 7.3230588084003255e-06,
+      "loss": 0.8744,
+      "step": 3109
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7564155917609552,
+      "learning_rate": 7.317049925202049e-06,
+      "loss": 0.9292,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6621374603603378,
+      "learning_rate": 7.3110420856863195e-06,
+      "loss": 0.9404,
+      "step": 3111
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.688653962251391,
+      "learning_rate": 7.305035292190219e-06,
+      "loss": 0.8483,
+      "step": 3112
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4807877860138388,
+      "learning_rate": 7.299029547050427e-06,
+      "loss": 0.8801,
+      "step": 3113
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6605624381020474,
+      "learning_rate": 7.293024852603208e-06,
+      "loss": 0.9024,
+      "step": 3114
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8151500323766774,
+      "learning_rate": 7.287021211184421e-06,
+      "loss": 0.9215,
+      "step": 3115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7143663493141423,
+      "learning_rate": 7.281018625129521e-06,
+      "loss": 0.8909,
+      "step": 3116
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4632239268780975,
+      "learning_rate": 7.275017096773543e-06,
+      "loss": 0.8895,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7757202689549778,
+      "learning_rate": 7.269016628451114e-06,
+      "loss": 0.9199,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7986046794283217,
+      "learning_rate": 7.263017222496453e-06,
+      "loss": 0.9421,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5923282854035992,
+      "learning_rate": 7.257018881243363e-06,
+      "loss": 0.8345,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.678183223439245,
+      "learning_rate": 7.25102160702523e-06,
+      "loss": 0.9211,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7109990678509022,
+      "learning_rate": 7.245025402175026e-06,
+      "loss": 0.8292,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6271271916219008,
+      "learning_rate": 7.239030269025311e-06,
+      "loss": 0.8898,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5761700535405883,
+      "learning_rate": 7.233036209908226e-06,
+      "loss": 0.8437,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7270184271627644,
+      "learning_rate": 7.227043227155489e-06,
+      "loss": 0.8261,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7401574260276096,
+      "learning_rate": 7.221051323098407e-06,
+      "loss": 0.8546,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.688458905620417,
+      "learning_rate": 7.215060500067864e-06,
+      "loss": 0.9339,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4897489034438953,
+      "learning_rate": 7.209070760394323e-06,
+      "loss": 0.8219,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7059085306277206,
+      "learning_rate": 7.203082106407824e-06,
+      "loss": 0.8552,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.554045668280347,
+      "learning_rate": 7.19709454043799e-06,
+      "loss": 0.8572,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5640664192719858,
+      "learning_rate": 7.191108064814017e-06,
+      "loss": 0.8245,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7491840561711343,
+      "learning_rate": 7.185122681864675e-06,
+      "loss": 0.971,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5214258105175522,
+      "learning_rate": 7.179138393918315e-06,
+      "loss": 0.9046,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7617859920431787,
+      "learning_rate": 7.173155203302859e-06,
+      "loss": 0.9706,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0810051683169288,
+      "learning_rate": 7.167173112345796e-06,
+      "loss": 0.8348,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7003560509418185,
+      "learning_rate": 7.161192123374194e-06,
+      "loss": 0.9151,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6919431990598726,
+      "learning_rate": 7.1552122387146996e-06,
+      "loss": 0.875,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5889010020551524,
+      "learning_rate": 7.14923346069351e-06,
+      "loss": 0.8934,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5397752467580414,
+      "learning_rate": 7.14325579163641e-06,
+      "loss": 0.8475,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6498084571140035,
+      "learning_rate": 7.137279233868741e-06,
+      "loss": 0.837,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5480172215515395,
+      "learning_rate": 7.131303789715422e-06,
+      "loss": 0.8182,
+      "step": 3141
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5349307650656652,
+      "learning_rate": 7.125329461500931e-06,
+      "loss": 0.8282,
+      "step": 3142
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6423692120655142,
+      "learning_rate": 7.1193562515493184e-06,
+      "loss": 0.9101,
+      "step": 3143
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6129524910185176,
+      "learning_rate": 7.1133841621841935e-06,
+      "loss": 0.8083,
+      "step": 3144
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5208295551209883,
+      "learning_rate": 7.1074131957287325e-06,
+      "loss": 0.8286,
+      "step": 3145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6768181915969202,
+      "learning_rate": 7.1014433545056785e-06,
+      "loss": 0.9416,
+      "step": 3146
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5856038065954947,
+      "learning_rate": 7.095474640837326e-06,
+      "loss": 0.8527,
+      "step": 3147
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.687063244043259,
+      "learning_rate": 7.089507057045543e-06,
+      "loss": 0.8691,
+      "step": 3148
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.650268203545808,
+      "learning_rate": 7.0835406054517505e-06,
+      "loss": 0.8617,
+      "step": 3149
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5329666779525062,
+      "learning_rate": 7.077575288376936e-06,
+      "loss": 0.8437,
+      "step": 3150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7451257082010783,
+      "learning_rate": 7.071611108141636e-06,
+      "loss": 0.9623,
+      "step": 3151
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.629879062011099,
+      "learning_rate": 7.065648067065951e-06,
+      "loss": 0.8517,
+      "step": 3152
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.724227514860152,
+      "learning_rate": 7.059686167469539e-06,
+      "loss": 0.8501,
+      "step": 3153
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6360877952836748,
+      "learning_rate": 7.053725411671611e-06,
+      "loss": 0.9165,
+      "step": 3154
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7300408919523058,
+      "learning_rate": 7.0477658019909324e-06,
+      "loss": 0.9401,
+      "step": 3155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6437295722115854,
+      "learning_rate": 7.041807340745828e-06,
+      "loss": 0.8833,
+      "step": 3156
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.574411392317655,
+      "learning_rate": 7.035850030254172e-06,
+      "loss": 0.8927,
+      "step": 3157
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6473761591330518,
+      "learning_rate": 7.029893872833389e-06,
+      "loss": 0.9698,
+      "step": 3158
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5177661594707206,
+      "learning_rate": 7.0239388708004575e-06,
+      "loss": 0.873,
+      "step": 3159
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9910048878127302,
+      "learning_rate": 7.017985026471912e-06,
+      "loss": 0.8371,
+      "step": 3160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7149870915390946,
+      "learning_rate": 7.0120323421638235e-06,
+      "loss": 0.8948,
+      "step": 3161
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.528357270711752,
+      "learning_rate": 7.006080820191825e-06,
+      "loss": 0.851,
+      "step": 3162
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.596273468838947,
+      "learning_rate": 7.000130462871088e-06,
+      "loss": 0.9059,
+      "step": 3163
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.535953981342906,
+      "learning_rate": 6.994181272516338e-06,
+      "loss": 0.8408,
+      "step": 3164
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8975600724242803,
+      "learning_rate": 6.98823325144184e-06,
+      "loss": 0.9718,
+      "step": 3165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.704163142093305,
+      "learning_rate": 6.982286401961411e-06,
+      "loss": 0.9478,
+      "step": 3166
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5929743249342574,
+      "learning_rate": 6.976340726388406e-06,
+      "loss": 0.8833,
+      "step": 3167
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6418748053215968,
+      "learning_rate": 6.970396227035727e-06,
+      "loss": 0.9336,
+      "step": 3168
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7583285944266691,
+      "learning_rate": 6.964452906215815e-06,
+      "loss": 0.9023,
+      "step": 3169
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7055622669353865,
+      "learning_rate": 6.958510766240658e-06,
+      "loss": 0.8141,
+      "step": 3170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5092333795558799,
+      "learning_rate": 6.952569809421783e-06,
+      "loss": 0.9002,
+      "step": 3171
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.528342408154013,
+      "learning_rate": 6.94663003807025e-06,
+      "loss": 0.8342,
+      "step": 3172
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0028547102686474,
+      "learning_rate": 6.940691454496666e-06,
+      "loss": 0.8088,
+      "step": 3173
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5711250955353222,
+      "learning_rate": 6.934754061011176e-06,
+      "loss": 0.9376,
+      "step": 3174
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6140710563975675,
+      "learning_rate": 6.928817859923455e-06,
+      "loss": 0.8063,
+      "step": 3175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.486749645590974,
+      "learning_rate": 6.922882853542718e-06,
+      "loss": 0.7944,
+      "step": 3176
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7029664774300626,
+      "learning_rate": 6.916949044177719e-06,
+      "loss": 0.8995,
+      "step": 3177
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.636532416739665,
+      "learning_rate": 6.911016434136744e-06,
+      "loss": 0.8581,
+      "step": 3178
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6515779486477038,
+      "learning_rate": 6.905085025727605e-06,
+      "loss": 0.8548,
+      "step": 3179
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6673147382014908,
+      "learning_rate": 6.899154821257657e-06,
+      "loss": 0.7705,
+      "step": 3180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6499573611000455,
+      "learning_rate": 6.8932258230337844e-06,
+      "loss": 0.846,
+      "step": 3181
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5734018879012046,
+      "learning_rate": 6.887298033362397e-06,
+      "loss": 0.8637,
+      "step": 3182
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6901896584580383,
+      "learning_rate": 6.881371454549444e-06,
+      "loss": 0.9944,
+      "step": 3183
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.625231574148097,
+      "learning_rate": 6.875446088900388e-06,
+      "loss": 0.8991,
+      "step": 3184
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6234264212369682,
+      "learning_rate": 6.869521938720234e-06,
+      "loss": 0.9097,
+      "step": 3185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7471008686595277,
+      "learning_rate": 6.863599006313513e-06,
+      "loss": 0.9825,
+      "step": 3186
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5965259697177523,
+      "learning_rate": 6.857677293984273e-06,
+      "loss": 0.8563,
+      "step": 3187
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6932675554239935,
+      "learning_rate": 6.851756804036097e-06,
+      "loss": 0.8821,
+      "step": 3188
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.677274221504665,
+      "learning_rate": 6.845837538772087e-06,
+      "loss": 0.8971,
+      "step": 3189
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5997800326707308,
+      "learning_rate": 6.839919500494872e-06,
+      "loss": 0.8454,
+      "step": 3190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7258279291335552,
+      "learning_rate": 6.834002691506599e-06,
+      "loss": 0.8824,
+      "step": 3191
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7308861014198347,
+      "learning_rate": 6.8280871141089415e-06,
+      "loss": 0.8066,
+      "step": 3192
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6851276845363354,
+      "learning_rate": 6.822172770603094e-06,
+      "loss": 0.9917,
+      "step": 3193
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.015967223238646,
+      "learning_rate": 6.816259663289763e-06,
+      "loss": 0.8595,
+      "step": 3194
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8197862797842115,
+      "learning_rate": 6.810347794469188e-06,
+      "loss": 0.7768,
+      "step": 3195
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7103427807298617,
+      "learning_rate": 6.804437166441113e-06,
+      "loss": 0.8886,
+      "step": 3196
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6517237267809355,
+      "learning_rate": 6.798527781504809e-06,
+      "loss": 0.787,
+      "step": 3197
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5888457795319797,
+      "learning_rate": 6.792619641959056e-06,
+      "loss": 0.8714,
+      "step": 3198
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6628925979042317,
+      "learning_rate": 6.786712750102156e-06,
+      "loss": 0.7544,
+      "step": 3199
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6966798604250748,
+      "learning_rate": 6.780807108231925e-06,
+      "loss": 0.8923,
+      "step": 3200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7753771548694566,
+      "learning_rate": 6.774902718645685e-06,
+      "loss": 0.9228,
+      "step": 3201
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.605300234454333,
+      "learning_rate": 6.76899958364028e-06,
+      "loss": 0.8143,
+      "step": 3202
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6470521252350923,
+      "learning_rate": 6.76309770551206e-06,
+      "loss": 0.8651,
+      "step": 3203
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7121682614973732,
+      "learning_rate": 6.757197086556895e-06,
+      "loss": 0.9056,
+      "step": 3204
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6616386039494564,
+      "learning_rate": 6.751297729070151e-06,
+      "loss": 0.8444,
+      "step": 3205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6169506148675503,
+      "learning_rate": 6.745399635346715e-06,
+      "loss": 0.9285,
+      "step": 3206
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.831777134341909,
+      "learning_rate": 6.739502807680977e-06,
+      "loss": 0.9574,
+      "step": 3207
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7298109015499632,
+      "learning_rate": 6.733607248366836e-06,
+      "loss": 0.9602,
+      "step": 3208
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5759985119579727,
+      "learning_rate": 6.727712959697698e-06,
+      "loss": 0.8672,
+      "step": 3209
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5799081282450538,
+      "learning_rate": 6.721819943966472e-06,
+      "loss": 0.8454,
+      "step": 3210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.671881260425936,
+      "learning_rate": 6.71592820346558e-06,
+      "loss": 0.8684,
+      "step": 3211
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6106058533987733,
+      "learning_rate": 6.710037740486936e-06,
+      "loss": 0.8608,
+      "step": 3212
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8997431295519474,
+      "learning_rate": 6.704148557321964e-06,
+      "loss": 0.8606,
+      "step": 3213
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6398396830988429,
+      "learning_rate": 6.69826065626159e-06,
+      "loss": 0.8023,
+      "step": 3214
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.34628931439458,
+      "learning_rate": 6.692374039596241e-06,
+      "loss": 0.8758,
+      "step": 3215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6509096302673407,
+      "learning_rate": 6.686488709615841e-06,
+      "loss": 0.9425,
+      "step": 3216
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6403698146777557,
+      "learning_rate": 6.68060466860982e-06,
+      "loss": 0.7854,
+      "step": 3217
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5074783111637953,
+      "learning_rate": 6.674721918867102e-06,
+      "loss": 0.8217,
+      "step": 3218
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7430425847497941,
+      "learning_rate": 6.668840462676106e-06,
+      "loss": 0.9003,
+      "step": 3219
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.652144722693054,
+      "learning_rate": 6.662960302324755e-06,
+      "loss": 0.8294,
+      "step": 3220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8179963498187857,
+      "learning_rate": 6.657081440100464e-06,
+      "loss": 0.9889,
+      "step": 3221
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7413457287175298,
+      "learning_rate": 6.651203878290139e-06,
+      "loss": 0.8677,
+      "step": 3222
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6661954215925858,
+      "learning_rate": 6.6453276191801885e-06,
+      "loss": 0.8672,
+      "step": 3223
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7309119650922056,
+      "learning_rate": 6.639452665056509e-06,
+      "loss": 0.874,
+      "step": 3224
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6406401466127527,
+      "learning_rate": 6.633579018204493e-06,
+      "loss": 0.7827,
+      "step": 3225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4630888919368088,
+      "learning_rate": 6.627706680909016e-06,
+      "loss": 0.8061,
+      "step": 3226
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5557001415033243,
+      "learning_rate": 6.621835655454454e-06,
+      "loss": 0.8039,
+      "step": 3227
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7196079666492075,
+      "learning_rate": 6.61596594412467e-06,
+      "loss": 0.8783,
+      "step": 3228
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7507132768942424,
+      "learning_rate": 6.610097549203009e-06,
+      "loss": 0.8483,
+      "step": 3229
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7751329322726324,
+      "learning_rate": 6.604230472972317e-06,
+      "loss": 0.9408,
+      "step": 3230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.665460905350575,
+      "learning_rate": 6.598364717714914e-06,
+      "loss": 0.9062,
+      "step": 3231
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.066148157923597,
+      "learning_rate": 6.592500285712609e-06,
+      "loss": 0.8775,
+      "step": 3232
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7196706680020162,
+      "learning_rate": 6.586637179246707e-06,
+      "loss": 0.8944,
+      "step": 3233
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.749502756998033,
+      "learning_rate": 6.580775400597981e-06,
+      "loss": 0.7865,
+      "step": 3234
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6154764970669386,
+      "learning_rate": 6.574914952046699e-06,
+      "loss": 0.9796,
+      "step": 3235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7269889074187739,
+      "learning_rate": 6.569055835872609e-06,
+      "loss": 0.9327,
+      "step": 3236
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7084701009001728,
+      "learning_rate": 6.5631980543549416e-06,
+      "loss": 0.8505,
+      "step": 3237
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7231693453904087,
+      "learning_rate": 6.5573416097724e-06,
+      "loss": 0.8605,
+      "step": 3238
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5803143327402303,
+      "learning_rate": 6.551486504403181e-06,
+      "loss": 0.8456,
+      "step": 3239
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6222957524118382,
+      "learning_rate": 6.545632740524947e-06,
+      "loss": 0.9133,
+      "step": 3240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.612956365324901,
+      "learning_rate": 6.539780320414852e-06,
+      "loss": 0.8613,
+      "step": 3241
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6870133541565844,
+      "learning_rate": 6.533929246349514e-06,
+      "loss": 0.856,
+      "step": 3242
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5091124719511138,
+      "learning_rate": 6.528079520605035e-06,
+      "loss": 0.9183,
+      "step": 3243
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6948083111093037,
+      "learning_rate": 6.522231145456993e-06,
+      "loss": 0.7897,
+      "step": 3244
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5569603510947116,
+      "learning_rate": 6.516384123180435e-06,
+      "loss": 0.8081,
+      "step": 3245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7338475589750406,
+      "learning_rate": 6.510538456049889e-06,
+      "loss": 0.8799,
+      "step": 3246
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.658556644653592,
+      "learning_rate": 6.504694146339348e-06,
+      "loss": 0.8445,
+      "step": 3247
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9392428264409952,
+      "learning_rate": 6.498851196322287e-06,
+      "loss": 0.7859,
+      "step": 3248
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6257491471116412,
+      "learning_rate": 6.493009608271639e-06,
+      "loss": 0.8882,
+      "step": 3249
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5802433025522178,
+      "learning_rate": 6.48716938445982e-06,
+      "loss": 0.9058,
+      "step": 3250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7588717592943552,
+      "learning_rate": 6.481330527158708e-06,
+      "loss": 0.8012,
+      "step": 3251
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8388663434870292,
+      "learning_rate": 6.475493038639648e-06,
+      "loss": 0.9562,
+      "step": 3252
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6497560729246346,
+      "learning_rate": 6.4696569211734605e-06,
+      "loss": 0.8183,
+      "step": 3253
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6209720176609959,
+      "learning_rate": 6.463822177030425e-06,
+      "loss": 0.882,
+      "step": 3254
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0146031598556424,
+      "learning_rate": 6.457988808480291e-06,
+      "loss": 0.7903,
+      "step": 3255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8052207880768631,
+      "learning_rate": 6.452156817792269e-06,
+      "loss": 0.8091,
+      "step": 3256
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6805071843300246,
+      "learning_rate": 6.4463262072350376e-06,
+      "loss": 0.8455,
+      "step": 3257
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0087181615331984,
+      "learning_rate": 6.440496979076738e-06,
+      "loss": 0.8281,
+      "step": 3258
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5712928346748192,
+      "learning_rate": 6.43466913558497e-06,
+      "loss": 0.8788,
+      "step": 3259
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6799734036245515,
+      "learning_rate": 6.428842679026796e-06,
+      "loss": 0.8649,
+      "step": 3260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6054120912303245,
+      "learning_rate": 6.423017611668745e-06,
+      "loss": 0.8931,
+      "step": 3261
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6643292595171393,
+      "learning_rate": 6.4171939357768e-06,
+      "loss": 0.7616,
+      "step": 3262
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6607697281854277,
+      "learning_rate": 6.411371653616398e-06,
+      "loss": 0.7805,
+      "step": 3263
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6136848878260208,
+      "learning_rate": 6.4055507674524434e-06,
+      "loss": 0.8527,
+      "step": 3264
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7491254228057402,
+      "learning_rate": 6.399731279549295e-06,
+      "loss": 0.9479,
+      "step": 3265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5910966038631957,
+      "learning_rate": 6.393913192170761e-06,
+      "loss": 0.7979,
+      "step": 3266
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.705746627387083,
+      "learning_rate": 6.3880965075801126e-06,
+      "loss": 0.7379,
+      "step": 3267
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0012842754767453,
+      "learning_rate": 6.382281228040074e-06,
+      "loss": 0.7954,
+      "step": 3268
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6346064676288288,
+      "learning_rate": 6.3764673558128206e-06,
+      "loss": 0.8378,
+      "step": 3269
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.520851925290432,
+      "learning_rate": 6.3706548931599795e-06,
+      "loss": 0.9169,
+      "step": 3270
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8319849896101923,
+      "learning_rate": 6.364843842342632e-06,
+      "loss": 0.9023,
+      "step": 3271
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.651272786453872,
+      "learning_rate": 6.3590342056213104e-06,
+      "loss": 0.8588,
+      "step": 3272
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5832481876297977,
+      "learning_rate": 6.353225985255994e-06,
+      "loss": 0.9414,
+      "step": 3273
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.639437834423186,
+      "learning_rate": 6.347419183506114e-06,
+      "loss": 0.9014,
+      "step": 3274
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8009372479630623,
+      "learning_rate": 6.341613802630546e-06,
+      "loss": 0.95,
+      "step": 3275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7180603802562666,
+      "learning_rate": 6.3358098448876246e-06,
+      "loss": 0.8896,
+      "step": 3276
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8610914001468855,
+      "learning_rate": 6.330007312535111e-06,
+      "loss": 0.9216,
+      "step": 3277
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.369676418310214,
+      "learning_rate": 6.324206207830227e-06,
+      "loss": 0.802,
+      "step": 3278
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6707717953281056,
+      "learning_rate": 6.318406533029633e-06,
+      "loss": 0.8258,
+      "step": 3279
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7231479313107185,
+      "learning_rate": 6.312608290389437e-06,
+      "loss": 0.8685,
+      "step": 3280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5571134258088268,
+      "learning_rate": 6.306811482165189e-06,
+      "loss": 0.8306,
+      "step": 3281
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6004722169024093,
+      "learning_rate": 6.301016110611875e-06,
+      "loss": 0.8241,
+      "step": 3282
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5898000756424482,
+      "learning_rate": 6.295222177983929e-06,
+      "loss": 0.8732,
+      "step": 3283
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0174912353364762,
+      "learning_rate": 6.289429686535226e-06,
+      "loss": 0.8443,
+      "step": 3284
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7149756538106462,
+      "learning_rate": 6.283638638519075e-06,
+      "loss": 0.9069,
+      "step": 3285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0369869829359608,
+      "learning_rate": 6.277849036188225e-06,
+      "loss": 0.8624,
+      "step": 3286
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.523116952648023,
+      "learning_rate": 6.272060881794866e-06,
+      "loss": 0.8609,
+      "step": 3287
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5373392592010882,
+      "learning_rate": 6.266274177590621e-06,
+      "loss": 0.8622,
+      "step": 3288
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.687213816888126,
+      "learning_rate": 6.260488925826549e-06,
+      "loss": 0.8604,
+      "step": 3289
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6491579434031054,
+      "learning_rate": 6.2547051287531465e-06,
+      "loss": 0.8931,
+      "step": 3290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7130702200041086,
+      "learning_rate": 6.248922788620346e-06,
+      "loss": 0.8086,
+      "step": 3291
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5707092315095053,
+      "learning_rate": 6.243141907677505e-06,
+      "loss": 0.9268,
+      "step": 3292
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6606006458324556,
+      "learning_rate": 6.23736248817342e-06,
+      "loss": 0.8594,
+      "step": 3293
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6521889570099735,
+      "learning_rate": 6.231584532356319e-06,
+      "loss": 0.8392,
+      "step": 3294
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6259332985273904,
+      "learning_rate": 6.225808042473857e-06,
+      "loss": 0.9428,
+      "step": 3295
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6741317306625867,
+      "learning_rate": 6.220033020773122e-06,
+      "loss": 0.8155,
+      "step": 3296
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6955435584502687,
+      "learning_rate": 6.214259469500629e-06,
+      "loss": 0.829,
+      "step": 3297
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6794381243202086,
+      "learning_rate": 6.208487390902323e-06,
+      "loss": 0.777,
+      "step": 3298
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6144491480845333,
+      "learning_rate": 6.202716787223573e-06,
+      "loss": 0.8537,
+      "step": 3299
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5824266171004489,
+      "learning_rate": 6.196947660709176e-06,
+      "loss": 0.8953,
+      "step": 3300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7794246852274598,
+      "learning_rate": 6.1911800136033555e-06,
+      "loss": 0.8437,
+      "step": 3301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4669601847204055,
+      "learning_rate": 6.185413848149761e-06,
+      "loss": 0.8664,
+      "step": 3302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7223080203742889,
+      "learning_rate": 6.179649166591457e-06,
+      "loss": 0.8604,
+      "step": 3303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6189360147210166,
+      "learning_rate": 6.17388597117094e-06,
+      "loss": 0.894,
+      "step": 3304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5624690267000818,
+      "learning_rate": 6.1681242641301265e-06,
+      "loss": 0.8895,
+      "step": 3305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0706223648543807,
+      "learning_rate": 6.162364047710351e-06,
+      "loss": 0.8461,
+      "step": 3306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0107963623632832,
+      "learning_rate": 6.156605324152369e-06,
+      "loss": 0.8386,
+      "step": 3307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.758426156118175,
+      "learning_rate": 6.150848095696357e-06,
+      "loss": 0.8975,
+      "step": 3308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6554266499333847,
+      "learning_rate": 6.145092364581911e-06,
+      "loss": 0.8804,
+      "step": 3309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7754326824189635,
+      "learning_rate": 6.1393381330480394e-06,
+      "loss": 0.8684,
+      "step": 3310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5600291685555385,
+      "learning_rate": 6.133585403333171e-06,
+      "loss": 0.8187,
+      "step": 3311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6558674614696918,
+      "learning_rate": 6.127834177675154e-06,
+      "loss": 0.8876,
+      "step": 3312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6228264910845556,
+      "learning_rate": 6.12208445831124e-06,
+      "loss": 0.8999,
+      "step": 3313
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6619627916348558,
+      "learning_rate": 6.116336247478106e-06,
+      "loss": 0.9502,
+      "step": 3314
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7188349345831475,
+      "learning_rate": 6.110589547411838e-06,
+      "loss": 0.9595,
+      "step": 3315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6921124923962365,
+      "learning_rate": 6.104844360347936e-06,
+      "loss": 0.7812,
+      "step": 3316
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7104424707808128,
+      "learning_rate": 6.099100688521305e-06,
+      "loss": 0.8309,
+      "step": 3317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6108289460868688,
+      "learning_rate": 6.09335853416627e-06,
+      "loss": 0.7679,
+      "step": 3318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7621549611437217,
+      "learning_rate": 6.08761789951656e-06,
+      "loss": 0.7848,
+      "step": 3319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.738016026586915,
+      "learning_rate": 6.081878786805311e-06,
+      "loss": 0.7709,
+      "step": 3320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.828420064093399,
+      "learning_rate": 6.076141198265073e-06,
+      "loss": 0.8806,
+      "step": 3321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6852385951022801,
+      "learning_rate": 6.070405136127802e-06,
+      "loss": 0.8177,
+      "step": 3322
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8264530789731466,
+      "learning_rate": 6.064670602624852e-06,
+      "loss": 0.948,
+      "step": 3323
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6336788807725109,
+      "learning_rate": 6.0589375999869935e-06,
+      "loss": 0.8728,
+      "step": 3324
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0657906124978056,
+      "learning_rate": 6.053206130444393e-06,
+      "loss": 0.8228,
+      "step": 3325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8088337944296866,
+      "learning_rate": 6.047476196226626e-06,
+      "loss": 0.9149,
+      "step": 3326
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.672375500539844,
+      "learning_rate": 6.041747799562668e-06,
+      "loss": 0.8303,
+      "step": 3327
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6847006004737952,
+      "learning_rate": 6.0360209426809e-06,
+      "loss": 0.8903,
+      "step": 3328
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6617148081605224,
+      "learning_rate": 6.030295627809099e-06,
+      "loss": 0.8132,
+      "step": 3329
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8266891322588126,
+      "learning_rate": 6.024571857174443e-06,
+      "loss": 0.8877,
+      "step": 3330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7132079304979002,
+      "learning_rate": 6.018849633003515e-06,
+      "loss": 0.8923,
+      "step": 3331
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7331666930371998,
+      "learning_rate": 6.013128957522287e-06,
+      "loss": 0.9225,
+      "step": 3332
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5784115110976922,
+      "learning_rate": 6.007409832956137e-06,
+      "loss": 0.8893,
+      "step": 3333
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6894572787288455,
+      "learning_rate": 6.001692261529835e-06,
+      "loss": 0.8566,
+      "step": 3334
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4500648238512173,
+      "learning_rate": 5.995976245467552e-06,
+      "loss": 0.7681,
+      "step": 3335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5661040362940781,
+      "learning_rate": 5.990261786992844e-06,
+      "loss": 0.8625,
+      "step": 3336
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6877881495597586,
+      "learning_rate": 5.9845488883286695e-06,
+      "loss": 0.8238,
+      "step": 3337
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7107764448410836,
+      "learning_rate": 5.978837551697381e-06,
+      "loss": 0.7973,
+      "step": 3338
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6601597602650011,
+      "learning_rate": 5.973127779320715e-06,
+      "loss": 0.9012,
+      "step": 3339
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7506818836770115,
+      "learning_rate": 5.967419573419809e-06,
+      "loss": 0.8598,
+      "step": 3340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6976022385679548,
+      "learning_rate": 5.961712936215185e-06,
+      "loss": 0.881,
+      "step": 3341
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7324534445318824,
+      "learning_rate": 5.956007869926761e-06,
+      "loss": 0.8771,
+      "step": 3342
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6900137837691214,
+      "learning_rate": 5.950304376773833e-06,
+      "loss": 0.777,
+      "step": 3343
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5804001676081891,
+      "learning_rate": 5.9446024589750925e-06,
+      "loss": 0.8683,
+      "step": 3344
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1050288679360947,
+      "learning_rate": 5.938902118748624e-06,
+      "loss": 0.8649,
+      "step": 3345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5988789689661567,
+      "learning_rate": 5.933203358311884e-06,
+      "loss": 0.8826,
+      "step": 3346
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7427036692643512,
+      "learning_rate": 5.927506179881726e-06,
+      "loss": 0.8595,
+      "step": 3347
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.547926209770693,
+      "learning_rate": 5.921810585674383e-06,
+      "loss": 0.9562,
+      "step": 3348
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.686059136812411,
+      "learning_rate": 5.916116577905475e-06,
+      "loss": 0.8691,
+      "step": 3349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6906690845638421,
+      "learning_rate": 5.910424158789999e-06,
+      "loss": 0.8244,
+      "step": 3350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6896334351963331,
+      "learning_rate": 5.904733330542341e-06,
+      "loss": 0.9291,
+      "step": 3351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.719952005490152,
+      "learning_rate": 5.8990440953762615e-06,
+      "loss": 0.8878,
+      "step": 3352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7632312857067636,
+      "learning_rate": 5.893356455504911e-06,
+      "loss": 0.8925,
+      "step": 3353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.748807578045438,
+      "learning_rate": 5.887670413140807e-06,
+      "loss": 0.8856,
+      "step": 3354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7589628316556065,
+      "learning_rate": 5.881985970495853e-06,
+      "loss": 0.8843,
+      "step": 3355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9934784449564216,
+      "learning_rate": 5.8763031297813335e-06,
+      "loss": 0.7972,
+      "step": 3356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6375593557258121,
+      "learning_rate": 5.870621893207897e-06,
+      "loss": 0.8507,
+      "step": 3357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4521032055351442,
+      "learning_rate": 5.864942262985583e-06,
+      "loss": 0.7803,
+      "step": 3358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4848319601622657,
+      "learning_rate": 5.859264241323796e-06,
+      "loss": 0.8243,
+      "step": 3359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4405786173250585,
+      "learning_rate": 5.853587830431322e-06,
+      "loss": 0.9234,
+      "step": 3360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7018548048042814,
+      "learning_rate": 5.84791303251631e-06,
+      "loss": 0.9002,
+      "step": 3361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.594653592479301,
+      "learning_rate": 5.842239849786293e-06,
+      "loss": 0.8197,
+      "step": 3362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5756583473960974,
+      "learning_rate": 5.83656828444817e-06,
+      "loss": 0.8085,
+      "step": 3363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7626131511202743,
+      "learning_rate": 5.830898338708211e-06,
+      "loss": 0.9344,
+      "step": 3364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.649917610910777,
+      "learning_rate": 5.825230014772058e-06,
+      "loss": 0.8363,
+      "step": 3365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.569302437878805,
+      "learning_rate": 5.819563314844719e-06,
+      "loss": 0.8351,
+      "step": 3366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6719922167167833,
+      "learning_rate": 5.81389824113057e-06,
+      "loss": 0.9018,
+      "step": 3367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6528935395887316,
+      "learning_rate": 5.8082347958333625e-06,
+      "loss": 0.8079,
+      "step": 3368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8786800084633193,
+      "learning_rate": 5.802572981156201e-06,
+      "loss": 0.924,
+      "step": 3369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.576421826255542,
+      "learning_rate": 5.796912799301564e-06,
+      "loss": 0.8707,
+      "step": 3370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6834120714876446,
+      "learning_rate": 5.791254252471298e-06,
+      "loss": 0.9089,
+      "step": 3371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6765426988268832,
+      "learning_rate": 5.785597342866607e-06,
+      "loss": 0.9261,
+      "step": 3372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.532651480702941,
+      "learning_rate": 5.779942072688064e-06,
+      "loss": 0.8086,
+      "step": 3373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.559219416736841,
+      "learning_rate": 5.774288444135594e-06,
+      "loss": 0.7578,
+      "step": 3374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4952470576827597,
+      "learning_rate": 5.768636459408492e-06,
+      "loss": 0.8629,
+      "step": 3375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5428806987649826,
+      "learning_rate": 5.7629861207054135e-06,
+      "loss": 0.827,
+      "step": 3376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6729096074471197,
+      "learning_rate": 5.757337430224373e-06,
+      "loss": 0.9359,
+      "step": 3377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0267608609288024,
+      "learning_rate": 5.751690390162738e-06,
+      "loss": 0.8381,
+      "step": 3378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6002916007627612,
+      "learning_rate": 5.746045002717244e-06,
+      "loss": 0.7845,
+      "step": 3379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9886814366288212,
+      "learning_rate": 5.740401270083978e-06,
+      "loss": 0.8517,
+      "step": 3380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.810084397268728,
+      "learning_rate": 5.734759194458378e-06,
+      "loss": 0.8358,
+      "step": 3381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0397697148497778,
+      "learning_rate": 5.729118778035245e-06,
+      "loss": 0.8738,
+      "step": 3382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6481239636251044,
+      "learning_rate": 5.723480023008735e-06,
+      "loss": 0.86,
+      "step": 3383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7487067313771325,
+      "learning_rate": 5.7178429315723526e-06,
+      "loss": 0.9039,
+      "step": 3384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6222567221816466,
+      "learning_rate": 5.7122075059189585e-06,
+      "loss": 0.8239,
+      "step": 3385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5509934630182858,
+      "learning_rate": 5.7065737482407655e-06,
+      "loss": 0.8945,
+      "step": 3386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7025672476139628,
+      "learning_rate": 5.70094166072934e-06,
+      "loss": 0.9455,
+      "step": 3387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6018065729312914,
+      "learning_rate": 5.695311245575589e-06,
+      "loss": 0.8122,
+      "step": 3388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.676325029000261,
+      "learning_rate": 5.689682504969778e-06,
+      "loss": 0.8149,
+      "step": 3389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.582590392063282,
+      "learning_rate": 5.684055441101517e-06,
+      "loss": 0.872,
+      "step": 3390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5998988095223343,
+      "learning_rate": 5.6784300561597685e-06,
+      "loss": 0.9153,
+      "step": 3391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.582939181666072,
+      "learning_rate": 5.672806352332836e-06,
+      "loss": 0.8483,
+      "step": 3392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5218903408239695,
+      "learning_rate": 5.667184331808373e-06,
+      "loss": 0.886,
+      "step": 3393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.714083650093975,
+      "learning_rate": 5.6615639967733804e-06,
+      "loss": 0.8952,
+      "step": 3394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6626433152876583,
+      "learning_rate": 5.655945349414194e-06,
+      "loss": 0.85,
+      "step": 3395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.693403455147878,
+      "learning_rate": 5.650328391916498e-06,
+      "loss": 0.9065,
+      "step": 3396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7803476992222576,
+      "learning_rate": 5.644713126465324e-06,
+      "loss": 0.8742,
+      "step": 3397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4665228566280242,
+      "learning_rate": 5.6390995552450425e-06,
+      "loss": 0.7853,
+      "step": 3398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6908178885357066,
+      "learning_rate": 5.633487680439362e-06,
+      "loss": 0.8518,
+      "step": 3399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7984440582413503,
+      "learning_rate": 5.627877504231332e-06,
+      "loss": 0.8781,
+      "step": 3400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5579641099851758,
+      "learning_rate": 5.622269028803349e-06,
+      "loss": 0.8492,
+      "step": 3401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7810903685222328,
+      "learning_rate": 5.616662256337131e-06,
+      "loss": 0.9173,
+      "step": 3402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.660410226764821,
+      "learning_rate": 5.611057189013749e-06,
+      "loss": 0.7611,
+      "step": 3403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6391997234796118,
+      "learning_rate": 5.605453829013605e-06,
+      "loss": 0.8698,
+      "step": 3404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5711109096935474,
+      "learning_rate": 5.599852178516438e-06,
+      "loss": 0.8428,
+      "step": 3405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.651124338946114,
+      "learning_rate": 5.594252239701321e-06,
+      "loss": 0.811,
+      "step": 3406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9303380868419815,
+      "learning_rate": 5.58865401474666e-06,
+      "loss": 0.8888,
+      "step": 3407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5523025416654284,
+      "learning_rate": 5.5830575058301985e-06,
+      "loss": 0.9,
+      "step": 3408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.56084207871062,
+      "learning_rate": 5.5774627151290115e-06,
+      "loss": 0.8854,
+      "step": 3409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8061440366710004,
+      "learning_rate": 5.571869644819496e-06,
+      "loss": 0.8413,
+      "step": 3410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7302065237233202,
+      "learning_rate": 5.566278297077394e-06,
+      "loss": 0.7463,
+      "step": 3411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6799036046486864,
+      "learning_rate": 5.560688674077768e-06,
+      "loss": 0.8802,
+      "step": 3412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5469782950158764,
+      "learning_rate": 5.555100777995013e-06,
+      "loss": 0.9331,
+      "step": 3413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.544937552349973,
+      "learning_rate": 5.549514611002855e-06,
+      "loss": 0.9276,
+      "step": 3414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9965577228542513,
+      "learning_rate": 5.543930175274339e-06,
+      "loss": 0.8481,
+      "step": 3415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0644037756286373,
+      "learning_rate": 5.538347472981844e-06,
+      "loss": 0.8654,
+      "step": 3416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5132409466428756,
+      "learning_rate": 5.532766506297077e-06,
+      "loss": 0.8589,
+      "step": 3417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.660492516705413,
+      "learning_rate": 5.527187277391055e-06,
+      "loss": 0.8435,
+      "step": 3418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6793898309004203,
+      "learning_rate": 5.521609788434137e-06,
+      "loss": 0.918,
+      "step": 3419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0218106694762028,
+      "learning_rate": 5.516034041595992e-06,
+      "loss": 0.8376,
+      "step": 3420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5657276723643492,
+      "learning_rate": 5.510460039045622e-06,
+      "loss": 0.8507,
+      "step": 3421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6830897515946668,
+      "learning_rate": 5.504887782951343e-06,
+      "loss": 0.9091,
+      "step": 3422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.638545207336349,
+      "learning_rate": 5.499317275480792e-06,
+      "loss": 0.7693,
+      "step": 3423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8374830962975814,
+      "learning_rate": 5.4937485188009345e-06,
+      "loss": 0.9255,
+      "step": 3424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6975730054025413,
+      "learning_rate": 5.488181515078037e-06,
+      "loss": 0.9261,
+      "step": 3425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5857338199004463,
+      "learning_rate": 5.4826162664777005e-06,
+      "loss": 0.8385,
+      "step": 3426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6622286703464972,
+      "learning_rate": 5.477052775164838e-06,
+      "loss": 0.8334,
+      "step": 3427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6214412282663229,
+      "learning_rate": 5.471491043303678e-06,
+      "loss": 0.8495,
+      "step": 3428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8379337465153043,
+      "learning_rate": 5.465931073057765e-06,
+      "loss": 0.8735,
+      "step": 3429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7088624471518803,
+      "learning_rate": 5.46037286658996e-06,
+      "loss": 0.9142,
+      "step": 3430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6402757557090903,
+      "learning_rate": 5.454816426062439e-06,
+      "loss": 0.8691,
+      "step": 3431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.838696395524443,
+      "learning_rate": 5.449261753636681e-06,
+      "loss": 0.9217,
+      "step": 3432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.647911499092526,
+      "learning_rate": 5.443708851473487e-06,
+      "loss": 0.7731,
+      "step": 3433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6348139851967465,
+      "learning_rate": 5.438157721732968e-06,
+      "loss": 0.9094,
+      "step": 3434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4861274594874034,
+      "learning_rate": 5.432608366574546e-06,
+      "loss": 0.837,
+      "step": 3435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6569719103923588,
+      "learning_rate": 5.427060788156951e-06,
+      "loss": 0.8475,
+      "step": 3436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6063272057843794,
+      "learning_rate": 5.421514988638217e-06,
+      "loss": 0.8145,
+      "step": 3437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8113745286027272,
+      "learning_rate": 5.4159709701757016e-06,
+      "loss": 0.8747,
+      "step": 3438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5748294158977107,
+      "learning_rate": 5.410428734926045e-06,
+      "loss": 0.9165,
+      "step": 3439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6032715959314903,
+      "learning_rate": 5.404888285045217e-06,
+      "loss": 0.9263,
+      "step": 3440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.649843781550223,
+      "learning_rate": 5.399349622688479e-06,
+      "loss": 0.9144,
+      "step": 3441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7078544240729918,
+      "learning_rate": 5.393812750010406e-06,
+      "loss": 0.9383,
+      "step": 3442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8320981631171191,
+      "learning_rate": 5.388277669164867e-06,
+      "loss": 0.8919,
+      "step": 3443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.588907363889368,
+      "learning_rate": 5.382744382305045e-06,
+      "loss": 0.8889,
+      "step": 3444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6356438612319488,
+      "learning_rate": 5.377212891583419e-06,
+      "loss": 0.8471,
+      "step": 3445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5644831980489304,
+      "learning_rate": 5.371683199151765e-06,
+      "loss": 0.9575,
+      "step": 3446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6958019616002429,
+      "learning_rate": 5.366155307161165e-06,
+      "loss": 0.9588,
+      "step": 3447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8108451975152278,
+      "learning_rate": 5.360629217762003e-06,
+      "loss": 0.7796,
+      "step": 3448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5603938761965273,
+      "learning_rate": 5.355104933103956e-06,
+      "loss": 0.8465,
+      "step": 3449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6153523895456097,
+      "learning_rate": 5.3495824553360036e-06,
+      "loss": 0.8841,
+      "step": 3450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5471501897033066,
+      "learning_rate": 5.344061786606418e-06,
+      "loss": 0.8412,
+      "step": 3451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.594680191899028,
+      "learning_rate": 5.338542929062778e-06,
+      "loss": 0.7719,
+      "step": 3452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7617012895105801,
+      "learning_rate": 5.333025884851938e-06,
+      "loss": 0.9171,
+      "step": 3453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5909627813329759,
+      "learning_rate": 5.327510656120065e-06,
+      "loss": 0.8838,
+      "step": 3454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6341758757943798,
+      "learning_rate": 5.321997245012613e-06,
+      "loss": 0.8184,
+      "step": 3455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7143193253583573,
+      "learning_rate": 5.316485653674331e-06,
+      "loss": 0.8761,
+      "step": 3456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9852129501508331,
+      "learning_rate": 5.310975884249256e-06,
+      "loss": 0.7838,
+      "step": 3457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6139694285374309,
+      "learning_rate": 5.305467938880722e-06,
+      "loss": 0.8515,
+      "step": 3458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.606123829050421,
+      "learning_rate": 5.299961819711353e-06,
+      "loss": 0.8199,
+      "step": 3459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7155254941329081,
+      "learning_rate": 5.294457528883052e-06,
+      "loss": 0.8586,
+      "step": 3460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6818251472437957,
+      "learning_rate": 5.28895506853703e-06,
+      "loss": 0.8265,
+      "step": 3461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6495628619682985,
+      "learning_rate": 5.28345444081376e-06,
+      "loss": 0.876,
+      "step": 3462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6327944361685691,
+      "learning_rate": 5.277955647853028e-06,
+      "loss": 0.8074,
+      "step": 3463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0196558888940344,
+      "learning_rate": 5.27245869179389e-06,
+      "loss": 0.8231,
+      "step": 3464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5129211617224263,
+      "learning_rate": 5.266963574774695e-06,
+      "loss": 0.8475,
+      "step": 3465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8766213337901607,
+      "learning_rate": 5.2614702989330745e-06,
+      "loss": 0.8998,
+      "step": 3466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.756290547166061,
+      "learning_rate": 5.255978866405942e-06,
+      "loss": 0.7845,
+      "step": 3467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6740182019760137,
+      "learning_rate": 5.250489279329501e-06,
+      "loss": 0.8462,
+      "step": 3468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5674872274602343,
+      "learning_rate": 5.245001539839223e-06,
+      "loss": 0.8214,
+      "step": 3469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8654795803627764,
+      "learning_rate": 5.239515650069871e-06,
+      "loss": 0.9309,
+      "step": 3470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6342037638788427,
+      "learning_rate": 5.23403161215549e-06,
+      "loss": 0.8973,
+      "step": 3471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7990386364475455,
+      "learning_rate": 5.2285494282293995e-06,
+      "loss": 0.9241,
+      "step": 3472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.60018533708385,
+      "learning_rate": 5.223069100424199e-06,
+      "loss": 0.8498,
+      "step": 3473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6831975257554828,
+      "learning_rate": 5.21759063087177e-06,
+      "loss": 0.7983,
+      "step": 3474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6285058728392026,
+      "learning_rate": 5.212114021703267e-06,
+      "loss": 0.8774,
+      "step": 3475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.531233133903309,
+      "learning_rate": 5.206639275049117e-06,
+      "loss": 0.8809,
+      "step": 3476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5422963052303207,
+      "learning_rate": 5.20116639303903e-06,
+      "loss": 0.8435,
+      "step": 3477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6803646209104026,
+      "learning_rate": 5.195695377801988e-06,
+      "loss": 0.8456,
+      "step": 3478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6288687470407628,
+      "learning_rate": 5.190226231466246e-06,
+      "loss": 0.8459,
+      "step": 3479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.720536855630722,
+      "learning_rate": 5.184758956159332e-06,
+      "loss": 0.8171,
+      "step": 3480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6250557117401936,
+      "learning_rate": 5.179293554008047e-06,
+      "loss": 0.8079,
+      "step": 3481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6796066602307127,
+      "learning_rate": 5.173830027138467e-06,
+      "loss": 0.7988,
+      "step": 3482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5726187742823352,
+      "learning_rate": 5.168368377675925e-06,
+      "loss": 0.9157,
+      "step": 3483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7690592111962589,
+      "learning_rate": 5.16290860774504e-06,
+      "loss": 0.8125,
+      "step": 3484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6006828309978147,
+      "learning_rate": 5.157450719469691e-06,
+      "loss": 0.877,
+      "step": 3485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8387590691450166,
+      "learning_rate": 5.1519947149730275e-06,
+      "loss": 0.9497,
+      "step": 3486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7437194623462446,
+      "learning_rate": 5.146540596377465e-06,
+      "loss": 0.8794,
+      "step": 3487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5260556664639773,
+      "learning_rate": 5.141088365804687e-06,
+      "loss": 0.9062,
+      "step": 3488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6121369814146513,
+      "learning_rate": 5.135638025375646e-06,
+      "loss": 0.8685,
+      "step": 3489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6307356264139352,
+      "learning_rate": 5.1301895772105455e-06,
+      "loss": 0.8101,
+      "step": 3490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6001379349482612,
+      "learning_rate": 5.124743023428867e-06,
+      "loss": 0.7936,
+      "step": 3491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6400861111456146,
+      "learning_rate": 5.11929836614935e-06,
+      "loss": 0.8655,
+      "step": 3492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5199373054992331,
+      "learning_rate": 5.1138556074899966e-06,
+      "loss": 0.8173,
+      "step": 3493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4163277782324815,
+      "learning_rate": 5.108414749568071e-06,
+      "loss": 0.8532,
+      "step": 3494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.769043248958,
+      "learning_rate": 5.102975794500098e-06,
+      "loss": 0.7277,
+      "step": 3495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5821581529209696,
+      "learning_rate": 5.097538744401862e-06,
+      "loss": 0.878,
+      "step": 3496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5668088978845927,
+      "learning_rate": 5.092103601388401e-06,
+      "loss": 0.8674,
+      "step": 3497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.619044331707455,
+      "learning_rate": 5.086670367574018e-06,
+      "loss": 0.9003,
+      "step": 3498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6856372620575115,
+      "learning_rate": 5.081239045072273e-06,
+      "loss": 0.8882,
+      "step": 3499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5350860654851521,
+      "learning_rate": 5.075809635995979e-06,
+      "loss": 0.9194,
+      "step": 3500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.579493551103141,
+      "learning_rate": 5.070382142457208e-06,
+      "loss": 0.8897,
+      "step": 3501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.783458467703695,
+      "learning_rate": 5.064956566567284e-06,
+      "loss": 0.9274,
+      "step": 3502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6579318803193228,
+      "learning_rate": 5.059532910436791e-06,
+      "loss": 0.8246,
+      "step": 3503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.706132243296039,
+      "learning_rate": 5.054111176175554e-06,
+      "loss": 0.8749,
+      "step": 3504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.84331292160113,
+      "learning_rate": 5.048691365892662e-06,
+      "loss": 0.8929,
+      "step": 3505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6877147835278588,
+      "learning_rate": 5.0432734816964494e-06,
+      "loss": 0.9108,
+      "step": 3506
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6795411157577118,
+      "learning_rate": 5.037857525694508e-06,
+      "loss": 0.8363,
+      "step": 3507
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8058021809794138,
+      "learning_rate": 5.0324434999936686e-06,
+      "loss": 0.8512,
+      "step": 3508
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7256199363811393,
+      "learning_rate": 5.02703140670002e-06,
+      "loss": 0.8911,
+      "step": 3509
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6593215838344397,
+      "learning_rate": 5.021621247918898e-06,
+      "loss": 0.8484,
+      "step": 3510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8635093903732782,
+      "learning_rate": 5.016213025754888e-06,
+      "loss": 0.8861,
+      "step": 3511
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6927155954888042,
+      "learning_rate": 5.0108067423118086e-06,
+      "loss": 0.825,
+      "step": 3512
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5601035153714797,
+      "learning_rate": 5.00540239969274e-06,
+      "loss": 0.8463,
+      "step": 3513
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6286917157661844,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.8938,
+      "step": 3514
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0357648819685474,
+      "learning_rate": 4.994599545335158e-06,
+      "loss": 0.8104,
+      "step": 3515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6208246241494164,
+      "learning_rate": 4.9892010377990145e-06,
+      "loss": 0.8712,
+      "step": 3516
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5990300944610627,
+      "learning_rate": 4.983804479491621e-06,
+      "loss": 0.8438,
+      "step": 3517
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0502307892507927,
+      "learning_rate": 4.9784098725122685e-06,
+      "loss": 0.896,
+      "step": 3518
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.704333376666638,
+      "learning_rate": 4.973017218959495e-06,
+      "loss": 0.9087,
+      "step": 3519
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.714287913965975,
+      "learning_rate": 4.967626520931062e-06,
+      "loss": 0.8513,
+      "step": 3520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6008993143743324,
+      "learning_rate": 4.962237780523987e-06,
+      "loss": 0.7947,
+      "step": 3521
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9966460959307214,
+      "learning_rate": 4.95685099983452e-06,
+      "loss": 0.8228,
+      "step": 3522
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.530872785095733,
+      "learning_rate": 4.95146618095815e-06,
+      "loss": 0.8465,
+      "step": 3523
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6699983529680114,
+      "learning_rate": 4.946083325989598e-06,
+      "loss": 0.8367,
+      "step": 3524
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7464610509453784,
+      "learning_rate": 4.940702437022828e-06,
+      "loss": 0.8455,
+      "step": 3525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5891005212689537,
+      "learning_rate": 4.935323516151037e-06,
+      "loss": 0.8342,
+      "step": 3526
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6057812638023987,
+      "learning_rate": 4.92994656546665e-06,
+      "loss": 0.8212,
+      "step": 3527
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8061583485266182,
+      "learning_rate": 4.9245715870613334e-06,
+      "loss": 0.783,
+      "step": 3528
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7138043676677468,
+      "learning_rate": 4.919198583025983e-06,
+      "loss": 0.8392,
+      "step": 3529
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5243807415062884,
+      "learning_rate": 4.913827555450729e-06,
+      "loss": 0.8299,
+      "step": 3530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5356461467261178,
+      "learning_rate": 4.908458506424929e-06,
+      "loss": 0.7894,
+      "step": 3531
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.624680240452524,
+      "learning_rate": 4.9030914380371765e-06,
+      "loss": 0.8429,
+      "step": 3532
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5075770951500866,
+      "learning_rate": 4.897726352375292e-06,
+      "loss": 0.8539,
+      "step": 3533
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5464342962933284,
+      "learning_rate": 4.8923632515263155e-06,
+      "loss": 0.8004,
+      "step": 3534
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9465852273753712,
+      "learning_rate": 4.88700213757653e-06,
+      "loss": 0.8025,
+      "step": 3535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5921089573869478,
+      "learning_rate": 4.8816430126114355e-06,
+      "loss": 0.8041,
+      "step": 3536
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5854399951475144,
+      "learning_rate": 4.876285878715764e-06,
+      "loss": 0.7974,
+      "step": 3537
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.515472042860783,
+      "learning_rate": 4.87093073797347e-06,
+      "loss": 0.8032,
+      "step": 3538
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6786174361879456,
+      "learning_rate": 4.865577592467732e-06,
+      "loss": 0.9424,
+      "step": 3539
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6844188001472047,
+      "learning_rate": 4.860226444280958e-06,
+      "loss": 0.9211,
+      "step": 3540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7737856993378502,
+      "learning_rate": 4.8548772954947666e-06,
+      "loss": 0.8425,
+      "step": 3541
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5794836834320127,
+      "learning_rate": 4.84953014819001e-06,
+      "loss": 0.8703,
+      "step": 3542
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7364043574329118,
+      "learning_rate": 4.844185004446761e-06,
+      "loss": 0.9545,
+      "step": 3543
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8090743942591776,
+      "learning_rate": 4.838841866344306e-06,
+      "loss": 0.9326,
+      "step": 3544
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5315191619480077,
+      "learning_rate": 4.833500735961158e-06,
+      "loss": 0.8277,
+      "step": 3545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.554763716925827,
+      "learning_rate": 4.8281616153750475e-06,
+      "loss": 0.7729,
+      "step": 3546
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6458448857724612,
+      "learning_rate": 4.822824506662924e-06,
+      "loss": 0.7925,
+      "step": 3547
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0569375747177285,
+      "learning_rate": 4.817489411900946e-06,
+      "loss": 0.8864,
+      "step": 3548
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5894729361683813,
+      "learning_rate": 4.812156333164498e-06,
+      "loss": 0.9899,
+      "step": 3549
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9834867318092824,
+      "learning_rate": 4.806825272528178e-06,
+      "loss": 0.8148,
+      "step": 3550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9800290717802019,
+      "learning_rate": 4.801496232065799e-06,
+      "loss": 0.7389,
+      "step": 3551
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7360144668109772,
+      "learning_rate": 4.796169213850387e-06,
+      "loss": 0.9231,
+      "step": 3552
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.469464853686856,
+      "learning_rate": 4.790844219954184e-06,
+      "loss": 0.8362,
+      "step": 3553
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.781461637312385,
+      "learning_rate": 4.7855212524486375e-06,
+      "loss": 0.8677,
+      "step": 3554
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.678541777882069,
+      "learning_rate": 4.780200313404417e-06,
+      "loss": 0.825,
+      "step": 3555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5377121209860265,
+      "learning_rate": 4.774881404891391e-06,
+      "loss": 0.7923,
+      "step": 3556
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6515637624260644,
+      "learning_rate": 4.7695645289786484e-06,
+      "loss": 0.9347,
+      "step": 3557
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5882412338816558,
+      "learning_rate": 4.764249687734482e-06,
+      "loss": 0.8714,
+      "step": 3558
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6514004611453184,
+      "learning_rate": 4.758936883226397e-06,
+      "loss": 0.934,
+      "step": 3559
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.074101218998927,
+      "learning_rate": 4.753626117521103e-06,
+      "loss": 0.8285,
+      "step": 3560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6708861727435662,
+      "learning_rate": 4.748317392684517e-06,
+      "loss": 0.891,
+      "step": 3561
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7645948946554029,
+      "learning_rate": 4.743010710781767e-06,
+      "loss": 0.7962,
+      "step": 3562
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.494160851850829,
+      "learning_rate": 4.7377060738771745e-06,
+      "loss": 0.7814,
+      "step": 3563
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7478862037192993,
+      "learning_rate": 4.7324034840342745e-06,
+      "loss": 0.8411,
+      "step": 3564
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5794785787402035,
+      "learning_rate": 4.727102943315808e-06,
+      "loss": 0.8624,
+      "step": 3565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7140839702199169,
+      "learning_rate": 4.7218044537837115e-06,
+      "loss": 0.9207,
+      "step": 3566
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5808657460922244,
+      "learning_rate": 4.716508017499127e-06,
+      "loss": 0.8936,
+      "step": 3567
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7189824450333906,
+      "learning_rate": 4.7112136365224014e-06,
+      "loss": 0.8876,
+      "step": 3568
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.780073280573899,
+      "learning_rate": 4.705921312913077e-06,
+      "loss": 0.9343,
+      "step": 3569
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5612946707072566,
+      "learning_rate": 4.7006310487298946e-06,
+      "loss": 0.8755,
+      "step": 3570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7815992853106772,
+      "learning_rate": 4.695342846030797e-06,
+      "loss": 0.854,
+      "step": 3571
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.575242459236206,
+      "learning_rate": 4.690056706872926e-06,
+      "loss": 0.7887,
+      "step": 3572
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5671637814907244,
+      "learning_rate": 4.6847726333126184e-06,
+      "loss": 0.8507,
+      "step": 3573
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6491720547280737,
+      "learning_rate": 4.6794906274054084e-06,
+      "loss": 0.8789,
+      "step": 3574
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7150410748419826,
+      "learning_rate": 4.674210691206027e-06,
+      "loss": 0.8368,
+      "step": 3575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6958327132034345,
+      "learning_rate": 4.668932826768401e-06,
+      "loss": 0.8928,
+      "step": 3576
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5943839982483177,
+      "learning_rate": 4.663657036145643e-06,
+      "loss": 0.8247,
+      "step": 3577
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5823965565363844,
+      "learning_rate": 4.658383321390067e-06,
+      "loss": 0.8435,
+      "step": 3578
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8969954015270543,
+      "learning_rate": 4.653111684553179e-06,
+      "loss": 0.9222,
+      "step": 3579
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7921762219153188,
+      "learning_rate": 4.647842127685675e-06,
+      "loss": 0.9342,
+      "step": 3580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5750581020477474,
+      "learning_rate": 4.6425746528374405e-06,
+      "loss": 0.8798,
+      "step": 3581
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6281849054923896,
+      "learning_rate": 4.637309262057555e-06,
+      "loss": 0.7863,
+      "step": 3582
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6556330434333513,
+      "learning_rate": 4.632045957394286e-06,
+      "loss": 0.8788,
+      "step": 3583
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7120047767550282,
+      "learning_rate": 4.626784740895081e-06,
+      "loss": 0.7725,
+      "step": 3584
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6387035438265312,
+      "learning_rate": 4.6215256146065905e-06,
+      "loss": 0.8123,
+      "step": 3585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4972208358393009,
+      "learning_rate": 4.61626858057464e-06,
+      "loss": 0.7642,
+      "step": 3586
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6629441101497722,
+      "learning_rate": 4.611013640844245e-06,
+      "loss": 0.8789,
+      "step": 3587
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0294475875600821,
+      "learning_rate": 4.6057607974596086e-06,
+      "loss": 0.8315,
+      "step": 3588
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5547376860771147,
+      "learning_rate": 4.600510052464115e-06,
+      "loss": 0.842,
+      "step": 3589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6102850295153976,
+      "learning_rate": 4.595261407900337e-06,
+      "loss": 0.8795,
+      "step": 3590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7073932502054254,
+      "learning_rate": 4.590014865810019e-06,
+      "loss": 0.9184,
+      "step": 3591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6307552513548709,
+      "learning_rate": 4.584770428234099e-06,
+      "loss": 0.9247,
+      "step": 3592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7511844477164964,
+      "learning_rate": 4.579528097212692e-06,
+      "loss": 0.9172,
+      "step": 3593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5617667596859601,
+      "learning_rate": 4.574287874785095e-06,
+      "loss": 0.8266,
+      "step": 3594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.677301021211091,
+      "learning_rate": 4.569049762989783e-06,
+      "loss": 0.9007,
+      "step": 3595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5183032971065564,
+      "learning_rate": 4.563813763864412e-06,
+      "loss": 0.7801,
+      "step": 3596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.53100220080267,
+      "learning_rate": 4.5585798794458166e-06,
+      "loss": 0.7586,
+      "step": 3597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5301323494577879,
+      "learning_rate": 4.553348111770002e-06,
+      "loss": 0.8137,
+      "step": 3598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.734826698439897,
+      "learning_rate": 4.54811846287216e-06,
+      "loss": 0.8898,
+      "step": 3599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9946874519591455,
+      "learning_rate": 4.542890934786647e-06,
+      "loss": 0.8159,
+      "step": 3600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7093360789782295,
+      "learning_rate": 4.537665529547004e-06,
+      "loss": 0.8273,
+      "step": 3601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.800766332666574,
+      "learning_rate": 4.5324422491859435e-06,
+      "loss": 0.7969,
+      "step": 3602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5793010103380125,
+      "learning_rate": 4.52722109573535e-06,
+      "loss": 0.8865,
+      "step": 3603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7775192681597463,
+      "learning_rate": 4.522002071226281e-06,
+      "loss": 0.8583,
+      "step": 3604
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6689435979838343,
+      "learning_rate": 4.516785177688966e-06,
+      "loss": 0.8858,
+      "step": 3605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.751447441691371,
+      "learning_rate": 4.5115704171528105e-06,
+      "loss": 0.8538,
+      "step": 3606
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4809265108759713,
+      "learning_rate": 4.506357791646379e-06,
+      "loss": 0.7901,
+      "step": 3607
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8211320016981851,
+      "learning_rate": 4.501147303197412e-06,
+      "loss": 0.8027,
+      "step": 3608
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5798854122832224,
+      "learning_rate": 4.495938953832822e-06,
+      "loss": 0.7727,
+      "step": 3609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6990068176357884,
+      "learning_rate": 4.490732745578684e-06,
+      "loss": 0.8391,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6314982243477787,
+      "learning_rate": 4.4855286804602425e-06,
+      "loss": 0.8845,
+      "step": 3611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.68624305000348,
+      "learning_rate": 4.480326760501908e-06,
+      "loss": 0.8387,
+      "step": 3612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.615480084278947,
+      "learning_rate": 4.475126987727262e-06,
+      "loss": 0.8975,
+      "step": 3613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5780535468426964,
+      "learning_rate": 4.469929364159035e-06,
+      "loss": 0.8175,
+      "step": 3614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6321944865368159,
+      "learning_rate": 4.464733891819135e-06,
+      "loss": 0.8683,
+      "step": 3615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.611239441817823,
+      "learning_rate": 4.459540572728631e-06,
+      "loss": 0.9075,
+      "step": 3616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4398520580351637,
+      "learning_rate": 4.4543494089077546e-06,
+      "loss": 0.8752,
+      "step": 3617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5933701051225586,
+      "learning_rate": 4.4491604023758936e-06,
+      "loss": 0.8255,
+      "step": 3618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.570569661620049,
+      "learning_rate": 4.443973555151605e-06,
+      "loss": 0.8632,
+      "step": 3619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6149601500472923,
+      "learning_rate": 4.438788869252601e-06,
+      "loss": 0.817,
+      "step": 3620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6051134447174518,
+      "learning_rate": 4.433606346695748e-06,
+      "loss": 0.8078,
+      "step": 3621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.642899810703513,
+      "learning_rate": 4.42842598949708e-06,
+      "loss": 0.9042,
+      "step": 3622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5651938987982812,
+      "learning_rate": 4.423247799671785e-06,
+      "loss": 0.9108,
+      "step": 3623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7166252860629612,
+      "learning_rate": 4.418071779234207e-06,
+      "loss": 0.8608,
+      "step": 3624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5507454195258559,
+      "learning_rate": 4.412897930197846e-06,
+      "loss": 0.9057,
+      "step": 3625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5890203619534435,
+      "learning_rate": 4.40772625457536e-06,
+      "loss": 0.8001,
+      "step": 3626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0096640774147496,
+      "learning_rate": 4.402556754378561e-06,
+      "loss": 0.814,
+      "step": 3627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5945954513125487,
+      "learning_rate": 4.3973894316184085e-06,
+      "loss": 0.9325,
+      "step": 3628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.631201081346321,
+      "learning_rate": 4.3922242883050226e-06,
+      "loss": 0.9118,
+      "step": 3629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6204836664340523,
+      "learning_rate": 4.3870613264476734e-06,
+      "loss": 0.8244,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.731478871103944,
+      "learning_rate": 4.38190054805478e-06,
+      "loss": 0.8113,
+      "step": 3631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5123977711011158,
+      "learning_rate": 4.3767419551339165e-06,
+      "loss": 0.8077,
+      "step": 3632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6673652270095751,
+      "learning_rate": 4.371585549691802e-06,
+      "loss": 0.9099,
+      "step": 3633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6379159211993135,
+      "learning_rate": 4.3664313337343135e-06,
+      "loss": 0.9126,
+      "step": 3634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6769383085355374,
+      "learning_rate": 4.36127930926646e-06,
+      "loss": 0.9803,
+      "step": 3635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6410152542846548,
+      "learning_rate": 4.356129478292411e-06,
+      "loss": 0.8565,
+      "step": 3636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.046570927844233,
+      "learning_rate": 4.350981842815481e-06,
+      "loss": 0.8607,
+      "step": 3637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6736540113121303,
+      "learning_rate": 4.345836404838129e-06,
+      "loss": 0.8205,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6642867538845576,
+      "learning_rate": 4.340693166361959e-06,
+      "loss": 0.8764,
+      "step": 3639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5993763089244322,
+      "learning_rate": 4.335552129387719e-06,
+      "loss": 0.8403,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5965514069669275,
+      "learning_rate": 4.330413295915304e-06,
+      "loss": 0.8799,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5616955027312898,
+      "learning_rate": 4.325276667943743e-06,
+      "loss": 0.8377,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5459849536945933,
+      "learning_rate": 4.320142247471215e-06,
+      "loss": 0.8279,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6135717791683195,
+      "learning_rate": 4.31501003649504e-06,
+      "loss": 0.8815,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4489686482856812,
+      "learning_rate": 4.30988003701168e-06,
+      "loss": 0.7829,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5828748610020011,
+      "learning_rate": 4.304752251016724e-06,
+      "loss": 0.8245,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5654637897986812,
+      "learning_rate": 4.299626680504916e-06,
+      "loss": 0.8715,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.556816543081121,
+      "learning_rate": 4.2945033274701296e-06,
+      "loss": 0.9032,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6833201323986993,
+      "learning_rate": 4.289382193905378e-06,
+      "loss": 0.9032,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7260315763667864,
+      "learning_rate": 4.284263281802815e-06,
+      "loss": 0.8688,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6561593102073475,
+      "learning_rate": 4.279146593153719e-06,
+      "loss": 0.8633,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6244981862029058,
+      "learning_rate": 4.274032129948512e-06,
+      "loss": 0.8406,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.770235850375749,
+      "learning_rate": 4.2689198941767515e-06,
+      "loss": 0.9147,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7254911771370334,
+      "learning_rate": 4.263809887827124e-06,
+      "loss": 0.855,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5957736633415454,
+      "learning_rate": 4.25870211288745e-06,
+      "loss": 0.7852,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6479798030839143,
+      "learning_rate": 4.253596571344684e-06,
+      "loss": 0.8592,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5773883931274022,
+      "learning_rate": 4.248493265184913e-06,
+      "loss": 0.8208,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7001173339824243,
+      "learning_rate": 4.243392196393344e-06,
+      "loss": 0.8333,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7367017462947945,
+      "learning_rate": 4.238293366954326e-06,
+      "loss": 0.8596,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7237695599274405,
+      "learning_rate": 4.2331967788513295e-06,
+      "loss": 0.9093,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.371858799604009,
+      "learning_rate": 4.22810243406696e-06,
+      "loss": 0.8721,
+      "step": 3661
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.473307945567007,
+      "learning_rate": 4.223010334582941e-06,
+      "loss": 0.8293,
+      "step": 3662
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6816983244712878,
+      "learning_rate": 4.217920482380131e-06,
+      "loss": 0.9443,
+      "step": 3663
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.607649892833428,
+      "learning_rate": 4.212832879438514e-06,
+      "loss": 0.8092,
+      "step": 3664
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6442336325854259,
+      "learning_rate": 4.207747527737189e-06,
+      "loss": 0.8246,
+      "step": 3665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8188788520103418,
+      "learning_rate": 4.202664429254389e-06,
+      "loss": 0.8026,
+      "step": 3666
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.693056306567913,
+      "learning_rate": 4.197583585967466e-06,
+      "loss": 0.7964,
+      "step": 3667
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6348642452366027,
+      "learning_rate": 4.1925049998529e-06,
+      "loss": 0.9411,
+      "step": 3668
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.516580412297516,
+      "learning_rate": 4.187428672886287e-06,
+      "loss": 0.8406,
+      "step": 3669
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4541942942790653,
+      "learning_rate": 4.182354607042347e-06,
+      "loss": 0.7673,
+      "step": 3670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6872921810233177,
+      "learning_rate": 4.177282804294925e-06,
+      "loss": 0.8997,
+      "step": 3671
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7302352490966475,
+      "learning_rate": 4.172213266616969e-06,
+      "loss": 0.8819,
+      "step": 3672
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4672097929274464,
+      "learning_rate": 4.167145995980566e-06,
+      "loss": 0.8024,
+      "step": 3673
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5657432344022562,
+      "learning_rate": 4.16208099435691e-06,
+      "loss": 0.8601,
+      "step": 3674
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.646397412188089,
+      "learning_rate": 4.1570182637163155e-06,
+      "loss": 0.8211,
+      "step": 3675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5970961425229546,
+      "learning_rate": 4.151957806028213e-06,
+      "loss": 0.9024,
+      "step": 3676
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7328867284181302,
+      "learning_rate": 4.146899623261148e-06,
+      "loss": 0.8666,
+      "step": 3677
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4524374598188179,
+      "learning_rate": 4.141843717382786e-06,
+      "loss": 0.8546,
+      "step": 3678
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.562223643823434,
+      "learning_rate": 4.136790090359897e-06,
+      "loss": 0.8487,
+      "step": 3679
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6171571576723736,
+      "learning_rate": 4.131738744158371e-06,
+      "loss": 0.8795,
+      "step": 3680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5505750342924478,
+      "learning_rate": 4.126689680743211e-06,
+      "loss": 0.8092,
+      "step": 3681
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8081947189654775,
+      "learning_rate": 4.121642902078531e-06,
+      "loss": 0.7959,
+      "step": 3682
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6662716307818053,
+      "learning_rate": 4.116598410127556e-06,
+      "loss": 0.8922,
+      "step": 3683
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5745333359330091,
+      "learning_rate": 4.11155620685262e-06,
+      "loss": 0.8653,
+      "step": 3684
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8784274240890808,
+      "learning_rate": 4.106516294215175e-06,
+      "loss": 0.8729,
+      "step": 3685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.020779471781759,
+      "learning_rate": 4.101478674175766e-06,
+      "loss": 0.8194,
+      "step": 3686
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7011976126214658,
+      "learning_rate": 4.096443348694058e-06,
+      "loss": 0.9455,
+      "step": 3687
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5877077169408984,
+      "learning_rate": 4.091410319728822e-06,
+      "loss": 0.9195,
+      "step": 3688
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5786814927915547,
+      "learning_rate": 4.086379589237936e-06,
+      "loss": 0.788,
+      "step": 3689
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5788787737407128,
+      "learning_rate": 4.0813511591783775e-06,
+      "loss": 0.9397,
+      "step": 3690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7082145960334039,
+      "learning_rate": 4.076325031506238e-06,
+      "loss": 0.8355,
+      "step": 3691
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.643176569789302,
+      "learning_rate": 4.071301208176711e-06,
+      "loss": 0.8903,
+      "step": 3692
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4689644371042252,
+      "learning_rate": 4.066279691144089e-06,
+      "loss": 0.7907,
+      "step": 3693
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6054150864047054,
+      "learning_rate": 4.061260482361766e-06,
+      "loss": 0.9166,
+      "step": 3694
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5988945398101166,
+      "learning_rate": 4.056243583782246e-06,
+      "loss": 0.7755,
+      "step": 3695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4781123122931703,
+      "learning_rate": 4.0512289973571314e-06,
+      "loss": 0.8449,
+      "step": 3696
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.58515813434214,
+      "learning_rate": 4.046216725037122e-06,
+      "loss": 0.7994,
+      "step": 3697
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9607463601990908,
+      "learning_rate": 4.041206768772023e-06,
+      "loss": 0.8363,
+      "step": 3698
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.623519917604812,
+      "learning_rate": 4.036199130510732e-06,
+      "loss": 0.8843,
+      "step": 3699
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6567885109097125,
+      "learning_rate": 4.031193812201254e-06,
+      "loss": 0.8348,
+      "step": 3700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5536187740034921,
+      "learning_rate": 4.026190815790677e-06,
+      "loss": 0.7633,
+      "step": 3701
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.710181361417983,
+      "learning_rate": 4.0211901432252e-06,
+      "loss": 0.8775,
+      "step": 3702
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6453753362376642,
+      "learning_rate": 4.016191796450111e-06,
+      "loss": 0.831,
+      "step": 3703
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5516933397355017,
+      "learning_rate": 4.011195777409795e-06,
+      "loss": 0.8298,
+      "step": 3704
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5549757496602066,
+      "learning_rate": 4.006202088047732e-06,
+      "loss": 0.9061,
+      "step": 3705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6628915789308327,
+      "learning_rate": 4.001210730306494e-06,
+      "loss": 0.8825,
+      "step": 3706
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4519102568862665,
+      "learning_rate": 3.996221706127751e-06,
+      "loss": 0.8991,
+      "step": 3707
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6641168879996435,
+      "learning_rate": 3.991235017452255e-06,
+      "loss": 0.8447,
+      "step": 3708
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.542751102825101,
+      "learning_rate": 3.986250666219858e-06,
+      "loss": 0.8738,
+      "step": 3709
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6167448554825485,
+      "learning_rate": 3.9812686543695015e-06,
+      "loss": 0.8904,
+      "step": 3710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9850629938707715,
+      "learning_rate": 3.976288983839215e-06,
+      "loss": 0.8032,
+      "step": 3711
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7272732831155586,
+      "learning_rate": 3.971311656566118e-06,
+      "loss": 0.8758,
+      "step": 3712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.736474968863131,
+      "learning_rate": 3.96633667448642e-06,
+      "loss": 0.8151,
+      "step": 3713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6128744724465163,
+      "learning_rate": 3.961364039535419e-06,
+      "loss": 0.802,
+      "step": 3714
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6067134485955124,
+      "learning_rate": 3.95639375364749e-06,
+      "loss": 0.7512,
+      "step": 3715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7557109457594247,
+      "learning_rate": 3.9514258187561084e-06,
+      "loss": 0.9069,
+      "step": 3716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6147266686235102,
+      "learning_rate": 3.946460236793825e-06,
+      "loss": 0.8692,
+      "step": 3717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7381934359851836,
+      "learning_rate": 3.9414970096922825e-06,
+      "loss": 0.8532,
+      "step": 3718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.591714477247225,
+      "learning_rate": 3.9365361393822e-06,
+      "loss": 0.815,
+      "step": 3719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5874456998465467,
+      "learning_rate": 3.931577627793387e-06,
+      "loss": 0.8587,
+      "step": 3720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.678635840094841,
+      "learning_rate": 3.9266214768547335e-06,
+      "loss": 0.8194,
+      "step": 3721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7282277608760357,
+      "learning_rate": 3.921667688494205e-06,
+      "loss": 0.9041,
+      "step": 3722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9455896130989909,
+      "learning_rate": 3.916716264638855e-06,
+      "loss": 0.796,
+      "step": 3723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6722875883271124,
+      "learning_rate": 3.911767207214814e-06,
+      "loss": 0.8908,
+      "step": 3724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7224516320240288,
+      "learning_rate": 3.906820518147294e-06,
+      "loss": 0.7916,
+      "step": 3725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6551523984520358,
+      "learning_rate": 3.901876199360585e-06,
+      "loss": 0.8661,
+      "step": 3726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6391669972253693,
+      "learning_rate": 3.896934252778054e-06,
+      "loss": 0.9176,
+      "step": 3727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5095170300473781,
+      "learning_rate": 3.891994680322151e-06,
+      "loss": 0.8023,
+      "step": 3728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6665459226814674,
+      "learning_rate": 3.887057483914387e-06,
+      "loss": 0.8955,
+      "step": 3729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7711827456538418,
+      "learning_rate": 3.882122665475365e-06,
+      "loss": 0.8596,
+      "step": 3730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5534564840422997,
+      "learning_rate": 3.8771902269247565e-06,
+      "loss": 0.8482,
+      "step": 3731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.74123742278698,
+      "learning_rate": 3.872260170181307e-06,
+      "loss": 0.8529,
+      "step": 3732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5281423345356357,
+      "learning_rate": 3.867332497162836e-06,
+      "loss": 0.8426,
+      "step": 3733
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.52946367949895,
+      "learning_rate": 3.862407209786236e-06,
+      "loss": 0.7698,
+      "step": 3734
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8745145995267125,
+      "learning_rate": 3.8574843099674764e-06,
+      "loss": 0.9299,
+      "step": 3735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6194029634694833,
+      "learning_rate": 3.852563799621582e-06,
+      "loss": 0.7878,
+      "step": 3736
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.722153309268582,
+      "learning_rate": 3.847645680662666e-06,
+      "loss": 0.8381,
+      "step": 3737
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7089182506782774,
+      "learning_rate": 3.842729955003906e-06,
+      "loss": 0.8167,
+      "step": 3738
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1121379831126952,
+      "learning_rate": 3.837816624557539e-06,
+      "loss": 0.8693,
+      "step": 3739
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6673732996903403,
+      "learning_rate": 3.832905691234883e-06,
+      "loss": 0.8735,
+      "step": 3740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5124908373089034,
+      "learning_rate": 3.827997156946317e-06,
+      "loss": 0.8129,
+      "step": 3741
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.507406772319593,
+      "learning_rate": 3.82309102360129e-06,
+      "loss": 0.8299,
+      "step": 3742
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7083475599965583,
+      "learning_rate": 3.818187293108313e-06,
+      "loss": 0.8536,
+      "step": 3743
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5173017099325348,
+      "learning_rate": 3.8132859673749688e-06,
+      "loss": 0.8059,
+      "step": 3744
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6589419512311503,
+      "learning_rate": 3.8083870483078955e-06,
+      "loss": 0.9221,
+      "step": 3745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6040194110981338,
+      "learning_rate": 3.8034905378128005e-06,
+      "loss": 0.8781,
+      "step": 3746
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6195276541610926,
+      "learning_rate": 3.798596437794455e-06,
+      "loss": 0.8355,
+      "step": 3747
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7839310768535381,
+      "learning_rate": 3.7937047501566916e-06,
+      "loss": 0.9086,
+      "step": 3748
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7533421915681255,
+      "learning_rate": 3.7888154768024043e-06,
+      "loss": 0.7984,
+      "step": 3749
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6895078199724958,
+      "learning_rate": 3.7839286196335455e-06,
+      "loss": 0.883,
+      "step": 3750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.641148347109113,
+      "learning_rate": 3.7790441805511367e-06,
+      "loss": 0.759,
+      "step": 3751
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7022158384532617,
+      "learning_rate": 3.7741621614552413e-06,
+      "loss": 0.8435,
+      "step": 3752
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7354494318784945,
+      "learning_rate": 3.769282564244999e-06,
+      "loss": 0.8629,
+      "step": 3753
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0732641327248178,
+      "learning_rate": 3.7644053908185987e-06,
+      "loss": 0.8981,
+      "step": 3754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.789991246442255,
+      "learning_rate": 3.759530643073287e-06,
+      "loss": 0.7584,
+      "step": 3755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7303734100933015,
+      "learning_rate": 3.7546583229053703e-06,
+      "loss": 0.8206,
+      "step": 3756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9870262746909141,
+      "learning_rate": 3.749788432210206e-06,
+      "loss": 0.7975,
+      "step": 3757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7862862272454605,
+      "learning_rate": 3.7449209728822147e-06,
+      "loss": 0.9571,
+      "step": 3758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.80646485010094,
+      "learning_rate": 3.7400559468148555e-06,
+      "loss": 0.8979,
+      "step": 3759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6371183665943017,
+      "learning_rate": 3.735193355900656e-06,
+      "loss": 0.8777,
+      "step": 3760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5439950806704743,
+      "learning_rate": 3.730333202031191e-06,
+      "loss": 0.857,
+      "step": 3761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7162619260418963,
+      "learning_rate": 3.7254754870970868e-06,
+      "loss": 0.9348,
+      "step": 3762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6039119722854573,
+      "learning_rate": 3.720620212988022e-06,
+      "loss": 0.7838,
+      "step": 3763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.844967106184359,
+      "learning_rate": 3.7157673815927263e-06,
+      "loss": 0.8057,
+      "step": 3764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6325179107742063,
+      "learning_rate": 3.71091699479898e-06,
+      "loss": 0.9258,
+      "step": 3765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6455721866633424,
+      "learning_rate": 3.706069054493604e-06,
+      "loss": 0.9435,
+      "step": 3766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.582473827452383,
+      "learning_rate": 3.701223562562478e-06,
+      "loss": 0.9475,
+      "step": 3767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7658946570717726,
+      "learning_rate": 3.6963805208905255e-06,
+      "loss": 0.8961,
+      "step": 3768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.707710539058806,
+      "learning_rate": 3.691539931361717e-06,
+      "loss": 0.8742,
+      "step": 3769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4562446770693516,
+      "learning_rate": 3.686701795859069e-06,
+      "loss": 0.8628,
+      "step": 3770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5747224659759664,
+      "learning_rate": 3.6818661162646416e-06,
+      "loss": 0.76,
+      "step": 3771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.738613933831526,
+      "learning_rate": 3.6770328944595457e-06,
+      "loss": 0.8528,
+      "step": 3772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5684927933516946,
+      "learning_rate": 3.6722021323239256e-06,
+      "loss": 0.8373,
+      "step": 3773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.841772610463016,
+      "learning_rate": 3.6673738317369757e-06,
+      "loss": 0.9335,
+      "step": 3774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4905192952418835,
+      "learning_rate": 3.662547994576934e-06,
+      "loss": 0.8297,
+      "step": 3775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6972491376309515,
+      "learning_rate": 3.6577246227210773e-06,
+      "loss": 0.8969,
+      "step": 3776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5974917505620612,
+      "learning_rate": 3.652903718045724e-06,
+      "loss": 0.7912,
+      "step": 3777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.707692452295001,
+      "learning_rate": 3.648085282426234e-06,
+      "loss": 0.8251,
+      "step": 3778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6428449840293473,
+      "learning_rate": 3.643269317737007e-06,
+      "loss": 0.778,
+      "step": 3779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6427354438274013,
+      "learning_rate": 3.6384558258514758e-06,
+      "loss": 0.8204,
+      "step": 3780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6466126584796252,
+      "learning_rate": 3.633644808642116e-06,
+      "loss": 0.8098,
+      "step": 3781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7645947349475741,
+      "learning_rate": 3.6288362679804434e-06,
+      "loss": 0.937,
+      "step": 3782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0286715466028724,
+      "learning_rate": 3.6240302057370048e-06,
+      "loss": 0.805,
+      "step": 3783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7174969287717576,
+      "learning_rate": 3.61922662378139e-06,
+      "loss": 0.856,
+      "step": 3784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6206767255791732,
+      "learning_rate": 3.6144255239822114e-06,
+      "loss": 0.8756,
+      "step": 3785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6872542704017086,
+      "learning_rate": 3.6096269082071266e-06,
+      "loss": 0.8148,
+      "step": 3786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5260433703163867,
+      "learning_rate": 3.6048307783228265e-06,
+      "loss": 0.86,
+      "step": 3787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6932224425476137,
+      "learning_rate": 3.6000371361950326e-06,
+      "loss": 0.8271,
+      "step": 3788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.621813676832059,
+      "learning_rate": 3.5952459836884933e-06,
+      "loss": 0.8552,
+      "step": 3789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6965597002497659,
+      "learning_rate": 3.590457322666997e-06,
+      "loss": 0.8251,
+      "step": 3790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7519703980141557,
+      "learning_rate": 3.58567115499336e-06,
+      "loss": 0.9033,
+      "step": 3791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6538459481291712,
+      "learning_rate": 3.580887482529426e-06,
+      "loss": 0.8388,
+      "step": 3792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4961949537471149,
+      "learning_rate": 3.5761063071360725e-06,
+      "loss": 0.8385,
+      "step": 3793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5595250456192469,
+      "learning_rate": 3.5713276306732024e-06,
+      "loss": 0.8794,
+      "step": 3794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5438241795023642,
+      "learning_rate": 3.5665514549997527e-06,
+      "loss": 0.8158,
+      "step": 3795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.990998643986908,
+      "learning_rate": 3.5617777819736734e-06,
+      "loss": 0.8339,
+      "step": 3796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4745982264251327,
+      "learning_rate": 3.5570066134519532e-06,
+      "loss": 0.9235,
+      "step": 3797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6299461331425866,
+      "learning_rate": 3.5522379512906045e-06,
+      "loss": 0.8523,
+      "step": 3798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.578662812243374,
+      "learning_rate": 3.5474717973446636e-06,
+      "loss": 0.8172,
+      "step": 3799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6217734157806942,
+      "learning_rate": 3.5427081534681896e-06,
+      "loss": 0.8479,
+      "step": 3800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.779308410235123,
+      "learning_rate": 3.5379470215142674e-06,
+      "loss": 0.9811,
+      "step": 3801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.553704377445108,
+      "learning_rate": 3.533188403335008e-06,
+      "loss": 0.8121,
+      "step": 3802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6199464103390704,
+      "learning_rate": 3.5284323007815334e-06,
+      "loss": 0.8894,
+      "step": 3803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6326132715456005,
+      "learning_rate": 3.5236787157039964e-06,
+      "loss": 0.8839,
+      "step": 3804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7185900030512984,
+      "learning_rate": 3.518927649951569e-06,
+      "loss": 0.8782,
+      "step": 3805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7212522658441494,
+      "learning_rate": 3.5141791053724405e-06,
+      "loss": 0.7825,
+      "step": 3806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8858222738252498,
+      "learning_rate": 3.5094330838138247e-06,
+      "loss": 0.7842,
+      "step": 3807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7366120478795295,
+      "learning_rate": 3.5046895871219466e-06,
+      "loss": 0.8652,
+      "step": 3808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5910390826732626,
+      "learning_rate": 3.4999486171420595e-06,
+      "loss": 0.9079,
+      "step": 3809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6377722734498026,
+      "learning_rate": 3.495210175718419e-06,
+      "loss": 0.8737,
+      "step": 3810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.571384655415289,
+      "learning_rate": 3.490474264694308e-06,
+      "loss": 0.8761,
+      "step": 3811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5516979470302725,
+      "learning_rate": 3.4857408859120223e-06,
+      "loss": 0.8871,
+      "step": 3812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6485757901157507,
+      "learning_rate": 3.4810100412128743e-06,
+      "loss": 0.7898,
+      "step": 3813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6453766258768676,
+      "learning_rate": 3.476281732437188e-06,
+      "loss": 0.8967,
+      "step": 3814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9560314261910495,
+      "learning_rate": 3.4715559614243023e-06,
+      "loss": 0.8368,
+      "step": 3815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5843611446317554,
+      "learning_rate": 3.4668327300125703e-06,
+      "loss": 0.8656,
+      "step": 3816
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.555759423350743,
+      "learning_rate": 3.4621120400393505e-06,
+      "loss": 0.8728,
+      "step": 3817
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6839509334917986,
+      "learning_rate": 3.45739389334102e-06,
+      "loss": 0.8307,
+      "step": 3818
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.631211279757716,
+      "learning_rate": 3.4526782917529646e-06,
+      "loss": 0.8456,
+      "step": 3819
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6169407690314532,
+      "learning_rate": 3.4479652371095797e-06,
+      "loss": 0.7547,
+      "step": 3820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6788480088399391,
+      "learning_rate": 3.443254731244269e-06,
+      "loss": 0.7947,
+      "step": 3821
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5234100245161795,
+      "learning_rate": 3.4385467759894464e-06,
+      "loss": 0.8472,
+      "step": 3822
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6552413369468861,
+      "learning_rate": 3.433841373176536e-06,
+      "loss": 0.8527,
+      "step": 3823
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7409668325207284,
+      "learning_rate": 3.4291385246359586e-06,
+      "loss": 0.9258,
+      "step": 3824
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5304052532197978,
+      "learning_rate": 3.4244382321971524e-06,
+      "loss": 0.8291,
+      "step": 3825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6310445997895089,
+      "learning_rate": 3.4197404976885585e-06,
+      "loss": 0.8458,
+      "step": 3826
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9411611085348174,
+      "learning_rate": 3.415045322937619e-06,
+      "loss": 0.8484,
+      "step": 3827
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6055926835520848,
+      "learning_rate": 3.4103527097707846e-06,
+      "loss": 0.8535,
+      "step": 3828
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.766977634546353,
+      "learning_rate": 3.405662660013509e-06,
+      "loss": 0.8631,
+      "step": 3829
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6683591584594866,
+      "learning_rate": 3.4009751754902487e-06,
+      "loss": 0.9121,
+      "step": 3830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5356923524328252,
+      "learning_rate": 3.3962902580244606e-06,
+      "loss": 0.7507,
+      "step": 3831
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8321383939880476,
+      "learning_rate": 3.391607909438599e-06,
+      "loss": 0.8847,
+      "step": 3832
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0012655360659892,
+      "learning_rate": 3.3869281315541258e-06,
+      "loss": 0.8157,
+      "step": 3833
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.691185005392317,
+      "learning_rate": 3.382250926191504e-06,
+      "loss": 0.8865,
+      "step": 3834
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5966049817433412,
+      "learning_rate": 3.3775762951701886e-06,
+      "loss": 0.8584,
+      "step": 3835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.651835303698378,
+      "learning_rate": 3.37290424030864e-06,
+      "loss": 0.8302,
+      "step": 3836
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9735907512109399,
+      "learning_rate": 3.368234763424313e-06,
+      "loss": 0.7906,
+      "step": 3837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7078311353040243,
+      "learning_rate": 3.3635678663336636e-06,
+      "loss": 0.9326,
+      "step": 3838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5114173469645062,
+      "learning_rate": 3.3589035508521326e-06,
+      "loss": 0.8386,
+      "step": 3839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4577024823644131,
+      "learning_rate": 3.3542418187941695e-06,
+      "loss": 0.8264,
+      "step": 3840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4932493271795817,
+      "learning_rate": 3.349582671973215e-06,
+      "loss": 0.8417,
+      "step": 3841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6065157657708715,
+      "learning_rate": 3.3449261122017008e-06,
+      "loss": 0.8308,
+      "step": 3842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6594977567441667,
+      "learning_rate": 3.3402721412910554e-06,
+      "loss": 0.8015,
+      "step": 3843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6276978451906654,
+      "learning_rate": 3.335620761051701e-06,
+      "loss": 0.8588,
+      "step": 3844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5751942663747105,
+      "learning_rate": 3.3309719732930525e-06,
+      "loss": 0.8758,
+      "step": 3845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.496461504737587,
+      "learning_rate": 3.3263257798235073e-06,
+      "loss": 0.8194,
+      "step": 3846
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7254799713911608,
+      "learning_rate": 3.321682182450465e-06,
+      "loss": 0.9304,
+      "step": 3847
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8039182857737357,
+      "learning_rate": 3.3170411829803105e-06,
+      "loss": 0.8239,
+      "step": 3848
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.524961052257668,
+      "learning_rate": 3.312402783218418e-06,
+      "loss": 0.8486,
+      "step": 3849
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9047167997868268,
+      "learning_rate": 3.3077669849691505e-06,
+      "loss": 0.7957,
+      "step": 3850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6740488860584501,
+      "learning_rate": 3.3031337900358606e-06,
+      "loss": 0.8412,
+      "step": 3851
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6511386334837166,
+      "learning_rate": 3.2985032002208863e-06,
+      "loss": 0.8268,
+      "step": 3852
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7034491526131301,
+      "learning_rate": 3.293875217325556e-06,
+      "loss": 0.9011,
+      "step": 3853
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6081847050588725,
+      "learning_rate": 3.289249843150175e-06,
+      "loss": 0.8297,
+      "step": 3854
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7499357627593553,
+      "learning_rate": 3.2846270794940425e-06,
+      "loss": 0.9172,
+      "step": 3855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7553521842518884,
+      "learning_rate": 3.280006928155439e-06,
+      "loss": 0.856,
+      "step": 3856
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5545021495020346,
+      "learning_rate": 3.2753893909316302e-06,
+      "loss": 0.8964,
+      "step": 3857
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0246614539280072,
+      "learning_rate": 3.2707744696188627e-06,
+      "loss": 0.8427,
+      "step": 3858
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7931560969656708,
+      "learning_rate": 3.2661621660123666e-06,
+      "loss": 0.8563,
+      "step": 3859
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.566807743520784,
+      "learning_rate": 3.261552481906358e-06,
+      "loss": 0.8669,
+      "step": 3860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5511149620535438,
+      "learning_rate": 3.256945419094023e-06,
+      "loss": 0.8512,
+      "step": 3861
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5714892659806674,
+      "learning_rate": 3.2523409793675376e-06,
+      "loss": 0.8607,
+      "step": 3862
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.434417407043539,
+      "learning_rate": 3.2477391645180547e-06,
+      "loss": 0.819,
+      "step": 3863
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6531598780983792,
+      "learning_rate": 3.243139976335705e-06,
+      "loss": 0.9044,
+      "step": 3864
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5263440818357996,
+      "learning_rate": 3.2385434166096e-06,
+      "loss": 0.8341,
+      "step": 3865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4323928653349074,
+      "learning_rate": 3.2339494871278267e-06,
+      "loss": 0.8189,
+      "step": 3866
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8428007358497673,
+      "learning_rate": 3.229358189677451e-06,
+      "loss": 0.8364,
+      "step": 3867
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.669801917393374,
+      "learning_rate": 3.224769526044509e-06,
+      "loss": 0.8314,
+      "step": 3868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.624308371909389,
+      "learning_rate": 3.2201834980140177e-06,
+      "loss": 0.8718,
+      "step": 3869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.633407758383068,
+      "learning_rate": 3.215600107369968e-06,
+      "loss": 0.8698,
+      "step": 3870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.025601753843546,
+      "learning_rate": 3.2110193558953264e-06,
+      "loss": 0.8276,
+      "step": 3871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.840922540778764,
+      "learning_rate": 3.2064412453720282e-06,
+      "loss": 1.0059,
+      "step": 3872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6254181421711398,
+      "learning_rate": 3.201865777580986e-06,
+      "loss": 0.7569,
+      "step": 3873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6139774128414464,
+      "learning_rate": 3.1972929543020836e-06,
+      "loss": 0.8293,
+      "step": 3874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.898553234206285,
+      "learning_rate": 3.19272277731417e-06,
+      "loss": 0.914,
+      "step": 3875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8612503640284934,
+      "learning_rate": 3.188155248395075e-06,
+      "loss": 0.8767,
+      "step": 3876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7373833703411445,
+      "learning_rate": 3.1835903693215862e-06,
+      "loss": 0.9241,
+      "step": 3877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6780787296670276,
+      "learning_rate": 3.179028141869471e-06,
+      "loss": 0.8648,
+      "step": 3878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.642955296678899,
+      "learning_rate": 3.174468567813461e-06,
+      "loss": 0.7847,
+      "step": 3879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5921890824364622,
+      "learning_rate": 3.1699116489272553e-06,
+      "loss": 0.843,
+      "step": 3880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7872413093574824,
+      "learning_rate": 3.1653573869835217e-06,
+      "loss": 0.8637,
+      "step": 3881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7145252077021904,
+      "learning_rate": 3.1608057837538976e-06,
+      "loss": 0.8104,
+      "step": 3882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3895503645799376,
+      "learning_rate": 3.1562568410089733e-06,
+      "loss": 0.8659,
+      "step": 3883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4559935687610588,
+      "learning_rate": 3.1517105605183173e-06,
+      "loss": 0.7999,
+      "step": 3884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6601840270062065,
+      "learning_rate": 3.1471669440504595e-06,
+      "loss": 0.8092,
+      "step": 3885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8065679489460145,
+      "learning_rate": 3.142625993372889e-06,
+      "loss": 0.8513,
+      "step": 3886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6812855586897617,
+      "learning_rate": 3.1380877102520636e-06,
+      "loss": 0.8206,
+      "step": 3887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5835142255619359,
+      "learning_rate": 3.1335520964533983e-06,
+      "loss": 0.8462,
+      "step": 3888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.56221579966197,
+      "learning_rate": 3.1290191537412774e-06,
+      "loss": 0.8188,
+      "step": 3889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5866884521617106,
+      "learning_rate": 3.124488883879034e-06,
+      "loss": 0.794,
+      "step": 3890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6419553384857244,
+      "learning_rate": 3.119961288628972e-06,
+      "loss": 0.7803,
+      "step": 3891
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.568809618104107,
+      "learning_rate": 3.1154363697523503e-06,
+      "loss": 0.8366,
+      "step": 3892
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7618416364659482,
+      "learning_rate": 3.110914129009388e-06,
+      "loss": 0.8697,
+      "step": 3893
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.796353724990638,
+      "learning_rate": 3.106394568159261e-06,
+      "loss": 0.8361,
+      "step": 3894
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5219230213689614,
+      "learning_rate": 3.1018776889601053e-06,
+      "loss": 0.7655,
+      "step": 3895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.64262011653539,
+      "learning_rate": 3.097363493169014e-06,
+      "loss": 0.8717,
+      "step": 3896
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6416953226391338,
+      "learning_rate": 3.09285198254203e-06,
+      "loss": 0.8244,
+      "step": 3897
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6565718013862478,
+      "learning_rate": 3.0883431588341574e-06,
+      "loss": 0.8201,
+      "step": 3898
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.669328752572013,
+      "learning_rate": 3.083837023799353e-06,
+      "loss": 0.8355,
+      "step": 3899
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.555262588161237,
+      "learning_rate": 3.079333579190531e-06,
+      "loss": 0.8263,
+      "step": 3900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0280544499564335,
+      "learning_rate": 3.0748328267595552e-06,
+      "loss": 0.8518,
+      "step": 3901
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.650987015261678,
+      "learning_rate": 3.070334768257244e-06,
+      "loss": 0.8679,
+      "step": 3902
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.536311822330493,
+      "learning_rate": 3.0658394054333684e-06,
+      "loss": 0.8579,
+      "step": 3903
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5404267122970303,
+      "learning_rate": 3.061346740036646e-06,
+      "loss": 0.8154,
+      "step": 3904
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5942751540839963,
+      "learning_rate": 3.0568567738147505e-06,
+      "loss": 0.9141,
+      "step": 3905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6729395018375404,
+      "learning_rate": 3.0523695085143033e-06,
+      "loss": 0.8295,
+      "step": 3906
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5386351438318264,
+      "learning_rate": 3.047884945880877e-06,
+      "loss": 0.8733,
+      "step": 3907
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.608293596125965,
+      "learning_rate": 3.0434030876589906e-06,
+      "loss": 0.8579,
+      "step": 3908
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.684477825102657,
+      "learning_rate": 3.038923935592112e-06,
+      "loss": 0.8782,
+      "step": 3909
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6765663895485428,
+      "learning_rate": 3.0344474914226586e-06,
+      "loss": 0.8348,
+      "step": 3910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6302157456029664,
+      "learning_rate": 3.0299737568919887e-06,
+      "loss": 0.824,
+      "step": 3911
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9518826152128164,
+      "learning_rate": 3.02550273374041e-06,
+      "loss": 0.8254,
+      "step": 3912
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6197618230987454,
+      "learning_rate": 3.0210344237071764e-06,
+      "loss": 0.9695,
+      "step": 3913
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7714728432186924,
+      "learning_rate": 3.0165688285304875e-06,
+      "loss": 0.8983,
+      "step": 3914
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6333253810590913,
+      "learning_rate": 3.0121059499474815e-06,
+      "loss": 0.8796,
+      "step": 3915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.744512338435159,
+      "learning_rate": 3.0076457896942445e-06,
+      "loss": 0.878,
+      "step": 3916
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6245451495368446,
+      "learning_rate": 3.0031883495058077e-06,
+      "loss": 0.8194,
+      "step": 3917
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6429039998422517,
+      "learning_rate": 2.998733631116133e-06,
+      "loss": 0.8683,
+      "step": 3918
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9782555767599034,
+      "learning_rate": 2.994281636258134e-06,
+      "loss": 0.7625,
+      "step": 3919
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.543268084254842,
+      "learning_rate": 2.9898323666636618e-06,
+      "loss": 0.8496,
+      "step": 3920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7147909134974455,
+      "learning_rate": 2.985385824063507e-06,
+      "loss": 0.8144,
+      "step": 3921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.615580594285689,
+      "learning_rate": 2.9809420101873987e-06,
+      "loss": 0.8576,
+      "step": 3922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5668551143173437,
+      "learning_rate": 2.976500926764009e-06,
+      "loss": 0.8653,
+      "step": 3923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5324184979979887,
+      "learning_rate": 2.9720625755209386e-06,
+      "loss": 0.7026,
+      "step": 3924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5085553646829553,
+      "learning_rate": 2.9676269581847327e-06,
+      "loss": 0.7798,
+      "step": 3925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6268880952546954,
+      "learning_rate": 2.9631940764808754e-06,
+      "loss": 0.8335,
+      "step": 3926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7456795722604157,
+      "learning_rate": 2.958763932133775e-06,
+      "loss": 0.8532,
+      "step": 3927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7140035777036833,
+      "learning_rate": 2.9543365268667866e-06,
+      "loss": 0.7947,
+      "step": 3928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6529705887478314,
+      "learning_rate": 2.949911862402195e-06,
+      "loss": 0.8377,
+      "step": 3929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0205093391859472,
+      "learning_rate": 2.945489940461219e-06,
+      "loss": 0.7964,
+      "step": 3930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.515685976240434,
+      "learning_rate": 2.9410707627640124e-06,
+      "loss": 0.8312,
+      "step": 3931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7660112414211953,
+      "learning_rate": 2.936654331029659e-06,
+      "loss": 0.87,
+      "step": 3932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7105637694685787,
+      "learning_rate": 2.932240646976178e-06,
+      "loss": 0.9917,
+      "step": 3933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5641638783653147,
+      "learning_rate": 2.9278297123205113e-06,
+      "loss": 0.8388,
+      "step": 3934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7608749258919396,
+      "learning_rate": 2.9234215287785406e-06,
+      "loss": 0.8193,
+      "step": 3935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.633553901702133,
+      "learning_rate": 2.919016098065073e-06,
+      "loss": 0.8468,
+      "step": 3936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5787681868671055,
+      "learning_rate": 2.914613421893847e-06,
+      "loss": 0.8513,
+      "step": 3937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6976417629886829,
+      "learning_rate": 2.9102135019775267e-06,
+      "loss": 0.8856,
+      "step": 3938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6068659250546018,
+      "learning_rate": 2.905816340027706e-06,
+      "loss": 0.875,
+      "step": 3939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.651967798284204,
+      "learning_rate": 2.901421937754909e-06,
+      "loss": 0.9143,
+      "step": 3940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7488634469977615,
+      "learning_rate": 2.897030296868577e-06,
+      "loss": 0.9152,
+      "step": 3941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6201300757977672,
+      "learning_rate": 2.8926414190770847e-06,
+      "loss": 0.8137,
+      "step": 3942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.528264578370851,
+      "learning_rate": 2.8882553060877296e-06,
+      "loss": 0.8202,
+      "step": 3943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7321413330435798,
+      "learning_rate": 2.8838719596067357e-06,
+      "loss": 0.8767,
+      "step": 3944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6919255906734783,
+      "learning_rate": 2.879491381339248e-06,
+      "loss": 0.8908,
+      "step": 3945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7077918579835913,
+      "learning_rate": 2.8751135729893378e-06,
+      "loss": 0.8425,
+      "step": 3946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5772461163848126,
+      "learning_rate": 2.870738536259997e-06,
+      "loss": 0.9118,
+      "step": 3947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6282911443553783,
+      "learning_rate": 2.866366272853135e-06,
+      "loss": 0.9092,
+      "step": 3948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.597259950499336,
+      "learning_rate": 2.8619967844695896e-06,
+      "loss": 0.8641,
+      "step": 3949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8321827305623852,
+      "learning_rate": 2.8576300728091166e-06,
+      "loss": 0.9581,
+      "step": 3950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.707180123915409,
+      "learning_rate": 2.853266139570391e-06,
+      "loss": 0.7967,
+      "step": 3951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7084549423190638,
+      "learning_rate": 2.8489049864510053e-06,
+      "loss": 0.8719,
+      "step": 3952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5344670076320632,
+      "learning_rate": 2.8445466151474755e-06,
+      "loss": 0.8407,
+      "step": 3953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5568170570674382,
+      "learning_rate": 2.8401910273552334e-06,
+      "loss": 0.8111,
+      "step": 3954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6248313064543114,
+      "learning_rate": 2.835838224768621e-06,
+      "loss": 0.8417,
+      "step": 3955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6982726534154837,
+      "learning_rate": 2.831488209080906e-06,
+      "loss": 0.797,
+      "step": 3956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4588971386617062,
+      "learning_rate": 2.8271409819842687e-06,
+      "loss": 0.7634,
+      "step": 3957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.695505314759919,
+      "learning_rate": 2.822796545169806e-06,
+      "loss": 0.8883,
+      "step": 3958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6085577245269673,
+      "learning_rate": 2.818454900327526e-06,
+      "loss": 0.9072,
+      "step": 3959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6591439053068358,
+      "learning_rate": 2.8141160491463537e-06,
+      "loss": 0.8666,
+      "step": 3960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.665123254034379,
+      "learning_rate": 2.80977999331413e-06,
+      "loss": 0.9007,
+      "step": 3961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.778377362867319,
+      "learning_rate": 2.8054467345175983e-06,
+      "loss": 0.8994,
+      "step": 3962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5834109956446196,
+      "learning_rate": 2.8011162744424224e-06,
+      "loss": 0.7475,
+      "step": 3963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7383557599814872,
+      "learning_rate": 2.7967886147731772e-06,
+      "loss": 0.911,
+      "step": 3964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.690375643242245,
+      "learning_rate": 2.7924637571933455e-06,
+      "loss": 0.7715,
+      "step": 3965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5138481117869569,
+      "learning_rate": 2.788141703385321e-06,
+      "loss": 0.8607,
+      "step": 3966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7671502347398909,
+      "learning_rate": 2.783822455030407e-06,
+      "loss": 0.798,
+      "step": 3967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5969290067214517,
+      "learning_rate": 2.779506013808818e-06,
+      "loss": 0.8639,
+      "step": 3968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8610869071017357,
+      "learning_rate": 2.7751923813996686e-06,
+      "loss": 0.8561,
+      "step": 3969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5607784418468138,
+      "learning_rate": 2.770881559480991e-06,
+      "loss": 0.843,
+      "step": 3970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6728863965775613,
+      "learning_rate": 2.766573549729712e-06,
+      "loss": 0.8566,
+      "step": 3971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5829000179453876,
+      "learning_rate": 2.7622683538216767e-06,
+      "loss": 0.8641,
+      "step": 3972
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.634786455095246,
+      "learning_rate": 2.7579659734316287e-06,
+      "loss": 0.7931,
+      "step": 3973
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6191756389924223,
+      "learning_rate": 2.7536664102332177e-06,
+      "loss": 0.811,
+      "step": 3974
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9849713541307549,
+      "learning_rate": 2.749369665899e-06,
+      "loss": 0.8048,
+      "step": 3975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.581592598618574,
+      "learning_rate": 2.7450757421004302e-06,
+      "loss": 0.8617,
+      "step": 3976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5064147259069443,
+      "learning_rate": 2.740784640507874e-06,
+      "loss": 0.8487,
+      "step": 3977
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9344588079430771,
+      "learning_rate": 2.736496362790586e-06,
+      "loss": 0.7782,
+      "step": 3978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.715311367904715,
+      "learning_rate": 2.732210910616735e-06,
+      "loss": 0.8296,
+      "step": 3979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5953548792367103,
+      "learning_rate": 2.727928285653384e-06,
+      "loss": 0.7571,
+      "step": 3980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5213949544062197,
+      "learning_rate": 2.7236484895664995e-06,
+      "loss": 0.8582,
+      "step": 3981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7245287343212736,
+      "learning_rate": 2.7193715240209452e-06,
+      "loss": 0.8914,
+      "step": 3982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5882206727488049,
+      "learning_rate": 2.7150973906804844e-06,
+      "loss": 0.8295,
+      "step": 3983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.665671333051446,
+      "learning_rate": 2.7108260912077834e-06,
+      "loss": 0.9335,
+      "step": 3984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6739836552617349,
+      "learning_rate": 2.706557627264392e-06,
+      "loss": 0.7713,
+      "step": 3985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7897326351088907,
+      "learning_rate": 2.7022920005107733e-06,
+      "loss": 0.8919,
+      "step": 3986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0063957999158306,
+      "learning_rate": 2.698029212606279e-06,
+      "loss": 0.8114,
+      "step": 3987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6216498912980306,
+      "learning_rate": 2.6937692652091563e-06,
+      "loss": 0.8593,
+      "step": 3988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7635258605654442,
+      "learning_rate": 2.6895121599765485e-06,
+      "loss": 0.8153,
+      "step": 3989
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8286307686033851,
+      "learning_rate": 2.685257898564494e-06,
+      "loss": 0.8629,
+      "step": 3990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.987494811500732,
+      "learning_rate": 2.6810064826279257e-06,
+      "loss": 0.7917,
+      "step": 3991
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6085858133461552,
+      "learning_rate": 2.6767579138206644e-06,
+      "loss": 0.8783,
+      "step": 3992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8155270471899105,
+      "learning_rate": 2.6725121937954303e-06,
+      "loss": 0.9501,
+      "step": 3993
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6847436427653715,
+      "learning_rate": 2.668269324203829e-06,
+      "loss": 0.8551,
+      "step": 3994
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5536844898603523,
+      "learning_rate": 2.6640293066963628e-06,
+      "loss": 0.8492,
+      "step": 3995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5730228503986523,
+      "learning_rate": 2.6597921429224227e-06,
+      "loss": 0.8311,
+      "step": 3996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.651690151684828,
+      "learning_rate": 2.655557834530288e-06,
+      "loss": 0.8627,
+      "step": 3997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5946480254471995,
+      "learning_rate": 2.6513263831671298e-06,
+      "loss": 0.8562,
+      "step": 3998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.649150578183313,
+      "learning_rate": 2.647097790479003e-06,
+      "loss": 0.8708,
+      "step": 3999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.629937165924763,
+      "learning_rate": 2.6428720581108557e-06,
+      "loss": 0.8536,
+      "step": 4000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9750985468160686,
+      "learning_rate": 2.638649187706521e-06,
+      "loss": 0.8304,
+      "step": 4001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7489494673972723,
+      "learning_rate": 2.6344291809087195e-06,
+      "loss": 0.9284,
+      "step": 4002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6860453054328925,
+      "learning_rate": 2.630212039359057e-06,
+      "loss": 0.9002,
+      "step": 4003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6777749845768535,
+      "learning_rate": 2.625997764698025e-06,
+      "loss": 0.8969,
+      "step": 4004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5460791854169103,
+      "learning_rate": 2.621786358565004e-06,
+      "loss": 0.8236,
+      "step": 4005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5423774827538266,
+      "learning_rate": 2.6175778225982475e-06,
+      "loss": 0.7759,
+      "step": 4006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6493142540818466,
+      "learning_rate": 2.6133721584349013e-06,
+      "loss": 0.8416,
+      "step": 4007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6854653313243428,
+      "learning_rate": 2.6091693677109943e-06,
+      "loss": 0.8573,
+      "step": 4008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6578232721517279,
+      "learning_rate": 2.6049694520614355e-06,
+      "loss": 0.8429,
+      "step": 4009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6139833922330244,
+      "learning_rate": 2.600772413120014e-06,
+      "loss": 0.8326,
+      "step": 4010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.750987155204775,
+      "learning_rate": 2.5965782525194016e-06,
+      "loss": 0.8482,
+      "step": 4011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7329958572271473,
+      "learning_rate": 2.5923869718911543e-06,
+      "loss": 0.8838,
+      "step": 4012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6143331898925106,
+      "learning_rate": 2.5881985728656977e-06,
+      "loss": 0.8803,
+      "step": 4013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6875051966454624,
+      "learning_rate": 2.5840130570723434e-06,
+      "loss": 0.8836,
+      "step": 4014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6778119658593789,
+      "learning_rate": 2.579830426139286e-06,
+      "loss": 0.841,
+      "step": 4015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5250848064568832,
+      "learning_rate": 2.5756506816935843e-06,
+      "loss": 0.8104,
+      "step": 4016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6857031577872985,
+      "learning_rate": 2.571473825361186e-06,
+      "loss": 0.7928,
+      "step": 4017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7698418692944227,
+      "learning_rate": 2.5672998587669105e-06,
+      "loss": 0.8894,
+      "step": 4018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5466879456392286,
+      "learning_rate": 2.5631287835344566e-06,
+      "loss": 0.8276,
+      "step": 4019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5374121100366827,
+      "learning_rate": 2.5589606012863968e-06,
+      "loss": 0.8593,
+      "step": 4020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6114041268471362,
+      "learning_rate": 2.554795313644172e-06,
+      "loss": 0.9032,
+      "step": 4021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6646435390716499,
+      "learning_rate": 2.5506329222281034e-06,
+      "loss": 0.8401,
+      "step": 4022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6525402105733686,
+      "learning_rate": 2.5464734286573876e-06,
+      "loss": 0.8313,
+      "step": 4023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.666593815124738,
+      "learning_rate": 2.5423168345500883e-06,
+      "loss": 0.8557,
+      "step": 4024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6653188317199388,
+      "learning_rate": 2.5381631415231455e-06,
+      "loss": 0.7962,
+      "step": 4025
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5579400467436564,
+      "learning_rate": 2.5340123511923686e-06,
+      "loss": 0.8525,
+      "step": 4026
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6398282287598558,
+      "learning_rate": 2.52986446517244e-06,
+      "loss": 0.8114,
+      "step": 4027
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7152468117735018,
+      "learning_rate": 2.5257194850769064e-06,
+      "loss": 0.9185,
+      "step": 4028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7290377080995591,
+      "learning_rate": 2.5215774125181893e-06,
+      "loss": 0.9536,
+      "step": 4029
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9572673637206633,
+      "learning_rate": 2.5174382491075787e-06,
+      "loss": 0.7616,
+      "step": 4030
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.71164777338371,
+      "learning_rate": 2.513301996455233e-06,
+      "loss": 0.8677,
+      "step": 4031
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0055922946021847,
+      "learning_rate": 2.509168656170177e-06,
+      "loss": 0.7883,
+      "step": 4032
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6461259455612434,
+      "learning_rate": 2.505038229860303e-06,
+      "loss": 0.8268,
+      "step": 4033
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.717810269968421,
+      "learning_rate": 2.5009107191323723e-06,
+      "loss": 0.8488,
+      "step": 4034
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5870071817331732,
+      "learning_rate": 2.4967861255920044e-06,
+      "loss": 0.9029,
+      "step": 4035
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.744809933393664,
+      "learning_rate": 2.4926644508436925e-06,
+      "loss": 0.953,
+      "step": 4036
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6506397906199426,
+      "learning_rate": 2.4885456964907905e-06,
+      "loss": 0.923,
+      "step": 4037
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.63340596736324,
+      "learning_rate": 2.484429864135518e-06,
+      "loss": 0.7941,
+      "step": 4038
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7113019525253217,
+      "learning_rate": 2.480316955378954e-06,
+      "loss": 0.862,
+      "step": 4039
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9178802216240571,
+      "learning_rate": 2.4762069718210456e-06,
+      "loss": 0.7481,
+      "step": 4040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6316671540423857,
+      "learning_rate": 2.4720999150606016e-06,
+      "loss": 0.8324,
+      "step": 4041
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7090204246482104,
+      "learning_rate": 2.467995786695283e-06,
+      "loss": 0.7771,
+      "step": 4042
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6144510334883178,
+      "learning_rate": 2.4638945883216236e-06,
+      "loss": 0.8359,
+      "step": 4043
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6937341547401072,
+      "learning_rate": 2.459796321535011e-06,
+      "loss": 0.8228,
+      "step": 4044
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9553792886752742,
+      "learning_rate": 2.4557009879296935e-06,
+      "loss": 0.8394,
+      "step": 4045
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.630941546812395,
+      "learning_rate": 2.4516085890987794e-06,
+      "loss": 0.7427,
+      "step": 4046
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.616508608565966,
+      "learning_rate": 2.4475191266342345e-06,
+      "loss": 0.8607,
+      "step": 4047
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.496771032114934,
+      "learning_rate": 2.443432602126886e-06,
+      "loss": 0.8382,
+      "step": 4048
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6542917918157012,
+      "learning_rate": 2.4393490171664093e-06,
+      "loss": 0.9011,
+      "step": 4049
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6115033504688236,
+      "learning_rate": 2.435268373341343e-06,
+      "loss": 0.8156,
+      "step": 4050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6040056976215498,
+      "learning_rate": 2.43119067223908e-06,
+      "loss": 0.8624,
+      "step": 4051
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.587047264047987,
+      "learning_rate": 2.427115915445871e-06,
+      "loss": 0.8462,
+      "step": 4052
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4142952837605076,
+      "learning_rate": 2.4230441045468176e-06,
+      "loss": 0.7697,
+      "step": 4053
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5301257694257,
+      "learning_rate": 2.4189752411258783e-06,
+      "loss": 0.7757,
+      "step": 4054
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7828693346790805,
+      "learning_rate": 2.414909326765864e-06,
+      "loss": 0.8361,
+      "step": 4055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8880380003261805,
+      "learning_rate": 2.410846363048436e-06,
+      "loss": 0.9493,
+      "step": 4056
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6599249345232954,
+      "learning_rate": 2.406786351554109e-06,
+      "loss": 0.923,
+      "step": 4057
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6133859917849176,
+      "learning_rate": 2.4027292938622526e-06,
+      "loss": 0.8461,
+      "step": 4058
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9580808723696416,
+      "learning_rate": 2.398675191551083e-06,
+      "loss": 0.7711,
+      "step": 4059
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6568351157204293,
+      "learning_rate": 2.394624046197671e-06,
+      "loss": 0.9109,
+      "step": 4060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8523626110783842,
+      "learning_rate": 2.390575859377935e-06,
+      "loss": 0.9126,
+      "step": 4061
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.744285916832801,
+      "learning_rate": 2.3865306326666383e-06,
+      "loss": 0.8598,
+      "step": 4062
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8092716406481215,
+      "learning_rate": 2.3824883676374e-06,
+      "loss": 0.9196,
+      "step": 4063
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5490720962625313,
+      "learning_rate": 2.378449065862685e-06,
+      "loss": 0.8439,
+      "step": 4064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5235174487868732,
+      "learning_rate": 2.3744127289137987e-06,
+      "loss": 0.7315,
+      "step": 4065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6016891365173485,
+      "learning_rate": 2.3703793583609013e-06,
+      "loss": 0.8659,
+      "step": 4066
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5650539332224847,
+      "learning_rate": 2.3663489557729992e-06,
+      "loss": 0.9074,
+      "step": 4067
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6410695706131928,
+      "learning_rate": 2.3623215227179377e-06,
+      "loss": 0.8126,
+      "step": 4068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.514127147314867,
+      "learning_rate": 2.3582970607624134e-06,
+      "loss": 0.8635,
+      "step": 4069
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5261112247598625,
+      "learning_rate": 2.354275571471962e-06,
+      "loss": 0.8631,
+      "step": 4070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6584632165754962,
+      "learning_rate": 2.3502570564109717e-06,
+      "loss": 0.6964,
+      "step": 4071
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6284620710824946,
+      "learning_rate": 2.346241517142658e-06,
+      "loss": 0.8738,
+      "step": 4072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.781747153946245,
+      "learning_rate": 2.3422289552290935e-06,
+      "loss": 0.82,
+      "step": 4073
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6207095112924301,
+      "learning_rate": 2.3382193722311852e-06,
+      "loss": 0.8,
+      "step": 4074
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.626869848877414,
+      "learning_rate": 2.3342127697086846e-06,
+      "loss": 0.825,
+      "step": 4075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5579991867381682,
+      "learning_rate": 2.3302091492201818e-06,
+      "loss": 0.8916,
+      "step": 4076
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6533913087638532,
+      "learning_rate": 2.326208512323107e-06,
+      "loss": 0.8681,
+      "step": 4077
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0078006730521327,
+      "learning_rate": 2.3222108605737348e-06,
+      "loss": 0.8381,
+      "step": 4078
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.968042420380422,
+      "learning_rate": 2.3182161955271676e-06,
+      "loss": 0.7764,
+      "step": 4079
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.972024210033977,
+      "learning_rate": 2.3142245187373545e-06,
+      "loss": 0.8132,
+      "step": 4080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4566202983861185,
+      "learning_rate": 2.3102358317570806e-06,
+      "loss": 0.8019,
+      "step": 4081
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6390510048583182,
+      "learning_rate": 2.3062501361379662e-06,
+      "loss": 0.8058,
+      "step": 4082
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5855385474350372,
+      "learning_rate": 2.3022674334304718e-06,
+      "loss": 0.7997,
+      "step": 4083
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6643876492382073,
+      "learning_rate": 2.2982877251838887e-06,
+      "loss": 0.9267,
+      "step": 4084
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7290064877181581,
+      "learning_rate": 2.2943110129463496e-06,
+      "loss": 0.8943,
+      "step": 4085
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6738065128703445,
+      "learning_rate": 2.2903372982648108e-06,
+      "loss": 0.8438,
+      "step": 4086
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.578826390164398,
+      "learning_rate": 2.2863665826850734e-06,
+      "loss": 0.7739,
+      "step": 4087
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5991027615541162,
+      "learning_rate": 2.282398867751768e-06,
+      "loss": 0.8686,
+      "step": 4088
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6142410501770876,
+      "learning_rate": 2.2784341550083577e-06,
+      "loss": 0.8419,
+      "step": 4089
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4737932992551817,
+      "learning_rate": 2.274472445997138e-06,
+      "loss": 0.7822,
+      "step": 4090
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7197650467341905,
+      "learning_rate": 2.2705137422592338e-06,
+      "loss": 0.8593,
+      "step": 4091
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.554039804037853,
+      "learning_rate": 2.266558045334609e-06,
+      "loss": 0.9028,
+      "step": 4092
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6660949642343355,
+      "learning_rate": 2.2626053567620443e-06,
+      "loss": 0.8692,
+      "step": 4093
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7016143223720606,
+      "learning_rate": 2.2586556780791603e-06,
+      "loss": 0.8727,
+      "step": 4094
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.565836329379734,
+      "learning_rate": 2.2547090108224047e-06,
+      "loss": 0.8128,
+      "step": 4095
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6091900686719338,
+      "learning_rate": 2.2507653565270525e-06,
+      "loss": 0.8099,
+      "step": 4096
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5596321586477313,
+      "learning_rate": 2.246824716727207e-06,
+      "loss": 0.9062,
+      "step": 4097
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.745696206632949,
+      "learning_rate": 2.2428870929558012e-06,
+      "loss": 0.859,
+      "step": 4098
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6900148771838046,
+      "learning_rate": 2.2389524867445922e-06,
+      "loss": 0.9086,
+      "step": 4099
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0384951811909353,
+      "learning_rate": 2.235020899624162e-06,
+      "loss": 0.8446,
+      "step": 4100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.987118619482762,
+      "learning_rate": 2.2310923331239185e-06,
+      "loss": 0.8289,
+      "step": 4101
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6689297362284206,
+      "learning_rate": 2.227166788772099e-06,
+      "loss": 0.9144,
+      "step": 4102
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6986427209801311,
+      "learning_rate": 2.223244268095761e-06,
+      "loss": 0.7504,
+      "step": 4103
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.572077793452484,
+      "learning_rate": 2.2193247726207876e-06,
+      "loss": 0.7821,
+      "step": 4104
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.552423223735056,
+      "learning_rate": 2.2154083038718834e-06,
+      "loss": 0.8266,
+      "step": 4105
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.763622601293992,
+      "learning_rate": 2.2114948633725806e-06,
+      "loss": 0.8951,
+      "step": 4106
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.637063063847859,
+      "learning_rate": 2.207584452645223e-06,
+      "loss": 0.9103,
+      "step": 4107
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7990475832492232,
+      "learning_rate": 2.203677073210988e-06,
+      "loss": 0.8815,
+      "step": 4108
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9535211362135779,
+      "learning_rate": 2.1997727265898615e-06,
+      "loss": 0.8427,
+      "step": 4109
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5619090985995323,
+      "learning_rate": 2.195871414300661e-06,
+      "loss": 0.8059,
+      "step": 4110
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5228710772933862,
+      "learning_rate": 2.1919731378610164e-06,
+      "loss": 0.886,
+      "step": 4111
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9564165204090196,
+      "learning_rate": 2.1880778987873806e-06,
+      "loss": 0.7742,
+      "step": 4112
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5307603913955545,
+      "learning_rate": 2.1841856985950226e-06,
+      "loss": 0.9008,
+      "step": 4113
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.553283791655139,
+      "learning_rate": 2.18029653879803e-06,
+      "loss": 0.8891,
+      "step": 4114
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.69179341691389,
+      "learning_rate": 2.1764104209093116e-06,
+      "loss": 0.8626,
+      "step": 4115
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.566307259843035,
+      "learning_rate": 2.1725273464405815e-06,
+      "loss": 0.884,
+      "step": 4116
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.719463493021518,
+      "learning_rate": 2.1686473169023815e-06,
+      "loss": 0.8274,
+      "step": 4117
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5302320648431127,
+      "learning_rate": 2.164770333804065e-06,
+      "loss": 0.8471,
+      "step": 4118
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6884629211301279,
+      "learning_rate": 2.160896398653798e-06,
+      "loss": 0.9017,
+      "step": 4119
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.611976098632058,
+      "learning_rate": 2.157025512958565e-06,
+      "loss": 0.8331,
+      "step": 4120
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5173698267314715,
+      "learning_rate": 2.1531576782241605e-06,
+      "loss": 0.8722,
+      "step": 4121
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.732043199436019,
+      "learning_rate": 2.1492928959551983e-06,
+      "loss": 0.887,
+      "step": 4122
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5256371409345997,
+      "learning_rate": 2.145431167655093e-06,
+      "loss": 0.786,
+      "step": 4123
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6724944411728144,
+      "learning_rate": 2.1415724948260828e-06,
+      "loss": 0.9018,
+      "step": 4124
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.585996790215057,
+      "learning_rate": 2.1377168789692114e-06,
+      "loss": 0.8082,
+      "step": 4125
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5868473513946162,
+      "learning_rate": 2.1338643215843368e-06,
+      "loss": 0.8681,
+      "step": 4126
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6350104550155353,
+      "learning_rate": 2.1300148241701235e-06,
+      "loss": 0.838,
+      "step": 4127
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.645054680359249,
+      "learning_rate": 2.126168388224048e-06,
+      "loss": 0.7968,
+      "step": 4128
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6356682372608564,
+      "learning_rate": 2.122325015242398e-06,
+      "loss": 0.8871,
+      "step": 4129
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.522367076780162,
+      "learning_rate": 2.118484706720263e-06,
+      "loss": 0.8234,
+      "step": 4130
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6540354096680119,
+      "learning_rate": 2.1146474641515446e-06,
+      "loss": 0.8794,
+      "step": 4131
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.69199232062196,
+      "learning_rate": 2.1108132890289533e-06,
+      "loss": 0.8914,
+      "step": 4132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6303261592521145,
+      "learning_rate": 2.106982182844005e-06,
+      "loss": 0.8499,
+      "step": 4133
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6227008562156693,
+      "learning_rate": 2.1031541470870198e-06,
+      "loss": 0.7955,
+      "step": 4134
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6596481589588663,
+      "learning_rate": 2.099329183247126e-06,
+      "loss": 0.8679,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.553838583602103,
+      "learning_rate": 2.0955072928122587e-06,
+      "loss": 0.8335,
+      "step": 4136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6363635847477036,
+      "learning_rate": 2.091688477269149e-06,
+      "loss": 0.8967,
+      "step": 4137
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6264565873359476,
+      "learning_rate": 2.0878727381033392e-06,
+      "loss": 0.7912,
+      "step": 4138
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.577505780952043,
+      "learning_rate": 2.084060076799175e-06,
+      "loss": 0.8208,
+      "step": 4139
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.767352671272326,
+      "learning_rate": 2.0802504948398006e-06,
+      "loss": 0.8687,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.733824282408085,
+      "learning_rate": 2.0764439937071666e-06,
+      "loss": 0.8579,
+      "step": 4141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.673918446079239,
+      "learning_rate": 2.0726405748820233e-06,
+      "loss": 0.8828,
+      "step": 4142
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.619983402983795,
+      "learning_rate": 2.068840239843922e-06,
+      "loss": 0.7744,
+      "step": 4143
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6255381575702414,
+      "learning_rate": 2.065042990071212e-06,
+      "loss": 0.7796,
+      "step": 4144
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.800828610807988,
+      "learning_rate": 2.0612488270410447e-06,
+      "loss": 0.8639,
+      "step": 4145
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4331666743173797,
+      "learning_rate": 2.0574577522293726e-06,
+      "loss": 0.8631,
+      "step": 4146
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.702540881187758,
+      "learning_rate": 2.053669767110944e-06,
+      "loss": 0.7642,
+      "step": 4147
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6372903650932455,
+      "learning_rate": 2.0498848731593067e-06,
+      "loss": 0.845,
+      "step": 4148
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.465865153203751,
+      "learning_rate": 2.0461030718468055e-06,
+      "loss": 0.8294,
+      "step": 4149
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5783135738070329,
+      "learning_rate": 2.0423243646445846e-06,
+      "loss": 0.8551,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6727395283448956,
+      "learning_rate": 2.0385487530225757e-06,
+      "loss": 0.8149,
+      "step": 4151
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8671378319703242,
+      "learning_rate": 2.0347762384495182e-06,
+      "loss": 0.912,
+      "step": 4152
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5983191896329303,
+      "learning_rate": 2.031006822392939e-06,
+      "loss": 0.9018,
+      "step": 4153
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5952924827835413,
+      "learning_rate": 2.027240506319166e-06,
+      "loss": 0.9322,
+      "step": 4154
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6825529018185432,
+      "learning_rate": 2.023477291693311e-06,
+      "loss": 0.822,
+      "step": 4155
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8226563942244558,
+      "learning_rate": 2.019717179979288e-06,
+      "loss": 0.8919,
+      "step": 4156
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6766644435468039,
+      "learning_rate": 2.0159601726398027e-06,
+      "loss": 0.872,
+      "step": 4157
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6574111661778033,
+      "learning_rate": 2.012206271136353e-06,
+      "loss": 0.8258,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7529037554326659,
+      "learning_rate": 2.008455476929224e-06,
+      "loss": 0.8757,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6178540332590046,
+      "learning_rate": 2.004707791477497e-06,
+      "loss": 0.7926,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9932326143042078,
+      "learning_rate": 2.000963216239041e-06,
+      "loss": 0.8097,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6756737117048939,
+      "learning_rate": 1.997221752670521e-06,
+      "loss": 0.8625,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.679154709137768,
+      "learning_rate": 1.993483402227383e-06,
+      "loss": 0.8294,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6333120820047409,
+      "learning_rate": 1.9897481663638674e-06,
+      "loss": 0.9028,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9633120910091779,
+      "learning_rate": 1.9860160465330057e-06,
+      "loss": 0.7665,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4951178942127465,
+      "learning_rate": 1.982287044186607e-06,
+      "loss": 0.8197,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4368760390183593,
+      "learning_rate": 1.9785611607752763e-06,
+      "loss": 0.8591,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7094553951120224,
+      "learning_rate": 1.9748383977484044e-06,
+      "loss": 0.8758,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6762282728076576,
+      "learning_rate": 1.9711187565541657e-06,
+      "loss": 0.9373,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4213848736645005,
+      "learning_rate": 1.9674022386395207e-06,
+      "loss": 0.8648,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7151557538573141,
+      "learning_rate": 1.963688845450218e-06,
+      "loss": 0.8077,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7551839993113323,
+      "learning_rate": 1.959978578430789e-06,
+      "loss": 0.902,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6889078189010633,
+      "learning_rate": 1.956271439024544e-06,
+      "loss": 0.8312,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6777833172499128,
+      "learning_rate": 1.952567428673582e-06,
+      "loss": 0.8768,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6044349176441102,
+      "learning_rate": 1.948866548818785e-06,
+      "loss": 0.8642,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.737775941064995,
+      "learning_rate": 1.945168800899816e-06,
+      "loss": 0.9193,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8411680509298494,
+      "learning_rate": 1.941474186355119e-06,
+      "loss": 0.9066,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.698486473889725,
+      "learning_rate": 1.93778270662192e-06,
+      "loss": 0.9119,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.750234707272132,
+      "learning_rate": 1.9340943631362285e-06,
+      "loss": 0.8409,
+      "step": 4179
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.630790257126567,
+      "learning_rate": 1.9304091573328232e-06,
+      "loss": 0.8594,
+      "step": 4180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5169349570487973,
+      "learning_rate": 1.926727090645275e-06,
+      "loss": 0.8414,
+      "step": 4181
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7750838185933542,
+      "learning_rate": 1.9230481645059263e-06,
+      "loss": 0.9201,
+      "step": 4182
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4916162559995068,
+      "learning_rate": 1.919372380345902e-06,
+      "loss": 0.7608,
+      "step": 4183
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6555032241939007,
+      "learning_rate": 1.9156997395951005e-06,
+      "loss": 0.9071,
+      "step": 4184
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6351673454589335,
+      "learning_rate": 1.9120302436822e-06,
+      "loss": 0.9004,
+      "step": 4185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.664953630697614,
+      "learning_rate": 1.9083638940346563e-06,
+      "loss": 0.8554,
+      "step": 4186
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.751598320140283,
+      "learning_rate": 1.9047006920787013e-06,
+      "loss": 0.8228,
+      "step": 4187
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.625473745353837,
+      "learning_rate": 1.9010406392393355e-06,
+      "loss": 0.8257,
+      "step": 4188
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7768286546633896,
+      "learning_rate": 1.8973837369403437e-06,
+      "loss": 0.8727,
+      "step": 4189
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.579782036734651,
+      "learning_rate": 1.8937299866042802e-06,
+      "loss": 0.8118,
+      "step": 4190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.68770533817507,
+      "learning_rate": 1.890079389652475e-06,
+      "loss": 0.8244,
+      "step": 4191
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6838238739323992,
+      "learning_rate": 1.8864319475050296e-06,
+      "loss": 0.8745,
+      "step": 4192
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6633499558639326,
+      "learning_rate": 1.8827876615808205e-06,
+      "loss": 0.8599,
+      "step": 4193
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.713397977397952,
+      "learning_rate": 1.8791465332974967e-06,
+      "loss": 0.8399,
+      "step": 4194
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8307761550724668,
+      "learning_rate": 1.8755085640714744e-06,
+      "loss": 0.8782,
+      "step": 4195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5241826249991033,
+      "learning_rate": 1.8718737553179444e-06,
+      "loss": 0.798,
+      "step": 4196
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6318742452063342,
+      "learning_rate": 1.8682421084508672e-06,
+      "loss": 0.8697,
+      "step": 4197
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6313336860750327,
+      "learning_rate": 1.8646136248829761e-06,
+      "loss": 0.9088,
+      "step": 4198
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6912166929606829,
+      "learning_rate": 1.8609883060257695e-06,
+      "loss": 0.8369,
+      "step": 4199
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6184288990400657,
+      "learning_rate": 1.8573661532895204e-06,
+      "loss": 0.9071,
+      "step": 4200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6185845134694796,
+      "learning_rate": 1.8537471680832597e-06,
+      "loss": 0.8759,
+      "step": 4201
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6119499303248346,
+      "learning_rate": 1.8501313518148002e-06,
+      "loss": 0.8561,
+      "step": 4202
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5603778968229378,
+      "learning_rate": 1.8465187058907075e-06,
+      "loss": 0.8693,
+      "step": 4203
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.697902511983673,
+      "learning_rate": 1.8429092317163244e-06,
+      "loss": 0.801,
+      "step": 4204
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.657562913131861,
+      "learning_rate": 1.8393029306957566e-06,
+      "loss": 0.8354,
+      "step": 4205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.013967106055655,
+      "learning_rate": 1.8356998042318742e-06,
+      "loss": 0.8558,
+      "step": 4206
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6015107071363857,
+      "learning_rate": 1.8320998537263157e-06,
+      "loss": 0.8986,
+      "step": 4207
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.934416702896104,
+      "learning_rate": 1.8285030805794801e-06,
+      "loss": 0.7617,
+      "step": 4208
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.564124226047107,
+      "learning_rate": 1.8249094861905348e-06,
+      "loss": 0.8844,
+      "step": 4209
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6818962284663344,
+      "learning_rate": 1.8213190719574037e-06,
+      "loss": 0.8679,
+      "step": 4210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5437790834770664,
+      "learning_rate": 1.8177318392767807e-06,
+      "loss": 0.8557,
+      "step": 4211
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5607510120969386,
+      "learning_rate": 1.8141477895441195e-06,
+      "loss": 0.7949,
+      "step": 4212
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7986035002346878,
+      "learning_rate": 1.810566924153634e-06,
+      "loss": 0.8406,
+      "step": 4213
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7094228969604546,
+      "learning_rate": 1.8069892444983016e-06,
+      "loss": 0.8515,
+      "step": 4214
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6569380206882343,
+      "learning_rate": 1.8034147519698597e-06,
+      "loss": 0.9553,
+      "step": 4215
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6181981239208851,
+      "learning_rate": 1.799843447958809e-06,
+      "loss": 0.8635,
+      "step": 4216
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9689118436043647,
+      "learning_rate": 1.7962753338544016e-06,
+      "loss": 0.8179,
+      "step": 4217
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6179077470547958,
+      "learning_rate": 1.7927104110446536e-06,
+      "loss": 0.8416,
+      "step": 4218
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.793608052607049,
+      "learning_rate": 1.7891486809163439e-06,
+      "loss": 0.8373,
+      "step": 4219
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6094276407729822,
+      "learning_rate": 1.7855901448550028e-06,
+      "loss": 0.8824,
+      "step": 4220
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5662726427033464,
+      "learning_rate": 1.782034804244921e-06,
+      "loss": 0.7967,
+      "step": 4221
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.642746744584633,
+      "learning_rate": 1.7784826604691474e-06,
+      "loss": 0.8016,
+      "step": 4222
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7081726001244744,
+      "learning_rate": 1.7749337149094858e-06,
+      "loss": 0.876,
+      "step": 4223
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5320756544692784,
+      "learning_rate": 1.7713879689464931e-06,
+      "loss": 0.8316,
+      "step": 4224
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6997635156240871,
+      "learning_rate": 1.7678454239594844e-06,
+      "loss": 0.8101,
+      "step": 4225
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5758251194290973,
+      "learning_rate": 1.7643060813265322e-06,
+      "loss": 0.8087,
+      "step": 4226
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6557339121434707,
+      "learning_rate": 1.7607699424244583e-06,
+      "loss": 0.8373,
+      "step": 4227
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5007565633418054,
+      "learning_rate": 1.7572370086288416e-06,
+      "loss": 0.7968,
+      "step": 4228
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5041501846460212,
+      "learning_rate": 1.7537072813140121e-06,
+      "loss": 0.7657,
+      "step": 4229
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7190944904709233,
+      "learning_rate": 1.7501807618530574e-06,
+      "loss": 0.8224,
+      "step": 4230
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7002929025635432,
+      "learning_rate": 1.746657451617807e-06,
+      "loss": 0.9036,
+      "step": 4231
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5936608775633896,
+      "learning_rate": 1.7431373519788507e-06,
+      "loss": 0.8743,
+      "step": 4232
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5878949718641973,
+      "learning_rate": 1.7396204643055281e-06,
+      "loss": 0.8177,
+      "step": 4233
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7389622865395324,
+      "learning_rate": 1.7361067899659267e-06,
+      "loss": 0.8429,
+      "step": 4234
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.67044524992391,
+      "learning_rate": 1.7325963303268867e-06,
+      "loss": 0.9071,
+      "step": 4235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6704663880215562,
+      "learning_rate": 1.7290890867539956e-06,
+      "loss": 0.783,
+      "step": 4236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.681738664780346,
+      "learning_rate": 1.7255850606115932e-06,
+      "loss": 0.8906,
+      "step": 4237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7628584924491535,
+      "learning_rate": 1.7220842532627602e-06,
+      "loss": 0.9066,
+      "step": 4238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.643576717234901,
+      "learning_rate": 1.718586666069334e-06,
+      "loss": 0.8756,
+      "step": 4239
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6298883405552507,
+      "learning_rate": 1.715092300391894e-06,
+      "loss": 0.9299,
+      "step": 4240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5176902317710959,
+      "learning_rate": 1.7116011575897684e-06,
+      "loss": 0.8039,
+      "step": 4241
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5034234194976928,
+      "learning_rate": 1.7081132390210308e-06,
+      "loss": 0.8607,
+      "step": 4242
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7265518111617897,
+      "learning_rate": 1.7046285460425028e-06,
+      "loss": 0.9065,
+      "step": 4243
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.597350223641049,
+      "learning_rate": 1.7011470800097496e-06,
+      "loss": 0.8635,
+      "step": 4244
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.722001449284894,
+      "learning_rate": 1.697668842277077e-06,
+      "loss": 0.9172,
+      "step": 4245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5400651488436177,
+      "learning_rate": 1.6941938341975438e-06,
+      "loss": 0.8452,
+      "step": 4246
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5446456777200521,
+      "learning_rate": 1.6907220571229422e-06,
+      "loss": 0.8464,
+      "step": 4247
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5299818784480728,
+      "learning_rate": 1.6872535124038158e-06,
+      "loss": 0.8621,
+      "step": 4248
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7037901553639654,
+      "learning_rate": 1.683788201389448e-06,
+      "loss": 0.7951,
+      "step": 4249
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7488408534293958,
+      "learning_rate": 1.6803261254278635e-06,
+      "loss": 0.9241,
+      "step": 4250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7427601481042996,
+      "learning_rate": 1.6768672858658297e-06,
+      "loss": 0.8736,
+      "step": 4251
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.616787523927124,
+      "learning_rate": 1.6734116840488545e-06,
+      "loss": 0.8107,
+      "step": 4252
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9557225411422805,
+      "learning_rate": 1.6699593213211884e-06,
+      "loss": 0.8125,
+      "step": 4253
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.817376045804014,
+      "learning_rate": 1.6665101990258147e-06,
+      "loss": 0.8783,
+      "step": 4254
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5621629635958216,
+      "learning_rate": 1.6630643185044638e-06,
+      "loss": 0.8349,
+      "step": 4255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9237864895315417,
+      "learning_rate": 1.6596216810976019e-06,
+      "loss": 0.8922,
+      "step": 4256
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5714637354036116,
+      "learning_rate": 1.6561822881444357e-06,
+      "loss": 0.8062,
+      "step": 4257
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8834177106710497,
+      "learning_rate": 1.6527461409829071e-06,
+      "loss": 0.8185,
+      "step": 4258
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.852231216702262,
+      "learning_rate": 1.649313240949697e-06,
+      "loss": 0.8795,
+      "step": 4259
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7446735311105743,
+      "learning_rate": 1.645883589380225e-06,
+      "loss": 0.9172,
+      "step": 4260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5049348216888374,
+      "learning_rate": 1.6424571876086393e-06,
+      "loss": 0.9061,
+      "step": 4261
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.637244628956277,
+      "learning_rate": 1.6390340369678327e-06,
+      "loss": 0.8079,
+      "step": 4262
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8218546477485633,
+      "learning_rate": 1.6356141387894287e-06,
+      "loss": 0.8651,
+      "step": 4263
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6174128565949117,
+      "learning_rate": 1.6321974944037876e-06,
+      "loss": 0.8758,
+      "step": 4264
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6719545413009034,
+      "learning_rate": 1.6287841051400033e-06,
+      "loss": 0.8987,
+      "step": 4265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7107482370538083,
+      "learning_rate": 1.625373972325902e-06,
+      "loss": 0.819,
+      "step": 4266
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6698814704421532,
+      "learning_rate": 1.6219670972880487e-06,
+      "loss": 0.8778,
+      "step": 4267
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5338267158353407,
+      "learning_rate": 1.6185634813517303e-06,
+      "loss": 0.8708,
+      "step": 4268
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0189923851492153,
+      "learning_rate": 1.6151631258409762e-06,
+      "loss": 0.826,
+      "step": 4269
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.812523990980582,
+      "learning_rate": 1.6117660320785422e-06,
+      "loss": 0.8743,
+      "step": 4270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5519912613423321,
+      "learning_rate": 1.608372201385917e-06,
+      "loss": 0.8017,
+      "step": 4271
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.746217444999071,
+      "learning_rate": 1.6049816350833214e-06,
+      "loss": 0.8111,
+      "step": 4272
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6536108636235152,
+      "learning_rate": 1.6015943344897022e-06,
+      "loss": 0.7914,
+      "step": 4273
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5362544084953247,
+      "learning_rate": 1.5982103009227423e-06,
+      "loss": 0.8385,
+      "step": 4274
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6318517495242428,
+      "learning_rate": 1.5948295356988452e-06,
+      "loss": 0.886,
+      "step": 4275
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6732706154071055,
+      "learning_rate": 1.5914520401331479e-06,
+      "loss": 0.8427,
+      "step": 4276
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6460170247091113,
+      "learning_rate": 1.588077815539516e-06,
+      "loss": 0.8635,
+      "step": 4277
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6769953594720446,
+      "learning_rate": 1.5847068632305419e-06,
+      "loss": 0.8479,
+      "step": 4278
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5810239007620464,
+      "learning_rate": 1.5813391845175452e-06,
+      "loss": 0.8351,
+      "step": 4279
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5979160117258007,
+      "learning_rate": 1.5779747807105706e-06,
+      "loss": 0.8672,
+      "step": 4280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7380750747890665,
+      "learning_rate": 1.5746136531183932e-06,
+      "loss": 0.8449,
+      "step": 4281
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6343867825589604,
+      "learning_rate": 1.5712558030485047e-06,
+      "loss": 0.792,
+      "step": 4282
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.589193240066642,
+      "learning_rate": 1.5679012318071307e-06,
+      "loss": 0.8797,
+      "step": 4283
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5739106558084588,
+      "learning_rate": 1.5645499406992193e-06,
+      "loss": 0.8152,
+      "step": 4284
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7162071203051543,
+      "learning_rate": 1.5612019310284388e-06,
+      "loss": 0.9034,
+      "step": 4285
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7064188394513413,
+      "learning_rate": 1.557857204097185e-06,
+      "loss": 0.8056,
+      "step": 4286
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6452568085599142,
+      "learning_rate": 1.5545157612065752e-06,
+      "loss": 0.8841,
+      "step": 4287
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8118946525234785,
+      "learning_rate": 1.551177603656453e-06,
+      "loss": 0.8707,
+      "step": 4288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5131858384022683,
+      "learning_rate": 1.5478427327453738e-06,
+      "loss": 0.9185,
+      "step": 4289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9870720709969131,
+      "learning_rate": 1.5445111497706233e-06,
+      "loss": 0.7891,
+      "step": 4290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6552693925584094,
+      "learning_rate": 1.541182856028206e-06,
+      "loss": 0.8394,
+      "step": 4291
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6225933678967386,
+      "learning_rate": 1.5378578528128506e-06,
+      "loss": 0.8205,
+      "step": 4292
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6355555286533208,
+      "learning_rate": 1.534536141417996e-06,
+      "loss": 0.7485,
+      "step": 4293
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7195409177054761,
+      "learning_rate": 1.5312177231358093e-06,
+      "loss": 0.9338,
+      "step": 4294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6162539472701516,
+      "learning_rate": 1.5279025992571728e-06,
+      "loss": 0.8307,
+      "step": 4295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6275619531823249,
+      "learning_rate": 1.5245907710716912e-06,
+      "loss": 0.9069,
+      "step": 4296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5222816132073522,
+      "learning_rate": 1.5212822398676796e-06,
+      "loss": 0.8299,
+      "step": 4297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6616626053369064,
+      "learning_rate": 1.5179770069321764e-06,
+      "loss": 0.7872,
+      "step": 4298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6713819310344615,
+      "learning_rate": 1.5146750735509375e-06,
+      "loss": 0.8515,
+      "step": 4299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6069465214905874,
+      "learning_rate": 1.511376441008432e-06,
+      "loss": 0.8502,
+      "step": 4300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6748761109440198,
+      "learning_rate": 1.5080811105878467e-06,
+      "loss": 0.8749,
+      "step": 4301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4584266785791618,
+      "learning_rate": 1.5047890835710843e-06,
+      "loss": 0.8453,
+      "step": 4302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5634136417281823,
+      "learning_rate": 1.5015003612387602e-06,
+      "loss": 0.8932,
+      "step": 4303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9397372116477118,
+      "learning_rate": 1.4982149448702087e-06,
+      "loss": 0.7396,
+      "step": 4304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5292908742465148,
+      "learning_rate": 1.4949328357434701e-06,
+      "loss": 0.7989,
+      "step": 4305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7436376742691757,
+      "learning_rate": 1.491654035135306e-06,
+      "loss": 0.8615,
+      "step": 4306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7509806980635338,
+      "learning_rate": 1.4883785443211884e-06,
+      "loss": 0.8475,
+      "step": 4307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5971251121355534,
+      "learning_rate": 1.4851063645753016e-06,
+      "loss": 0.8518,
+      "step": 4308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7664582554868036,
+      "learning_rate": 1.4818374971705407e-06,
+      "loss": 0.8449,
+      "step": 4309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6037136406371009,
+      "learning_rate": 1.478571943378515e-06,
+      "loss": 0.8052,
+      "step": 4310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7352330136276743,
+      "learning_rate": 1.4753097044695431e-06,
+      "loss": 0.8567,
+      "step": 4311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.666619311238721,
+      "learning_rate": 1.4720507817126517e-06,
+      "loss": 0.8317,
+      "step": 4312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7201421147823996,
+      "learning_rate": 1.4687951763755804e-06,
+      "loss": 0.8236,
+      "step": 4313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9623689672594291,
+      "learning_rate": 1.4655428897247803e-06,
+      "loss": 0.8055,
+      "step": 4314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9638174968389658,
+      "learning_rate": 1.4622939230254064e-06,
+      "loss": 0.7513,
+      "step": 4315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7756562757427046,
+      "learning_rate": 1.459048277541326e-06,
+      "loss": 0.8605,
+      "step": 4316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5831767715162313,
+      "learning_rate": 1.4558059545351144e-06,
+      "loss": 0.8419,
+      "step": 4317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.84481977282202,
+      "learning_rate": 1.4525669552680533e-06,
+      "loss": 0.8677,
+      "step": 4318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.633397732564203,
+      "learning_rate": 1.4493312810001293e-06,
+      "loss": 0.9163,
+      "step": 4319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.459253751811932,
+      "learning_rate": 1.4460989329900377e-06,
+      "loss": 0.8194,
+      "step": 4320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.772858570266449,
+      "learning_rate": 1.4428699124951817e-06,
+      "loss": 0.8145,
+      "step": 4321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6945730972018598,
+      "learning_rate": 1.4396442207716676e-06,
+      "loss": 0.878,
+      "step": 4322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6094814768309065,
+      "learning_rate": 1.4364218590743084e-06,
+      "loss": 0.8182,
+      "step": 4323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6097700365602874,
+      "learning_rate": 1.433202828656619e-06,
+      "loss": 0.739,
+      "step": 4324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8476006217073462,
+      "learning_rate": 1.4299871307708235e-06,
+      "loss": 0.8643,
+      "step": 4325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.80162222996367,
+      "learning_rate": 1.4267747666678423e-06,
+      "loss": 0.7859,
+      "step": 4326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6525317284519665,
+      "learning_rate": 1.4235657375973045e-06,
+      "loss": 0.8454,
+      "step": 4327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6401423504510815,
+      "learning_rate": 1.4203600448075395e-06,
+      "loss": 0.8908,
+      "step": 4328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6815404646660566,
+      "learning_rate": 1.4171576895455819e-06,
+      "loss": 0.8572,
+      "step": 4329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6654157933004743,
+      "learning_rate": 1.4139586730571642e-06,
+      "loss": 0.9033,
+      "step": 4330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6846689876733776,
+      "learning_rate": 1.4107629965867208e-06,
+      "loss": 0.7843,
+      "step": 4331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9336103223240242,
+      "learning_rate": 1.407570661377392e-06,
+      "loss": 0.8075,
+      "step": 4332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6466407406514192,
+      "learning_rate": 1.4043816686710087e-06,
+      "loss": 0.7425,
+      "step": 4333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6084338043623005,
+      "learning_rate": 1.4011960197081086e-06,
+      "loss": 0.8308,
+      "step": 4334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9302723512143318,
+      "learning_rate": 1.3980137157279273e-06,
+      "loss": 0.7508,
+      "step": 4335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6836488185385154,
+      "learning_rate": 1.3948347579683985e-06,
+      "loss": 0.7776,
+      "step": 4336
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5487301287745663,
+      "learning_rate": 1.3916591476661557e-06,
+      "loss": 0.7153,
+      "step": 4337
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5836042112929758,
+      "learning_rate": 1.388486886056527e-06,
+      "loss": 0.9365,
+      "step": 4338
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6690369618247025,
+      "learning_rate": 1.3853179743735435e-06,
+      "loss": 0.8204,
+      "step": 4339
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5742845986707346,
+      "learning_rate": 1.3821524138499276e-06,
+      "loss": 0.8514,
+      "step": 4340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6914090333044636,
+      "learning_rate": 1.3789902057170968e-06,
+      "loss": 0.8679,
+      "step": 4341
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6508941448478838,
+      "learning_rate": 1.3758313512051702e-06,
+      "loss": 0.8897,
+      "step": 4342
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6498943116829727,
+      "learning_rate": 1.3726758515429606e-06,
+      "loss": 0.8618,
+      "step": 4343
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6665136954851711,
+      "learning_rate": 1.3695237079579738e-06,
+      "loss": 0.8604,
+      "step": 4344
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5968932620213256,
+      "learning_rate": 1.366374921676411e-06,
+      "loss": 0.8312,
+      "step": 4345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4919440899847343,
+      "learning_rate": 1.3632294939231694e-06,
+      "loss": 0.8307,
+      "step": 4346
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6158959793546552,
+      "learning_rate": 1.3600874259218388e-06,
+      "loss": 0.7715,
+      "step": 4347
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6958288799594565,
+      "learning_rate": 1.3569487188946972e-06,
+      "loss": 0.8459,
+      "step": 4348
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6418023572707168,
+      "learning_rate": 1.3538133740627212e-06,
+      "loss": 0.799,
+      "step": 4349
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6013663411591732,
+      "learning_rate": 1.3506813926455775e-06,
+      "loss": 0.8904,
+      "step": 4350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4610402569656513,
+      "learning_rate": 1.3475527758616236e-06,
+      "loss": 0.8919,
+      "step": 4351
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5307694225767354,
+      "learning_rate": 1.3444275249279103e-06,
+      "loss": 0.9144,
+      "step": 4352
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5562160595740888,
+      "learning_rate": 1.3413056410601766e-06,
+      "loss": 0.8232,
+      "step": 4353
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6471658520790176,
+      "learning_rate": 1.3381871254728563e-06,
+      "loss": 0.8316,
+      "step": 4354
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4916898215900993,
+      "learning_rate": 1.3350719793790634e-06,
+      "loss": 0.8613,
+      "step": 4355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.652688053137387,
+      "learning_rate": 1.3319602039906087e-06,
+      "loss": 0.8613,
+      "step": 4356
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5613250178247624,
+      "learning_rate": 1.3288518005179928e-06,
+      "loss": 0.7757,
+      "step": 4357
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.56171311764221,
+      "learning_rate": 1.3257467701703996e-06,
+      "loss": 0.851,
+      "step": 4358
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.564961169825869,
+      "learning_rate": 1.322645114155705e-06,
+      "loss": 0.8445,
+      "step": 4359
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0428484747804532,
+      "learning_rate": 1.3195468336804696e-06,
+      "loss": 0.7914,
+      "step": 4360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5625972092936242,
+      "learning_rate": 1.3164519299499434e-06,
+      "loss": 0.8862,
+      "step": 4361
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6610838916370576,
+      "learning_rate": 1.313360404168058e-06,
+      "loss": 0.8101,
+      "step": 4362
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6651012876838398,
+      "learning_rate": 1.310272257537435e-06,
+      "loss": 0.9458,
+      "step": 4363
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7263316061658478,
+      "learning_rate": 1.3071874912593829e-06,
+      "loss": 0.8474,
+      "step": 4364
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.449193351194833,
+      "learning_rate": 1.30410610653389e-06,
+      "loss": 0.7726,
+      "step": 4365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.627081010596026,
+      "learning_rate": 1.301028104559635e-06,
+      "loss": 0.8746,
+      "step": 4366
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6265402368128339,
+      "learning_rate": 1.297953486533976e-06,
+      "loss": 0.7558,
+      "step": 4367
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5897401362510504,
+      "learning_rate": 1.2948822536529593e-06,
+      "loss": 0.7586,
+      "step": 4368
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5936325842801757,
+      "learning_rate": 1.291814407111308e-06,
+      "loss": 0.8612,
+      "step": 4369
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.770651685807801,
+      "learning_rate": 1.2887499481024323e-06,
+      "loss": 0.8655,
+      "step": 4370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9240525614388679,
+      "learning_rate": 1.285688877818425e-06,
+      "loss": 0.7532,
+      "step": 4371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7421203566605523,
+      "learning_rate": 1.2826311974500582e-06,
+      "loss": 0.8616,
+      "step": 4372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6507216581322597,
+      "learning_rate": 1.2795769081867882e-06,
+      "loss": 0.9044,
+      "step": 4373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9064051593031216,
+      "learning_rate": 1.27652601121675e-06,
+      "loss": 0.7755,
+      "step": 4374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.647271869243701,
+      "learning_rate": 1.2734785077267609e-06,
+      "loss": 0.8238,
+      "step": 4375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.64234963084613,
+      "learning_rate": 1.2704343989023137e-06,
+      "loss": 0.8827,
+      "step": 4376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6411706578576473,
+      "learning_rate": 1.267393685927585e-06,
+      "loss": 0.8705,
+      "step": 4377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7742181758694806,
+      "learning_rate": 1.264356369985429e-06,
+      "loss": 0.9358,
+      "step": 4378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6963004827349129,
+      "learning_rate": 1.2613224522573797e-06,
+      "loss": 0.8272,
+      "step": 4379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7545281904315895,
+      "learning_rate": 1.2582919339236466e-06,
+      "loss": 0.8151,
+      "step": 4380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5640211311238734,
+      "learning_rate": 1.2552648161631198e-06,
+      "loss": 0.8131,
+      "step": 4381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7700974672369651,
+      "learning_rate": 1.2522411001533651e-06,
+      "loss": 0.8246,
+      "step": 4382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.39399693816039,
+      "learning_rate": 1.249220787070622e-06,
+      "loss": 0.7769,
+      "step": 4383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7098670769550388,
+      "learning_rate": 1.2462038780898112e-06,
+      "loss": 0.9023,
+      "step": 4384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.56991590376335,
+      "learning_rate": 1.2431903743845286e-06,
+      "loss": 0.8621,
+      "step": 4385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7459063772562895,
+      "learning_rate": 1.240180277127041e-06,
+      "loss": 0.8742,
+      "step": 4386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8222022986331625,
+      "learning_rate": 1.237173587488294e-06,
+      "loss": 0.9033,
+      "step": 4387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6636875445253645,
+      "learning_rate": 1.2341703066379073e-06,
+      "loss": 0.8592,
+      "step": 4388
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.516463002969594,
+      "learning_rate": 1.2311704357441722e-06,
+      "loss": 0.904,
+      "step": 4389
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8833292479334316,
+      "learning_rate": 1.2281739759740575e-06,
+      "loss": 0.7882,
+      "step": 4390
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7804524655990048,
+      "learning_rate": 1.2251809284932048e-06,
+      "loss": 0.8787,
+      "step": 4391
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5407877221598578,
+      "learning_rate": 1.2221912944659197e-06,
+      "loss": 0.8347,
+      "step": 4392
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8432295057836263,
+      "learning_rate": 1.219205075055191e-06,
+      "loss": 0.8447,
+      "step": 4393
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5716527692478683,
+      "learning_rate": 1.2162222714226734e-06,
+      "loss": 0.8454,
+      "step": 4394
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.699604852393047,
+      "learning_rate": 1.2132428847286948e-06,
+      "loss": 0.8674,
+      "step": 4395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5880508487438296,
+      "learning_rate": 1.2102669161322533e-06,
+      "loss": 0.7886,
+      "step": 4396
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.541863715279756,
+      "learning_rate": 1.2072943667910165e-06,
+      "loss": 0.8683,
+      "step": 4397
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9853460349538734,
+      "learning_rate": 1.2043252378613258e-06,
+      "loss": 0.8234,
+      "step": 4398
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6973160467496062,
+      "learning_rate": 1.201359530498184e-06,
+      "loss": 0.818,
+      "step": 4399
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7832473250129361,
+      "learning_rate": 1.1983972458552696e-06,
+      "loss": 0.8366,
+      "step": 4400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7544883497954733,
+      "learning_rate": 1.1954383850849293e-06,
+      "loss": 0.8754,
+      "step": 4401
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5676676844488326,
+      "learning_rate": 1.192482949338175e-06,
+      "loss": 0.7983,
+      "step": 4402
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.654178355561157,
+      "learning_rate": 1.1895309397646881e-06,
+      "loss": 0.8355,
+      "step": 4403
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5794555970325268,
+      "learning_rate": 1.1865823575128167e-06,
+      "loss": 0.8097,
+      "step": 4404
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.714684067948255,
+      "learning_rate": 1.1836372037295784e-06,
+      "loss": 0.8569,
+      "step": 4405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4778767443771448,
+      "learning_rate": 1.1806954795606496e-06,
+      "loss": 0.8064,
+      "step": 4406
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5771339720205801,
+      "learning_rate": 1.17775718615038e-06,
+      "loss": 0.9385,
+      "step": 4407
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.833409089106392,
+      "learning_rate": 1.1748223246417823e-06,
+      "loss": 0.9251,
+      "step": 4408
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.539901330366489,
+      "learning_rate": 1.1718908961765341e-06,
+      "loss": 0.8665,
+      "step": 4409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7093969684908097,
+      "learning_rate": 1.1689629018949756e-06,
+      "loss": 0.9015,
+      "step": 4410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5924414892535526,
+      "learning_rate": 1.1660383429361155e-06,
+      "loss": 0.8422,
+      "step": 4411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6341640359152323,
+      "learning_rate": 1.163117220437624e-06,
+      "loss": 0.873,
+      "step": 4412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.577217380135338,
+      "learning_rate": 1.1601995355358297e-06,
+      "loss": 0.8217,
+      "step": 4413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6578022087682756,
+      "learning_rate": 1.1572852893657317e-06,
+      "loss": 0.8499,
+      "step": 4414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8078607800621609,
+      "learning_rate": 1.154374483060987e-06,
+      "loss": 0.7952,
+      "step": 4415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7672911150902952,
+      "learning_rate": 1.151467117753915e-06,
+      "loss": 0.81,
+      "step": 4416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.551214104183218,
+      "learning_rate": 1.1485631945754994e-06,
+      "loss": 0.69,
+      "step": 4417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5654454311468475,
+      "learning_rate": 1.14566271465538e-06,
+      "loss": 0.7955,
+      "step": 4418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6343785590322206,
+      "learning_rate": 1.1427656791218622e-06,
+      "loss": 0.8704,
+      "step": 4419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6859327960281694,
+      "learning_rate": 1.1398720891019067e-06,
+      "loss": 0.7787,
+      "step": 4420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7431427496116794,
+      "learning_rate": 1.1369819457211362e-06,
+      "loss": 0.8799,
+      "step": 4421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6952005627321072,
+      "learning_rate": 1.1340952501038327e-06,
+      "loss": 0.8494,
+      "step": 4422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.741690674136835,
+      "learning_rate": 1.131212003372938e-06,
+      "loss": 0.8477,
+      "step": 4423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6966799554774559,
+      "learning_rate": 1.1283322066500501e-06,
+      "loss": 0.8783,
+      "step": 4424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5427455659887492,
+      "learning_rate": 1.1254558610554279e-06,
+      "loss": 0.91,
+      "step": 4425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7126135198408454,
+      "learning_rate": 1.1225829677079847e-06,
+      "loss": 0.7573,
+      "step": 4426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.957373016096781,
+      "learning_rate": 1.1197135277252913e-06,
+      "loss": 0.7708,
+      "step": 4427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5650832334881755,
+      "learning_rate": 1.1168475422235747e-06,
+      "loss": 0.799,
+      "step": 4428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4653603909467448,
+      "learning_rate": 1.113985012317722e-06,
+      "loss": 0.8121,
+      "step": 4429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6109537532466554,
+      "learning_rate": 1.1111259391212725e-06,
+      "loss": 0.8419,
+      "step": 4430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.572869697213479,
+      "learning_rate": 1.108270323746423e-06,
+      "loss": 0.8004,
+      "step": 4431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.61997907294903,
+      "learning_rate": 1.1054181673040198e-06,
+      "loss": 0.7694,
+      "step": 4432
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6975966158657543,
+      "learning_rate": 1.1025694709035705e-06,
+      "loss": 0.8735,
+      "step": 4433
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5240345097385706,
+      "learning_rate": 1.0997242356532335e-06,
+      "loss": 0.8002,
+      "step": 4434
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0498134701388133,
+      "learning_rate": 1.096882462659823e-06,
+      "loss": 0.8051,
+      "step": 4435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4544105167445185,
+      "learning_rate": 1.0940441530288004e-06,
+      "loss": 0.8181,
+      "step": 4436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.537898740190789,
+      "learning_rate": 1.0912093078642872e-06,
+      "loss": 0.7909,
+      "step": 4437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5252512123601627,
+      "learning_rate": 1.0883779282690522e-06,
+      "loss": 0.9167,
+      "step": 4438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7355729542900917,
+      "learning_rate": 1.0855500153445187e-06,
+      "loss": 0.7736,
+      "step": 4439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6040812858952032,
+      "learning_rate": 1.0827255701907601e-06,
+      "loss": 0.7504,
+      "step": 4440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5488295215336318,
+      "learning_rate": 1.0799045939065023e-06,
+      "loss": 0.8184,
+      "step": 4441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.734075928353889,
+      "learning_rate": 1.0770870875891225e-06,
+      "loss": 0.8345,
+      "step": 4442
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5801376343824856,
+      "learning_rate": 1.0742730523346423e-06,
+      "loss": 0.775,
+      "step": 4443
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5090404313759835,
+      "learning_rate": 1.0714624892377378e-06,
+      "loss": 0.7481,
+      "step": 4444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5551259584064296,
+      "learning_rate": 1.068655399391737e-06,
+      "loss": 0.7962,
+      "step": 4445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.056566937526707,
+      "learning_rate": 1.0658517838886095e-06,
+      "loss": 0.8027,
+      "step": 4446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9912042483277311,
+      "learning_rate": 1.0630516438189808e-06,
+      "loss": 0.7818,
+      "step": 4447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7416781406479296,
+      "learning_rate": 1.0602549802721184e-06,
+      "loss": 0.8123,
+      "step": 4448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.702295684210208,
+      "learning_rate": 1.0574617943359434e-06,
+      "loss": 0.7602,
+      "step": 4449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5154385595527293,
+      "learning_rate": 1.0546720870970174e-06,
+      "loss": 0.8576,
+      "step": 4450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5556049950136834,
+      "learning_rate": 1.0518858596405522e-06,
+      "loss": 0.8548,
+      "step": 4451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6621240261710906,
+      "learning_rate": 1.049103113050407e-06,
+      "loss": 0.9105,
+      "step": 4452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8517385751958417,
+      "learning_rate": 1.0463238484090853e-06,
+      "loss": 0.8656,
+      "step": 4453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5617830735612712,
+      "learning_rate": 1.043548066797737e-06,
+      "loss": 0.7509,
+      "step": 4454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5035927145676493,
+      "learning_rate": 1.0407757692961561e-06,
+      "loss": 0.8458,
+      "step": 4455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4740272653791686,
+      "learning_rate": 1.0380069569827845e-06,
+      "loss": 0.8104,
+      "step": 4456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4885806249156424,
+      "learning_rate": 1.0352416309347003e-06,
+      "loss": 0.8004,
+      "step": 4457
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5660833801364895,
+      "learning_rate": 1.0324797922276353e-06,
+      "loss": 0.86,
+      "step": 4458
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5597472265575503,
+      "learning_rate": 1.0297214419359568e-06,
+      "loss": 0.811,
+      "step": 4459
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6576635369410904,
+      "learning_rate": 1.0269665811326812e-06,
+      "loss": 0.821,
+      "step": 4460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5629992419217096,
+      "learning_rate": 1.0242152108894631e-06,
+      "loss": 0.795,
+      "step": 4461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5294767478046345,
+      "learning_rate": 1.0214673322766022e-06,
+      "loss": 0.8612,
+      "step": 4462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.695359131532769,
+      "learning_rate": 1.01872294636304e-06,
+      "loss": 0.8457,
+      "step": 4463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5577342461255246,
+      "learning_rate": 1.0159820542163534e-06,
+      "loss": 0.7357,
+      "step": 4464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6295529494071734,
+      "learning_rate": 1.0132446569027665e-06,
+      "loss": 0.8693,
+      "step": 4465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.651250710958029,
+      "learning_rate": 1.0105107554871442e-06,
+      "loss": 0.8292,
+      "step": 4466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4370475675735417,
+      "learning_rate": 1.0077803510329876e-06,
+      "loss": 0.7895,
+      "step": 4467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.559410229511489,
+      "learning_rate": 1.0050534446024384e-06,
+      "loss": 0.8734,
+      "step": 4468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7276994360855216,
+      "learning_rate": 1.0023300372562816e-06,
+      "loss": 0.8128,
+      "step": 4469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5849665222055656,
+      "learning_rate": 9.996101300539363e-07,
+      "loss": 0.8889,
+      "step": 4470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5441439308623104,
+      "learning_rate": 9.968937240534603e-07,
+      "loss": 0.8331,
+      "step": 4471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5098544672724046,
+      "learning_rate": 9.94180820311551e-07,
+      "loss": 0.8328,
+      "step": 4472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6883501121479696,
+      "learning_rate": 9.91471419883544e-07,
+      "loss": 0.9008,
+      "step": 4473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4924003148742895,
+      "learning_rate": 9.887655238234106e-07,
+      "loss": 0.7525,
+      "step": 4474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8874762246105932,
+      "learning_rate": 9.8606313318376e-07,
+      "loss": 0.9309,
+      "step": 4475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6036935283802969,
+      "learning_rate": 9.833642490158368e-07,
+      "loss": 0.8576,
+      "step": 4476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7975476549672167,
+      "learning_rate": 9.806688723695236e-07,
+      "loss": 0.8986,
+      "step": 4477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5988616056160807,
+      "learning_rate": 9.779770042933356e-07,
+      "loss": 0.8717,
+      "step": 4478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6547262424773457,
+      "learning_rate": 9.752886458344236e-07,
+      "loss": 0.8331,
+      "step": 4479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5094676585430962,
+      "learning_rate": 9.72603798038574e-07,
+      "loss": 0.756,
+      "step": 4480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5325925328000547,
+      "learning_rate": 9.69922461950208e-07,
+      "loss": 0.8242,
+      "step": 4481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7071431511887294,
+      "learning_rate": 9.672446386123813e-07,
+      "loss": 0.8925,
+      "step": 4482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.638219340812007,
+      "learning_rate": 9.645703290667819e-07,
+      "loss": 0.807,
+      "step": 4483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5825072685972157,
+      "learning_rate": 9.618995343537297e-07,
+      "loss": 0.8356,
+      "step": 4484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5918910375564583,
+      "learning_rate": 9.592322555121825e-07,
+      "loss": 0.8459,
+      "step": 4485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6383908833154675,
+      "learning_rate": 9.565684935797215e-07,
+      "loss": 0.7859,
+      "step": 4486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.636787530738983,
+      "learning_rate": 9.539082495925678e-07,
+      "loss": 0.7725,
+      "step": 4487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6770619075969655,
+      "learning_rate": 9.512515245855714e-07,
+      "loss": 0.8478,
+      "step": 4488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6359824398012264,
+      "learning_rate": 9.485983195922111e-07,
+      "loss": 0.8708,
+      "step": 4489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5901916558435076,
+      "learning_rate": 9.459486356446012e-07,
+      "loss": 0.901,
+      "step": 4490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7003724177239694,
+      "learning_rate": 9.433024737734808e-07,
+      "loss": 0.8992,
+      "step": 4491
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.743395071148314,
+      "learning_rate": 9.406598350082263e-07,
+      "loss": 0.8039,
+      "step": 4492
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5948808474067564,
+      "learning_rate": 9.380207203768333e-07,
+      "loss": 0.7533,
+      "step": 4493
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6626631382243278,
+      "learning_rate": 9.353851309059336e-07,
+      "loss": 0.8088,
+      "step": 4494
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6749592591271818,
+      "learning_rate": 9.327530676207885e-07,
+      "loss": 0.7742,
+      "step": 4495
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6330816835522501,
+      "learning_rate": 9.301245315452823e-07,
+      "loss": 0.8444,
+      "step": 4496
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6796177031981108,
+      "learning_rate": 9.274995237019324e-07,
+      "loss": 0.8635,
+      "step": 4497
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6737646647499338,
+      "learning_rate": 9.248780451118811e-07,
+      "loss": 0.7737,
+      "step": 4498
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.500287652498723,
+      "learning_rate": 9.222600967949002e-07,
+      "loss": 0.8205,
+      "step": 4499
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5999065998288446,
+      "learning_rate": 9.19645679769382e-07,
+      "loss": 0.8809,
+      "step": 4500
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.594906527805702,
+      "learning_rate": 9.170347950523506e-07,
+      "loss": 0.8415,
+      "step": 4501
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6839026327500048,
+      "learning_rate": 9.144274436594558e-07,
+      "loss": 0.8076,
+      "step": 4502
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5852937015977455,
+      "learning_rate": 9.118236266049707e-07,
+      "loss": 0.8393,
+      "step": 4503
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.482828251223841,
+      "learning_rate": 9.092233449017962e-07,
+      "loss": 0.8594,
+      "step": 4504
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.454387680780983,
+      "learning_rate": 9.066265995614554e-07,
+      "loss": 0.7132,
+      "step": 4505
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6223084567765689,
+      "learning_rate": 9.040333915940991e-07,
+      "loss": 0.7961,
+      "step": 4506
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6762400840034626,
+      "learning_rate": 9.014437220084948e-07,
+      "loss": 0.8379,
+      "step": 4507
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5407728498704398,
+      "learning_rate": 8.988575918120413e-07,
+      "loss": 0.7942,
+      "step": 4508
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8192484009488412,
+      "learning_rate": 8.96275002010758e-07,
+      "loss": 0.8772,
+      "step": 4509
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7400925959795057,
+      "learning_rate": 8.93695953609286e-07,
+      "loss": 0.8795,
+      "step": 4510
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6922514370220625,
+      "learning_rate": 8.911204476108892e-07,
+      "loss": 0.8421,
+      "step": 4511
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5665789156211634,
+      "learning_rate": 8.885484850174541e-07,
+      "loss": 0.7943,
+      "step": 4512
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4469490149019075,
+      "learning_rate": 8.859800668294916e-07,
+      "loss": 0.8074,
+      "step": 4513
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.753138597801452,
+      "learning_rate": 8.834151940461255e-07,
+      "loss": 0.8427,
+      "step": 4514
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.76427896737675,
+      "learning_rate": 8.808538676651079e-07,
+      "loss": 0.8607,
+      "step": 4515
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6686741522110797,
+      "learning_rate": 8.782960886828084e-07,
+      "loss": 0.8144,
+      "step": 4516
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5827402549651326,
+      "learning_rate": 8.75741858094219e-07,
+      "loss": 0.8336,
+      "step": 4517
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6691552735982413,
+      "learning_rate": 8.73191176892948e-07,
+      "loss": 0.7777,
+      "step": 4518
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.569135781232643,
+      "learning_rate": 8.706440460712251e-07,
+      "loss": 0.8517,
+      "step": 4519
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7146933616106483,
+      "learning_rate": 8.681004666199011e-07,
+      "loss": 0.8775,
+      "step": 4520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6239984857466807,
+      "learning_rate": 8.655604395284378e-07,
+      "loss": 0.8086,
+      "step": 4521
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.656663129562715,
+      "learning_rate": 8.630239657849215e-07,
+      "loss": 0.8586,
+      "step": 4522
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6751159708848744,
+      "learning_rate": 8.604910463760585e-07,
+      "loss": 0.8371,
+      "step": 4523
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5923042899442412,
+      "learning_rate": 8.579616822871628e-07,
+      "loss": 0.8247,
+      "step": 4524
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6631348630525378,
+      "learning_rate": 8.554358745021741e-07,
+      "loss": 0.8895,
+      "step": 4525
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7292854021155024,
+      "learning_rate": 8.529136240036439e-07,
+      "loss": 0.8041,
+      "step": 4526
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6301954531826386,
+      "learning_rate": 8.503949317727444e-07,
+      "loss": 0.8314,
+      "step": 4527
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8784379544296637,
+      "learning_rate": 8.478797987892595e-07,
+      "loss": 0.7901,
+      "step": 4528
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5767654640020843,
+      "learning_rate": 8.45368226031592e-07,
+      "loss": 0.8215,
+      "step": 4529
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5956433850813396,
+      "learning_rate": 8.42860214476754e-07,
+      "loss": 0.7906,
+      "step": 4530
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9037041961425434,
+      "learning_rate": 8.403557651003779e-07,
+      "loss": 0.8136,
+      "step": 4531
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7133031905826972,
+      "learning_rate": 8.378548788767083e-07,
+      "loss": 0.7756,
+      "step": 4532
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6653051307639297,
+      "learning_rate": 8.353575567786032e-07,
+      "loss": 0.8682,
+      "step": 4533
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6041819040966523,
+      "learning_rate": 8.328637997775368e-07,
+      "loss": 0.8277,
+      "step": 4534
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.788934035577091,
+      "learning_rate": 8.303736088435921e-07,
+      "loss": 0.9114,
+      "step": 4535
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5686609721094282,
+      "learning_rate": 8.278869849454718e-07,
+      "loss": 0.809,
+      "step": 4536
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4676365965258953,
+      "learning_rate": 8.2540392905048e-07,
+      "loss": 0.8122,
+      "step": 4537
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6464117299937717,
+      "learning_rate": 8.22924442124543e-07,
+      "loss": 0.8284,
+      "step": 4538
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5900447488545841,
+      "learning_rate": 8.204485251321947e-07,
+      "loss": 0.8182,
+      "step": 4539
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5356577844841082,
+      "learning_rate": 8.179761790365803e-07,
+      "loss": 0.7832,
+      "step": 4540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.692243945185495,
+      "learning_rate": 8.15507404799456e-07,
+      "loss": 0.9026,
+      "step": 4541
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6824591329207468,
+      "learning_rate": 8.130422033811892e-07,
+      "loss": 0.8577,
+      "step": 4542
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6332709059097492,
+      "learning_rate": 8.105805757407592e-07,
+      "loss": 0.8176,
+      "step": 4543
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0034749651827926,
+      "learning_rate": 8.081225228357481e-07,
+      "loss": 0.7566,
+      "step": 4544
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5996492708651493,
+      "learning_rate": 8.056680456223553e-07,
+      "loss": 0.8032,
+      "step": 4545
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9290414679374163,
+      "learning_rate": 8.03217145055385e-07,
+      "loss": 0.7635,
+      "step": 4546
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.637370705811582,
+      "learning_rate": 8.007698220882521e-07,
+      "loss": 0.8748,
+      "step": 4547
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6635441096211938,
+      "learning_rate": 7.983260776729773e-07,
+      "loss": 0.8472,
+      "step": 4548
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5998342354299384,
+      "learning_rate": 7.958859127601937e-07,
+      "loss": 0.8708,
+      "step": 4549
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9729592102274293,
+      "learning_rate": 7.934493282991373e-07,
+      "loss": 0.7931,
+      "step": 4550
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6456794482605581,
+      "learning_rate": 7.910163252376524e-07,
+      "loss": 0.7578,
+      "step": 4551
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5501096019336218,
+      "learning_rate": 7.885869045221917e-07,
+      "loss": 0.8705,
+      "step": 4552
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5009277862429256,
+      "learning_rate": 7.861610670978126e-07,
+      "loss": 0.8719,
+      "step": 4553
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6530036454212895,
+      "learning_rate": 7.837388139081803e-07,
+      "loss": 0.839,
+      "step": 4554
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.728410895192731,
+      "learning_rate": 7.813201458955644e-07,
+      "loss": 0.7931,
+      "step": 4555
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.645971164332409,
+      "learning_rate": 7.789050640008411e-07,
+      "loss": 0.8329,
+      "step": 4556
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8456320281211538,
+      "learning_rate": 7.7649356916349e-07,
+      "loss": 0.8624,
+      "step": 4557
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5518984995792304,
+      "learning_rate": 7.740856623215953e-07,
+      "loss": 0.7869,
+      "step": 4558
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.701746462823959,
+      "learning_rate": 7.716813444118476e-07,
+      "loss": 0.8131,
+      "step": 4559
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5597681704026225,
+      "learning_rate": 7.692806163695377e-07,
+      "loss": 0.8239,
+      "step": 4560
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6461590533877206,
+      "learning_rate": 7.668834791285651e-07,
+      "loss": 0.9086,
+      "step": 4561
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7135544196441648,
+      "learning_rate": 7.644899336214273e-07,
+      "loss": 0.899,
+      "step": 4562
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6928781909018462,
+      "learning_rate": 7.620999807792284e-07,
+      "loss": 0.892,
+      "step": 4563
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7778128461398606,
+      "learning_rate": 7.597136215316737e-07,
+      "loss": 0.8981,
+      "step": 4564
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5417553658687158,
+      "learning_rate": 7.573308568070681e-07,
+      "loss": 0.8443,
+      "step": 4565
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.645822549835426,
+      "learning_rate": 7.549516875323215e-07,
+      "loss": 0.7982,
+      "step": 4566
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5611574387528526,
+      "learning_rate": 7.525761146329447e-07,
+      "loss": 0.9295,
+      "step": 4567
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.611365281760663,
+      "learning_rate": 7.502041390330472e-07,
+      "loss": 0.7989,
+      "step": 4568
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4307881929378798,
+      "learning_rate": 7.478357616553433e-07,
+      "loss": 0.8531,
+      "step": 4569
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6613458669647516,
+      "learning_rate": 7.454709834211438e-07,
+      "loss": 0.8417,
+      "step": 4570
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5988643310648785,
+      "learning_rate": 7.431098052503594e-07,
+      "loss": 0.8731,
+      "step": 4571
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.520700797069022,
+      "learning_rate": 7.40752228061502e-07,
+      "loss": 0.8015,
+      "step": 4572
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6569071538578897,
+      "learning_rate": 7.383982527716848e-07,
+      "loss": 0.9232,
+      "step": 4573
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6385053303931487,
+      "learning_rate": 7.360478802966131e-07,
+      "loss": 0.7837,
+      "step": 4574
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.58320715655672,
+      "learning_rate": 7.337011115505976e-07,
+      "loss": 0.7945,
+      "step": 4575
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6038227379699823,
+      "learning_rate": 7.313579474465437e-07,
+      "loss": 0.8834,
+      "step": 4576
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.701399427841146,
+      "learning_rate": 7.290183888959557e-07,
+      "loss": 0.8813,
+      "step": 4577
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.671771680207799,
+      "learning_rate": 7.266824368089342e-07,
+      "loss": 0.7681,
+      "step": 4578
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9405244410682378,
+      "learning_rate": 7.243500920941793e-07,
+      "loss": 0.8179,
+      "step": 4579
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6971553092278484,
+      "learning_rate": 7.220213556589851e-07,
+      "loss": 0.8041,
+      "step": 4580
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6092722366159369,
+      "learning_rate": 7.196962284092423e-07,
+      "loss": 0.906,
+      "step": 4581
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5790028990958835,
+      "learning_rate": 7.173747112494389e-07,
+      "loss": 0.9017,
+      "step": 4582
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6613994467242639,
+      "learning_rate": 7.150568050826579e-07,
+      "loss": 0.7846,
+      "step": 4583
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7813916794522948,
+      "learning_rate": 7.12742510810579e-07,
+      "loss": 0.8844,
+      "step": 4584
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6133888168185018,
+      "learning_rate": 7.104318293334733e-07,
+      "loss": 0.8234,
+      "step": 4585
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6638283699302128,
+      "learning_rate": 7.081247615502107e-07,
+      "loss": 0.8494,
+      "step": 4586
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.713196915699892,
+      "learning_rate": 7.058213083582532e-07,
+      "loss": 0.7312,
+      "step": 4587
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5004791256135879,
+      "learning_rate": 7.035214706536564e-07,
+      "loss": 0.7935,
+      "step": 4588
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5682227642510727,
+      "learning_rate": 7.012252493310689e-07,
+      "loss": 0.762,
+      "step": 4589
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6640041441800169,
+      "learning_rate": 6.989326452837364e-07,
+      "loss": 0.7774,
+      "step": 4590
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7428000494600797,
+      "learning_rate": 6.96643659403492e-07,
+      "loss": 0.863,
+      "step": 4591
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7553837389304079,
+      "learning_rate": 6.943582925807646e-07,
+      "loss": 0.8168,
+      "step": 4592
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5911234884094496,
+      "learning_rate": 6.920765457045753e-07,
+      "loss": 0.7925,
+      "step": 4593
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6509026867265733,
+      "learning_rate": 6.897984196625385e-07,
+      "loss": 0.871,
+      "step": 4594
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9510394161604407,
+      "learning_rate": 6.875239153408541e-07,
+      "loss": 0.8161,
+      "step": 4595
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.536910091669346,
+      "learning_rate": 6.852530336243179e-07,
+      "loss": 0.8231,
+      "step": 4596
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6302778327684713,
+      "learning_rate": 6.829857753963154e-07,
+      "loss": 0.8091,
+      "step": 4597
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6899195310445059,
+      "learning_rate": 6.807221415388243e-07,
+      "loss": 0.8516,
+      "step": 4598
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6688393652366873,
+      "learning_rate": 6.784621329324104e-07,
+      "loss": 0.8116,
+      "step": 4599
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6968648933180437,
+      "learning_rate": 6.76205750456228e-07,
+      "loss": 0.8566,
+      "step": 4600
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.699073788028988,
+      "learning_rate": 6.739529949880263e-07,
+      "loss": 0.9198,
+      "step": 4601
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0303363034885953,
+      "learning_rate": 6.717038674041354e-07,
+      "loss": 0.8691,
+      "step": 4602
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7700745086786087,
+      "learning_rate": 6.694583685794798e-07,
+      "loss": 0.8631,
+      "step": 4603
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7156230999545747,
+      "learning_rate": 6.672164993875707e-07,
+      "loss": 0.8802,
+      "step": 4604
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7587124281022528,
+      "learning_rate": 6.649782607005095e-07,
+      "loss": 0.8197,
+      "step": 4605
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5890448851947154,
+      "learning_rate": 6.62743653388982e-07,
+      "loss": 0.7421,
+      "step": 4606
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6941583093864598,
+      "learning_rate": 6.605126783222637e-07,
+      "loss": 0.9055,
+      "step": 4607
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9345631981398188,
+      "learning_rate": 6.582853363682184e-07,
+      "loss": 0.8264,
+      "step": 4608
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5727895120331519,
+      "learning_rate": 6.560616283932897e-07,
+      "loss": 0.8729,
+      "step": 4609
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7270552928379281,
+      "learning_rate": 6.538415552625143e-07,
+      "loss": 0.839,
+      "step": 4610
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6423156022324654,
+      "learning_rate": 6.51625117839515e-07,
+      "loss": 0.9414,
+      "step": 4611
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.588099616516975,
+      "learning_rate": 6.494123169864964e-07,
+      "loss": 0.9061,
+      "step": 4612
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6908422858313596,
+      "learning_rate": 6.472031535642509e-07,
+      "loss": 0.8322,
+      "step": 4613
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6507473606262408,
+      "learning_rate": 6.449976284321547e-07,
+      "loss": 0.8885,
+      "step": 4614
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6452922413130318,
+      "learning_rate": 6.427957424481724e-07,
+      "loss": 0.7852,
+      "step": 4615
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5906868583803813,
+      "learning_rate": 6.405974964688477e-07,
+      "loss": 0.8616,
+      "step": 4616
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8525527893589988,
+      "learning_rate": 6.38402891349309e-07,
+      "loss": 0.8991,
+      "step": 4617
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.77877164245849,
+      "learning_rate": 6.36211927943271e-07,
+      "loss": 0.92,
+      "step": 4618
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.679959528587235,
+      "learning_rate": 6.34024607103032e-07,
+      "loss": 0.8461,
+      "step": 4619
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.650275475813174,
+      "learning_rate": 6.318409296794703e-07,
+      "loss": 0.7834,
+      "step": 4620
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7103844161253232,
+      "learning_rate": 6.296608965220496e-07,
+      "loss": 0.8533,
+      "step": 4621
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5560480879884606,
+      "learning_rate": 6.274845084788151e-07,
+      "loss": 0.8134,
+      "step": 4622
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6758532716694152,
+      "learning_rate": 6.253117663963948e-07,
+      "loss": 0.8064,
+      "step": 4623
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.703648351870057,
+      "learning_rate": 6.23142671119995e-07,
+      "loss": 0.8758,
+      "step": 4624
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7443635783563851,
+      "learning_rate": 6.209772234934075e-07,
+      "loss": 0.9093,
+      "step": 4625
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6349407103004254,
+      "learning_rate": 6.188154243590028e-07,
+      "loss": 0.8695,
+      "step": 4626
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7170681228786828,
+      "learning_rate": 6.166572745577326e-07,
+      "loss": 0.8604,
+      "step": 4627
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6149701805074526,
+      "learning_rate": 6.145027749291299e-07,
+      "loss": 0.7834,
+      "step": 4628
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9629179919446664,
+      "learning_rate": 6.123519263113075e-07,
+      "loss": 0.7807,
+      "step": 4629
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4310449112443988,
+      "learning_rate": 6.102047295409585e-07,
+      "loss": 0.8408,
+      "step": 4630
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.431591253406188,
+      "learning_rate": 6.080611854533514e-07,
+      "loss": 0.7416,
+      "step": 4631
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.452024940046649,
+      "learning_rate": 6.059212948823379e-07,
+      "loss": 0.7764,
+      "step": 4632
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6641309102131858,
+      "learning_rate": 6.037850586603478e-07,
+      "loss": 0.9695,
+      "step": 4633
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5963484121260585,
+      "learning_rate": 6.016524776183886e-07,
+      "loss": 0.77,
+      "step": 4634
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7710370822654729,
+      "learning_rate": 5.995235525860466e-07,
+      "loss": 0.835,
+      "step": 4635
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9238054758388379,
+      "learning_rate": 5.97398284391486e-07,
+      "loss": 0.8299,
+      "step": 4636
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.732216941168927,
+      "learning_rate": 5.952766738614468e-07,
+      "loss": 0.8778,
+      "step": 4637
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.618882221464819,
+      "learning_rate": 5.931587218212498e-07,
+      "loss": 0.7453,
+      "step": 4638
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7275497499921202,
+      "learning_rate": 5.91044429094787e-07,
+      "loss": 0.8808,
+      "step": 4639
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6947725598628667,
+      "learning_rate": 5.889337965045305e-07,
+      "loss": 0.817,
+      "step": 4640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.645115300544802,
+      "learning_rate": 5.868268248715292e-07,
+      "loss": 0.8123,
+      "step": 4641
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.552338918938507,
+      "learning_rate": 5.847235150154074e-07,
+      "loss": 0.8338,
+      "step": 4642
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6695952907682257,
+      "learning_rate": 5.826238677543628e-07,
+      "loss": 0.8792,
+      "step": 4643
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.768852515536146,
+      "learning_rate": 5.805278839051709e-07,
+      "loss": 0.8514,
+      "step": 4644
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4671139405698723,
+      "learning_rate": 5.784355642831829e-07,
+      "loss": 0.771,
+      "step": 4645
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7568780187384645,
+      "learning_rate": 5.763469097023189e-07,
+      "loss": 0.8898,
+      "step": 4646
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9703185609130268,
+      "learning_rate": 5.742619209750788e-07,
+      "loss": 0.8261,
+      "step": 4647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.552615942105755,
+      "learning_rate": 5.721805989125362e-07,
+      "loss": 0.84,
+      "step": 4648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.566663525662729,
+      "learning_rate": 5.701029443243345e-07,
+      "loss": 0.7267,
+      "step": 4649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8205065610738194,
+      "learning_rate": 5.680289580186959e-07,
+      "loss": 0.8879,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.654652335755769,
+      "learning_rate": 5.659586408024098e-07,
+      "loss": 0.8391,
+      "step": 4651
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6905002178206892,
+      "learning_rate": 5.63891993480844e-07,
+      "loss": 0.8697,
+      "step": 4652
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9546336563075242,
+      "learning_rate": 5.618290168579333e-07,
+      "loss": 0.7947,
+      "step": 4653
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6926962772920793,
+      "learning_rate": 5.597697117361877e-07,
+      "loss": 0.8929,
+      "step": 4654
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7332662858890597,
+      "learning_rate": 5.57714078916689e-07,
+      "loss": 0.9356,
+      "step": 4655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8157951920257163,
+      "learning_rate": 5.556621191990907e-07,
+      "loss": 0.9756,
+      "step": 4656
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7425133604802003,
+      "learning_rate": 5.536138333816144e-07,
+      "loss": 0.7792,
+      "step": 4657
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5621644307282403,
+      "learning_rate": 5.515692222610558e-07,
+      "loss": 0.7705,
+      "step": 4658
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6356106516646511,
+      "learning_rate": 5.495282866327812e-07,
+      "loss": 0.8203,
+      "step": 4659
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7357420054931105,
+      "learning_rate": 5.474910272907242e-07,
+      "loss": 0.8684,
+      "step": 4660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.644574615849858,
+      "learning_rate": 5.454574450273898e-07,
+      "loss": 0.8575,
+      "step": 4661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5828708989675613,
+      "learning_rate": 5.434275406338552e-07,
+      "loss": 0.8204,
+      "step": 4662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4186623546018264,
+      "learning_rate": 5.414013148997621e-07,
+      "loss": 0.7704,
+      "step": 4663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6833154380110336,
+      "learning_rate": 5.393787686133234e-07,
+      "loss": 0.8795,
+      "step": 4664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0095255383815676,
+      "learning_rate": 5.37359902561323e-07,
+      "loss": 0.86,
+      "step": 4665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5918336391342192,
+      "learning_rate": 5.353447175291104e-07,
+      "loss": 0.8537,
+      "step": 4666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6224700248960138,
+      "learning_rate": 5.333332143006043e-07,
+      "loss": 0.8482,
+      "step": 4667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6704298241763658,
+      "learning_rate": 5.313253936582885e-07,
+      "loss": 0.7602,
+      "step": 4668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9501917910467306,
+      "learning_rate": 5.293212563832173e-07,
+      "loss": 0.7765,
+      "step": 4669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.527283488559378,
+      "learning_rate": 5.273208032550126e-07,
+      "loss": 0.7678,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6792519676900624,
+      "learning_rate": 5.253240350518607e-07,
+      "loss": 0.7365,
+      "step": 4671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.64356865145821,
+      "learning_rate": 5.233309525505159e-07,
+      "loss": 0.8504,
+      "step": 4672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4855516463960838,
+      "learning_rate": 5.213415565262981e-07,
+      "loss": 0.7894,
+      "step": 4673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.623938335284566,
+      "learning_rate": 5.193558477530958e-07,
+      "loss": 0.8903,
+      "step": 4674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6525083687168214,
+      "learning_rate": 5.173738270033568e-07,
+      "loss": 0.7674,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.471112511849222,
+      "learning_rate": 5.153954950481e-07,
+      "loss": 0.7344,
+      "step": 4676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6775328255178097,
+      "learning_rate": 5.134208526569084e-07,
+      "loss": 0.9133,
+      "step": 4677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.687127162343473,
+      "learning_rate": 5.114499005979279e-07,
+      "loss": 0.8031,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.749454069288342,
+      "learning_rate": 5.094826396378694e-07,
+      "loss": 0.7052,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7090967580697272,
+      "learning_rate": 5.075190705420097e-07,
+      "loss": 0.8282,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6219991440125505,
+      "learning_rate": 5.0555919407419e-07,
+      "loss": 0.7687,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.712762287461227,
+      "learning_rate": 5.036030109968082e-07,
+      "loss": 0.9188,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6465939602965196,
+      "learning_rate": 5.016505220708334e-07,
+      "loss": 0.91,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6876805121272327,
+      "learning_rate": 4.997017280557936e-07,
+      "loss": 0.8541,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5718236529911196,
+      "learning_rate": 4.977566297097824e-07,
+      "loss": 0.8048,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7317068845916241,
+      "learning_rate": 4.958152277894523e-07,
+      "loss": 0.8608,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5823403939243572,
+      "learning_rate": 4.938775230500192e-07,
+      "loss": 0.8097,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7201556233252429,
+      "learning_rate": 4.919435162452635e-07,
+      "loss": 0.8875,
+      "step": 4688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5195980744396738,
+      "learning_rate": 4.900132081275222e-07,
+      "loss": 0.7755,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6020899057466513,
+      "learning_rate": 4.880865994476958e-07,
+      "loss": 0.8562,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.806839673319644,
+      "learning_rate": 4.861636909552469e-07,
+      "loss": 0.7896,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7078013031874621,
+      "learning_rate": 4.842444833981985e-07,
+      "loss": 0.8869,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7614360589223628,
+      "learning_rate": 4.82328977523131e-07,
+      "loss": 0.8127,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7649355206668866,
+      "learning_rate": 4.804171740751873e-07,
+      "loss": 0.9398,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7582056140597686,
+      "learning_rate": 4.785090737980725e-07,
+      "loss": 0.9227,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5040542605572478,
+      "learning_rate": 4.766046774340438e-07,
+      "loss": 0.8656,
+      "step": 4696
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6405321539227562,
+      "learning_rate": 4.747039857239244e-07,
+      "loss": 0.797,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9701252958307068,
+      "learning_rate": 4.7280699940709275e-07,
+      "loss": 0.8364,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.632269414247232,
+      "learning_rate": 4.7091371922148966e-07,
+      "loss": 0.8506,
+      "step": 4699
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8161203221894053,
+      "learning_rate": 4.6902414590360935e-07,
+      "loss": 0.9253,
+      "step": 4700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.591192668567659,
+      "learning_rate": 4.6713828018850606e-07,
+      "loss": 0.8823,
+      "step": 4701
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.753644938899693,
+      "learning_rate": 4.6525612280979514e-07,
+      "loss": 0.8554,
+      "step": 4702
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7057079269067104,
+      "learning_rate": 4.6337767449964323e-07,
+      "loss": 0.8982,
+      "step": 4703
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8081618193792555,
+      "learning_rate": 4.6150293598877793e-07,
+      "loss": 0.8193,
+      "step": 4704
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7086633815503671,
+      "learning_rate": 4.596319080064837e-07,
+      "loss": 0.8331,
+      "step": 4705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6974864863612644,
+      "learning_rate": 4.577645912806006e-07,
+      "loss": 0.7984,
+      "step": 4706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.643569617189755,
+      "learning_rate": 4.559009865375241e-07,
+      "loss": 0.8059,
+      "step": 4707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5728461903585766,
+      "learning_rate": 4.5404109450221e-07,
+      "loss": 0.7617,
+      "step": 4708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5168007933423657,
+      "learning_rate": 4.521849158981628e-07,
+      "loss": 0.7905,
+      "step": 4709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6654829931616189,
+      "learning_rate": 4.503324514474483e-07,
+      "loss": 0.8554,
+      "step": 4710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.601183658916333,
+      "learning_rate": 4.484837018706867e-07,
+      "loss": 0.8237,
+      "step": 4711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7492383350646525,
+      "learning_rate": 4.4663866788704823e-07,
+      "loss": 0.9066,
+      "step": 4712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6652859699269564,
+      "learning_rate": 4.447973502142644e-07,
+      "loss": 0.8496,
+      "step": 4713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5176071936844036,
+      "learning_rate": 4.4295974956861664e-07,
+      "loss": 0.9279,
+      "step": 4714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7104988370209713,
+      "learning_rate": 4.4112586666494317e-07,
+      "loss": 0.8779,
+      "step": 4715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.466684809772613,
+      "learning_rate": 4.3929570221663444e-07,
+      "loss": 0.8849,
+      "step": 4716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5064332377635345,
+      "learning_rate": 4.3746925693563314e-07,
+      "loss": 0.7313,
+      "step": 4717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6662312130875077,
+      "learning_rate": 4.3564653153243875e-07,
+      "loss": 0.8869,
+      "step": 4718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.669135117581719,
+      "learning_rate": 4.3382752671610075e-07,
+      "loss": 0.8753,
+      "step": 4719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5739945493482388,
+      "learning_rate": 4.3201224319422084e-07,
+      "loss": 0.8615,
+      "step": 4720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6975091587403395,
+      "learning_rate": 4.3020068167295646e-07,
+      "loss": 0.8725,
+      "step": 4721
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.559228645302732,
+      "learning_rate": 4.283928428570139e-07,
+      "loss": 0.8113,
+      "step": 4722
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4789161304081258,
+      "learning_rate": 4.2658872744965273e-07,
+      "loss": 0.8488,
+      "step": 4723
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.762455128063593,
+      "learning_rate": 4.2478833615268386e-07,
+      "loss": 0.8784,
+      "step": 4724
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6819388276178813,
+      "learning_rate": 4.2299166966647154e-07,
+      "loss": 0.9057,
+      "step": 4725
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5074004293968402,
+      "learning_rate": 4.211987286899255e-07,
+      "loss": 0.7763,
+      "step": 4726
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6874112613914176,
+      "learning_rate": 4.194095139205123e-07,
+      "loss": 0.8848,
+      "step": 4727
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4389099026076047,
+      "learning_rate": 4.1762402605424526e-07,
+      "loss": 0.8292,
+      "step": 4728
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7523147070895562,
+      "learning_rate": 4.1584226578568977e-07,
+      "loss": 0.8041,
+      "step": 4729
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5440277838586294,
+      "learning_rate": 4.1406423380796037e-07,
+      "loss": 0.842,
+      "step": 4730
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.683555768036824,
+      "learning_rate": 4.122899308127215e-07,
+      "loss": 0.7741,
+      "step": 4731
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5537148224409472,
+      "learning_rate": 4.105193574901878e-07,
+      "loss": 0.8457,
+      "step": 4732
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5393656799619777,
+      "learning_rate": 4.087525145291205e-07,
+      "loss": 0.8413,
+      "step": 4733
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5918678095739118,
+      "learning_rate": 4.0698940261683197e-07,
+      "loss": 0.807,
+      "step": 4734
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7728363838958756,
+      "learning_rate": 4.052300224391825e-07,
+      "loss": 0.7864,
+      "step": 4735
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7482045193198998,
+      "learning_rate": 4.0347437468058026e-07,
+      "loss": 0.877,
+      "step": 4736
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7525840635202183,
+      "learning_rate": 4.017224600239833e-07,
+      "loss": 0.8769,
+      "step": 4737
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5071276409577965,
+      "learning_rate": 3.999742791508965e-07,
+      "loss": 0.8583,
+      "step": 4738
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5494750382317546,
+      "learning_rate": 3.9822983274137137e-07,
+      "loss": 0.8042,
+      "step": 4739
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7760013173719795,
+      "learning_rate": 3.964891214740063e-07,
+      "loss": 0.7957,
+      "step": 4740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.616107440677217,
+      "learning_rate": 3.9475214602594844e-07,
+      "loss": 0.9369,
+      "step": 4741
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6297856162850126,
+      "learning_rate": 3.930189070728907e-07,
+      "loss": 0.8166,
+      "step": 4742
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5261404817116855,
+      "learning_rate": 3.9128940528907256e-07,
+      "loss": 0.882,
+      "step": 4743
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6171829450573758,
+      "learning_rate": 3.895636413472803e-07,
+      "loss": 0.8086,
+      "step": 4744
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.65460452013808,
+      "learning_rate": 3.878416159188458e-07,
+      "loss": 0.9052,
+      "step": 4745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.634016027582634,
+      "learning_rate": 3.8612332967364776e-07,
+      "loss": 0.8796,
+      "step": 4746
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4705543286456317,
+      "learning_rate": 3.844087832801069e-07,
+      "loss": 0.8437,
+      "step": 4747
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7151010171281107,
+      "learning_rate": 3.826979774051909e-07,
+      "loss": 0.8337,
+      "step": 4748
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5843930520104088,
+      "learning_rate": 3.809909127144151e-07,
+      "loss": 0.8299,
+      "step": 4749
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.526416336262696,
+      "learning_rate": 3.7928758987183624e-07,
+      "loss": 0.7759,
+      "step": 4750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6428339947689987,
+      "learning_rate": 3.7758800954005636e-07,
+      "loss": 0.8362,
+      "step": 4751
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6391771092396883,
+      "learning_rate": 3.7589217238022113e-07,
+      "loss": 0.8826,
+      "step": 4752
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7667209649913396,
+      "learning_rate": 3.7420007905202283e-07,
+      "loss": 0.8631,
+      "step": 4753
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5839302857640825,
+      "learning_rate": 3.7251173021369156e-07,
+      "loss": 0.8494,
+      "step": 4754
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5713058747975237,
+      "learning_rate": 3.708271265220087e-07,
+      "loss": 0.7535,
+      "step": 4755
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4581919015342857,
+      "learning_rate": 3.6914626863229e-07,
+      "loss": 0.7839,
+      "step": 4756
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7593764793503315,
+      "learning_rate": 3.674691571984013e-07,
+      "loss": 0.8261,
+      "step": 4757
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0229678649372016,
+      "learning_rate": 3.657957928727474e-07,
+      "loss": 0.8256,
+      "step": 4758
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9535118258403272,
+      "learning_rate": 3.6412617630627755e-07,
+      "loss": 0.7838,
+      "step": 4759
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7347847279444861,
+      "learning_rate": 3.624603081484812e-07,
+      "loss": 0.8139,
+      "step": 4760
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.622160056422886,
+      "learning_rate": 3.6079818904738884e-07,
+      "loss": 0.7831,
+      "step": 4761
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9716276915514496,
+      "learning_rate": 3.5913981964957766e-07,
+      "loss": 0.8886,
+      "step": 4762
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6441177659837407,
+      "learning_rate": 3.5748520060015944e-07,
+      "loss": 0.8599,
+      "step": 4763
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6551197231737023,
+      "learning_rate": 3.5583433254279155e-07,
+      "loss": 0.8248,
+      "step": 4764
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9760164605599994,
+      "learning_rate": 3.541872161196691e-07,
+      "loss": 0.7797,
+      "step": 4765
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4504786189252332,
+      "learning_rate": 3.5254385197153297e-07,
+      "loss": 0.8549,
+      "step": 4766
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6089778666465842,
+      "learning_rate": 3.5090424073765725e-07,
+      "loss": 0.9239,
+      "step": 4767
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6161196157246704,
+      "learning_rate": 3.4926838305586295e-07,
+      "loss": 0.9207,
+      "step": 4768
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6606225882806407,
+      "learning_rate": 3.4763627956250654e-07,
+      "loss": 0.7786,
+      "step": 4769
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9505629813721634,
+      "learning_rate": 3.4600793089248465e-07,
+      "loss": 0.8017,
+      "step": 4770
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8036875245945276,
+      "learning_rate": 3.443833376792349e-07,
+      "loss": 0.8881,
+      "step": 4771
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.694822329392725,
+      "learning_rate": 3.4276250055473304e-07,
+      "loss": 0.8805,
+      "step": 4772
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5982949061577643,
+      "learning_rate": 3.411454201494935e-07,
+      "loss": 0.8112,
+      "step": 4773
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.575896533752538,
+      "learning_rate": 3.3953209709256975e-07,
+      "loss": 0.7669,
+      "step": 4774
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7289475032328157,
+      "learning_rate": 3.3792253201155313e-07,
+      "loss": 0.8515,
+      "step": 4775
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6772017195281679,
+      "learning_rate": 3.36316725532575e-07,
+      "loss": 0.9792,
+      "step": 4776
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6483114346919296,
+      "learning_rate": 3.3471467828030013e-07,
+      "loss": 0.8247,
+      "step": 4777
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7981315587309108,
+      "learning_rate": 3.3311639087793556e-07,
+      "loss": 0.9334,
+      "step": 4778
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.798395276634243,
+      "learning_rate": 3.3152186394722506e-07,
+      "loss": 0.8194,
+      "step": 4779
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6884601725851665,
+      "learning_rate": 3.299310981084469e-07,
+      "loss": 0.8869,
+      "step": 4780
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6532311253737053,
+      "learning_rate": 3.283440939804172e-07,
+      "loss": 0.7801,
+      "step": 4781
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4639294831852823,
+      "learning_rate": 3.2676085218049215e-07,
+      "loss": 0.8718,
+      "step": 4782
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.688433245042595,
+      "learning_rate": 3.251813733245601e-07,
+      "loss": 0.7258,
+      "step": 4783
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5282375927340333,
+      "learning_rate": 3.2360565802704634e-07,
+      "loss": 0.7776,
+      "step": 4784
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4939423179271425,
+      "learning_rate": 3.2203370690091385e-07,
+      "loss": 0.8289,
+      "step": 4785
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.686477793286507,
+      "learning_rate": 3.2046552055766014e-07,
+      "loss": 0.8194,
+      "step": 4786
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6118000752000259,
+      "learning_rate": 3.189010996073183e-07,
+      "loss": 0.8074,
+      "step": 4787
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7914929045154324,
+      "learning_rate": 3.17340444658456e-07,
+      "loss": 0.846,
+      "step": 4788
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6823572069118011,
+      "learning_rate": 3.157835563181788e-07,
+      "loss": 0.8335,
+      "step": 4789
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6606169497019847,
+      "learning_rate": 3.1423043519212546e-07,
+      "loss": 0.7818,
+      "step": 4790
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5313722890918302,
+      "learning_rate": 3.12681081884465e-07,
+      "loss": 0.8774,
+      "step": 4791
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.729672334609458,
+      "learning_rate": 3.1113549699790745e-07,
+      "loss": 0.7877,
+      "step": 4792
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7275981350830443,
+      "learning_rate": 3.0959368113369305e-07,
+      "loss": 0.8518,
+      "step": 4793
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5431849570033087,
+      "learning_rate": 3.080556348915964e-07,
+      "loss": 0.8513,
+      "step": 4794
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5549587703130405,
+      "learning_rate": 3.0652135886992674e-07,
+      "loss": 0.845,
+      "step": 4795
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6150632070900868,
+      "learning_rate": 3.049908536655266e-07,
+      "loss": 0.8748,
+      "step": 4796
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7712407035313649,
+      "learning_rate": 3.0346411987377087e-07,
+      "loss": 0.8705,
+      "step": 4797
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7296781843055429,
+      "learning_rate": 3.019411580885656e-07,
+      "loss": 0.8107,
+      "step": 4798
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7321411729256602,
+      "learning_rate": 3.0042196890235244e-07,
+      "loss": 0.8312,
+      "step": 4799
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6667705278133578,
+      "learning_rate": 2.9890655290610524e-07,
+      "loss": 0.8439,
+      "step": 4800
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6183317675327349,
+      "learning_rate": 2.9739491068932924e-07,
+      "loss": 0.8938,
+      "step": 4801
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6711648840775002,
+      "learning_rate": 2.9588704284006176e-07,
+      "loss": 0.9273,
+      "step": 4802
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5800037148285875,
+      "learning_rate": 2.9438294994487025e-07,
+      "loss": 0.7868,
+      "step": 4803
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.674519887401665,
+      "learning_rate": 2.9288263258885564e-07,
+      "loss": 0.8108,
+      "step": 4804
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.685941013734219,
+      "learning_rate": 2.913860913556521e-07,
+      "loss": 0.8894,
+      "step": 4805
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6992304001777496,
+      "learning_rate": 2.8989332682741953e-07,
+      "loss": 0.9151,
+      "step": 4806
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5681189714737205,
+      "learning_rate": 2.8840433958485346e-07,
+      "loss": 0.828,
+      "step": 4807
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.73459313527006,
+      "learning_rate": 2.869191302071772e-07,
+      "loss": 0.8636,
+      "step": 4808
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5708889933524866,
+      "learning_rate": 2.8543769927214635e-07,
+      "loss": 0.8016,
+      "step": 4809
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6556950571345788,
+      "learning_rate": 2.8396004735604556e-07,
+      "loss": 0.8362,
+      "step": 4810
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5939396280990095,
+      "learning_rate": 2.8248617503368934e-07,
+      "loss": 0.8214,
+      "step": 4811
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6445737537214664,
+      "learning_rate": 2.8101608287842255e-07,
+      "loss": 0.8101,
+      "step": 4812
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.530900385278659,
+      "learning_rate": 2.795497714621198e-07,
+      "loss": 0.8873,
+      "step": 4813
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6440147626071886,
+      "learning_rate": 2.7808724135518275e-07,
+      "loss": 0.8459,
+      "step": 4814
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5967301329211467,
+      "learning_rate": 2.76628493126545e-07,
+      "loss": 0.8309,
+      "step": 4815
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9060266887554025,
+      "learning_rate": 2.751735273436673e-07,
+      "loss": 0.7643,
+      "step": 4816
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7546073748438287,
+      "learning_rate": 2.737223445725401e-07,
+      "loss": 0.8663,
+      "step": 4817
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6244540188287706,
+      "learning_rate": 2.7227494537768075e-07,
+      "loss": 0.765,
+      "step": 4818
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.760202352517511,
+      "learning_rate": 2.708313303221377e-07,
+      "loss": 0.9034,
+      "step": 4819
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7666061320096818,
+      "learning_rate": 2.693914999674818e-07,
+      "loss": 0.8755,
+      "step": 4820
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5613714635005946,
+      "learning_rate": 2.6795545487381724e-07,
+      "loss": 0.7883,
+      "step": 4821
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5200286900067441,
+      "learning_rate": 2.665231955997738e-07,
+      "loss": 0.8187,
+      "step": 4822
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.588152135460304,
+      "learning_rate": 2.650947227025069e-07,
+      "loss": 0.814,
+      "step": 4823
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5266137304631693,
+      "learning_rate": 2.6367003673770207e-07,
+      "loss": 0.7984,
+      "step": 4824
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5381375133627766,
+      "learning_rate": 2.6224913825956933e-07,
+      "loss": 0.868,
+      "step": 4825
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.649187576200236,
+      "learning_rate": 2.608320278208465e-07,
+      "loss": 0.8201,
+      "step": 4826
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8076495675440554,
+      "learning_rate": 2.5941870597279705e-07,
+      "loss": 0.8381,
+      "step": 4827
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7049459231631452,
+      "learning_rate": 2.5800917326521013e-07,
+      "loss": 0.8039,
+      "step": 4828
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9538485271628983,
+      "learning_rate": 2.566034302464027e-07,
+      "loss": 0.7581,
+      "step": 4829
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6724880354656986,
+      "learning_rate": 2.552014774632172e-07,
+      "loss": 0.8479,
+      "step": 4830
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8133591830513744,
+      "learning_rate": 2.538033154610209e-07,
+      "loss": 0.8705,
+      "step": 4831
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6579651660565642,
+      "learning_rate": 2.524089447837064e-07,
+      "loss": 0.8498,
+      "step": 4832
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6548338743351285,
+      "learning_rate": 2.51018365973692e-07,
+      "loss": 0.8748,
+      "step": 4833
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.525397157490326,
+      "learning_rate": 2.496315795719195e-07,
+      "loss": 0.8361,
+      "step": 4834
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5920176646750865,
+      "learning_rate": 2.482485861178563e-07,
+      "loss": 0.8105,
+      "step": 4835
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6429465465349558,
+      "learning_rate": 2.4686938614949643e-07,
+      "loss": 0.8905,
+      "step": 4836
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7832098610576956,
+      "learning_rate": 2.4549398020335625e-07,
+      "loss": 0.8103,
+      "step": 4837
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5345252846573398,
+      "learning_rate": 2.441223688144745e-07,
+      "loss": 0.8195,
+      "step": 4838
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5290031237515256,
+      "learning_rate": 2.4275455251641653e-07,
+      "loss": 0.8786,
+      "step": 4839
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.58164443431506,
+      "learning_rate": 2.4139053184127237e-07,
+      "loss": 0.827,
+      "step": 4840
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.610597724461763,
+      "learning_rate": 2.400303073196508e-07,
+      "loss": 0.8535,
+      "step": 4841
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3834549233361646,
+      "learning_rate": 2.3867387948068865e-07,
+      "loss": 0.799,
+      "step": 4842
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8398115746259007,
+      "learning_rate": 2.3732124885204266e-07,
+      "loss": 0.8797,
+      "step": 4843
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5963950703347418,
+      "learning_rate": 2.3597241595989417e-07,
+      "loss": 0.7603,
+      "step": 4844
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5744137147113135,
+      "learning_rate": 2.346273813289468e-07,
+      "loss": 0.8245,
+      "step": 4845
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4577636380990016,
+      "learning_rate": 2.3328614548242646e-07,
+      "loss": 0.8599,
+      "step": 4846
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6929866898934611,
+      "learning_rate": 2.3194870894208244e-07,
+      "loss": 0.8307,
+      "step": 4847
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5446532896341592,
+      "learning_rate": 2.3061507222818303e-07,
+      "loss": 0.8035,
+      "step": 4848
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6311185078352026,
+      "learning_rate": 2.2928523585952323e-07,
+      "loss": 0.8205,
+      "step": 4849
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5943013003594422,
+      "learning_rate": 2.2795920035341258e-07,
+      "loss": 0.9206,
+      "step": 4850
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6240632118828648,
+      "learning_rate": 2.2663696622568955e-07,
+      "loss": 0.877,
+      "step": 4851
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4939145654179917,
+      "learning_rate": 2.2531853399070936e-07,
+      "loss": 0.8568,
+      "step": 4852
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5981329832684763,
+      "learning_rate": 2.2400390416134953e-07,
+      "loss": 0.8412,
+      "step": 4853
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.570133236687216,
+      "learning_rate": 2.226930772490088e-07,
+      "loss": 0.845,
+      "step": 4854
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.743169482849413,
+      "learning_rate": 2.213860537636059e-07,
+      "loss": 0.9209,
+      "step": 4855
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6958034405374705,
+      "learning_rate": 2.200828342135819e-07,
+      "loss": 0.8887,
+      "step": 4856
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5922805104822988,
+      "learning_rate": 2.187834191058935e-07,
+      "loss": 0.8502,
+      "step": 4857
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.518497807992221,
+      "learning_rate": 2.1748780894602194e-07,
+      "loss": 0.8574,
+      "step": 4858
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.686973118522028,
+      "learning_rate": 2.1619600423796628e-07,
+      "loss": 0.8445,
+      "step": 4859
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6772664760048672,
+      "learning_rate": 2.1490800548424672e-07,
+      "loss": 0.8219,
+      "step": 4860
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5617253901249581,
+      "learning_rate": 2.136238131859014e-07,
+      "loss": 0.8314,
+      "step": 4861
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6075469863562961,
+      "learning_rate": 2.123434278424885e-07,
+      "loss": 0.8681,
+      "step": 4862
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.789235541183758,
+      "learning_rate": 2.1106684995208626e-07,
+      "loss": 0.8466,
+      "step": 4863
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6447974400912926,
+      "learning_rate": 2.0979408001128743e-07,
+      "loss": 0.8522,
+      "step": 4864
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.765136819247867,
+      "learning_rate": 2.085251185152093e-07,
+      "loss": 0.8604,
+      "step": 4865
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6266851116541106,
+      "learning_rate": 2.0725996595748366e-07,
+      "loss": 0.8664,
+      "step": 4866
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.67819672473352,
+      "learning_rate": 2.059986228302624e-07,
+      "loss": 0.7761,
+      "step": 4867
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.65969174055999,
+      "learning_rate": 2.0474108962421524e-07,
+      "loss": 0.8423,
+      "step": 4868
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5079462034803506,
+      "learning_rate": 2.0348736682852864e-07,
+      "loss": 0.7754,
+      "step": 4869
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6052162598092945,
+      "learning_rate": 2.022374549309103e-07,
+      "loss": 0.8417,
+      "step": 4870
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7703126547521626,
+      "learning_rate": 2.00991354417579e-07,
+      "loss": 0.9103,
+      "step": 4871
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7526946610459049,
+      "learning_rate": 1.9974906577327813e-07,
+      "loss": 0.8527,
+      "step": 4872
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9231864358048003,
+      "learning_rate": 1.9851058948126333e-07,
+      "loss": 0.8165,
+      "step": 4873
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6876394188237012,
+      "learning_rate": 1.9727592602330926e-07,
+      "loss": 0.8579,
+      "step": 4874
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.546043471808283,
+      "learning_rate": 1.9604507587970612e-07,
+      "loss": 0.7516,
+      "step": 4875
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8330467033982552,
+      "learning_rate": 1.9481803952926314e-07,
+      "loss": 0.8899,
+      "step": 4876
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6924749568421145,
+      "learning_rate": 1.9359481744930297e-07,
+      "loss": 0.831,
+      "step": 4877
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7045527025533789,
+      "learning_rate": 1.9237541011566606e-07,
+      "loss": 0.8909,
+      "step": 4878
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6507040101580346,
+      "learning_rate": 1.9115981800270855e-07,
+      "loss": 0.8436,
+      "step": 4879
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.510467164138072,
+      "learning_rate": 1.8994804158330327e-07,
+      "loss": 0.7436,
+      "step": 4880
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4970798059476809,
+      "learning_rate": 1.8874008132883646e-07,
+      "loss": 0.8396,
+      "step": 4881
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7338623941791689,
+      "learning_rate": 1.875359377092134e-07,
+      "loss": 0.8812,
+      "step": 4882
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6861050115544627,
+      "learning_rate": 1.863356111928516e-07,
+      "loss": 0.8854,
+      "step": 4883
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.526204807597603,
+      "learning_rate": 1.851391022466853e-07,
+      "loss": 0.858,
+      "step": 4884
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6976895766365234,
+      "learning_rate": 1.839464113361611e-07,
+      "loss": 0.7955,
+      "step": 4885
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6559912626039093,
+      "learning_rate": 1.827575389252456e-07,
+      "loss": 0.8226,
+      "step": 4886
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.622119688685812,
+      "learning_rate": 1.815724854764145e-07,
+      "loss": 0.8671,
+      "step": 4887
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5437991175471397,
+      "learning_rate": 1.8039125145066115e-07,
+      "loss": 0.8028,
+      "step": 4888
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.723305371104265,
+      "learning_rate": 1.7921383730749143e-07,
+      "loss": 0.8234,
+      "step": 4889
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6710362217257508,
+      "learning_rate": 1.7804024350492778e-07,
+      "loss": 0.8893,
+      "step": 4890
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6106969878587862,
+      "learning_rate": 1.768704704995028e-07,
+      "loss": 0.8756,
+      "step": 4891
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6065165768298206,
+      "learning_rate": 1.7570451874626583e-07,
+      "loss": 0.7772,
+      "step": 4892
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.629084990697285,
+      "learning_rate": 1.745423886987785e-07,
+      "loss": 0.8493,
+      "step": 4893
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7264309930868569,
+      "learning_rate": 1.7338408080911473e-07,
+      "loss": 0.8663,
+      "step": 4894
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7192802669576013,
+      "learning_rate": 1.7222959552786412e-07,
+      "loss": 0.8826,
+      "step": 4895
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5667756101128578,
+      "learning_rate": 1.7107893330412738e-07,
+      "loss": 0.7617,
+      "step": 4896
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6970738721179803,
+      "learning_rate": 1.6993209458551763e-07,
+      "loss": 0.9408,
+      "step": 4897
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5125985162571438,
+      "learning_rate": 1.687890798181635e-07,
+      "loss": 0.8234,
+      "step": 4898
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5611681895794751,
+      "learning_rate": 1.6764988944670158e-07,
+      "loss": 0.8612,
+      "step": 4899
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.725702892372231,
+      "learning_rate": 1.6651452391428736e-07,
+      "loss": 0.9418,
+      "step": 4900
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6391013469838338,
+      "learning_rate": 1.6538298366257975e-07,
+      "loss": 0.7918,
+      "step": 4901
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.661758037273611,
+      "learning_rate": 1.6425526913175672e-07,
+      "loss": 0.8399,
+      "step": 4902
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5954579453399507,
+      "learning_rate": 1.6313138076050505e-07,
+      "loss": 0.8521,
+      "step": 4903
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.602569704425456,
+      "learning_rate": 1.6201131898602284e-07,
+      "loss": 0.8762,
+      "step": 4904
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6726770139591156,
+      "learning_rate": 1.6089508424402156e-07,
+      "loss": 0.8862,
+      "step": 4905
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.720165892486497,
+      "learning_rate": 1.5978267696872274e-07,
+      "loss": 0.8524,
+      "step": 4906
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5200991395166976,
+      "learning_rate": 1.5867409759285802e-07,
+      "loss": 0.8561,
+      "step": 4907
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6590015797171926,
+      "learning_rate": 1.5756934654767243e-07,
+      "loss": 0.8831,
+      "step": 4908
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.556760071346181,
+      "learning_rate": 1.5646842426291775e-07,
+      "loss": 0.8126,
+      "step": 4909
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6475807742135788,
+      "learning_rate": 1.5537133116686142e-07,
+      "loss": 0.7223,
+      "step": 4910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6864491977631575,
+      "learning_rate": 1.542780676862776e-07,
+      "loss": 0.8824,
+      "step": 4911
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4541071162068364,
+      "learning_rate": 1.531886342464517e-07,
+      "loss": 0.7289,
+      "step": 4912
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6553024079033145,
+      "learning_rate": 1.521030312711802e-07,
+      "loss": 0.8028,
+      "step": 4913
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5381855897941494,
+      "learning_rate": 1.5102125918276978e-07,
+      "loss": 0.8452,
+      "step": 4914
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.81594511213681,
+      "learning_rate": 1.499433184020327e-07,
+      "loss": 0.9022,
+      "step": 4915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6259255253716556,
+      "learning_rate": 1.4886920934829574e-07,
+      "loss": 0.8288,
+      "step": 4916
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6591298059831057,
+      "learning_rate": 1.4779893243939358e-07,
+      "loss": 0.8641,
+      "step": 4917
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6529473509522685,
+      "learning_rate": 1.4673248809166984e-07,
+      "loss": 0.7588,
+      "step": 4918
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6004992764307049,
+      "learning_rate": 1.45669876719976e-07,
+      "loss": 0.803,
+      "step": 4919
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6114035564591866,
+      "learning_rate": 1.4461109873767587e-07,
+      "loss": 0.8716,
+      "step": 4920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6912543405825373,
+      "learning_rate": 1.4355615455663884e-07,
+      "loss": 0.7688,
+      "step": 4921
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9525310032028005,
+      "learning_rate": 1.4250504458724336e-07,
+      "loss": 0.8375,
+      "step": 4922
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8051419365888983,
+      "learning_rate": 1.4145776923837895e-07,
+      "loss": 0.9364,
+      "step": 4923
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7084739920204828,
+      "learning_rate": 1.4041432891743978e-07,
+      "loss": 0.9089,
+      "step": 4924
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6253733176066043,
+      "learning_rate": 1.3937472403033002e-07,
+      "loss": 0.8488,
+      "step": 4925
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7310556112020248,
+      "learning_rate": 1.3833895498146287e-07,
+      "loss": 0.9087,
+      "step": 4926
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6576546514999582,
+      "learning_rate": 1.3730702217375825e-07,
+      "loss": 0.8847,
+      "step": 4927
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6200150482231441,
+      "learning_rate": 1.362789260086439e-07,
+      "loss": 0.7445,
+      "step": 4928
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5773998945496541,
+      "learning_rate": 1.3525466688605328e-07,
+      "loss": 0.7787,
+      "step": 4929
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.629333696277938,
+      "learning_rate": 1.3423424520443095e-07,
+      "loss": 0.9095,
+      "step": 4930
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6469354057372534,
+      "learning_rate": 1.33217661360725e-07,
+      "loss": 0.8278,
+      "step": 4931
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5712895075456599,
+      "learning_rate": 1.3220491575039352e-07,
+      "loss": 0.8249,
+      "step": 4932
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6775311526534245,
+      "learning_rate": 1.3119600876739913e-07,
+      "loss": 0.8519,
+      "step": 4933
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7074824689906087,
+      "learning_rate": 1.3019094080421236e-07,
+      "loss": 0.8262,
+      "step": 4934
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5978239435800343,
+      "learning_rate": 1.2918971225181265e-07,
+      "loss": 0.8392,
+      "step": 4935
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0103492039735638,
+      "learning_rate": 1.2819232349967958e-07,
+      "loss": 0.8086,
+      "step": 4936
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5515066332509084,
+      "learning_rate": 1.2719877493580613e-07,
+      "loss": 0.8259,
+      "step": 4937
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9456664714615861,
+      "learning_rate": 1.2620906694668533e-07,
+      "loss": 0.8409,
+      "step": 4938
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.716993131432421,
+      "learning_rate": 1.2522319991732256e-07,
+      "loss": 0.8065,
+      "step": 4939
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.519584786287795,
+      "learning_rate": 1.242411742312233e-07,
+      "loss": 0.7987,
+      "step": 4940
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.814988108224232,
+      "learning_rate": 1.232629902704008e-07,
+      "loss": 0.7686,
+      "step": 4941
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.587242203219095,
+      "learning_rate": 1.2228864841537512e-07,
+      "loss": 0.8539,
+      "step": 4942
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5219598591187173,
+      "learning_rate": 1.2131814904517093e-07,
+      "loss": 0.8002,
+      "step": 4943
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7067301828543635,
+      "learning_rate": 1.2035149253731615e-07,
+      "loss": 0.8167,
+      "step": 4944
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4780687729376636,
+      "learning_rate": 1.1938867926784782e-07,
+      "loss": 0.8022,
+      "step": 4945
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8245418722982043,
+      "learning_rate": 1.1842970961130406e-07,
+      "loss": 0.9316,
+      "step": 4946
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8012418885011572,
+      "learning_rate": 1.1747458394072986e-07,
+      "loss": 0.8605,
+      "step": 4947
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6095664936474214,
+      "learning_rate": 1.1652330262767353e-07,
+      "loss": 0.8166,
+      "step": 4948
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7146979682040404,
+      "learning_rate": 1.1557586604219129e-07,
+      "loss": 0.8968,
+      "step": 4949
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5895056850751075,
+      "learning_rate": 1.1463227455283943e-07,
+      "loss": 0.8926,
+      "step": 4950
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5541333797103272,
+      "learning_rate": 1.13692528526681e-07,
+      "loss": 0.7857,
+      "step": 4951
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5896127496722006,
+      "learning_rate": 1.1275662832928135e-07,
+      "loss": 0.8841,
+      "step": 4952
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4811800456998214,
+      "learning_rate": 1.118245743247115e-07,
+      "loss": 0.7616,
+      "step": 4953
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4505754590481381,
+      "learning_rate": 1.1089636687554583e-07,
+      "loss": 0.837,
+      "step": 4954
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5661718657308172,
+      "learning_rate": 1.099720063428622e-07,
+      "loss": 0.8273,
+      "step": 4955
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.950799488866897,
+      "learning_rate": 1.0905149308624185e-07,
+      "loss": 0.7992,
+      "step": 4956
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6311388423421525,
+      "learning_rate": 1.0813482746377058e-07,
+      "loss": 0.8573,
+      "step": 4957
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5762042142140522,
+      "learning_rate": 1.0722200983203423e-07,
+      "loss": 0.7506,
+      "step": 4958
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7399002723266332,
+      "learning_rate": 1.0631304054612545e-07,
+      "loss": 0.8777,
+      "step": 4959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.64896309366328,
+      "learning_rate": 1.0540791995963917e-07,
+      "loss": 0.8091,
+      "step": 4960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6609498470045294,
+      "learning_rate": 1.045066484246704e-07,
+      "loss": 0.7811,
+      "step": 4961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0063699978672147,
+      "learning_rate": 1.036092262918198e-07,
+      "loss": 0.8015,
+      "step": 4962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6084066802033907,
+      "learning_rate": 1.0271565391018922e-07,
+      "loss": 0.8932,
+      "step": 4963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6165487237067675,
+      "learning_rate": 1.0182593162738508e-07,
+      "loss": 0.8962,
+      "step": 4964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6049658161860338,
+      "learning_rate": 1.009400597895116e-07,
+      "loss": 0.8494,
+      "step": 4965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6258440115836938,
+      "learning_rate": 1.0005803874117981e-07,
+      "loss": 0.8588,
+      "step": 4966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6540136388403845,
+      "learning_rate": 9.917986882549968e-08,
+      "loss": 0.8433,
+      "step": 4967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6530462586563348,
+      "learning_rate": 9.830555038408463e-08,
+      "loss": 0.856,
+      "step": 4968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4860983635785385,
+      "learning_rate": 9.743508375704924e-08,
+      "loss": 0.6871,
+      "step": 4969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5507562812035725,
+      "learning_rate": 9.656846928300934e-08,
+      "loss": 0.8492,
+      "step": 4970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6346673988974738,
+      "learning_rate": 9.570570729908412e-08,
+      "loss": 0.8396,
+      "step": 4971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7368013133597024,
+      "learning_rate": 9.484679814089293e-08,
+      "loss": 0.9203,
+      "step": 4972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6340143749831773,
+      "learning_rate": 9.399174214255513e-08,
+      "loss": 0.8939,
+      "step": 4973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.707005458041116,
+      "learning_rate": 9.314053963669245e-08,
+      "loss": 0.7957,
+      "step": 4974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6392193096320475,
+      "learning_rate": 9.22931909544278e-08,
+      "loss": 0.8707,
+      "step": 4975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.587703446312483,
+      "learning_rate": 9.144969642538415e-08,
+      "loss": 0.768,
+      "step": 4976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.635947063395266,
+      "learning_rate": 9.061005637768571e-08,
+      "loss": 0.8609,
+      "step": 4977
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6263184836466396,
+      "learning_rate": 8.977427113795678e-08,
+      "loss": 0.9195,
+      "step": 4978
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6023902021963157,
+      "learning_rate": 8.894234103132394e-08,
+      "loss": 0.8682,
+      "step": 4979
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6563911158006064,
+      "learning_rate": 8.811426638141163e-08,
+      "loss": 0.8095,
+      "step": 4980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7609044106395593,
+      "learning_rate": 8.729004751034443e-08,
+      "loss": 0.8567,
+      "step": 4981
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4790167691488278,
+      "learning_rate": 8.646968473874917e-08,
+      "loss": 0.8171,
+      "step": 4982
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6269543335652379,
+      "learning_rate": 8.565317838575171e-08,
+      "loss": 0.8258,
+      "step": 4983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5583935374943,
+      "learning_rate": 8.484052876897797e-08,
+      "loss": 0.8265,
+      "step": 4984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4943474875070648,
+      "learning_rate": 8.403173620455174e-08,
+      "loss": 0.7888,
+      "step": 4985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.594762702711398,
+      "learning_rate": 8.322680100710023e-08,
+      "loss": 0.8098,
+      "step": 4986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6022569198318086,
+      "learning_rate": 8.24257234897452e-08,
+      "loss": 0.7609,
+      "step": 4987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6467693510359207,
+      "learning_rate": 8.162850396411182e-08,
+      "loss": 0.8345,
+      "step": 4988
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.751101015490329,
+      "learning_rate": 8.08351427403209e-08,
+      "loss": 0.8488,
+      "step": 4989
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5795700884294022,
+      "learning_rate": 8.004564012699667e-08,
+      "loss": 0.8266,
+      "step": 4990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6452428063118647,
+      "learning_rate": 7.925999643125904e-08,
+      "loss": 0.7832,
+      "step": 4991
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6461661712590487,
+      "learning_rate": 7.847821195872796e-08,
+      "loss": 0.9808,
+      "step": 4992
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5657255170988507,
+      "learning_rate": 7.770028701352239e-08,
+      "loss": 0.8519,
+      "step": 4993
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6852755964063066,
+      "learning_rate": 7.692622189825916e-08,
+      "loss": 0.8425,
+      "step": 4994
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9054468178480464,
+      "learning_rate": 7.615601691405405e-08,
+      "loss": 0.8855,
+      "step": 4995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4668506651688153,
+      "learning_rate": 7.538967236051963e-08,
+      "loss": 0.8569,
+      "step": 4996
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7272847342044952,
+      "learning_rate": 7.462718853576966e-08,
+      "loss": 0.8052,
+      "step": 4997
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7114468698535032,
+      "learning_rate": 7.386856573641576e-08,
+      "loss": 0.7993,
+      "step": 4998
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.573996192998908,
+      "learning_rate": 7.311380425756298e-08,
+      "loss": 0.8533,
+      "step": 4999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6580048196864743,
+      "learning_rate": 7.236290439282089e-08,
+      "loss": 0.8353,
+      "step": 5000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7654954750565555,
+      "learning_rate": 7.161586643429363e-08,
+      "loss": 0.8921,
+      "step": 5001
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6145501684763122,
+      "learning_rate": 7.087269067258096e-08,
+      "loss": 0.8783,
+      "step": 5002
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9118114667109425,
+      "learning_rate": 7.013337739678384e-08,
+      "loss": 0.8452,
+      "step": 5003
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5736946290540468,
+      "learning_rate": 6.939792689449887e-08,
+      "loss": 0.7785,
+      "step": 5004
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.655560944591741,
+      "learning_rate": 6.866633945182056e-08,
+      "loss": 0.8096,
+      "step": 5005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5978020130154031,
+      "learning_rate": 6.793861535334123e-08,
+      "loss": 0.7086,
+      "step": 5006
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5052218419363015,
+      "learning_rate": 6.721475488214779e-08,
+      "loss": 0.8228,
+      "step": 5007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6533228152329014,
+      "learning_rate": 6.649475831982832e-08,
+      "loss": 0.8797,
+      "step": 5008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6787394183685973,
+      "learning_rate": 6.577862594646323e-08,
+      "loss": 0.8118,
+      "step": 5009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5411752551911844,
+      "learning_rate": 6.506635804063299e-08,
+      "loss": 0.8224,
+      "step": 5010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.688450557931162,
+      "learning_rate": 6.435795487941376e-08,
+      "loss": 0.8157,
+      "step": 5011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6425588483398081,
+      "learning_rate": 6.365341673837844e-08,
+      "loss": 0.8502,
+      "step": 5012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6119876126491235,
+      "learning_rate": 6.295274389159666e-08,
+      "loss": 0.8813,
+      "step": 5013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7214892802938053,
+      "learning_rate": 6.225593661163376e-08,
+      "loss": 0.8453,
+      "step": 5014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5256952210050458,
+      "learning_rate": 6.156299516955178e-08,
+      "loss": 0.8145,
+      "step": 5015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.508922977482306,
+      "learning_rate": 6.087391983490843e-08,
+      "loss": 0.8324,
+      "step": 5016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.649823571705899,
+      "learning_rate": 6.018871087575706e-08,
+      "loss": 0.8625,
+      "step": 5017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6228795975521746,
+      "learning_rate": 5.950736855864892e-08,
+      "loss": 0.7047,
+      "step": 5018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.554732455772496,
+      "learning_rate": 5.882989314863086e-08,
+      "loss": 0.8763,
+      "step": 5019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6282802143773705,
+      "learning_rate": 5.815628490924319e-08,
+      "loss": 0.8891,
+      "step": 5020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4826657295670957,
+      "learning_rate": 5.7486544102522966e-08,
+      "loss": 0.8672,
+      "step": 5021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6076111337935641,
+      "learning_rate": 5.682067098900401e-08,
+      "loss": 0.8568,
+      "step": 5022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5827901484220512,
+      "learning_rate": 5.6158665827714676e-08,
+      "loss": 0.8224,
+      "step": 5023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6421253258867103,
+      "learning_rate": 5.550052887617785e-08,
+      "loss": 0.7737,
+      "step": 5024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6332704153447963,
+      "learning_rate": 5.48462603904143e-08,
+      "loss": 0.8125,
+      "step": 5025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5726502131794666,
+      "learning_rate": 5.419586062493487e-08,
+      "loss": 0.8096,
+      "step": 5026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5968877523552603,
+      "learning_rate": 5.354932983275163e-08,
+      "loss": 0.834,
+      "step": 5027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5907294651763357,
+      "learning_rate": 5.290666826536783e-08,
+      "loss": 0.8861,
+      "step": 5028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7127662547875466,
+      "learning_rate": 5.226787617278351e-08,
+      "loss": 0.81,
+      "step": 5029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5965697594445905,
+      "learning_rate": 5.1632953803489873e-08,
+      "loss": 0.7972,
+      "step": 5030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8419305334707237,
+      "learning_rate": 5.100190140447603e-08,
+      "loss": 0.8995,
+      "step": 5031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8379876509641018,
+      "learning_rate": 5.037471922122561e-08,
+      "loss": 0.8115,
+      "step": 5032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7083041995397021,
+      "learning_rate": 4.9751407497716785e-08,
+      "loss": 0.8653,
+      "step": 5033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5417880730960605,
+      "learning_rate": 4.913196647641894e-08,
+      "loss": 0.762,
+      "step": 5034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6238524114520576,
+      "learning_rate": 4.8516396398299307e-08,
+      "loss": 0.8216,
+      "step": 5035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6746641291015834,
+      "learning_rate": 4.790469750281857e-08,
+      "loss": 0.7293,
+      "step": 5036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.603942617262543,
+      "learning_rate": 4.729687002793082e-08,
+      "loss": 0.7568,
+      "step": 5037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8032201477135261,
+      "learning_rate": 4.6692914210084704e-08,
+      "loss": 0.867,
+      "step": 5038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.613870404648611,
+      "learning_rate": 4.609283028422118e-08,
+      "loss": 0.7986,
+      "step": 5039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6279806909052765,
+      "learning_rate": 4.5496618483776844e-08,
+      "loss": 0.8508,
+      "step": 5040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7216084429710294,
+      "learning_rate": 4.490427904068173e-08,
+      "loss": 0.8321,
+      "step": 5041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6595239989758406,
+      "learning_rate": 4.431581218535819e-08,
+      "loss": 0.8461,
+      "step": 5042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6467603184605435,
+      "learning_rate": 4.3731218146724205e-08,
+      "loss": 0.8069,
+      "step": 5043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6047223897402323,
+      "learning_rate": 4.3150497152190104e-08,
+      "loss": 0.8527,
+      "step": 5044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5565508205041574,
+      "learning_rate": 4.257364942765962e-08,
+      "loss": 0.8968,
+      "step": 5045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6198805652761188,
+      "learning_rate": 4.2000675197527705e-08,
+      "loss": 0.8613,
+      "step": 5046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5747109313858192,
+      "learning_rate": 4.143157468468717e-08,
+      "loss": 0.779,
+      "step": 5047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7218217942329022,
+      "learning_rate": 4.086634811052093e-08,
+      "loss": 0.8046,
+      "step": 5048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7749428420271092,
+      "learning_rate": 4.0304995694904206e-08,
+      "loss": 0.8753,
+      "step": 5049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5052709074365256,
+      "learning_rate": 3.974751765620566e-08,
+      "loss": 0.8693,
+      "step": 5050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.703107345853086,
+      "learning_rate": 3.9193914211289595e-08,
+      "loss": 0.8621,
+      "step": 5051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9309859559371372,
+      "learning_rate": 3.8644185575509304e-08,
+      "loss": 0.8399,
+      "step": 5052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6304712280786575,
+      "learning_rate": 3.809833196271262e-08,
+      "loss": 0.8844,
+      "step": 5053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7012787192726428,
+      "learning_rate": 3.755635358523968e-08,
+      "loss": 0.8641,
+      "step": 5054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6312800733769852,
+      "learning_rate": 3.701825065392184e-08,
+      "loss": 0.8646,
+      "step": 5055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6561177972776637,
+      "learning_rate": 3.6484023378087207e-08,
+      "loss": 0.9099,
+      "step": 5056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6494112533499055,
+      "learning_rate": 3.5953671965549556e-08,
+      "loss": 0.8814,
+      "step": 5057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6005921553975702,
+      "learning_rate": 3.542719662262162e-08,
+      "loss": 0.8799,
+      "step": 5058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5993174398296366,
+      "learning_rate": 3.4904597554102916e-08,
+      "loss": 0.8295,
+      "step": 5059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5803482933456774,
+      "learning_rate": 3.438587496328971e-08,
+      "loss": 0.7949,
+      "step": 5060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5293500483630196,
+      "learning_rate": 3.387102905196504e-08,
+      "loss": 0.8407,
+      "step": 5061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5728658963313757,
+      "learning_rate": 3.336006002040981e-08,
+      "loss": 0.8043,
+      "step": 5062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6537775564074648,
+      "learning_rate": 3.285296806739169e-08,
+      "loss": 0.9022,
+      "step": 5063
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9480932759109553,
+      "learning_rate": 3.2349753390172875e-08,
+      "loss": 0.8044,
+      "step": 5064
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6509451730940818,
+      "learning_rate": 3.185041618450791e-08,
+      "loss": 0.8602,
+      "step": 5065
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6383268325158629,
+      "learning_rate": 3.135495664464028e-08,
+      "loss": 0.8491,
+      "step": 5066
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8052044072850135,
+      "learning_rate": 3.0863374963306937e-08,
+      "loss": 0.9483,
+      "step": 5067
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6475072164800793,
+      "learning_rate": 3.037567133173491e-08,
+      "loss": 0.8118,
+      "step": 5068
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7332152988329468,
+      "learning_rate": 2.9891845939645784e-08,
+      "loss": 0.9038,
+      "step": 5069
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7999343950708404,
+      "learning_rate": 2.9411898975248986e-08,
+      "loss": 0.8353,
+      "step": 5070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.854215710005796,
+      "learning_rate": 2.8935830625246298e-08,
+      "loss": 0.8407,
+      "step": 5071
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.549691186480608,
+      "learning_rate": 2.8463641074831795e-08,
+      "loss": 0.7982,
+      "step": 5072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7358625186265642,
+      "learning_rate": 2.7995330507689654e-08,
+      "loss": 0.8464,
+      "step": 5073
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.563700543248081,
+      "learning_rate": 2.753089910599527e-08,
+      "loss": 0.7923,
+      "step": 5074
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0515256215730302,
+      "learning_rate": 2.7070347050414116e-08,
+      "loss": 0.8461,
+      "step": 5075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7741478284170926,
+      "learning_rate": 2.6613674520103993e-08,
+      "loss": 0.8226,
+      "step": 5076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.724609452883939,
+      "learning_rate": 2.6160881692715022e-08,
+      "loss": 0.861,
+      "step": 5077
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.565425406538292,
+      "learning_rate": 2.5711968744382975e-08,
+      "loss": 0.8323,
+      "step": 5078
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6609451728076836,
+      "learning_rate": 2.526693584973927e-08,
+      "loss": 0.7217,
+      "step": 5079
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6241790912713077,
+      "learning_rate": 2.4825783181904316e-08,
+      "loss": 0.8674,
+      "step": 5080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.690972598745633,
+      "learning_rate": 2.4388510912488616e-08,
+      "loss": 0.8584,
+      "step": 5081
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.740374592153776,
+      "learning_rate": 2.395511921159388e-08,
+      "loss": 0.8418,
+      "step": 5082
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7464268750640428,
+      "learning_rate": 2.3525608247810804e-08,
+      "loss": 0.8143,
+      "step": 5083
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4885365270139768,
+      "learning_rate": 2.3099978188222405e-08,
+      "loss": 0.8288,
+      "step": 5084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.692331717964315,
+      "learning_rate": 2.267822919840068e-08,
+      "loss": 0.8872,
+      "step": 5085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7073103239035485,
+      "learning_rate": 2.226036144240884e-08,
+      "loss": 0.8419,
+      "step": 5086
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6344892997572353,
+      "learning_rate": 2.1846375082799075e-08,
+      "loss": 0.8914,
+      "step": 5087
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7455944887727222,
+      "learning_rate": 2.143627028061479e-08,
+      "loss": 0.8782,
+      "step": 5088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7734954614990008,
+      "learning_rate": 2.103004719539059e-08,
+      "loss": 0.8708,
+      "step": 5089
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.623441631078164,
+      "learning_rate": 2.062770598514674e-08,
+      "loss": 0.8876,
+      "step": 5090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6465556905807972,
+      "learning_rate": 2.0229246806398038e-08,
+      "loss": 0.9057,
+      "step": 5091
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.653369543030162,
+      "learning_rate": 1.9834669814147167e-08,
+      "loss": 0.7653,
+      "step": 5092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7081306457412124,
+      "learning_rate": 1.9443975161885788e-08,
+      "loss": 0.8659,
+      "step": 5093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6432939622440883,
+      "learning_rate": 1.905716300159677e-08,
+      "loss": 0.8421,
+      "step": 5094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7667255391963297,
+      "learning_rate": 1.867423348375308e-08,
+      "loss": 0.7795,
+      "step": 5095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9320083179845093,
+      "learning_rate": 1.8295186757315564e-08,
+      "loss": 0.811,
+      "step": 5096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8562440730099243,
+      "learning_rate": 1.7920022969735164e-08,
+      "loss": 0.8449,
+      "step": 5097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6650818602783475,
+      "learning_rate": 1.754874226695402e-08,
+      "loss": 0.8192,
+      "step": 5098
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.591780877478698,
+      "learning_rate": 1.7181344793401055e-08,
+      "loss": 0.9129,
+      "step": 5099
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6354706976221034,
+      "learning_rate": 1.6817830691996385e-08,
+      "loss": 0.8056,
+      "step": 5100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7235408126957397,
+      "learning_rate": 1.6458200104149115e-08,
+      "loss": 0.8615,
+      "step": 5101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6661822512455184,
+      "learning_rate": 1.6102453169757337e-08,
+      "loss": 0.8825,
+      "step": 5102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.759564492398854,
+      "learning_rate": 1.5750590027209244e-08,
+      "loss": 0.8853,
+      "step": 5103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.59989224405631,
+      "learning_rate": 1.54026108133809e-08,
+      "loss": 0.7578,
+      "step": 5104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.573909842104531,
+      "learning_rate": 1.5058515663638473e-08,
+      "loss": 0.8162,
+      "step": 5105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5889184849238762,
+      "learning_rate": 1.4718304711837106e-08,
+      "loss": 0.8347,
+      "step": 5106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5829717172650974,
+      "learning_rate": 1.4381978090319825e-08,
+      "loss": 0.8046,
+      "step": 5107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6915179448141584,
+      "learning_rate": 1.4049535929920866e-08,
+      "loss": 0.8854,
+      "step": 5108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5875188336651425,
+      "learning_rate": 1.3720978359961223e-08,
+      "loss": 0.8951,
+      "step": 5109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.881345862043747,
+      "learning_rate": 1.3396305508251995e-08,
+      "loss": 0.8567,
+      "step": 5110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.641138427179114,
+      "learning_rate": 1.307551750109215e-08,
+      "loss": 0.8283,
+      "step": 5111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.566645039933604,
+      "learning_rate": 1.2758614463269647e-08,
+      "loss": 0.854,
+      "step": 5112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.630184522216553,
+      "learning_rate": 1.2445596518063652e-08,
+      "loss": 0.8447,
+      "step": 5113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6343335121137554,
+      "learning_rate": 1.2136463787237873e-08,
+      "loss": 0.8477,
+      "step": 5114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5971187299515386,
+      "learning_rate": 1.1831216391047229e-08,
+      "loss": 0.8627,
+      "step": 5115
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5508386494216162,
+      "learning_rate": 1.1529854448234512e-08,
+      "loss": 0.8263,
+      "step": 5116
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6482742697102761,
+      "learning_rate": 1.1232378076032613e-08,
+      "loss": 0.8037,
+      "step": 5117
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7093490984569042,
+      "learning_rate": 1.0938787390158967e-08,
+      "loss": 0.8181,
+      "step": 5118
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9114038176631246,
+      "learning_rate": 1.0649082504824438e-08,
+      "loss": 0.854,
+      "step": 5119
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5118071087111942,
+      "learning_rate": 1.0363263532724433e-08,
+      "loss": 0.7517,
+      "step": 5120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7291453537426127,
+      "learning_rate": 1.0081330585044458e-08,
+      "loss": 0.7582,
+      "step": 5121
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4866449715148402,
+      "learning_rate": 9.803283771457894e-09,
+      "loss": 0.821,
+      "step": 5122
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.844592526370849,
+      "learning_rate": 9.529123200126e-09,
+      "loss": 0.8861,
+      "step": 5123
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.665918236829414,
+      "learning_rate": 9.25884897770013e-09,
+      "loss": 0.9235,
+      "step": 5124
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.715931657309746,
+      "learning_rate": 8.992461209317293e-09,
+      "loss": 0.8231,
+      "step": 5125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4955005155353707,
+      "learning_rate": 8.729959998604598e-09,
+      "loss": 0.8368,
+      "step": 5126
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.77563611808697,
+      "learning_rate": 8.471345447675916e-09,
+      "loss": 0.8615,
+      "step": 5127
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5835590146137293,
+      "learning_rate": 8.216617657134107e-09,
+      "loss": 0.8523,
+      "step": 5128
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.574452618002154,
+      "learning_rate": 7.965776726069907e-09,
+      "loss": 0.7325,
+      "step": 5129
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.850592131174334,
+      "learning_rate": 7.718822752060818e-09,
+      "loss": 0.8744,
+      "step": 5130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9710698074816303,
+      "learning_rate": 7.475755831173326e-09,
+      "loss": 0.8804,
+      "step": 5131
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6180211661057113,
+      "learning_rate": 7.2365760579640174e-09,
+      "loss": 0.9059,
+      "step": 5132
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5694597320085724,
+      "learning_rate": 7.001283525472913e-09,
+      "loss": 0.8977,
+      "step": 5133
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.616611521861758,
+      "learning_rate": 6.769878325230128e-09,
+      "loss": 0.8727,
+      "step": 5134
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8143085265899568,
+      "learning_rate": 6.542360547254767e-09,
+      "loss": 0.7946,
+      "step": 5135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.774321525857253,
+      "learning_rate": 6.318730280052699e-09,
+      "loss": 0.7916,
+      "step": 5136
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8114381557865862,
+      "learning_rate": 6.098987610615448e-09,
+      "loss": 0.8519,
+      "step": 5137
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5144282562601932,
+      "learning_rate": 5.8831326244268575e-09,
+      "loss": 0.812,
+      "step": 5138
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7824645219315305,
+      "learning_rate": 5.671165405453094e-09,
+      "loss": 0.844,
+      "step": 5139
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6399204992947645,
+      "learning_rate": 5.4630860361515324e-09,
+      "loss": 0.7886,
+      "step": 5140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8207594207869748,
+      "learning_rate": 5.258894597467423e-09,
+      "loss": 0.8766,
+      "step": 5141
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.991625244902772,
+      "learning_rate": 5.058591168830562e-09,
+      "loss": 0.8613,
+      "step": 5142
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7312053968795718,
+      "learning_rate": 4.8621758281608426e-09,
+      "loss": 0.8603,
+      "step": 5143
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.698241931859162,
+      "learning_rate": 4.669648651864922e-09,
+      "loss": 0.8037,
+      "step": 5144
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6908540589117786,
+      "learning_rate": 4.481009714837337e-09,
+      "loss": 0.8546,
+      "step": 5145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.790975338751888,
+      "learning_rate": 4.296259090459387e-09,
+      "loss": 0.8762,
+      "step": 5146
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7757705743109653,
+      "learning_rate": 4.11539685059914e-09,
+      "loss": 0.8598,
+      "step": 5147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7243442269052214,
+      "learning_rate": 3.938423065614761e-09,
+      "loss": 0.822,
+      "step": 5148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9632648701084365,
+      "learning_rate": 3.765337804348956e-09,
+      "loss": 0.791,
+      "step": 5149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9813652622548175,
+      "learning_rate": 3.5961411341334242e-09,
+      "loss": 0.822,
+      "step": 5150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5630357443677856,
+      "learning_rate": 3.430833120785515e-09,
+      "loss": 0.84,
+      "step": 5151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7492883390211051,
+      "learning_rate": 3.2694138286115675e-09,
+      "loss": 0.9079,
+      "step": 5152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6309410954171883,
+      "learning_rate": 3.111883320405795e-09,
+      "loss": 0.8473,
+      "step": 5153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.596877739821343,
+      "learning_rate": 2.9582416574469586e-09,
+      "loss": 0.8461,
+      "step": 5154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5925287611946466,
+      "learning_rate": 2.808488899502804e-09,
+      "loss": 0.8105,
+      "step": 5155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6492552852360387,
+      "learning_rate": 2.662625104828953e-09,
+      "loss": 0.8085,
+      "step": 5156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6328585020726267,
+      "learning_rate": 2.520650330165575e-09,
+      "loss": 0.9675,
+      "step": 5157
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7251759419660853,
+      "learning_rate": 2.382564630742934e-09,
+      "loss": 0.8282,
+      "step": 5158
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7082303055482375,
+      "learning_rate": 2.2483680602758405e-09,
+      "loss": 0.9106,
+      "step": 5159
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7016339257944202,
+      "learning_rate": 2.118060670969202e-09,
+      "loss": 0.8177,
+      "step": 5160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.874284129232015,
+      "learning_rate": 1.9916425135124705e-09,
+      "loss": 0.9121,
+      "step": 5161
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6236162084713028,
+      "learning_rate": 1.8691136370829753e-09,
+      "loss": 0.8313,
+      "step": 5162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6740781383851866,
+      "learning_rate": 1.7504740893448114e-09,
+      "loss": 0.8147,
+      "step": 5163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9672322600019123,
+      "learning_rate": 1.6357239164499494e-09,
+      "loss": 0.8352,
+      "step": 5164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9668260069330462,
+      "learning_rate": 1.5248631630371269e-09,
+      "loss": 0.8004,
+      "step": 5165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6209283876716638,
+      "learning_rate": 1.4178918722307368e-09,
+      "loss": 0.7747,
+      "step": 5166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.413046755338035,
+      "learning_rate": 1.3148100856452683e-09,
+      "loss": 0.6984,
+      "step": 5167
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9613978459669573,
+      "learning_rate": 1.2156178433775367e-09,
+      "loss": 0.8455,
+      "step": 5168
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6700346943228064,
+      "learning_rate": 1.1203151840144532e-09,
+      "loss": 0.8398,
+      "step": 5169
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6359624496211949,
+      "learning_rate": 1.0289021446308057e-09,
+      "loss": 0.8803,
+      "step": 5170
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8651244442136061,
+      "learning_rate": 9.41378760785927e-10,
+      "loss": 0.8589,
+      "step": 5171
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5911907887138388,
+      "learning_rate": 8.577450665259168e-10,
+      "loss": 0.8057,
+      "step": 5172
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5554905803158499,
+      "learning_rate": 7.780010943869709e-10,
+      "loss": 0.8975,
+      "step": 5173
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4970547995831882,
+      "learning_rate": 7.021468753876104e-10,
+      "loss": 0.8237,
+      "step": 5174
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5530252279042271,
+      "learning_rate": 6.301824390364531e-10,
+      "loss": 0.76,
+      "step": 5175
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9978355972737286,
+      "learning_rate": 5.621078133288826e-10,
+      "loss": 0.8682,
+      "step": 5176
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.562981919130239,
+      "learning_rate": 4.979230247448285e-10,
+      "loss": 0.9366,
+      "step": 5177
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5726159078415627,
+      "learning_rate": 4.3762809825431683e-10,
+      "loss": 0.8269,
+      "step": 5178
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7359554516016393,
+      "learning_rate": 3.812230573108089e-10,
+      "loss": 0.8227,
+      "step": 5179
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7266499350927984,
+      "learning_rate": 3.287079238556423e-10,
+      "loss": 0.9035,
+      "step": 5180
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.484723383772358,
+      "learning_rate": 2.8008271832025146e-10,
+      "loss": 0.8493,
+      "step": 5181
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5940697942187512,
+      "learning_rate": 2.353474596172856e-10,
+      "loss": 0.7817,
+      "step": 5182
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6369499183598217,
+      "learning_rate": 1.9450216515171095e-10,
+      "loss": 0.8108,
+      "step": 5183
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9527476097881272,
+      "learning_rate": 1.575468508097089e-10,
+      "loss": 0.7859,
+      "step": 5184
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6190775938265858,
+      "learning_rate": 1.2448153096866754e-10,
+      "loss": 0.8352,
+      "step": 5185
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.852738844719661,
+      "learning_rate": 9.530621849163091e-11,
+      "loss": 0.893,
+      "step": 5186
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8217311084116505,
+      "learning_rate": 7.002092472618849e-11,
+      "loss": 0.8475,
+      "step": 5187
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7199245087384853,
+      "learning_rate": 4.862565951002651e-11,
+      "loss": 0.7977,
+      "step": 5188
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5914966865108728,
+      "learning_rate": 3.1120431166487e-11,
+      "loss": 0.8064,
+      "step": 5189
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6126429604019001,
+      "learning_rate": 1.750524650345753e-11,
+      "loss": 0.8504,
+      "step": 5190
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9184052457101984,
+      "learning_rate": 7.780110817812159e-12,
+      "loss": 0.8093,
+      "step": 5191
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7925105827339654,
+      "learning_rate": 1.945027893190954e-12,
+      "loss": 0.8722,
+      "step": 5192
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2295351322457988,
+      "learning_rate": 0.0,
+      "loss": 0.8058,
+      "step": 5193
+    },
+    {
+      "epoch": 1.0,
+      "step": 5193,
+      "total_flos": 7015177953673216.0,
+      "train_loss": 0.9086582657805579,
+      "train_runtime": 33710.3918,
+      "train_samples_per_second": 19.718,
+      "train_steps_per_second": 0.154
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5193,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 7015177953673216.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}