diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,76953 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 10989,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.963009671393625,
+      "learning_rate": 3.0303030303030305e-08,
+      "loss": 0.7908,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.095744933985866,
+      "learning_rate": 6.060606060606061e-08,
+      "loss": 0.6598,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.69702268094129,
+      "learning_rate": 9.090909090909091e-08,
+      "loss": 0.753,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.044986917244724,
+      "learning_rate": 1.2121212121212122e-07,
+      "loss": 0.7035,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.812274623010673,
+      "learning_rate": 1.5151515151515152e-07,
+      "loss": 0.6897,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.60420722502775,
+      "learning_rate": 1.8181818181818183e-07,
+      "loss": 0.6663,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.887572617589537,
+      "learning_rate": 2.1212121212121216e-07,
+      "loss": 0.7324,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.1023251864958,
+      "learning_rate": 2.4242424242424244e-07,
+      "loss": 0.6696,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.88183998563793,
+      "learning_rate": 2.7272727272727274e-07,
+      "loss": 0.7208,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.53190036199524,
+      "learning_rate": 3.0303030303030305e-07,
+      "loss": 0.6542,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.172514534798532,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 0.6134,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.64664638504347,
+      "learning_rate": 3.6363636363636366e-07,
+      "loss": 0.6765,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.0733259638494,
+      "learning_rate": 3.9393939393939396e-07,
+      "loss": 0.6455,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.454225618738189,
+      "learning_rate": 4.242424242424243e-07,
+      "loss": 0.5932,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.586251337949811,
+      "learning_rate": 4.5454545454545457e-07,
+      "loss": 0.5715,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.675021673294397,
+      "learning_rate": 4.848484848484849e-07,
+      "loss": 0.5912,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.270919680918658,
+      "learning_rate": 5.151515151515152e-07,
+      "loss": 0.5805,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.306759889818276,
+      "learning_rate": 5.454545454545455e-07,
+      "loss": 0.5568,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.34247172442105,
+      "learning_rate": 5.757575757575758e-07,
+      "loss": 0.6023,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.58882846298745,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 0.5345,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.106751923119504,
+      "learning_rate": 6.363636363636364e-07,
+      "loss": 0.5005,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.46482346022856,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 0.5472,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.834331230088742,
+      "learning_rate": 6.969696969696971e-07,
+      "loss": 0.5666,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.025044482419066,
+      "learning_rate": 7.272727272727273e-07,
+      "loss": 0.545,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.509716871511943,
+      "learning_rate": 7.575757575757576e-07,
+      "loss": 0.6096,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.694280082090164,
+      "learning_rate": 7.878787878787879e-07,
+      "loss": 0.5522,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.000787108434309,
+      "learning_rate": 8.181818181818182e-07,
+      "loss": 0.5293,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.268597340283241,
+      "learning_rate": 8.484848484848486e-07,
+      "loss": 0.5282,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.682821541492869,
+      "learning_rate": 8.787878787878788e-07,
+      "loss": 0.4408,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.1023797679311205,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.4725,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.787773096933421,
+      "learning_rate": 9.393939393939395e-07,
+      "loss": 0.5228,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.357141474052988,
+      "learning_rate": 9.696969696969698e-07,
+      "loss": 0.4875,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.03764502344773,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.5957,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.559939494819406,
+      "learning_rate": 1.0303030303030304e-06,
+      "loss": 0.4828,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.933533714084318,
+      "learning_rate": 1.0606060606060608e-06,
+      "loss": 0.4173,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.4322447547597355,
+      "learning_rate": 1.090909090909091e-06,
+      "loss": 0.4351,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.8341909409978765,
+      "learning_rate": 1.1212121212121214e-06,
+      "loss": 0.4846,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.024296627960633,
+      "learning_rate": 1.1515151515151516e-06,
+      "loss": 0.4643,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.6892465548146225,
+      "learning_rate": 1.181818181818182e-06,
+      "loss": 0.4609,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.375897228313598,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 0.4192,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.239571486593917,
+      "learning_rate": 1.2424242424242424e-06,
+      "loss": 0.4234,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.189995411250279,
+      "learning_rate": 1.2727272727272728e-06,
+      "loss": 0.4106,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.251701313336269,
+      "learning_rate": 1.3030303030303032e-06,
+      "loss": 0.4673,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.416210270505744,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 0.48,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.956862036107494,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 0.4189,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.631065755223114,
+      "learning_rate": 1.3939393939393942e-06,
+      "loss": 0.469,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.143206429696774,
+      "learning_rate": 1.4242424242424244e-06,
+      "loss": 0.4748,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.6352843604807195,
+      "learning_rate": 1.4545454545454546e-06,
+      "loss": 0.4002,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.6503288524640585,
+      "learning_rate": 1.484848484848485e-06,
+      "loss": 0.4248,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.922470709983655,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 0.426,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.337156233437923,
+      "learning_rate": 1.5454545454545454e-06,
+      "loss": 0.4249,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.414700581820504,
+      "learning_rate": 1.5757575757575759e-06,
+      "loss": 0.3907,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.827231932409133,
+      "learning_rate": 1.6060606060606063e-06,
+      "loss": 0.4119,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.804412312797118,
+      "learning_rate": 1.6363636363636365e-06,
+      "loss": 0.3846,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.975492832662959,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.4187,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.749363114201106,
+      "learning_rate": 1.6969696969696973e-06,
+      "loss": 0.3686,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.693761925144835,
+      "learning_rate": 1.7272727272727275e-06,
+      "loss": 0.4609,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.040286815075631,
+      "learning_rate": 1.7575757575757577e-06,
+      "loss": 0.4155,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.575731911956731,
+      "learning_rate": 1.787878787878788e-06,
+      "loss": 0.3466,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.259506414275161,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.4287,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.583304081613292,
+      "learning_rate": 1.8484848484848487e-06,
+      "loss": 0.3393,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.260371222704529,
+      "learning_rate": 1.878787878787879e-06,
+      "loss": 0.4168,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.191282588576895,
+      "learning_rate": 1.9090909090909095e-06,
+      "loss": 0.4385,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.464209068139261,
+      "learning_rate": 1.9393939393939395e-06,
+      "loss": 0.4007,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.662211296069837,
+      "learning_rate": 1.96969696969697e-06,
+      "loss": 0.3577,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.775996102629911,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.3989,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.011144490054664,
+      "learning_rate": 2.0303030303030303e-06,
+      "loss": 0.3788,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.953400646654258,
+      "learning_rate": 2.0606060606060607e-06,
+      "loss": 0.4083,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.052280696703109,
+      "learning_rate": 2.090909090909091e-06,
+      "loss": 0.367,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.9004762598977125,
+      "learning_rate": 2.1212121212121216e-06,
+      "loss": 0.3817,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.762973742327527,
+      "learning_rate": 2.1515151515151515e-06,
+      "loss": 0.4092,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.940872146814278,
+      "learning_rate": 2.181818181818182e-06,
+      "loss": 0.3763,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.497505091944975,
+      "learning_rate": 2.2121212121212124e-06,
+      "loss": 0.3936,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.591034031060306,
+      "learning_rate": 2.2424242424242428e-06,
+      "loss": 0.3623,
+      "step": 74
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.879686854503531,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 0.4042,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.660394276940384,
+      "learning_rate": 2.303030303030303e-06,
+      "loss": 0.3604,
+      "step": 76
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.875393281807664,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 0.3743,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.948837592400559,
+      "learning_rate": 2.363636363636364e-06,
+      "loss": 0.3633,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.937301466289537,
+      "learning_rate": 2.393939393939394e-06,
+      "loss": 0.396,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.950484159292576,
+      "learning_rate": 2.4242424242424244e-06,
+      "loss": 0.3806,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.46072994196432,
+      "learning_rate": 2.454545454545455e-06,
+      "loss": 0.3086,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.390828680491305,
+      "learning_rate": 2.4848484848484848e-06,
+      "loss": 0.4507,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.557633452416991,
+      "learning_rate": 2.5151515151515156e-06,
+      "loss": 0.3172,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.382270030708564,
+      "learning_rate": 2.5454545454545456e-06,
+      "loss": 0.3918,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.234534643265852,
+      "learning_rate": 2.575757575757576e-06,
+      "loss": 0.382,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.057527792062359,
+      "learning_rate": 2.6060606060606064e-06,
+      "loss": 0.4154,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.577604785412596,
+      "learning_rate": 2.6363636363636364e-06,
+      "loss": 0.3612,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.645821291016337,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 0.3735,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.470086629186744,
+      "learning_rate": 2.6969696969696972e-06,
+      "loss": 0.3597,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.614806605418893,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.3958,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.491349384324725,
+      "learning_rate": 2.7575757575757576e-06,
+      "loss": 0.3258,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.601731049079032,
+      "learning_rate": 2.7878787878787885e-06,
+      "loss": 0.3725,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.427486992793585,
+      "learning_rate": 2.818181818181818e-06,
+      "loss": 0.343,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.027874471200769,
+      "learning_rate": 2.848484848484849e-06,
+      "loss": 0.4245,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.254502616168089,
+      "learning_rate": 2.8787878787878793e-06,
+      "loss": 0.3527,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.213165569079894,
+      "learning_rate": 2.9090909090909093e-06,
+      "loss": 0.336,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.295188374328629,
+      "learning_rate": 2.9393939393939397e-06,
+      "loss": 0.3424,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.391180680688915,
+      "learning_rate": 2.96969696969697e-06,
+      "loss": 0.4139,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.549063964880351,
+      "learning_rate": 3e-06,
+      "loss": 0.3561,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.377539803863616,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.358,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.049776567725781,
+      "learning_rate": 3.0606060606060605e-06,
+      "loss": 0.419,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.826287611008103,
+      "learning_rate": 3.090909090909091e-06,
+      "loss": 0.3643,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.734091355477589,
+      "learning_rate": 3.1212121212121217e-06,
+      "loss": 0.3581,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.545001951779309,
+      "learning_rate": 3.1515151515151517e-06,
+      "loss": 0.4023,
+      "step": 104
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.7114006714939105,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 0.4039,
+      "step": 105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.688865981710678,
+      "learning_rate": 3.2121212121212125e-06,
+      "loss": 0.3779,
+      "step": 106
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.360055231790772,
+      "learning_rate": 3.2424242424242425e-06,
+      "loss": 0.3828,
+      "step": 107
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.4478559409980365,
+      "learning_rate": 3.272727272727273e-06,
+      "loss": 0.3969,
+      "step": 108
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.437423794056705,
+      "learning_rate": 3.3030303030303033e-06,
+      "loss": 0.3817,
+      "step": 109
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.192241324454696,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.3216,
+      "step": 110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.518015664701024,
+      "learning_rate": 3.3636363636363637e-06,
+      "loss": 0.3551,
+      "step": 111
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.657918730500032,
+      "learning_rate": 3.3939393939393946e-06,
+      "loss": 0.3804,
+      "step": 112
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.330609247644378,
+      "learning_rate": 3.4242424242424246e-06,
+      "loss": 0.4238,
+      "step": 113
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.7867427615733025,
+      "learning_rate": 3.454545454545455e-06,
+      "loss": 0.356,
+      "step": 114
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.929152315305148,
+      "learning_rate": 3.4848484848484854e-06,
+      "loss": 0.3828,
+      "step": 115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.506698075720523,
+      "learning_rate": 3.5151515151515154e-06,
+      "loss": 0.3344,
+      "step": 116
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.770793727102921,
+      "learning_rate": 3.5454545454545458e-06,
+      "loss": 0.3798,
+      "step": 117
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.709331141733425,
+      "learning_rate": 3.575757575757576e-06,
+      "loss": 0.4183,
+      "step": 118
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.465020543307715,
+      "learning_rate": 3.606060606060606e-06,
+      "loss": 0.4153,
+      "step": 119
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.421646322435644,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.393,
+      "step": 120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.105192436627971,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 0.3354,
+      "step": 121
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.3952579878429106,
+      "learning_rate": 3.6969696969696974e-06,
+      "loss": 0.3512,
+      "step": 122
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.2473999885219635,
+      "learning_rate": 3.727272727272728e-06,
+      "loss": 0.3633,
+      "step": 123
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.96772057951783,
+      "learning_rate": 3.757575757575758e-06,
+      "loss": 0.3289,
+      "step": 124
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.818716856381312,
+      "learning_rate": 3.7878787878787882e-06,
+      "loss": 0.342,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.9962235047636225,
+      "learning_rate": 3.818181818181819e-06,
+      "loss": 0.2972,
+      "step": 126
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.345489946333712,
+      "learning_rate": 3.848484848484848e-06,
+      "loss": 0.3567,
+      "step": 127
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.388955400375534,
+      "learning_rate": 3.878787878787879e-06,
+      "loss": 0.3402,
+      "step": 128
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.051885892673309,
+      "learning_rate": 3.90909090909091e-06,
+      "loss": 0.3106,
+      "step": 129
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.372439122822092,
+      "learning_rate": 3.93939393939394e-06,
+      "loss": 0.3512,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.92207286736919,
+      "learning_rate": 3.96969696969697e-06,
+      "loss": 0.3798,
+      "step": 131
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.244384893104628,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.3848,
+      "step": 132
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.443894549057829,
+      "learning_rate": 4.030303030303031e-06,
+      "loss": 0.3594,
+      "step": 133
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.503245419426116,
+      "learning_rate": 4.060606060606061e-06,
+      "loss": 0.3909,
+      "step": 134
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.563029843998091,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 0.3797,
+      "step": 135
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.58358025111985,
+      "learning_rate": 4.1212121212121215e-06,
+      "loss": 0.3486,
+      "step": 136
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.488839068680304,
+      "learning_rate": 4.151515151515152e-06,
+      "loss": 0.3477,
+      "step": 137
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.759215057290953,
+      "learning_rate": 4.181818181818182e-06,
+      "loss": 0.4031,
+      "step": 138
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.233461690394125,
+      "learning_rate": 4.212121212121212e-06,
+      "loss": 0.3509,
+      "step": 139
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.58228048027132,
+      "learning_rate": 4.242424242424243e-06,
+      "loss": 0.383,
+      "step": 140
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.493012254311687,
+      "learning_rate": 4.272727272727273e-06,
+      "loss": 0.4095,
+      "step": 141
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.02434810357748,
+      "learning_rate": 4.303030303030303e-06,
+      "loss": 0.3403,
+      "step": 142
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.674592942271979,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 0.3446,
+      "step": 143
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.553046792559494,
+      "learning_rate": 4.363636363636364e-06,
+      "loss": 0.3852,
+      "step": 144
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.907495033740153,
+      "learning_rate": 4.393939393939394e-06,
+      "loss": 0.3012,
+      "step": 145
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.9635217935573355,
+      "learning_rate": 4.424242424242425e-06,
+      "loss": 0.3147,
+      "step": 146
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.9466591027305125,
+      "learning_rate": 4.454545454545455e-06,
+      "loss": 0.4026,
+      "step": 147
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.028605788131155,
+      "learning_rate": 4.4848484848484855e-06,
+      "loss": 0.3698,
+      "step": 148
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.865447860113725,
+      "learning_rate": 4.5151515151515155e-06,
+      "loss": 0.3047,
+      "step": 149
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.677980463632439,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.4124,
+      "step": 150
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.0956202927225345,
+      "learning_rate": 4.575757575757576e-06,
+      "loss": 0.3653,
+      "step": 151
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.974959560587927,
+      "learning_rate": 4.606060606060606e-06,
+      "loss": 0.3514,
+      "step": 152
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.9707710431415215,
+      "learning_rate": 4.636363636363636e-06,
+      "loss": 0.338,
+      "step": 153
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.54767464342179,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.3571,
+      "step": 154
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.085419128808824,
+      "learning_rate": 4.696969696969698e-06,
+      "loss": 0.3209,
+      "step": 155
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.776400786739939,
+      "learning_rate": 4.727272727272728e-06,
+      "loss": 0.3478,
+      "step": 156
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.298048639918654,
+      "learning_rate": 4.757575757575758e-06,
+      "loss": 0.3278,
+      "step": 157
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.309330170318605,
+      "learning_rate": 4.787878787878788e-06,
+      "loss": 0.3445,
+      "step": 158
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.079416144435435,
+      "learning_rate": 4.818181818181819e-06,
+      "loss": 0.3588,
+      "step": 159
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.4053480245061465,
+      "learning_rate": 4.848484848484849e-06,
+      "loss": 0.3983,
+      "step": 160
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.1132658608232235,
+      "learning_rate": 4.878787878787879e-06,
+      "loss": 0.3352,
+      "step": 161
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.546328837286164,
+      "learning_rate": 4.90909090909091e-06,
+      "loss": 0.3965,
+      "step": 162
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.91262498808614,
+      "learning_rate": 4.93939393939394e-06,
+      "loss": 0.3584,
+      "step": 163
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.977515764826995,
+      "learning_rate": 4.9696969696969696e-06,
+      "loss": 0.3192,
+      "step": 164
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.576877435480275,
+      "learning_rate": 5e-06,
+      "loss": 0.3568,
+      "step": 165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.092478277961341,
+      "learning_rate": 5.030303030303031e-06,
+      "loss": 0.3644,
+      "step": 166
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.018685998108373,
+      "learning_rate": 5.060606060606061e-06,
+      "loss": 0.3669,
+      "step": 167
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.374406064793353,
+      "learning_rate": 5.090909090909091e-06,
+      "loss": 0.3466,
+      "step": 168
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.954392818884204,
+      "learning_rate": 5.121212121212121e-06,
+      "loss": 0.354,
+      "step": 169
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.02059888343161,
+      "learning_rate": 5.151515151515152e-06,
+      "loss": 0.3279,
+      "step": 170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.242788697474413,
+      "learning_rate": 5.181818181818182e-06,
+      "loss": 0.346,
+      "step": 171
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.065887809244624,
+      "learning_rate": 5.212121212121213e-06,
+      "loss": 0.3627,
+      "step": 172
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.359910442453347,
+      "learning_rate": 5.242424242424244e-06,
+      "loss": 0.3731,
+      "step": 173
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.243829282397834,
+      "learning_rate": 5.272727272727273e-06,
+      "loss": 0.3676,
+      "step": 174
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.197560167606785,
+      "learning_rate": 5.303030303030303e-06,
+      "loss": 0.3654,
+      "step": 175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.359166123924114,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.4098,
+      "step": 176
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.070099999781161,
+      "learning_rate": 5.3636363636363645e-06,
+      "loss": 0.3415,
+      "step": 177
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.890704309531856,
+      "learning_rate": 5.3939393939393945e-06,
+      "loss": 0.3629,
+      "step": 178
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.983436572084698,
+      "learning_rate": 5.424242424242425e-06,
+      "loss": 0.3819,
+      "step": 179
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.3023861417999,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.3627,
+      "step": 180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.239807160282111,
+      "learning_rate": 5.484848484848485e-06,
+      "loss": 0.3428,
+      "step": 181
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.128244901835555,
+      "learning_rate": 5.515151515151515e-06,
+      "loss": 0.3806,
+      "step": 182
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.042549030366478,
+      "learning_rate": 5.545454545454546e-06,
+      "loss": 0.3375,
+      "step": 183
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.802789116341605,
+      "learning_rate": 5.575757575757577e-06,
+      "loss": 0.3515,
+      "step": 184
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.657750733656597,
+      "learning_rate": 5.606060606060606e-06,
+      "loss": 0.3982,
+      "step": 185
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.861167868735887,
+      "learning_rate": 5.636363636363636e-06,
+      "loss": 0.3663,
+      "step": 186
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.746241905347231,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 0.3478,
+      "step": 187
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.101928408899225,
+      "learning_rate": 5.696969696969698e-06,
+      "loss": 0.3248,
+      "step": 188
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.338465728282297,
+      "learning_rate": 5.727272727272728e-06,
+      "loss": 0.3954,
+      "step": 189
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.082123966970521,
+      "learning_rate": 5.7575757575757586e-06,
+      "loss": 0.3599,
+      "step": 190
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.301358545475363,
+      "learning_rate": 5.787878787878788e-06,
+      "loss": 0.3891,
+      "step": 191
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.966892866760607,
+      "learning_rate": 5.8181818181818185e-06,
+      "loss": 0.3225,
+      "step": 192
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.23131765871254,
+      "learning_rate": 5.8484848484848485e-06,
+      "loss": 0.3697,
+      "step": 193
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.736444371073679,
+      "learning_rate": 5.878787878787879e-06,
+      "loss": 0.4188,
+      "step": 194
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.96307753674885,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 0.3409,
+      "step": 195
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.310009629385978,
+      "learning_rate": 5.93939393939394e-06,
+      "loss": 0.3507,
+      "step": 196
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.373497423064573,
+      "learning_rate": 5.96969696969697e-06,
+      "loss": 0.3731,
+      "step": 197
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.1067076209325135,
+      "learning_rate": 6e-06,
+      "loss": 0.3304,
+      "step": 198
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.228116827660428,
+      "learning_rate": 6.030303030303031e-06,
+      "loss": 0.3886,
+      "step": 199
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.8393324043112065,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.3224,
+      "step": 200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.351476056036347,
+      "learning_rate": 6.090909090909092e-06,
+      "loss": 0.3879,
+      "step": 201
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.2444602461543965,
+      "learning_rate": 6.121212121212121e-06,
+      "loss": 0.3782,
+      "step": 202
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.418002596987469,
+      "learning_rate": 6.151515151515152e-06,
+      "loss": 0.3348,
+      "step": 203
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.368707185290135,
+      "learning_rate": 6.181818181818182e-06,
+      "loss": 0.3974,
+      "step": 204
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.268123765918797,
+      "learning_rate": 6.212121212121213e-06,
+      "loss": 0.3684,
+      "step": 205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.381903351059237,
+      "learning_rate": 6.2424242424242434e-06,
+      "loss": 0.4258,
+      "step": 206
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.016630412052827,
+      "learning_rate": 6.2727272727272734e-06,
+      "loss": 0.3784,
+      "step": 207
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.589692142948399,
+      "learning_rate": 6.303030303030303e-06,
+      "loss": 0.3613,
+      "step": 208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.08075316271684,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 0.3641,
+      "step": 209
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.737607482154883,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.3245,
+      "step": 210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.764211700748343,
+      "learning_rate": 6.393939393939394e-06,
+      "loss": 0.3448,
+      "step": 211
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.144947326785369,
+      "learning_rate": 6.424242424242425e-06,
+      "loss": 0.4126,
+      "step": 212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.929264734380186,
+      "learning_rate": 6.454545454545456e-06,
+      "loss": 0.3385,
+      "step": 213
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.659585273092885,
+      "learning_rate": 6.484848484848485e-06,
+      "loss": 0.3543,
+      "step": 214
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.592493126489269,
+      "learning_rate": 6.515151515151516e-06,
+      "loss": 0.3184,
+      "step": 215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.260498402630153,
+      "learning_rate": 6.545454545454546e-06,
+      "loss": 0.4262,
+      "step": 216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.420814786187238,
+      "learning_rate": 6.575757575757577e-06,
+      "loss": 0.3014,
+      "step": 217
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.408942350340549,
+      "learning_rate": 6.606060606060607e-06,
+      "loss": 0.4314,
+      "step": 218
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.73209243587902,
+      "learning_rate": 6.6363636363636375e-06,
+      "loss": 0.3573,
+      "step": 219
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.7639149056367724,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.3731,
+      "step": 220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.74487748529022,
+      "learning_rate": 6.6969696969696975e-06,
+      "loss": 0.3848,
+      "step": 221
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.912156430473806,
+      "learning_rate": 6.7272727272727275e-06,
+      "loss": 0.3922,
+      "step": 222
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.490592197956874,
+      "learning_rate": 6.757575757575758e-06,
+      "loss": 0.3785,
+      "step": 223
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.8134193025728775,
+      "learning_rate": 6.787878787878789e-06,
+      "loss": 0.4372,
+      "step": 224
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.899403009837595,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 0.3728,
+      "step": 225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.162684147179018,
+      "learning_rate": 6.848484848484849e-06,
+      "loss": 0.3692,
+      "step": 226
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.405595687845643,
+      "learning_rate": 6.878787878787879e-06,
+      "loss": 0.3572,
+      "step": 227
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.756466609168483,
+      "learning_rate": 6.90909090909091e-06,
+      "loss": 0.3668,
+      "step": 228
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.583332237764576,
+      "learning_rate": 6.93939393939394e-06,
+      "loss": 0.325,
+      "step": 229
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.454990683408617,
+      "learning_rate": 6.969696969696971e-06,
+      "loss": 0.3558,
+      "step": 230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.627618101422334,
+      "learning_rate": 7e-06,
+      "loss": 0.3468,
+      "step": 231
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.6688383247015945,
+      "learning_rate": 7.030303030303031e-06,
+      "loss": 0.3307,
+      "step": 232
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.744443819557828,
+      "learning_rate": 7.060606060606061e-06,
+      "loss": 0.4343,
+      "step": 233
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.475080534609491,
+      "learning_rate": 7.0909090909090916e-06,
+      "loss": 0.3784,
+      "step": 234
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.089697275448076,
+      "learning_rate": 7.121212121212122e-06,
+      "loss": 0.3448,
+      "step": 235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.684068430012713,
+      "learning_rate": 7.151515151515152e-06,
+      "loss": 0.3515,
+      "step": 236
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.1266847779895715,
+      "learning_rate": 7.181818181818182e-06,
+      "loss": 0.4246,
+      "step": 237
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.640950040755782,
+      "learning_rate": 7.212121212121212e-06,
+      "loss": 0.3386,
+      "step": 238
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.3941029283087545,
+      "learning_rate": 7.242424242424243e-06,
+      "loss": 0.3164,
+      "step": 239
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.586561657989462,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.353,
+      "step": 240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.455869785834386,
+      "learning_rate": 7.303030303030304e-06,
+      "loss": 0.3816,
+      "step": 241
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.624605507469806,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.3219,
+      "step": 242
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.668622094803916,
+      "learning_rate": 7.363636363636364e-06,
+      "loss": 0.37,
+      "step": 243
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.0291412278716905,
+      "learning_rate": 7.393939393939395e-06,
+      "loss": 0.423,
+      "step": 244
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.399242096008631,
+      "learning_rate": 7.424242424242425e-06,
+      "loss": 0.3091,
+      "step": 245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.387874066312647,
+      "learning_rate": 7.454545454545456e-06,
+      "loss": 0.3559,
+      "step": 246
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.3071504268861815,
+      "learning_rate": 7.484848484848486e-06,
+      "loss": 0.3308,
+      "step": 247
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.704741476990014,
+      "learning_rate": 7.515151515151516e-06,
+      "loss": 0.364,
+      "step": 248
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.656372955262324,
+      "learning_rate": 7.545454545454546e-06,
+      "loss": 0.364,
+      "step": 249
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.84504276054378,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.3884,
+      "step": 250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.595365561220155,
+      "learning_rate": 7.606060606060606e-06,
+      "loss": 0.3817,
+      "step": 251
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.832724610279132,
+      "learning_rate": 7.636363636363638e-06,
+      "loss": 0.3586,
+      "step": 252
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.105327183133213,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 0.3436,
+      "step": 253
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.6333247611294315,
+      "learning_rate": 7.696969696969696e-06,
+      "loss": 0.3776,
+      "step": 254
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.8923194405407715,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 0.3456,
+      "step": 255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.654324255061061,
+      "learning_rate": 7.757575757575758e-06,
+      "loss": 0.3586,
+      "step": 256
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.216031059077568,
+      "learning_rate": 7.787878787878789e-06,
+      "loss": 0.33,
+      "step": 257
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.3421184941495445,
+      "learning_rate": 7.81818181818182e-06,
+      "loss": 0.3639,
+      "step": 258
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.7246835791480635,
+      "learning_rate": 7.848484848484849e-06,
+      "loss": 0.3668,
+      "step": 259
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.216895871590719,
+      "learning_rate": 7.87878787878788e-06,
+      "loss": 0.3334,
+      "step": 260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.865126708149083,
+      "learning_rate": 7.909090909090909e-06,
+      "loss": 0.418,
+      "step": 261
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.5352477939235,
+      "learning_rate": 7.93939393939394e-06,
+      "loss": 0.3544,
+      "step": 262
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.43017485905463,
+      "learning_rate": 7.96969696969697e-06,
+      "loss": 0.4023,
+      "step": 263
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.361207498380666,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.346,
+      "step": 264
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.2962554393520325,
+      "learning_rate": 8.03030303030303e-06,
+      "loss": 0.3771,
+      "step": 265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.542847793967784,
+      "learning_rate": 8.060606060606061e-06,
+      "loss": 0.3545,
+      "step": 266
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.754720356123425,
+      "learning_rate": 8.090909090909092e-06,
+      "loss": 0.3295,
+      "step": 267
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.753606387063651,
+      "learning_rate": 8.121212121212121e-06,
+      "loss": 0.3839,
+      "step": 268
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.819196567998008,
+      "learning_rate": 8.151515151515152e-06,
+      "loss": 0.3751,
+      "step": 269
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.270059328658117,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.3237,
+      "step": 270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.478757758383395,
+      "learning_rate": 8.212121212121212e-06,
+      "loss": 0.3253,
+      "step": 271
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.855333213634504,
+      "learning_rate": 8.242424242424243e-06,
+      "loss": 0.4513,
+      "step": 272
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.496352026923094,
+      "learning_rate": 8.272727272727274e-06,
+      "loss": 0.3842,
+      "step": 273
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.1830239779035745,
+      "learning_rate": 8.303030303030305e-06,
+      "loss": 0.3465,
+      "step": 274
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.506639392867929,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.3297,
+      "step": 275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.718371369676239,
+      "learning_rate": 8.363636363636365e-06,
+      "loss": 0.4098,
+      "step": 276
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.80973464900621,
+      "learning_rate": 8.393939393939394e-06,
+      "loss": 0.3611,
+      "step": 277
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.029258451064713,
+      "learning_rate": 8.424242424242425e-06,
+      "loss": 0.3592,
+      "step": 278
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.589949492587676,
+      "learning_rate": 8.454545454545455e-06,
+      "loss": 0.3478,
+      "step": 279
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.547167896276771,
+      "learning_rate": 8.484848484848486e-06,
+      "loss": 0.3337,
+      "step": 280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.585228209009267,
+      "learning_rate": 8.515151515151517e-06,
+      "loss": 0.3806,
+      "step": 281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.4117849020954765,
+      "learning_rate": 8.545454545454546e-06,
+      "loss": 0.3307,
+      "step": 282
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.063490479238941,
+      "learning_rate": 8.575757575757575e-06,
+      "loss": 0.3744,
+      "step": 283
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.714007266776934,
+      "learning_rate": 8.606060606060606e-06,
+      "loss": 0.3748,
+      "step": 284
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.568036150462816,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 0.3623,
+      "step": 285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.416935840344904,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.3391,
+      "step": 286
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.214651700781016,
+      "learning_rate": 8.696969696969699e-06,
+      "loss": 0.3329,
+      "step": 287
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.5773373099013455,
+      "learning_rate": 8.727272727272728e-06,
+      "loss": 0.3795,
+      "step": 288
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.34543023429054,
+      "learning_rate": 8.757575757575759e-06,
+      "loss": 0.3605,
+      "step": 289
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.675123049567969,
+      "learning_rate": 8.787878787878788e-06,
+      "loss": 0.431,
+      "step": 290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.560810561829422,
+      "learning_rate": 8.818181818181819e-06,
+      "loss": 0.3418,
+      "step": 291
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.415951836280444,
+      "learning_rate": 8.84848484848485e-06,
+      "loss": 0.3237,
+      "step": 292
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.195433119296457,
+      "learning_rate": 8.87878787878788e-06,
+      "loss": 0.3285,
+      "step": 293
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.347089817177498,
+      "learning_rate": 8.90909090909091e-06,
+      "loss": 0.3367,
+      "step": 294
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.896486143791702,
+      "learning_rate": 8.93939393939394e-06,
+      "loss": 0.3859,
+      "step": 295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.643803373471152,
+      "learning_rate": 8.969696969696971e-06,
+      "loss": 0.3505,
+      "step": 296
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.519047240443195,
+      "learning_rate": 9e-06,
+      "loss": 0.3372,
+      "step": 297
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.6753265490756535,
+      "learning_rate": 9.030303030303031e-06,
+      "loss": 0.3673,
+      "step": 298
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.211958443926448,
+      "learning_rate": 9.06060606060606e-06,
+      "loss": 0.3378,
+      "step": 299
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.469853288176385,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.3728,
+      "step": 300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.56166594783095,
+      "learning_rate": 9.121212121212122e-06,
+      "loss": 0.3689,
+      "step": 301
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.219168289352615,
+      "learning_rate": 9.151515151515153e-06,
+      "loss": 0.3646,
+      "step": 302
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.241187922680894,
+      "learning_rate": 9.181818181818184e-06,
+      "loss": 0.3405,
+      "step": 303
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.458178869933374,
+      "learning_rate": 9.212121212121213e-06,
+      "loss": 0.3507,
+      "step": 304
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.29934909908872,
+      "learning_rate": 9.242424242424244e-06,
+      "loss": 0.3287,
+      "step": 305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.923389714856165,
+      "learning_rate": 9.272727272727273e-06,
+      "loss": 0.4204,
+      "step": 306
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.47561834607557,
+      "learning_rate": 9.303030303030303e-06,
+      "loss": 0.3552,
+      "step": 307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.438159776380354,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.3242,
+      "step": 308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.080568527786985,
+      "learning_rate": 9.363636363636365e-06,
+      "loss": 0.3344,
+      "step": 309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.295039432974613,
+      "learning_rate": 9.393939393939396e-06,
+      "loss": 0.3392,
+      "step": 310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.2027332270961715,
+      "learning_rate": 9.424242424242425e-06,
+      "loss": 0.3182,
+      "step": 311
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.289971974457048,
+      "learning_rate": 9.454545454545456e-06,
+      "loss": 0.3451,
+      "step": 312
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.614472867798603,
+      "learning_rate": 9.484848484848485e-06,
+      "loss": 0.3813,
+      "step": 313
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.055910292621515,
+      "learning_rate": 9.515151515151516e-06,
+      "loss": 0.3189,
+      "step": 314
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.245295970012739,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 0.3289,
+      "step": 315
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.465327940507587,
+      "learning_rate": 9.575757575757576e-06,
+      "loss": 0.3532,
+      "step": 316
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.564700676924458,
+      "learning_rate": 9.606060606060607e-06,
+      "loss": 0.3611,
+      "step": 317
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.512417904019265,
+      "learning_rate": 9.636363636363638e-06,
+      "loss": 0.3892,
+      "step": 318
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.183674694569067,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 0.3209,
+      "step": 319
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.101289997697547,
+      "learning_rate": 9.696969696969698e-06,
+      "loss": 0.3422,
+      "step": 320
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.310839907356913,
+      "learning_rate": 9.727272727272728e-06,
+      "loss": 0.3293,
+      "step": 321
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.9054376600635585,
+      "learning_rate": 9.757575757575758e-06,
+      "loss": 0.3323,
+      "step": 322
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.888021079794081,
+      "learning_rate": 9.787878787878788e-06,
+      "loss": 0.3044,
+      "step": 323
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.273469837470886,
+      "learning_rate": 9.81818181818182e-06,
+      "loss": 0.3517,
+      "step": 324
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7144185914422505,
+      "learning_rate": 9.84848484848485e-06,
+      "loss": 0.3049,
+      "step": 325
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.603880691659557,
+      "learning_rate": 9.87878787878788e-06,
+      "loss": 0.4109,
+      "step": 326
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.334268112001053,
+      "learning_rate": 9.90909090909091e-06,
+      "loss": 0.3464,
+      "step": 327
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.341108772132298,
+      "learning_rate": 9.939393939393939e-06,
+      "loss": 0.3451,
+      "step": 328
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.160707272996998,
+      "learning_rate": 9.96969696969697e-06,
+      "loss": 0.3359,
+      "step": 329
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.673438573112379,
+      "learning_rate": 1e-05,
+      "loss": 0.4091,
+      "step": 330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.4089667880024255,
+      "learning_rate": 9.999999782826503e-06,
+      "loss": 0.3628,
+      "step": 331
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9446848867881914,
+      "learning_rate": 9.999999131306029e-06,
+      "loss": 0.3335,
+      "step": 332
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.435323063815749,
+      "learning_rate": 9.999998045438632e-06,
+      "loss": 0.3643,
+      "step": 333
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.320827685063797,
+      "learning_rate": 9.999996525224412e-06,
+      "loss": 0.3329,
+      "step": 334
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.36265608974536,
+      "learning_rate": 9.999994570663496e-06,
+      "loss": 0.3461,
+      "step": 335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.2592983722217,
+      "learning_rate": 9.999992181756056e-06,
+      "loss": 0.3766,
+      "step": 336
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.79646378296195,
+      "learning_rate": 9.999989358502298e-06,
+      "loss": 0.3659,
+      "step": 337
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.400343051257205,
+      "learning_rate": 9.99998610090247e-06,
+      "loss": 0.3507,
+      "step": 338
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.827448978040931,
+      "learning_rate": 9.999982408956851e-06,
+      "loss": 0.3366,
+      "step": 339
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.4231733343936055,
+      "learning_rate": 9.999978282665768e-06,
+      "loss": 0.346,
+      "step": 340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.370154878114904,
+      "learning_rate": 9.999973722029575e-06,
+      "loss": 0.4103,
+      "step": 341
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.20966197134485,
+      "learning_rate": 9.999968727048667e-06,
+      "loss": 0.3531,
+      "step": 342
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.448450851329632,
+      "learning_rate": 9.999963297723481e-06,
+      "loss": 0.3629,
+      "step": 343
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.6259560351897155,
+      "learning_rate": 9.999957434054487e-06,
+      "loss": 0.3382,
+      "step": 344
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.547978010284929,
+      "learning_rate": 9.999951136042194e-06,
+      "loss": 0.4058,
+      "step": 345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.142281305157911,
+      "learning_rate": 9.99994440368715e-06,
+      "loss": 0.3545,
+      "step": 346
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.740369583978333,
+      "learning_rate": 9.99993723698994e-06,
+      "loss": 0.3753,
+      "step": 347
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.343321713377683,
+      "learning_rate": 9.999929635951186e-06,
+      "loss": 0.3887,
+      "step": 348
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.2010583750758235,
+      "learning_rate": 9.99992160057155e-06,
+      "loss": 0.3212,
+      "step": 349
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.049524747453005,
+      "learning_rate": 9.999913130851726e-06,
+      "loss": 0.3324,
+      "step": 350
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.050123203528034,
+      "learning_rate": 9.999904226792453e-06,
+      "loss": 0.3308,
+      "step": 351
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.009035365811596,
+      "learning_rate": 9.999894888394505e-06,
+      "loss": 0.3417,
+      "step": 352
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.294750340037094,
+      "learning_rate": 9.999885115658693e-06,
+      "loss": 0.3578,
+      "step": 353
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.77967792592922,
+      "learning_rate": 9.999874908585864e-06,
+      "loss": 0.4079,
+      "step": 354
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.674857837086743,
+      "learning_rate": 9.999864267176904e-06,
+      "loss": 0.4115,
+      "step": 355
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8718618109598215,
+      "learning_rate": 9.999853191432741e-06,
+      "loss": 0.3152,
+      "step": 356
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.331597939165435,
+      "learning_rate": 9.999841681354334e-06,
+      "loss": 0.3871,
+      "step": 357
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.277256615205333,
+      "learning_rate": 9.999829736942686e-06,
+      "loss": 0.3239,
+      "step": 358
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.149331062618436,
+      "learning_rate": 9.999817358198831e-06,
+      "loss": 0.3668,
+      "step": 359
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.315716480649614,
+      "learning_rate": 9.999804545123847e-06,
+      "loss": 0.3782,
+      "step": 360
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.164901665057027,
+      "learning_rate": 9.999791297718844e-06,
+      "loss": 0.3532,
+      "step": 361
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.7676989097225055,
+      "learning_rate": 9.999777615984978e-06,
+      "loss": 0.3412,
+      "step": 362
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.336317032183553,
+      "learning_rate": 9.999763499923432e-06,
+      "loss": 0.3552,
+      "step": 363
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.334603725462001,
+      "learning_rate": 9.999748949535436e-06,
+      "loss": 0.3554,
+      "step": 364
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.131537987396616,
+      "learning_rate": 9.999733964822252e-06,
+      "loss": 0.3322,
+      "step": 365
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.513789284576847,
+      "learning_rate": 9.999718545785183e-06,
+      "loss": 0.3425,
+      "step": 366
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.170596768573397,
+      "learning_rate": 9.999702692425567e-06,
+      "loss": 0.3318,
+      "step": 367
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.391527149872584,
+      "learning_rate": 9.999686404744782e-06,
+      "loss": 0.3528,
+      "step": 368
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.025111572088874,
+      "learning_rate": 9.999669682744245e-06,
+      "loss": 0.3607,
+      "step": 369
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.515811782591134,
+      "learning_rate": 9.999652526425404e-06,
+      "loss": 0.3738,
+      "step": 370
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.376679293359408,
+      "learning_rate": 9.999634935789753e-06,
+      "loss": 0.3696,
+      "step": 371
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.669377622956597,
+      "learning_rate": 9.999616910838818e-06,
+      "loss": 0.3421,
+      "step": 372
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9162994981984234,
+      "learning_rate": 9.999598451574167e-06,
+      "loss": 0.364,
+      "step": 373
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.517952400999471,
+      "learning_rate": 9.999579557997402e-06,
+      "loss": 0.3449,
+      "step": 374
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.467483484309139,
+      "learning_rate": 9.999560230110165e-06,
+      "loss": 0.3194,
+      "step": 375
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.6303457867490785,
+      "learning_rate": 9.999540467914133e-06,
+      "loss": 0.3819,
+      "step": 376
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.3445923038232035,
+      "learning_rate": 9.999520271411026e-06,
+      "loss": 0.3334,
+      "step": 377
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.1673411257928565,
+      "learning_rate": 9.999499640602597e-06,
+      "loss": 0.3084,
+      "step": 378
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9273173382754276,
+      "learning_rate": 9.999478575490637e-06,
+      "loss": 0.3342,
+      "step": 379
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.07461273599012,
+      "learning_rate": 9.999457076076978e-06,
+      "loss": 0.3015,
+      "step": 380
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.594858339869631,
+      "learning_rate": 9.999435142363484e-06,
+      "loss": 0.3753,
+      "step": 381
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.10065742206399,
+      "learning_rate": 9.999412774352064e-06,
+      "loss": 0.3432,
+      "step": 382
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.210509226933416,
+      "learning_rate": 9.99938997204466e-06,
+      "loss": 0.3507,
+      "step": 383
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9802053693025004,
+      "learning_rate": 9.999366735443255e-06,
+      "loss": 0.3413,
+      "step": 384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.809766404833952,
+      "learning_rate": 9.999343064549862e-06,
+      "loss": 0.3257,
+      "step": 385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.008566832373406,
+      "learning_rate": 9.999318959366543e-06,
+      "loss": 0.335,
+      "step": 386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.539104446749643,
+      "learning_rate": 9.999294419895389e-06,
+      "loss": 0.3836,
+      "step": 387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.30428666363456,
+      "learning_rate": 9.999269446138533e-06,
+      "loss": 0.3678,
+      "step": 388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9484798240843726,
+      "learning_rate": 9.999244038098144e-06,
+      "loss": 0.3215,
+      "step": 389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.886010977410963,
+      "learning_rate": 9.999218195776428e-06,
+      "loss": 0.3347,
+      "step": 390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.361829511204051,
+      "learning_rate": 9.99919191917563e-06,
+      "loss": 0.3647,
+      "step": 391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.04963150428601,
+      "learning_rate": 9.999165208298034e-06,
+      "loss": 0.333,
+      "step": 392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.979505051692121,
+      "learning_rate": 9.999138063145962e-06,
+      "loss": 0.2983,
+      "step": 393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7829758596070415,
+      "learning_rate": 9.999110483721767e-06,
+      "loss": 0.3342,
+      "step": 394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.145699243065273,
+      "learning_rate": 9.99908247002785e-06,
+      "loss": 0.313,
+      "step": 395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.00672408957499,
+      "learning_rate": 9.999054022066643e-06,
+      "loss": 0.3357,
+      "step": 396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7406060827271075,
+      "learning_rate": 9.999025139840615e-06,
+      "loss": 0.3206,
+      "step": 397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.585124069819002,
+      "learning_rate": 9.998995823352276e-06,
+      "loss": 0.4021,
+      "step": 398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.209054176443017,
+      "learning_rate": 9.998966072604175e-06,
+      "loss": 0.3423,
+      "step": 399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.12076488626069,
+      "learning_rate": 9.998935887598894e-06,
+      "loss": 0.36,
+      "step": 400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.101877867787852,
+      "learning_rate": 9.998905268339056e-06,
+      "loss": 0.3212,
+      "step": 401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8115206401654493,
+      "learning_rate": 9.99887421482732e-06,
+      "loss": 0.2913,
+      "step": 402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.26716023870372,
+      "learning_rate": 9.998842727066385e-06,
+      "loss": 0.3433,
+      "step": 403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.186709441045844,
+      "learning_rate": 9.998810805058986e-06,
+      "loss": 0.335,
+      "step": 404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.815787007607476,
+      "learning_rate": 9.998778448807895e-06,
+      "loss": 0.3196,
+      "step": 405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.0619966376546826,
+      "learning_rate": 9.998745658315924e-06,
+      "loss": 0.2965,
+      "step": 406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.021023049120253,
+      "learning_rate": 9.998712433585919e-06,
+      "loss": 0.3255,
+      "step": 407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.195956615305363,
+      "learning_rate": 9.998678774620771e-06,
+      "loss": 0.3573,
+      "step": 408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.049007998677101,
+      "learning_rate": 9.9986446814234e-06,
+      "loss": 0.3219,
+      "step": 409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.123368548044139,
+      "learning_rate": 9.998610153996768e-06,
+      "loss": 0.3692,
+      "step": 410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9411023784619568,
+      "learning_rate": 9.998575192343877e-06,
+      "loss": 0.3117,
+      "step": 411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.212240106003559,
+      "learning_rate": 9.998539796467761e-06,
+      "loss": 0.3124,
+      "step": 412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9840644647494647,
+      "learning_rate": 9.998503966371496e-06,
+      "loss": 0.3674,
+      "step": 413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.1834298094248235,
+      "learning_rate": 9.998467702058194e-06,
+      "loss": 0.3571,
+      "step": 414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.165577776843734,
+      "learning_rate": 9.998431003531008e-06,
+      "loss": 0.3215,
+      "step": 415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6707088750477324,
+      "learning_rate": 9.99839387079312e-06,
+      "loss": 0.3326,
+      "step": 416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6256633108539034,
+      "learning_rate": 9.998356303847764e-06,
+      "loss": 0.3201,
+      "step": 417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.604686435985164,
+      "learning_rate": 9.998318302698198e-06,
+      "loss": 0.3844,
+      "step": 418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.014397880699301,
+      "learning_rate": 9.998279867347723e-06,
+      "loss": 0.3235,
+      "step": 419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.194092429654693,
+      "learning_rate": 9.998240997799677e-06,
+      "loss": 0.3497,
+      "step": 420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8496118027336963,
+      "learning_rate": 9.998201694057441e-06,
+      "loss": 0.3697,
+      "step": 421
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.881436284637087,
+      "learning_rate": 9.998161956124428e-06,
+      "loss": 0.3712,
+      "step": 422
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.046258792596131,
+      "learning_rate": 9.998121784004086e-06,
+      "loss": 0.4064,
+      "step": 423
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.165185372475956,
+      "learning_rate": 9.998081177699909e-06,
+      "loss": 0.3411,
+      "step": 424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8882062893973273,
+      "learning_rate": 9.998040137215423e-06,
+      "loss": 0.2749,
+      "step": 425
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.1282734757014845,
+      "learning_rate": 9.997998662554194e-06,
+      "loss": 0.352,
+      "step": 426
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8967391743955373,
+      "learning_rate": 9.997956753719821e-06,
+      "loss": 0.3202,
+      "step": 427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.435224973732747,
+      "learning_rate": 9.99791441071595e-06,
+      "loss": 0.3269,
+      "step": 428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9480633469602116,
+      "learning_rate": 9.997871633546257e-06,
+      "loss": 0.3644,
+      "step": 429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9663656432995764,
+      "learning_rate": 9.997828422214458e-06,
+      "loss": 0.3232,
+      "step": 430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8201026443227786,
+      "learning_rate": 9.997784776724306e-06,
+      "loss": 0.317,
+      "step": 431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9195793023804253,
+      "learning_rate": 9.997740697079595e-06,
+      "loss": 0.3004,
+      "step": 432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.480878096091991,
+      "learning_rate": 9.997696183284148e-06,
+      "loss": 0.3938,
+      "step": 433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.675436703113189,
+      "learning_rate": 9.997651235341842e-06,
+      "loss": 0.3086,
+      "step": 434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7856809338775257,
+      "learning_rate": 9.997605853256572e-06,
+      "loss": 0.3335,
+      "step": 435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5992673548883913,
+      "learning_rate": 9.997560037032283e-06,
+      "loss": 0.2924,
+      "step": 436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.189646847760906,
+      "learning_rate": 9.997513786672959e-06,
+      "loss": 0.3634,
+      "step": 437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.768981956623533,
+      "learning_rate": 9.997467102182614e-06,
+      "loss": 0.3204,
+      "step": 438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.725180570879781,
+      "learning_rate": 9.997419983565304e-06,
+      "loss": 0.3317,
+      "step": 439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8185595584113607,
+      "learning_rate": 9.997372430825125e-06,
+      "loss": 0.3579,
+      "step": 440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7944729920242954,
+      "learning_rate": 9.997324443966203e-06,
+      "loss": 0.3194,
+      "step": 441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9814502632301445,
+      "learning_rate": 9.997276022992709e-06,
+      "loss": 0.3642,
+      "step": 442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.759728753298408,
+      "learning_rate": 9.997227167908849e-06,
+      "loss": 0.3272,
+      "step": 443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.76763352101135,
+      "learning_rate": 9.99717787871887e-06,
+      "loss": 0.3346,
+      "step": 444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5291936249395457,
+      "learning_rate": 9.997128155427047e-06,
+      "loss": 0.3079,
+      "step": 445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.971559126992198,
+      "learning_rate": 9.997077998037707e-06,
+      "loss": 0.3383,
+      "step": 446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7248181192929812,
+      "learning_rate": 9.997027406555201e-06,
+      "loss": 0.3018,
+      "step": 447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5357343304449085,
+      "learning_rate": 9.996976380983927e-06,
+      "loss": 0.3082,
+      "step": 448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.538354569891425,
+      "learning_rate": 9.99692492132832e-06,
+      "loss": 0.3561,
+      "step": 449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6579649325136776,
+      "learning_rate": 9.996873027592844e-06,
+      "loss": 0.3217,
+      "step": 450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6764037888712315,
+      "learning_rate": 9.99682069978201e-06,
+      "loss": 0.3061,
+      "step": 451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.727210693944586,
+      "learning_rate": 9.996767937900367e-06,
+      "loss": 0.3459,
+      "step": 452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.610670266073993,
+      "learning_rate": 9.996714741952493e-06,
+      "loss": 0.3041,
+      "step": 453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6911017688401393,
+      "learning_rate": 9.99666111194301e-06,
+      "loss": 0.3275,
+      "step": 454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7116898131973532,
+      "learning_rate": 9.996607047876582e-06,
+      "loss": 0.3161,
+      "step": 455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.051156243428742,
+      "learning_rate": 9.9965525497579e-06,
+      "loss": 0.3463,
+      "step": 456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8719930210911286,
+      "learning_rate": 9.9964976175917e-06,
+      "loss": 0.3241,
+      "step": 457
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8748489302929148,
+      "learning_rate": 9.996442251382754e-06,
+      "loss": 0.3513,
+      "step": 458
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.533801535678403,
+      "learning_rate": 9.996386451135871e-06,
+      "loss": 0.3055,
+      "step": 459
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.100961801715447,
+      "learning_rate": 9.996330216855901e-06,
+      "loss": 0.358,
+      "step": 460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6075421336937636,
+      "learning_rate": 9.996273548547724e-06,
+      "loss": 0.3037,
+      "step": 461
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.521861829096483,
+      "learning_rate": 9.996216446216267e-06,
+      "loss": 0.3395,
+      "step": 462
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9016695749723147,
+      "learning_rate": 9.99615890986649e-06,
+      "loss": 0.3351,
+      "step": 463
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.835776071517302,
+      "learning_rate": 9.996100939503387e-06,
+      "loss": 0.3165,
+      "step": 464
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.427218450275447,
+      "learning_rate": 9.996042535132001e-06,
+      "loss": 0.3049,
+      "step": 465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9211356318146557,
+      "learning_rate": 9.9959836967574e-06,
+      "loss": 0.279,
+      "step": 466
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.759359701375455,
+      "learning_rate": 9.995924424384697e-06,
+      "loss": 0.3253,
+      "step": 467
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.4364516776188925,
+      "learning_rate": 9.995864718019042e-06,
+      "loss": 0.3325,
+      "step": 468
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3747990747650687,
+      "learning_rate": 9.995804577665617e-06,
+      "loss": 0.2497,
+      "step": 469
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.00746261820677,
+      "learning_rate": 9.995744003329655e-06,
+      "loss": 0.3352,
+      "step": 470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.65870207006295,
+      "learning_rate": 9.995682995016409e-06,
+      "loss": 0.3021,
+      "step": 471
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.498382997724767,
+      "learning_rate": 9.995621552731182e-06,
+      "loss": 0.3284,
+      "step": 472
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.553554526451541,
+      "learning_rate": 9.995559676479317e-06,
+      "loss": 0.2751,
+      "step": 473
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.10259191393736,
+      "learning_rate": 9.99549736626618e-06,
+      "loss": 0.3617,
+      "step": 474
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.825525380444603,
+      "learning_rate": 9.995434622097189e-06,
+      "loss": 0.3434,
+      "step": 475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.062851973559195,
+      "learning_rate": 9.995371443977794e-06,
+      "loss": 0.3673,
+      "step": 476
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.708189013193882,
+      "learning_rate": 9.995307831913483e-06,
+      "loss": 0.3416,
+      "step": 477
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.63564165057117,
+      "learning_rate": 9.995243785909782e-06,
+      "loss": 0.3429,
+      "step": 478
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6879984610544487,
+      "learning_rate": 9.995179305972256e-06,
+      "loss": 0.32,
+      "step": 479
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.787211983521054,
+      "learning_rate": 9.995114392106502e-06,
+      "loss": 0.3699,
+      "step": 480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.327141793920121,
+      "learning_rate": 9.995049044318164e-06,
+      "loss": 0.3235,
+      "step": 481
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.833920065352593,
+      "learning_rate": 9.994983262612916e-06,
+      "loss": 0.3613,
+      "step": 482
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.164456973598666,
+      "learning_rate": 9.994917046996472e-06,
+      "loss": 0.3743,
+      "step": 483
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.743020193674409,
+      "learning_rate": 9.994850397474588e-06,
+      "loss": 0.3439,
+      "step": 484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.515403754484158,
+      "learning_rate": 9.994783314053047e-06,
+      "loss": 0.2701,
+      "step": 485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.048462614319834,
+      "learning_rate": 9.994715796737683e-06,
+      "loss": 0.3045,
+      "step": 486
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.188690490763364,
+      "learning_rate": 9.994647845534357e-06,
+      "loss": 0.3726,
+      "step": 487
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.307582758326019,
+      "learning_rate": 9.994579460448975e-06,
+      "loss": 0.3283,
+      "step": 488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.922971038535889,
+      "learning_rate": 9.994510641487477e-06,
+      "loss": 0.3773,
+      "step": 489
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7130894724742296,
+      "learning_rate": 9.994441388655837e-06,
+      "loss": 0.3614,
+      "step": 490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5745668345635844,
+      "learning_rate": 9.994371701960077e-06,
+      "loss": 0.3129,
+      "step": 491
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.06846832131553,
+      "learning_rate": 9.994301581406247e-06,
+      "loss": 0.3356,
+      "step": 492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.051494997930581,
+      "learning_rate": 9.994231027000439e-06,
+      "loss": 0.3086,
+      "step": 493
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.45797411695485,
+      "learning_rate": 9.994160038748783e-06,
+      "loss": 0.3053,
+      "step": 494
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8783408338046015,
+      "learning_rate": 9.994088616657445e-06,
+      "loss": 0.3127,
+      "step": 495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7568169990750424,
+      "learning_rate": 9.99401676073263e-06,
+      "loss": 0.2881,
+      "step": 496
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7328241539684694,
+      "learning_rate": 9.993944470980576e-06,
+      "loss": 0.3311,
+      "step": 497
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8466499584832587,
+      "learning_rate": 9.99387174740757e-06,
+      "loss": 0.3281,
+      "step": 498
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7481130912508873,
+      "learning_rate": 9.993798590019924e-06,
+      "loss": 0.3589,
+      "step": 499
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5427552057156704,
+      "learning_rate": 9.993724998823995e-06,
+      "loss": 0.305,
+      "step": 500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.584783558606137,
+      "learning_rate": 9.993650973826177e-06,
+      "loss": 0.3003,
+      "step": 501
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.312690359942732,
+      "learning_rate": 9.993576515032896e-06,
+      "loss": 0.3202,
+      "step": 502
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.023569302547953,
+      "learning_rate": 9.993501622450626e-06,
+      "loss": 0.3255,
+      "step": 503
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.959083722012039,
+      "learning_rate": 9.99342629608587e-06,
+      "loss": 0.3449,
+      "step": 504
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6409605833348713,
+      "learning_rate": 9.993350535945172e-06,
+      "loss": 0.3122,
+      "step": 505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.881365220819395,
+      "learning_rate": 9.993274342035111e-06,
+      "loss": 0.3485,
+      "step": 506
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.875874738456161,
+      "learning_rate": 9.99319771436231e-06,
+      "loss": 0.3314,
+      "step": 507
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6973392319182214,
+      "learning_rate": 9.993120652933424e-06,
+      "loss": 0.3234,
+      "step": 508
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.3021385034497746,
+      "learning_rate": 9.993043157755145e-06,
+      "loss": 0.3478,
+      "step": 509
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.990629222446292,
+      "learning_rate": 9.992965228834208e-06,
+      "loss": 0.3751,
+      "step": 510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.790412874579726,
+      "learning_rate": 9.99288686617738e-06,
+      "loss": 0.3689,
+      "step": 511
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.011892983516215,
+      "learning_rate": 9.992808069791472e-06,
+      "loss": 0.3274,
+      "step": 512
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7001945245196604,
+      "learning_rate": 9.992728839683324e-06,
+      "loss": 0.289,
+      "step": 513
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9138478787249467,
+      "learning_rate": 9.992649175859822e-06,
+      "loss": 0.3401,
+      "step": 514
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.463325827559996,
+      "learning_rate": 9.992569078327886e-06,
+      "loss": 0.3335,
+      "step": 515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7471733764932624,
+      "learning_rate": 9.992488547094474e-06,
+      "loss": 0.3605,
+      "step": 516
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5089807155299573,
+      "learning_rate": 9.992407582166582e-06,
+      "loss": 0.3054,
+      "step": 517
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.693730588351396,
+      "learning_rate": 9.992326183551242e-06,
+      "loss": 0.2857,
+      "step": 518
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7079229821251385,
+      "learning_rate": 9.992244351255526e-06,
+      "loss": 0.316,
+      "step": 519
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.67808542959306,
+      "learning_rate": 9.992162085286543e-06,
+      "loss": 0.2683,
+      "step": 520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.809588995572531,
+      "learning_rate": 9.992079385651439e-06,
+      "loss": 0.3048,
+      "step": 521
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9332022803075697,
+      "learning_rate": 9.9919962523574e-06,
+      "loss": 0.3053,
+      "step": 522
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6569290266564947,
+      "learning_rate": 9.991912685411642e-06,
+      "loss": 0.2779,
+      "step": 523
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9510493670523834,
+      "learning_rate": 9.99182868482143e-06,
+      "loss": 0.3166,
+      "step": 524
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.1862387312553935,
+      "learning_rate": 9.99174425059406e-06,
+      "loss": 0.3299,
+      "step": 525
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9389947846094957,
+      "learning_rate": 9.991659382736864e-06,
+      "loss": 0.3245,
+      "step": 526
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.101612951270305,
+      "learning_rate": 9.991574081257219e-06,
+      "loss": 0.3047,
+      "step": 527
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.566958804625748,
+      "learning_rate": 9.99148834616253e-06,
+      "loss": 0.3102,
+      "step": 528
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.482376189919131,
+      "learning_rate": 9.99140217746025e-06,
+      "loss": 0.322,
+      "step": 529
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9368497200134085,
+      "learning_rate": 9.991315575157861e-06,
+      "loss": 0.3134,
+      "step": 530
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.476457092838848,
+      "learning_rate": 9.991228539262886e-06,
+      "loss": 0.3566,
+      "step": 531
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.137508305832424,
+      "learning_rate": 9.991141069782886e-06,
+      "loss": 0.3556,
+      "step": 532
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.766688493619562,
+      "learning_rate": 9.99105316672546e-06,
+      "loss": 0.3219,
+      "step": 533
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.457024072940931,
+      "learning_rate": 9.990964830098246e-06,
+      "loss": 0.2686,
+      "step": 534
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9299979659182602,
+      "learning_rate": 9.990876059908915e-06,
+      "loss": 0.3281,
+      "step": 535
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5873933053383884,
+      "learning_rate": 9.990786856165178e-06,
+      "loss": 0.3214,
+      "step": 536
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8995754008029153,
+      "learning_rate": 9.990697218874788e-06,
+      "loss": 0.3122,
+      "step": 537
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.100429866752299,
+      "learning_rate": 9.990607148045526e-06,
+      "loss": 0.3295,
+      "step": 538
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.746335512814467,
+      "learning_rate": 9.990516643685222e-06,
+      "loss": 0.3116,
+      "step": 539
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.492998895598886,
+      "learning_rate": 9.990425705801733e-06,
+      "loss": 0.3205,
+      "step": 540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8407713955138734,
+      "learning_rate": 9.990334334402964e-06,
+      "loss": 0.3363,
+      "step": 541
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4686474170387442,
+      "learning_rate": 9.990242529496848e-06,
+      "loss": 0.2929,
+      "step": 542
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.067101646488163,
+      "learning_rate": 9.990150291091363e-06,
+      "loss": 0.3225,
+      "step": 543
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6481539958661005,
+      "learning_rate": 9.990057619194517e-06,
+      "loss": 0.3372,
+      "step": 544
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5511909125879906,
+      "learning_rate": 9.989964513814368e-06,
+      "loss": 0.2825,
+      "step": 545
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7857716862895283,
+      "learning_rate": 9.989870974958997e-06,
+      "loss": 0.3556,
+      "step": 546
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.092788081965213,
+      "learning_rate": 9.989777002636533e-06,
+      "loss": 0.3153,
+      "step": 547
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.104019121056975,
+      "learning_rate": 9.989682596855138e-06,
+      "loss": 0.3178,
+      "step": 548
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.912919278414138,
+      "learning_rate": 9.989587757623015e-06,
+      "loss": 0.3099,
+      "step": 549
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.512015459175421,
+      "learning_rate": 9.9894924849484e-06,
+      "loss": 0.3071,
+      "step": 550
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8098330049714644,
+      "learning_rate": 9.989396778839572e-06,
+      "loss": 0.3322,
+      "step": 551
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6136184825769444,
+      "learning_rate": 9.989300639304843e-06,
+      "loss": 0.2779,
+      "step": 552
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.411006081757144,
+      "learning_rate": 9.989204066352565e-06,
+      "loss": 0.2884,
+      "step": 553
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4235555412349874,
+      "learning_rate": 9.989107059991127e-06,
+      "loss": 0.3342,
+      "step": 554
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4669876679296965,
+      "learning_rate": 9.989009620228957e-06,
+      "loss": 0.2853,
+      "step": 555
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9654896484091946,
+      "learning_rate": 9.988911747074518e-06,
+      "loss": 0.3398,
+      "step": 556
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.618634090408832,
+      "learning_rate": 9.988813440536314e-06,
+      "loss": 0.3031,
+      "step": 557
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.709794299197405,
+      "learning_rate": 9.988714700622882e-06,
+      "loss": 0.3045,
+      "step": 558
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6379338775764603,
+      "learning_rate": 9.988615527342801e-06,
+      "loss": 0.3287,
+      "step": 559
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.906541445398935,
+      "learning_rate": 9.988515920704689e-06,
+      "loss": 0.3446,
+      "step": 560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.802958218449508,
+      "learning_rate": 9.988415880717195e-06,
+      "loss": 0.294,
+      "step": 561
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5750411292773627,
+      "learning_rate": 9.988315407389009e-06,
+      "loss": 0.284,
+      "step": 562
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3997428129594223,
+      "learning_rate": 9.988214500728862e-06,
+      "loss": 0.2854,
+      "step": 563
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.852934108845612,
+      "learning_rate": 9.988113160745519e-06,
+      "loss": 0.3597,
+      "step": 564
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8019680400237608,
+      "learning_rate": 9.988011387447781e-06,
+      "loss": 0.323,
+      "step": 565
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8149719698103,
+      "learning_rate": 9.987909180844491e-06,
+      "loss": 0.3083,
+      "step": 566
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.079548503486936,
+      "learning_rate": 9.987806540944528e-06,
+      "loss": 0.3289,
+      "step": 567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.90149328690206,
+      "learning_rate": 9.987703467756807e-06,
+      "loss": 0.329,
+      "step": 568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.99169371798887,
+      "learning_rate": 9.987599961290283e-06,
+      "loss": 0.3305,
+      "step": 569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2957002191485714,
+      "learning_rate": 9.987496021553946e-06,
+      "loss": 0.2987,
+      "step": 570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3513096485139275,
+      "learning_rate": 9.987391648556828e-06,
+      "loss": 0.3049,
+      "step": 571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.631545057379283,
+      "learning_rate": 9.987286842307991e-06,
+      "loss": 0.2969,
+      "step": 572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3346306865300264,
+      "learning_rate": 9.987181602816545e-06,
+      "loss": 0.3115,
+      "step": 573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.168666835132431,
+      "learning_rate": 9.987075930091629e-06,
+      "loss": 0.3722,
+      "step": 574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.24943490302494,
+      "learning_rate": 9.986969824142424e-06,
+      "loss": 0.2941,
+      "step": 575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5940233382677063,
+      "learning_rate": 9.986863284978143e-06,
+      "loss": 0.3144,
+      "step": 576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.390409293378752,
+      "learning_rate": 9.986756312608048e-06,
+      "loss": 0.3172,
+      "step": 577
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5327688539081157,
+      "learning_rate": 9.986648907041428e-06,
+      "loss": 0.2818,
+      "step": 578
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6394050896129273,
+      "learning_rate": 9.986541068287612e-06,
+      "loss": 0.304,
+      "step": 579
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.940551438373569,
+      "learning_rate": 9.98643279635597e-06,
+      "loss": 0.3581,
+      "step": 580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.69009213992704,
+      "learning_rate": 9.986324091255908e-06,
+      "loss": 0.2925,
+      "step": 581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.95483477723251,
+      "learning_rate": 9.986214952996867e-06,
+      "loss": 0.3411,
+      "step": 582
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.514909402110056,
+      "learning_rate": 9.986105381588329e-06,
+      "loss": 0.2952,
+      "step": 583
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3243367049974073,
+      "learning_rate": 9.985995377039812e-06,
+      "loss": 0.2793,
+      "step": 584
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5604191206416402,
+      "learning_rate": 9.985884939360873e-06,
+      "loss": 0.3374,
+      "step": 585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6168118282133177,
+      "learning_rate": 9.985774068561102e-06,
+      "loss": 0.3224,
+      "step": 586
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4265991333209294,
+      "learning_rate": 9.985662764650138e-06,
+      "loss": 0.2998,
+      "step": 587
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4462723878531967,
+      "learning_rate": 9.98555102763764e-06,
+      "loss": 0.2972,
+      "step": 588
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.987835596012852,
+      "learning_rate": 9.985438857533323e-06,
+      "loss": 0.3782,
+      "step": 589
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5532748451596614,
+      "learning_rate": 9.985326254346928e-06,
+      "loss": 0.287,
+      "step": 590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.381271295289075,
+      "learning_rate": 9.985213218088234e-06,
+      "loss": 0.2783,
+      "step": 591
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5978459141959895,
+      "learning_rate": 9.985099748767065e-06,
+      "loss": 0.3263,
+      "step": 592
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.352022064648451,
+      "learning_rate": 9.984985846393276e-06,
+      "loss": 0.3487,
+      "step": 593
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4412290772492145,
+      "learning_rate": 9.98487151097676e-06,
+      "loss": 0.3153,
+      "step": 594
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7175627996193428,
+      "learning_rate": 9.984756742527451e-06,
+      "loss": 0.3069,
+      "step": 595
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.241936216144905,
+      "learning_rate": 9.98464154105532e-06,
+      "loss": 0.2914,
+      "step": 596
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3858799911347424,
+      "learning_rate": 9.984525906570372e-06,
+      "loss": 0.3229,
+      "step": 597
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.212120349376869,
+      "learning_rate": 9.984409839082654e-06,
+      "loss": 0.3308,
+      "step": 598
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.702787114368071,
+      "learning_rate": 9.984293338602249e-06,
+      "loss": 0.3236,
+      "step": 599
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5629364380620303,
+      "learning_rate": 9.984176405139275e-06,
+      "loss": 0.3187,
+      "step": 600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.51753497626045,
+      "learning_rate": 9.98405903870389e-06,
+      "loss": 0.3297,
+      "step": 601
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.536184171564669,
+      "learning_rate": 9.983941239306291e-06,
+      "loss": 0.2962,
+      "step": 602
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4298149652464835,
+      "learning_rate": 9.983823006956714e-06,
+      "loss": 0.3068,
+      "step": 603
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5493663971781544,
+      "learning_rate": 9.983704341665425e-06,
+      "loss": 0.3154,
+      "step": 604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6615116590414227,
+      "learning_rate": 9.983585243442733e-06,
+      "loss": 0.3078,
+      "step": 605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.027194215842722,
+      "learning_rate": 9.983465712298985e-06,
+      "loss": 0.357,
+      "step": 606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7860114628325077,
+      "learning_rate": 9.983345748244566e-06,
+      "loss": 0.3774,
+      "step": 607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.631378294109237,
+      "learning_rate": 9.983225351289896e-06,
+      "loss": 0.3072,
+      "step": 608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6392994212400533,
+      "learning_rate": 9.983104521445434e-06,
+      "loss": 0.3156,
+      "step": 609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7247714179757434,
+      "learning_rate": 9.982983258721675e-06,
+      "loss": 0.3245,
+      "step": 610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4410774696668964,
+      "learning_rate": 9.982861563129154e-06,
+      "loss": 0.3114,
+      "step": 611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5363650805327866,
+      "learning_rate": 9.982739434678444e-06,
+      "loss": 0.284,
+      "step": 612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5516954195413963,
+      "learning_rate": 9.982616873380151e-06,
+      "loss": 0.3116,
+      "step": 613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3188741657324146,
+      "learning_rate": 9.982493879244925e-06,
+      "loss": 0.2762,
+      "step": 614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5845694867073155,
+      "learning_rate": 9.982370452283451e-06,
+      "loss": 0.341,
+      "step": 615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8009145877378963,
+      "learning_rate": 9.982246592506446e-06,
+      "loss": 0.3373,
+      "step": 616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.691928228524045,
+      "learning_rate": 9.982122299924676e-06,
+      "loss": 0.2776,
+      "step": 617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3401087126991134,
+      "learning_rate": 9.981997574548933e-06,
+      "loss": 0.3045,
+      "step": 618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3507709179044514,
+      "learning_rate": 9.981872416390055e-06,
+      "loss": 0.2873,
+      "step": 619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4071140932586723,
+      "learning_rate": 9.981746825458914e-06,
+      "loss": 0.3195,
+      "step": 620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4645877328068493,
+      "learning_rate": 9.981620801766418e-06,
+      "loss": 0.2886,
+      "step": 621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1229118687038695,
+      "learning_rate": 9.981494345323516e-06,
+      "loss": 0.2735,
+      "step": 622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.202525560544861,
+      "learning_rate": 9.981367456141193e-06,
+      "loss": 0.3136,
+      "step": 623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4894889312577115,
+      "learning_rate": 9.981240134230473e-06,
+      "loss": 0.3027,
+      "step": 624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.438132474262501,
+      "learning_rate": 9.981112379602414e-06,
+      "loss": 0.2794,
+      "step": 625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7382894579174724,
+      "learning_rate": 9.980984192268116e-06,
+      "loss": 0.3309,
+      "step": 626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3036678910272803,
+      "learning_rate": 9.980855572238715e-06,
+      "loss": 0.3006,
+      "step": 627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.940467815872872,
+      "learning_rate": 9.980726519525382e-06,
+      "loss": 0.3199,
+      "step": 628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.274658661689854,
+      "learning_rate": 9.980597034139328e-06,
+      "loss": 0.2977,
+      "step": 629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4313120747684582,
+      "learning_rate": 9.980467116091803e-06,
+      "loss": 0.2858,
+      "step": 630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6889372362101156,
+      "learning_rate": 9.980336765394092e-06,
+      "loss": 0.3416,
+      "step": 631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.250325525490539,
+      "learning_rate": 9.980205982057517e-06,
+      "loss": 0.2786,
+      "step": 632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4879333564924555,
+      "learning_rate": 9.980074766093442e-06,
+      "loss": 0.3053,
+      "step": 633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.777489549643108,
+      "learning_rate": 9.979943117513265e-06,
+      "loss": 0.3085,
+      "step": 634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2209229366135164,
+      "learning_rate": 9.979811036328419e-06,
+      "loss": 0.287,
+      "step": 635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2635223841120466,
+      "learning_rate": 9.979678522550382e-06,
+      "loss": 0.3024,
+      "step": 636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.364567623220258,
+      "learning_rate": 9.979545576190662e-06,
+      "loss": 0.3159,
+      "step": 637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.345702800279604,
+      "learning_rate": 9.979412197260811e-06,
+      "loss": 0.3457,
+      "step": 638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2566655299204696,
+      "learning_rate": 9.979278385772414e-06,
+      "loss": 0.2547,
+      "step": 639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4134939093571446,
+      "learning_rate": 9.979144141737094e-06,
+      "loss": 0.2887,
+      "step": 640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7000426602020213,
+      "learning_rate": 9.979009465166515e-06,
+      "loss": 0.3404,
+      "step": 641
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.580331355465202,
+      "learning_rate": 9.978874356072376e-06,
+      "loss": 0.2834,
+      "step": 642
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5867870360792637,
+      "learning_rate": 9.978738814466411e-06,
+      "loss": 0.3177,
+      "step": 643
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.316895667360294,
+      "learning_rate": 9.978602840360398e-06,
+      "loss": 0.2988,
+      "step": 644
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.305031879529095,
+      "learning_rate": 9.978466433766148e-06,
+      "loss": 0.2752,
+      "step": 645
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9269551875751225,
+      "learning_rate": 9.978329594695508e-06,
+      "loss": 0.2957,
+      "step": 646
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.256096704933166,
+      "learning_rate": 9.978192323160368e-06,
+      "loss": 0.2645,
+      "step": 647
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.440585663831507,
+      "learning_rate": 9.978054619172652e-06,
+      "loss": 0.2684,
+      "step": 648
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6839411970396596,
+      "learning_rate": 9.977916482744323e-06,
+      "loss": 0.2976,
+      "step": 649
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.634028855036008,
+      "learning_rate": 9.977777913887379e-06,
+      "loss": 0.2793,
+      "step": 650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5660590265818954,
+      "learning_rate": 9.977638912613858e-06,
+      "loss": 0.3292,
+      "step": 651
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7364939252998695,
+      "learning_rate": 9.977499478935835e-06,
+      "loss": 0.3336,
+      "step": 652
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5216471106096296,
+      "learning_rate": 9.977359612865424e-06,
+      "loss": 0.2767,
+      "step": 653
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.484724921616459,
+      "learning_rate": 9.977219314414773e-06,
+      "loss": 0.3201,
+      "step": 654
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.0180796758846995,
+      "learning_rate": 9.977078583596071e-06,
+      "loss": 0.3285,
+      "step": 655
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4677323811189806,
+      "learning_rate": 9.976937420421543e-06,
+      "loss": 0.2674,
+      "step": 656
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3661415931659713,
+      "learning_rate": 9.976795824903451e-06,
+      "loss": 0.308,
+      "step": 657
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.353678458271599,
+      "learning_rate": 9.976653797054097e-06,
+      "loss": 0.2963,
+      "step": 658
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.982285147070446,
+      "learning_rate": 9.976511336885815e-06,
+      "loss": 0.3455,
+      "step": 659
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.995264844114513,
+      "learning_rate": 9.976368444410985e-06,
+      "loss": 0.2912,
+      "step": 660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5517290648902455,
+      "learning_rate": 9.976225119642018e-06,
+      "loss": 0.2731,
+      "step": 661
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.467066287909187,
+      "learning_rate": 9.976081362591365e-06,
+      "loss": 0.3061,
+      "step": 662
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4463278196686407,
+      "learning_rate": 9.975937173271513e-06,
+      "loss": 0.2998,
+      "step": 663
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7554152146835937,
+      "learning_rate": 9.975792551694988e-06,
+      "loss": 0.3322,
+      "step": 664
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.340958472323737,
+      "learning_rate": 9.975647497874354e-06,
+      "loss": 0.2882,
+      "step": 665
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.377911734877334,
+      "learning_rate": 9.975502011822212e-06,
+      "loss": 0.3085,
+      "step": 666
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4037774066888984,
+      "learning_rate": 9.975356093551198e-06,
+      "loss": 0.2641,
+      "step": 667
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6539226316218123,
+      "learning_rate": 9.97520974307399e-06,
+      "loss": 0.2875,
+      "step": 668
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.394718200753031,
+      "learning_rate": 9.975062960403303e-06,
+      "loss": 0.2769,
+      "step": 669
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.307075595532365,
+      "learning_rate": 9.974915745551882e-06,
+      "loss": 0.2774,
+      "step": 670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4330449767071975,
+      "learning_rate": 9.974768098532521e-06,
+      "loss": 0.2878,
+      "step": 671
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.528474114953311,
+      "learning_rate": 9.974620019358046e-06,
+      "loss": 0.298,
+      "step": 672
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1364271194911257,
+      "learning_rate": 9.974471508041317e-06,
+      "loss": 0.2823,
+      "step": 673
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6125744890901776,
+      "learning_rate": 9.974322564595236e-06,
+      "loss": 0.3083,
+      "step": 674
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7447606021937654,
+      "learning_rate": 9.974173189032744e-06,
+      "loss": 0.2994,
+      "step": 675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6914472124919238,
+      "learning_rate": 9.974023381366814e-06,
+      "loss": 0.3038,
+      "step": 676
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6122526597042492,
+      "learning_rate": 9.973873141610462e-06,
+      "loss": 0.3342,
+      "step": 677
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.793099827448773,
+      "learning_rate": 9.973722469776739e-06,
+      "loss": 0.2968,
+      "step": 678
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.809569548110065,
+      "learning_rate": 9.973571365878732e-06,
+      "loss": 0.3123,
+      "step": 679
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6698612347943005,
+      "learning_rate": 9.97341982992957e-06,
+      "loss": 0.3125,
+      "step": 680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.976927671573408,
+      "learning_rate": 9.973267861942416e-06,
+      "loss": 0.2685,
+      "step": 681
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.315185801763032,
+      "learning_rate": 9.973115461930469e-06,
+      "loss": 0.2547,
+      "step": 682
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7387169895798125,
+      "learning_rate": 9.97296262990697e-06,
+      "loss": 0.3174,
+      "step": 683
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.583815067524512,
+      "learning_rate": 9.972809365885197e-06,
+      "loss": 0.3034,
+      "step": 684
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.389417405378248,
+      "learning_rate": 9.972655669878462e-06,
+      "loss": 0.3176,
+      "step": 685
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.770990982203721,
+      "learning_rate": 9.972501541900115e-06,
+      "loss": 0.302,
+      "step": 686
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.162530491530101,
+      "learning_rate": 9.972346981963546e-06,
+      "loss": 0.2757,
+      "step": 687
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7118838342387286,
+      "learning_rate": 9.972191990082183e-06,
+      "loss": 0.2863,
+      "step": 688
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.380304387182788,
+      "learning_rate": 9.97203656626949e-06,
+      "loss": 0.2749,
+      "step": 689
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.589306708221887,
+      "learning_rate": 9.971880710538967e-06,
+      "loss": 0.2851,
+      "step": 690
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.310786736478933,
+      "learning_rate": 9.971724422904154e-06,
+      "loss": 0.2863,
+      "step": 691
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3741422903867284,
+      "learning_rate": 9.971567703378629e-06,
+      "loss": 0.282,
+      "step": 692
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6876230819120135,
+      "learning_rate": 9.971410551976001e-06,
+      "loss": 0.3207,
+      "step": 693
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.682734463658358,
+      "learning_rate": 9.971252968709927e-06,
+      "loss": 0.3094,
+      "step": 694
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.377390763626571,
+      "learning_rate": 9.971094953594095e-06,
+      "loss": 0.2732,
+      "step": 695
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4195438055886367,
+      "learning_rate": 9.970936506642232e-06,
+      "loss": 0.2991,
+      "step": 696
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.092369251801411,
+      "learning_rate": 9.9707776278681e-06,
+      "loss": 0.2981,
+      "step": 697
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.724113439290786,
+      "learning_rate": 9.970618317285501e-06,
+      "loss": 0.3086,
+      "step": 698
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.194815294806323,
+      "learning_rate": 9.970458574908277e-06,
+      "loss": 0.2704,
+      "step": 699
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.362049297021861,
+      "learning_rate": 9.970298400750303e-06,
+      "loss": 0.3081,
+      "step": 700
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.546960570040055,
+      "learning_rate": 9.970137794825491e-06,
+      "loss": 0.3052,
+      "step": 701
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5294080868743736,
+      "learning_rate": 9.969976757147795e-06,
+      "loss": 0.3013,
+      "step": 702
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.639605539281974,
+      "learning_rate": 9.969815287731205e-06,
+      "loss": 0.2826,
+      "step": 703
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4951388093222646,
+      "learning_rate": 9.969653386589749e-06,
+      "loss": 0.3022,
+      "step": 704
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.684031151838524,
+      "learning_rate": 9.969491053737487e-06,
+      "loss": 0.2935,
+      "step": 705
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5053122144135536,
+      "learning_rate": 9.969328289188522e-06,
+      "loss": 0.3145,
+      "step": 706
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6128578564831915,
+      "learning_rate": 9.969165092956996e-06,
+      "loss": 0.2857,
+      "step": 707
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4479859372531347,
+      "learning_rate": 9.969001465057084e-06,
+      "loss": 0.2982,
+      "step": 708
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.456346321232317,
+      "learning_rate": 9.968837405502998e-06,
+      "loss": 0.3118,
+      "step": 709
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6896387714171346,
+      "learning_rate": 9.968672914308995e-06,
+      "loss": 0.3434,
+      "step": 710
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7527125135135795,
+      "learning_rate": 9.968507991489358e-06,
+      "loss": 0.2592,
+      "step": 711
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7451778291708675,
+      "learning_rate": 9.968342637058418e-06,
+      "loss": 0.3107,
+      "step": 712
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3614085447903888,
+      "learning_rate": 9.96817685103054e-06,
+      "loss": 0.3142,
+      "step": 713
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2611284861341434,
+      "learning_rate": 9.968010633420122e-06,
+      "loss": 0.2655,
+      "step": 714
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5702815951786158,
+      "learning_rate": 9.967843984241606e-06,
+      "loss": 0.3223,
+      "step": 715
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2223114339106003,
+      "learning_rate": 9.967676903509467e-06,
+      "loss": 0.2823,
+      "step": 716
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.355306906446752,
+      "learning_rate": 9.967509391238218e-06,
+      "loss": 0.3132,
+      "step": 717
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.44546429194885,
+      "learning_rate": 9.967341447442418e-06,
+      "loss": 0.2485,
+      "step": 718
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1115232855635186,
+      "learning_rate": 9.967173072136647e-06,
+      "loss": 0.3042,
+      "step": 719
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.203475365793821,
+      "learning_rate": 9.967004265335536e-06,
+      "loss": 0.3106,
+      "step": 720
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5798124241562346,
+      "learning_rate": 9.96683502705375e-06,
+      "loss": 0.2534,
+      "step": 721
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2773676888747385,
+      "learning_rate": 9.966665357305988e-06,
+      "loss": 0.2713,
+      "step": 722
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.723228547540024,
+      "learning_rate": 9.966495256106991e-06,
+      "loss": 0.2826,
+      "step": 723
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4860363526907254,
+      "learning_rate": 9.966324723471535e-06,
+      "loss": 0.3045,
+      "step": 724
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3812284091208986,
+      "learning_rate": 9.966153759414434e-06,
+      "loss": 0.2848,
+      "step": 725
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.191936349781771,
+      "learning_rate": 9.96598236395054e-06,
+      "loss": 0.3016,
+      "step": 726
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8881091244639414,
+      "learning_rate": 9.965810537094741e-06,
+      "loss": 0.3334,
+      "step": 727
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4495767848787366,
+      "learning_rate": 9.965638278861966e-06,
+      "loss": 0.3044,
+      "step": 728
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7543822685702954,
+      "learning_rate": 9.965465589267176e-06,
+      "loss": 0.3069,
+      "step": 729
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4668386863288805,
+      "learning_rate": 9.965292468325373e-06,
+      "loss": 0.2744,
+      "step": 730
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1985198222144184,
+      "learning_rate": 9.965118916051597e-06,
+      "loss": 0.2557,
+      "step": 731
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.995622422707886,
+      "learning_rate": 9.964944932460923e-06,
+      "loss": 0.3204,
+      "step": 732
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2285352897398423,
+      "learning_rate": 9.964770517568467e-06,
+      "loss": 0.2864,
+      "step": 733
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.503607216309929,
+      "learning_rate": 9.964595671389379e-06,
+      "loss": 0.2887,
+      "step": 734
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.302761482883814,
+      "learning_rate": 9.964420393938848e-06,
+      "loss": 0.2739,
+      "step": 735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1116239504167753,
+      "learning_rate": 9.964244685232098e-06,
+      "loss": 0.2307,
+      "step": 736
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6026380225463503,
+      "learning_rate": 9.964068545284396e-06,
+      "loss": 0.3025,
+      "step": 737
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.829076461368195,
+      "learning_rate": 9.963891974111042e-06,
+      "loss": 0.2869,
+      "step": 738
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.155511060956363,
+      "learning_rate": 9.963714971727374e-06,
+      "loss": 0.2742,
+      "step": 739
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.507236817874012,
+      "learning_rate": 9.96353753814877e-06,
+      "loss": 0.2853,
+      "step": 740
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6058429077383973,
+      "learning_rate": 9.96335967339064e-06,
+      "loss": 0.2626,
+      "step": 741
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.790342132554691,
+      "learning_rate": 9.96318137746844e-06,
+      "loss": 0.3271,
+      "step": 742
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.565100051079855,
+      "learning_rate": 9.963002650397655e-06,
+      "loss": 0.3191,
+      "step": 743
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4529144863883072,
+      "learning_rate": 9.96282349219381e-06,
+      "loss": 0.3003,
+      "step": 744
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.372454447920437,
+      "learning_rate": 9.96264390287247e-06,
+      "loss": 0.283,
+      "step": 745
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3548046061817156,
+      "learning_rate": 9.962463882449238e-06,
+      "loss": 0.2864,
+      "step": 746
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1307353004840004,
+      "learning_rate": 9.96228343093975e-06,
+      "loss": 0.2326,
+      "step": 747
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3358542177937953,
+      "learning_rate": 9.96210254835968e-06,
+      "loss": 0.2748,
+      "step": 748
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2425270065380642,
+      "learning_rate": 9.961921234724743e-06,
+      "loss": 0.2511,
+      "step": 749
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.230679448919394,
+      "learning_rate": 9.96173949005069e-06,
+      "loss": 0.2731,
+      "step": 750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.8136454472507255,
+      "learning_rate": 9.961557314353309e-06,
+      "loss": 0.2846,
+      "step": 751
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.554129291982215,
+      "learning_rate": 9.961374707648424e-06,
+      "loss": 0.2745,
+      "step": 752
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6373577331973843,
+      "learning_rate": 9.9611916699519e-06,
+      "loss": 0.2961,
+      "step": 753
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6789201237509217,
+      "learning_rate": 9.961008201279636e-06,
+      "loss": 0.3321,
+      "step": 754
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6832397156715637,
+      "learning_rate": 9.960824301647569e-06,
+      "loss": 0.3356,
+      "step": 755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.432029081351122,
+      "learning_rate": 9.960639971071677e-06,
+      "loss": 0.2773,
+      "step": 756
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7951568034748164,
+      "learning_rate": 9.960455209567971e-06,
+      "loss": 0.2427,
+      "step": 757
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.141713511374909,
+      "learning_rate": 9.960270017152502e-06,
+      "loss": 0.3027,
+      "step": 758
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2027679396906596,
+      "learning_rate": 9.960084393841355e-06,
+      "loss": 0.2652,
+      "step": 759
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4805531515238735,
+      "learning_rate": 9.959898339650658e-06,
+      "loss": 0.2826,
+      "step": 760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.686151448013631,
+      "learning_rate": 9.959711854596573e-06,
+      "loss": 0.3276,
+      "step": 761
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2125987881135933,
+      "learning_rate": 9.959524938695296e-06,
+      "loss": 0.258,
+      "step": 762
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2359755147444935,
+      "learning_rate": 9.959337591963069e-06,
+      "loss": 0.2894,
+      "step": 763
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6188907281506673,
+      "learning_rate": 9.959149814416165e-06,
+      "loss": 0.2608,
+      "step": 764
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6435810728364535,
+      "learning_rate": 9.958961606070896e-06,
+      "loss": 0.2962,
+      "step": 765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3582571673054757,
+      "learning_rate": 9.958772966943612e-06,
+      "loss": 0.29,
+      "step": 766
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.34219987746046,
+      "learning_rate": 9.9585838970507e-06,
+      "loss": 0.2694,
+      "step": 767
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6100025686666433,
+      "learning_rate": 9.958394396408583e-06,
+      "loss": 0.267,
+      "step": 768
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.399622053459657,
+      "learning_rate": 9.958204465033726e-06,
+      "loss": 0.2862,
+      "step": 769
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0979020336646337,
+      "learning_rate": 9.958014102942623e-06,
+      "loss": 0.262,
+      "step": 770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.555138762200467,
+      "learning_rate": 9.957823310151816e-06,
+      "loss": 0.2562,
+      "step": 771
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8621948610660617,
+      "learning_rate": 9.957632086677876e-06,
+      "loss": 0.3156,
+      "step": 772
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.317185612079889,
+      "learning_rate": 9.957440432537415e-06,
+      "loss": 0.3039,
+      "step": 773
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5648043021430467,
+      "learning_rate": 9.957248347747083e-06,
+      "loss": 0.2993,
+      "step": 774
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3233896511158,
+      "learning_rate": 9.957055832323566e-06,
+      "loss": 0.2839,
+      "step": 775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.099205957694149,
+      "learning_rate": 9.956862886283586e-06,
+      "loss": 0.2632,
+      "step": 776
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4295710193400355,
+      "learning_rate": 9.956669509643904e-06,
+      "loss": 0.2774,
+      "step": 777
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7061755972324506,
+      "learning_rate": 9.95647570242132e-06,
+      "loss": 0.334,
+      "step": 778
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.416689351648574,
+      "learning_rate": 9.95628146463267e-06,
+      "loss": 0.3067,
+      "step": 779
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.305566131288089,
+      "learning_rate": 9.956086796294828e-06,
+      "loss": 0.2681,
+      "step": 780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4932129502735303,
+      "learning_rate": 9.955891697424704e-06,
+      "loss": 0.3035,
+      "step": 781
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8483997669145533,
+      "learning_rate": 9.955696168039244e-06,
+      "loss": 0.245,
+      "step": 782
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4241581489682256,
+      "learning_rate": 9.955500208155438e-06,
+      "loss": 0.3039,
+      "step": 783
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3286605374314213,
+      "learning_rate": 9.955303817790303e-06,
+      "loss": 0.2968,
+      "step": 784
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9719950010043386,
+      "learning_rate": 9.955106996960903e-06,
+      "loss": 0.286,
+      "step": 785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0243052146947793,
+      "learning_rate": 9.954909745684339e-06,
+      "loss": 0.2745,
+      "step": 786
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.097466339402794,
+      "learning_rate": 9.954712063977738e-06,
+      "loss": 0.3112,
+      "step": 787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3248046861463463,
+      "learning_rate": 9.954513951858279e-06,
+      "loss": 0.2847,
+      "step": 788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.204870899669546,
+      "learning_rate": 9.95431540934317e-06,
+      "loss": 0.2618,
+      "step": 789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3254016311962613,
+      "learning_rate": 9.954116436449656e-06,
+      "loss": 0.2929,
+      "step": 790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.246936263303662,
+      "learning_rate": 9.953917033195026e-06,
+      "loss": 0.2656,
+      "step": 791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1496320423146447,
+      "learning_rate": 9.953717199596598e-06,
+      "loss": 0.2153,
+      "step": 792
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.175948122557839,
+      "learning_rate": 9.953516935671734e-06,
+      "loss": 0.288,
+      "step": 793
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.36687264722415,
+      "learning_rate": 9.95331624143783e-06,
+      "loss": 0.2937,
+      "step": 794
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.423492599897248,
+      "learning_rate": 9.95311511691232e-06,
+      "loss": 0.2754,
+      "step": 795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.439266131874316,
+      "learning_rate": 9.952913562112675e-06,
+      "loss": 0.2964,
+      "step": 796
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.952737750541927,
+      "learning_rate": 9.952711577056403e-06,
+      "loss": 0.2723,
+      "step": 797
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7179796705645707,
+      "learning_rate": 9.952509161761056e-06,
+      "loss": 0.3098,
+      "step": 798
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.404768195466816,
+      "learning_rate": 9.95230631624421e-06,
+      "loss": 0.2729,
+      "step": 799
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3747019633380724,
+      "learning_rate": 9.952103040523493e-06,
+      "loss": 0.2692,
+      "step": 800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1322021548770578,
+      "learning_rate": 9.951899334616559e-06,
+      "loss": 0.2526,
+      "step": 801
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5184023033352556,
+      "learning_rate": 9.951695198541105e-06,
+      "loss": 0.3073,
+      "step": 802
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8511693229634583,
+      "learning_rate": 9.951490632314863e-06,
+      "loss": 0.2472,
+      "step": 803
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0923307213726834,
+      "learning_rate": 9.951285635955606e-06,
+      "loss": 0.2473,
+      "step": 804
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.510034222113154,
+      "learning_rate": 9.951080209481138e-06,
+      "loss": 0.2886,
+      "step": 805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7110009649027296,
+      "learning_rate": 9.95087435290931e-06,
+      "loss": 0.341,
+      "step": 806
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1540785530107054,
+      "learning_rate": 9.950668066258e-06,
+      "loss": 0.2444,
+      "step": 807
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0355556075817396,
+      "learning_rate": 9.950461349545131e-06,
+      "loss": 0.2393,
+      "step": 808
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1495907271209647,
+      "learning_rate": 9.950254202788655e-06,
+      "loss": 0.2486,
+      "step": 809
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2902495330650754,
+      "learning_rate": 9.950046626006575e-06,
+      "loss": 0.2636,
+      "step": 810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3813702500369938,
+      "learning_rate": 9.949838619216917e-06,
+      "loss": 0.3369,
+      "step": 811
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5617853787228464,
+      "learning_rate": 9.949630182437753e-06,
+      "loss": 0.2974,
+      "step": 812
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.386285071892533,
+      "learning_rate": 9.949421315687186e-06,
+      "loss": 0.2609,
+      "step": 813
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6558861125591653,
+      "learning_rate": 9.949212018983366e-06,
+      "loss": 0.3085,
+      "step": 814
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.1046420451903805,
+      "learning_rate": 9.94900229234447e-06,
+      "loss": 0.2951,
+      "step": 815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.667593788754319,
+      "learning_rate": 9.94879213578872e-06,
+      "loss": 0.3009,
+      "step": 816
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.104437788153663,
+      "learning_rate": 9.948581549334368e-06,
+      "loss": 0.2511,
+      "step": 817
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8860325625070287,
+      "learning_rate": 9.948370532999711e-06,
+      "loss": 0.3225,
+      "step": 818
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0756469664723327,
+      "learning_rate": 9.948159086803078e-06,
+      "loss": 0.2671,
+      "step": 819
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.690499751984658,
+      "learning_rate": 9.94794721076284e-06,
+      "loss": 0.315,
+      "step": 820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.078693142640425,
+      "learning_rate": 9.9477349048974e-06,
+      "loss": 0.2147,
+      "step": 821
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0451427390150854,
+      "learning_rate": 9.9475221692252e-06,
+      "loss": 0.2416,
+      "step": 822
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4963280008252466,
+      "learning_rate": 9.947309003764723e-06,
+      "loss": 0.2981,
+      "step": 823
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2465290072893165,
+      "learning_rate": 9.947095408534483e-06,
+      "loss": 0.2732,
+      "step": 824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.218986694396093,
+      "learning_rate": 9.94688138355304e-06,
+      "loss": 0.2907,
+      "step": 825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5567896355598005,
+      "learning_rate": 9.946666928838982e-06,
+      "loss": 0.2915,
+      "step": 826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.565878929589163,
+      "learning_rate": 9.946452044410941e-06,
+      "loss": 0.2851,
+      "step": 827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5318361932044806,
+      "learning_rate": 9.946236730287582e-06,
+      "loss": 0.2832,
+      "step": 828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2918080992345478,
+      "learning_rate": 9.94602098648761e-06,
+      "loss": 0.2585,
+      "step": 829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3094119631714847,
+      "learning_rate": 9.945804813029767e-06,
+      "loss": 0.3079,
+      "step": 830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9561930344791723,
+      "learning_rate": 9.94558820993283e-06,
+      "loss": 0.2417,
+      "step": 831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.169239609485324,
+      "learning_rate": 9.945371177215619e-06,
+      "loss": 0.2555,
+      "step": 832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7261798459193667,
+      "learning_rate": 9.945153714896982e-06,
+      "loss": 0.3168,
+      "step": 833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8886358861284704,
+      "learning_rate": 9.944935822995815e-06,
+      "loss": 0.2451,
+      "step": 834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.299749967695516,
+      "learning_rate": 9.944717501531045e-06,
+      "loss": 0.2478,
+      "step": 835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3127143993757824,
+      "learning_rate": 9.944498750521634e-06,
+      "loss": 0.2643,
+      "step": 836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2585992568654554,
+      "learning_rate": 9.944279569986588e-06,
+      "loss": 0.2957,
+      "step": 837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4082780034209983,
+      "learning_rate": 9.944059959944948e-06,
+      "loss": 0.3225,
+      "step": 838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.012256541341559,
+      "learning_rate": 9.943839920415787e-06,
+      "loss": 0.253,
+      "step": 839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.267056622102402,
+      "learning_rate": 9.943619451418225e-06,
+      "loss": 0.2514,
+      "step": 840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1673656637844374,
+      "learning_rate": 9.943398552971409e-06,
+      "loss": 0.3029,
+      "step": 841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8834986105335467,
+      "learning_rate": 9.943177225094532e-06,
+      "loss": 0.2973,
+      "step": 842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.117308380535242,
+      "learning_rate": 9.94295546780682e-06,
+      "loss": 0.2223,
+      "step": 843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2606961416618687,
+      "learning_rate": 9.942733281127536e-06,
+      "loss": 0.2638,
+      "step": 844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4167423699936785,
+      "learning_rate": 9.94251066507598e-06,
+      "loss": 0.2807,
+      "step": 845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3056984549563353,
+      "learning_rate": 9.942287619671494e-06,
+      "loss": 0.3129,
+      "step": 846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1739941932147806,
+      "learning_rate": 9.94206414493345e-06,
+      "loss": 0.2726,
+      "step": 847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4368729160803593,
+      "learning_rate": 9.941840240881265e-06,
+      "loss": 0.2781,
+      "step": 848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5578848531342517,
+      "learning_rate": 9.941615907534387e-06,
+      "loss": 0.313,
+      "step": 849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9482153702050877,
+      "learning_rate": 9.941391144912304e-06,
+      "loss": 0.2456,
+      "step": 850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0827951822313717,
+      "learning_rate": 9.94116595303454e-06,
+      "loss": 0.2722,
+      "step": 851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.129244654862273,
+      "learning_rate": 9.94094033192066e-06,
+      "loss": 0.2399,
+      "step": 852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4025518228901825,
+      "learning_rate": 9.94071428159026e-06,
+      "loss": 0.2997,
+      "step": 853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5116719524469544,
+      "learning_rate": 9.940487802062979e-06,
+      "loss": 0.3363,
+      "step": 854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1240186869932733,
+      "learning_rate": 9.940260893358493e-06,
+      "loss": 0.2955,
+      "step": 855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.453953164988273,
+      "learning_rate": 9.940033555496509e-06,
+      "loss": 0.306,
+      "step": 856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.461703476402533,
+      "learning_rate": 9.939805788496778e-06,
+      "loss": 0.2835,
+      "step": 857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1825409882710254,
+      "learning_rate": 9.939577592379088e-06,
+      "loss": 0.2549,
+      "step": 858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.338118785789188,
+      "learning_rate": 9.93934896716326e-06,
+      "loss": 0.2726,
+      "step": 859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3044553116495945,
+      "learning_rate": 9.939119912869155e-06,
+      "loss": 0.2638,
+      "step": 860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0747782437008375,
+      "learning_rate": 9.93889042951667e-06,
+      "loss": 0.2713,
+      "step": 861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9895779820807205,
+      "learning_rate": 9.93866051712574e-06,
+      "loss": 0.2636,
+      "step": 862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.363429208502432,
+      "learning_rate": 9.93843017571634e-06,
+      "loss": 0.2786,
+      "step": 863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.391370227576883,
+      "learning_rate": 9.938199405308475e-06,
+      "loss": 0.2772,
+      "step": 864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.216721416316368,
+      "learning_rate": 9.937968205922198e-06,
+      "loss": 0.2758,
+      "step": 865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.033388138988561,
+      "learning_rate": 9.937736577577587e-06,
+      "loss": 0.2684,
+      "step": 866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7052363377283357,
+      "learning_rate": 9.937504520294767e-06,
+      "loss": 0.3103,
+      "step": 867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.30732787785901,
+      "learning_rate": 9.937272034093897e-06,
+      "loss": 0.2968,
+      "step": 868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4051708530518447,
+      "learning_rate": 9.93703911899517e-06,
+      "loss": 0.2636,
+      "step": 869
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.276172911418095,
+      "learning_rate": 9.93680577501882e-06,
+      "loss": 0.2866,
+      "step": 870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6143399487223853,
+      "learning_rate": 9.93657200218512e-06,
+      "loss": 0.2885,
+      "step": 871
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4778218379995374,
+      "learning_rate": 9.936337800514377e-06,
+      "loss": 0.2913,
+      "step": 872
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.340457203848266,
+      "learning_rate": 9.936103170026934e-06,
+      "loss": 0.2513,
+      "step": 873
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.559938192438544,
+      "learning_rate": 9.935868110743175e-06,
+      "loss": 0.3287,
+      "step": 874
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3847174795028616,
+      "learning_rate": 9.935632622683518e-06,
+      "loss": 0.2891,
+      "step": 875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.056051455197346,
+      "learning_rate": 9.93539670586842e-06,
+      "loss": 0.228,
+      "step": 876
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5349518806084133,
+      "learning_rate": 9.935160360318376e-06,
+      "loss": 0.2646,
+      "step": 877
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.235116153391338,
+      "learning_rate": 9.934923586053917e-06,
+      "loss": 0.2633,
+      "step": 878
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4477474718655348,
+      "learning_rate": 9.93468638309561e-06,
+      "loss": 0.2708,
+      "step": 879
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4013210275500563,
+      "learning_rate": 9.934448751464064e-06,
+      "loss": 0.2771,
+      "step": 880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.088117903879566,
+      "learning_rate": 9.934210691179918e-06,
+      "loss": 0.2546,
+      "step": 881
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.382518928898099,
+      "learning_rate": 9.933972202263853e-06,
+      "loss": 0.2716,
+      "step": 882
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7159108465171986,
+      "learning_rate": 9.933733284736588e-06,
+      "loss": 0.2795,
+      "step": 883
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8373461775254762,
+      "learning_rate": 9.933493938618878e-06,
+      "loss": 0.2504,
+      "step": 884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3390583871704256,
+      "learning_rate": 9.933254163931512e-06,
+      "loss": 0.2943,
+      "step": 885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2670928887377344,
+      "learning_rate": 9.933013960695321e-06,
+      "loss": 0.276,
+      "step": 886
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3618465757078155,
+      "learning_rate": 9.932773328931171e-06,
+      "loss": 0.2638,
+      "step": 887
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1766026470572344,
+      "learning_rate": 9.932532268659966e-06,
+      "loss": 0.277,
+      "step": 888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.007161501151894,
+      "learning_rate": 9.932290779902648e-06,
+      "loss": 0.2496,
+      "step": 889
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3360649173645647,
+      "learning_rate": 9.93204886268019e-06,
+      "loss": 0.2541,
+      "step": 890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5294014099423774,
+      "learning_rate": 9.931806517013612e-06,
+      "loss": 0.3163,
+      "step": 891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7764181028997834,
+      "learning_rate": 9.931563742923967e-06,
+      "loss": 0.3239,
+      "step": 892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4927472302130713,
+      "learning_rate": 9.931320540432339e-06,
+      "loss": 0.3105,
+      "step": 893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.912071475306121,
+      "learning_rate": 9.93107690955986e-06,
+      "loss": 0.2616,
+      "step": 894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8986147938294793,
+      "learning_rate": 9.930832850327693e-06,
+      "loss": 0.2512,
+      "step": 895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.958875753553084,
+      "learning_rate": 9.930588362757038e-06,
+      "loss": 0.2461,
+      "step": 896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.580466817979422,
+      "learning_rate": 9.930343446869134e-06,
+      "loss": 0.2555,
+      "step": 897
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.063186430828628,
+      "learning_rate": 9.93009810268526e-06,
+      "loss": 0.297,
+      "step": 898
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.134047207459698,
+      "learning_rate": 9.929852330226723e-06,
+      "loss": 0.2636,
+      "step": 899
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.242664574250869,
+      "learning_rate": 9.929606129514875e-06,
+      "loss": 0.2932,
+      "step": 900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3303198387606514,
+      "learning_rate": 9.929359500571108e-06,
+      "loss": 0.2545,
+      "step": 901
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.865057469300579,
+      "learning_rate": 9.92911244341684e-06,
+      "loss": 0.3238,
+      "step": 902
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2447404443038073,
+      "learning_rate": 9.928864958073536e-06,
+      "loss": 0.2332,
+      "step": 903
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0135428087406364,
+      "learning_rate": 9.928617044562695e-06,
+      "loss": 0.2314,
+      "step": 904
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.617340165307831,
+      "learning_rate": 9.92836870290585e-06,
+      "loss": 0.3012,
+      "step": 905
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.47175086119539,
+      "learning_rate": 9.92811993312458e-06,
+      "loss": 0.2722,
+      "step": 906
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1705522076845174,
+      "learning_rate": 9.927870735240492e-06,
+      "loss": 0.2815,
+      "step": 907
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0077622773513553,
+      "learning_rate": 9.927621109275233e-06,
+      "loss": 0.2663,
+      "step": 908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2973207485974556,
+      "learning_rate": 9.927371055250489e-06,
+      "loss": 0.2544,
+      "step": 909
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1272857794777784,
+      "learning_rate": 9.927120573187981e-06,
+      "loss": 0.2598,
+      "step": 910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4522236094285303,
+      "learning_rate": 9.92686966310947e-06,
+      "loss": 0.283,
+      "step": 911
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9510192069679353,
+      "learning_rate": 9.926618325036752e-06,
+      "loss": 0.2777,
+      "step": 912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.657729761965793,
+      "learning_rate": 9.926366558991659e-06,
+      "loss": 0.2584,
+      "step": 913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.065916477202066,
+      "learning_rate": 9.926114364996063e-06,
+      "loss": 0.2407,
+      "step": 914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.082148895459199,
+      "learning_rate": 9.925861743071872e-06,
+      "loss": 0.2642,
+      "step": 915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.386573938151328,
+      "learning_rate": 9.925608693241031e-06,
+      "loss": 0.2665,
+      "step": 916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.262545697249871,
+      "learning_rate": 9.925355215525523e-06,
+      "loss": 0.2707,
+      "step": 917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.953267251908424,
+      "learning_rate": 9.925101309947365e-06,
+      "loss": 0.2235,
+      "step": 918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2239983903259875,
+      "learning_rate": 9.924846976528618e-06,
+      "loss": 0.26,
+      "step": 919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6974143702568694,
+      "learning_rate": 9.924592215291368e-06,
+      "loss": 0.2403,
+      "step": 920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.35068535670828,
+      "learning_rate": 9.924337026257756e-06,
+      "loss": 0.2523,
+      "step": 921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.221139260263104,
+      "learning_rate": 9.924081409449943e-06,
+      "loss": 0.2583,
+      "step": 922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5040974384723786,
+      "learning_rate": 9.923825364890137e-06,
+      "loss": 0.2846,
+      "step": 923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9705703472482505,
+      "learning_rate": 9.923568892600579e-06,
+      "loss": 0.2721,
+      "step": 924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1974979222629516,
+      "learning_rate": 9.92331199260355e-06,
+      "loss": 0.2918,
+      "step": 925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2392929177127923,
+      "learning_rate": 9.923054664921366e-06,
+      "loss": 0.2799,
+      "step": 926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3223309521614284,
+      "learning_rate": 9.922796909576383e-06,
+      "loss": 0.2851,
+      "step": 927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.002716741800669,
+      "learning_rate": 9.922538726590987e-06,
+      "loss": 0.3056,
+      "step": 928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.226612579952788,
+      "learning_rate": 9.92228011598761e-06,
+      "loss": 0.2518,
+      "step": 929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0239036497117975,
+      "learning_rate": 9.922021077788717e-06,
+      "loss": 0.2408,
+      "step": 930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1996834708257884,
+      "learning_rate": 9.92176161201681e-06,
+      "loss": 0.2859,
+      "step": 931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1060558343210904,
+      "learning_rate": 9.921501718694431e-06,
+      "loss": 0.2698,
+      "step": 932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.466907669704993,
+      "learning_rate": 9.921241397844153e-06,
+      "loss": 0.2788,
+      "step": 933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2084457047284394,
+      "learning_rate": 9.920980649488591e-06,
+      "loss": 0.2642,
+      "step": 934
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.279648066311588,
+      "learning_rate": 9.920719473650397e-06,
+      "loss": 0.2792,
+      "step": 935
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.106194111971018,
+      "learning_rate": 9.920457870352259e-06,
+      "loss": 0.2321,
+      "step": 936
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1934582290410654,
+      "learning_rate": 9.920195839616901e-06,
+      "loss": 0.2803,
+      "step": 937
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.292074374731251,
+      "learning_rate": 9.919933381467088e-06,
+      "loss": 0.2687,
+      "step": 938
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1994067582994976,
+      "learning_rate": 9.919670495925618e-06,
+      "loss": 0.2738,
+      "step": 939
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.415954983887142,
+      "learning_rate": 9.919407183015327e-06,
+      "loss": 0.2858,
+      "step": 940
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3381735172147082,
+      "learning_rate": 9.91914344275909e-06,
+      "loss": 0.2781,
+      "step": 941
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.61713607697877,
+      "learning_rate": 9.918879275179819e-06,
+      "loss": 0.2608,
+      "step": 942
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2805387980907814,
+      "learning_rate": 9.918614680300458e-06,
+      "loss": 0.296,
+      "step": 943
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2086112749813966,
+      "learning_rate": 9.918349658143997e-06,
+      "loss": 0.277,
+      "step": 944
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1151483528131627,
+      "learning_rate": 9.918084208733454e-06,
+      "loss": 0.2805,
+      "step": 945
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.903697337675332,
+      "learning_rate": 9.917818332091892e-06,
+      "loss": 0.2662,
+      "step": 946
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9618248080189367,
+      "learning_rate": 9.917552028242406e-06,
+      "loss": 0.2387,
+      "step": 947
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6805600533646814,
+      "learning_rate": 9.91728529720813e-06,
+      "loss": 0.234,
+      "step": 948
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.252167376273173,
+      "learning_rate": 9.917018139012236e-06,
+      "loss": 0.2739,
+      "step": 949
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.706046767878392,
+      "learning_rate": 9.916750553677929e-06,
+      "loss": 0.2392,
+      "step": 950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.35702234988839,
+      "learning_rate": 9.916482541228456e-06,
+      "loss": 0.3126,
+      "step": 951
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2029297541061816,
+      "learning_rate": 9.916214101687096e-06,
+      "loss": 0.2804,
+      "step": 952
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0002871366757327,
+      "learning_rate": 9.915945235077173e-06,
+      "loss": 0.2496,
+      "step": 953
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5733977270923636,
+      "learning_rate": 9.915675941422042e-06,
+      "loss": 0.2079,
+      "step": 954
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4243578854293064,
+      "learning_rate": 9.915406220745093e-06,
+      "loss": 0.3017,
+      "step": 955
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9458666508721474,
+      "learning_rate": 9.915136073069759e-06,
+      "loss": 0.2417,
+      "step": 956
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0923961214081066,
+      "learning_rate": 9.91486549841951e-06,
+      "loss": 0.2454,
+      "step": 957
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4901091314863693,
+      "learning_rate": 9.914594496817846e-06,
+      "loss": 0.2728,
+      "step": 958
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9784470235241427,
+      "learning_rate": 9.914323068288312e-06,
+      "loss": 0.252,
+      "step": 959
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0681060236739826,
+      "learning_rate": 9.914051212854484e-06,
+      "loss": 0.2962,
+      "step": 960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7016575028420533,
+      "learning_rate": 9.91377893053998e-06,
+      "loss": 0.2126,
+      "step": 961
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2050958677541117,
+      "learning_rate": 9.913506221368455e-06,
+      "loss": 0.2445,
+      "step": 962
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.126204998663041,
+      "learning_rate": 9.913233085363595e-06,
+      "loss": 0.2272,
+      "step": 963
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3519564433198252,
+      "learning_rate": 9.912959522549126e-06,
+      "loss": 0.2455,
+      "step": 964
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2199417712369023,
+      "learning_rate": 9.912685532948819e-06,
+      "loss": 0.234,
+      "step": 965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.468102420508789,
+      "learning_rate": 9.912411116586469e-06,
+      "loss": 0.31,
+      "step": 966
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2211199656799425,
+      "learning_rate": 9.912136273485917e-06,
+      "loss": 0.2797,
+      "step": 967
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.046776038463699,
+      "learning_rate": 9.91186100367104e-06,
+      "loss": 0.2551,
+      "step": 968
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0177188372682866,
+      "learning_rate": 9.911585307165747e-06,
+      "loss": 0.2377,
+      "step": 969
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0982130995189543,
+      "learning_rate": 9.911309183993988e-06,
+      "loss": 0.2672,
+      "step": 970
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8651599495662263,
+      "learning_rate": 9.911032634179754e-06,
+      "loss": 0.2588,
+      "step": 971
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.185402106015171,
+      "learning_rate": 9.910755657747064e-06,
+      "loss": 0.279,
+      "step": 972
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.014181446254031,
+      "learning_rate": 9.910478254719983e-06,
+      "loss": 0.2432,
+      "step": 973
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.501906145237206,
+      "learning_rate": 9.910200425122603e-06,
+      "loss": 0.3076,
+      "step": 974
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.353380746290943,
+      "learning_rate": 9.909922168979063e-06,
+      "loss": 0.2716,
+      "step": 975
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4491161780728565,
+      "learning_rate": 9.909643486313533e-06,
+      "loss": 0.3122,
+      "step": 976
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1748124863854725,
+      "learning_rate": 9.909364377150226e-06,
+      "loss": 0.2623,
+      "step": 977
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1485172269620842,
+      "learning_rate": 9.909084841513383e-06,
+      "loss": 0.2303,
+      "step": 978
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.003349431867429,
+      "learning_rate": 9.90880487942729e-06,
+      "loss": 0.2332,
+      "step": 979
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2354365471636934,
+      "learning_rate": 9.908524490916267e-06,
+      "loss": 0.2516,
+      "step": 980
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.234736477686663,
+      "learning_rate": 9.90824367600467e-06,
+      "loss": 0.2754,
+      "step": 981
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.069791878099338,
+      "learning_rate": 9.907962434716894e-06,
+      "loss": 0.2719,
+      "step": 982
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.967540685989675,
+      "learning_rate": 9.90768076707737e-06,
+      "loss": 0.2356,
+      "step": 983
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.565257987898602,
+      "learning_rate": 9.907398673110565e-06,
+      "loss": 0.3098,
+      "step": 984
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.409556460052405,
+      "learning_rate": 9.907116152840987e-06,
+      "loss": 0.2963,
+      "step": 985
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0627131576161557,
+      "learning_rate": 9.906833206293177e-06,
+      "loss": 0.2614,
+      "step": 986
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.143487194763398,
+      "learning_rate": 9.906549833491714e-06,
+      "loss": 0.2304,
+      "step": 987
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.525379915086094,
+      "learning_rate": 9.906266034461216e-06,
+      "loss": 0.2717,
+      "step": 988
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2134697424710996,
+      "learning_rate": 9.905981809226334e-06,
+      "loss": 0.2426,
+      "step": 989
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4192533850413236,
+      "learning_rate": 9.905697157811761e-06,
+      "loss": 0.2644,
+      "step": 990
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5101120655373146,
+      "learning_rate": 9.905412080242222e-06,
+      "loss": 0.2596,
+      "step": 991
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3518818908320664,
+      "learning_rate": 9.905126576542485e-06,
+      "loss": 0.2846,
+      "step": 992
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.108242466710249,
+      "learning_rate": 9.904840646737346e-06,
+      "loss": 0.2566,
+      "step": 993
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.253718175470579,
+      "learning_rate": 9.904554290851648e-06,
+      "loss": 0.2855,
+      "step": 994
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.172997631569075,
+      "learning_rate": 9.904267508910269e-06,
+      "loss": 0.2631,
+      "step": 995
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.304507528666179,
+      "learning_rate": 9.903980300938115e-06,
+      "loss": 0.2742,
+      "step": 996
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.928623139741482,
+      "learning_rate": 9.903692666960139e-06,
+      "loss": 0.2354,
+      "step": 997
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.217885043210312,
+      "learning_rate": 9.903404607001325e-06,
+      "loss": 0.2561,
+      "step": 998
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1953572178467553,
+      "learning_rate": 9.903116121086703e-06,
+      "loss": 0.2629,
+      "step": 999
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1947480136099347,
+      "learning_rate": 9.902827209241326e-06,
+      "loss": 0.249,
+      "step": 1000
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3561782992489233,
+      "learning_rate": 9.902537871490297e-06,
+      "loss": 0.2547,
+      "step": 1001
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8556978096661165,
+      "learning_rate": 9.902248107858747e-06,
+      "loss": 0.3023,
+      "step": 1002
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.935430286506648,
+      "learning_rate": 9.901957918371851e-06,
+      "loss": 0.2313,
+      "step": 1003
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.257278084056097,
+      "learning_rate": 9.901667303054814e-06,
+      "loss": 0.2312,
+      "step": 1004
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2722066456768424,
+      "learning_rate": 9.901376261932885e-06,
+      "loss": 0.29,
+      "step": 1005
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5281381534322533,
+      "learning_rate": 9.901084795031344e-06,
+      "loss": 0.267,
+      "step": 1006
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.193621687136732,
+      "learning_rate": 9.900792902375512e-06,
+      "loss": 0.2699,
+      "step": 1007
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7455931938218647,
+      "learning_rate": 9.900500583990744e-06,
+      "loss": 0.2959,
+      "step": 1008
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.613484087345668,
+      "learning_rate": 9.900207839902436e-06,
+      "loss": 0.2481,
+      "step": 1009
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0810597686054075,
+      "learning_rate": 9.899914670136016e-06,
+      "loss": 0.2623,
+      "step": 1010
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8907600619594698,
+      "learning_rate": 9.899621074716954e-06,
+      "loss": 0.2483,
+      "step": 1011
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0890777839868737,
+      "learning_rate": 9.899327053670751e-06,
+      "loss": 0.228,
+      "step": 1012
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9832103393375506,
+      "learning_rate": 9.899032607022952e-06,
+      "loss": 0.2415,
+      "step": 1013
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4087419607000222,
+      "learning_rate": 9.898737734799134e-06,
+      "loss": 0.2566,
+      "step": 1014
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3485029500827044,
+      "learning_rate": 9.89844243702491e-06,
+      "loss": 0.2733,
+      "step": 1015
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.799907698810353,
+      "learning_rate": 9.898146713725937e-06,
+      "loss": 0.2219,
+      "step": 1016
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.306839757064171,
+      "learning_rate": 9.8978505649279e-06,
+      "loss": 0.2681,
+      "step": 1017
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2337710986801267,
+      "learning_rate": 9.897553990656528e-06,
+      "loss": 0.2703,
+      "step": 1018
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.757213785073275,
+      "learning_rate": 9.897256990937583e-06,
+      "loss": 0.3208,
+      "step": 1019
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0881498453707765,
+      "learning_rate": 9.896959565796865e-06,
+      "loss": 0.2288,
+      "step": 1020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.033413308459793,
+      "learning_rate": 9.896661715260213e-06,
+      "loss": 0.2553,
+      "step": 1021
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4432396085433408,
+      "learning_rate": 9.896363439353499e-06,
+      "loss": 0.2794,
+      "step": 1022
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1218470624167893,
+      "learning_rate": 9.896064738102635e-06,
+      "loss": 0.2652,
+      "step": 1023
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3103158980156056,
+      "learning_rate": 9.895765611533568e-06,
+      "loss": 0.2933,
+      "step": 1024
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.40077800866619,
+      "learning_rate": 9.895466059672284e-06,
+      "loss": 0.2672,
+      "step": 1025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1978108022936005,
+      "learning_rate": 9.895166082544807e-06,
+      "loss": 0.2422,
+      "step": 1026
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.176225234842347,
+      "learning_rate": 9.89486568017719e-06,
+      "loss": 0.2486,
+      "step": 1027
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.139861385729894,
+      "learning_rate": 9.894564852595535e-06,
+      "loss": 0.2706,
+      "step": 1028
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.024639904274534,
+      "learning_rate": 9.89426359982597e-06,
+      "loss": 0.2248,
+      "step": 1029
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0409580415425523,
+      "learning_rate": 9.893961921894668e-06,
+      "loss": 0.2446,
+      "step": 1030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 29.481094777662676,
+      "learning_rate": 9.893659818827834e-06,
+      "loss": 0.2903,
+      "step": 1031
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3411641832779577,
+      "learning_rate": 9.893357290651712e-06,
+      "loss": 0.2682,
+      "step": 1032
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2154849374842884,
+      "learning_rate": 9.89305433739258e-06,
+      "loss": 0.2557,
+      "step": 1033
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.282477206424088,
+      "learning_rate": 9.89275095907676e-06,
+      "loss": 0.2586,
+      "step": 1034
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3687299584791632,
+      "learning_rate": 9.892447155730603e-06,
+      "loss": 0.2851,
+      "step": 1035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.11629746597544,
+      "learning_rate": 9.892142927380502e-06,
+      "loss": 0.2544,
+      "step": 1036
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.23760662814848,
+      "learning_rate": 9.891838274052882e-06,
+      "loss": 0.2473,
+      "step": 1037
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2476169173451708,
+      "learning_rate": 9.89153319577421e-06,
+      "loss": 0.2561,
+      "step": 1038
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.258902731779671,
+      "learning_rate": 9.89122769257099e-06,
+      "loss": 0.3059,
+      "step": 1039
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.652080752292637,
+      "learning_rate": 9.890921764469759e-06,
+      "loss": 0.2743,
+      "step": 1040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.452074246162822,
+      "learning_rate": 9.890615411497094e-06,
+      "loss": 0.2742,
+      "step": 1041
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.000403217156622,
+      "learning_rate": 9.890308633679604e-06,
+      "loss": 0.2455,
+      "step": 1042
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.19328378948334,
+      "learning_rate": 9.890001431043941e-06,
+      "loss": 0.2535,
+      "step": 1043
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4545661876685605,
+      "learning_rate": 9.889693803616793e-06,
+      "loss": 0.267,
+      "step": 1044
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8058250540756426,
+      "learning_rate": 9.889385751424882e-06,
+      "loss": 0.281,
+      "step": 1045
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.546375324065433,
+      "learning_rate": 9.889077274494967e-06,
+      "loss": 0.233,
+      "step": 1046
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9125999345132505,
+      "learning_rate": 9.888768372853849e-06,
+      "loss": 0.2319,
+      "step": 1047
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.01413418160032,
+      "learning_rate": 9.888459046528358e-06,
+      "loss": 0.2656,
+      "step": 1048
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6235588986227585,
+      "learning_rate": 9.888149295545367e-06,
+      "loss": 0.221,
+      "step": 1049
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.388174311186669,
+      "learning_rate": 9.887839119931783e-06,
+      "loss": 0.2877,
+      "step": 1050
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.25830270365746,
+      "learning_rate": 9.887528519714554e-06,
+      "loss": 0.2691,
+      "step": 1051
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.146037450476557,
+      "learning_rate": 9.887217494920655e-06,
+      "loss": 0.274,
+      "step": 1052
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4675280873830836,
+      "learning_rate": 9.886906045577111e-06,
+      "loss": 0.2902,
+      "step": 1053
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5954757774160786,
+      "learning_rate": 9.886594171710975e-06,
+      "loss": 0.252,
+      "step": 1054
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1538211304333394,
+      "learning_rate": 9.886281873349338e-06,
+      "loss": 0.2627,
+      "step": 1055
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.011755539575295,
+      "learning_rate": 9.885969150519332e-06,
+      "loss": 0.2458,
+      "step": 1056
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.434800138844234,
+      "learning_rate": 9.88565600324812e-06,
+      "loss": 0.2551,
+      "step": 1057
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.224425124524942,
+      "learning_rate": 9.885342431562907e-06,
+      "loss": 0.2516,
+      "step": 1058
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0796506590369255,
+      "learning_rate": 9.88502843549093e-06,
+      "loss": 0.23,
+      "step": 1059
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.59875495565925,
+      "learning_rate": 9.884714015059472e-06,
+      "loss": 0.27,
+      "step": 1060
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.446816424246436,
+      "learning_rate": 9.884399170295839e-06,
+      "loss": 0.2758,
+      "step": 1061
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6709401080186677,
+      "learning_rate": 9.884083901227387e-06,
+      "loss": 0.276,
+      "step": 1062
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.348570107464168,
+      "learning_rate": 9.883768207881498e-06,
+      "loss": 0.2543,
+      "step": 1063
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.323242320209527,
+      "learning_rate": 9.8834520902856e-06,
+      "loss": 0.2809,
+      "step": 1064
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2254075723025024,
+      "learning_rate": 9.883135548467155e-06,
+      "loss": 0.2396,
+      "step": 1065
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.234494384024653,
+      "learning_rate": 9.882818582453657e-06,
+      "loss": 0.2462,
+      "step": 1066
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0802528203355988,
+      "learning_rate": 9.882501192272642e-06,
+      "loss": 0.225,
+      "step": 1067
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.235901460182486,
+      "learning_rate": 9.882183377951683e-06,
+      "loss": 0.2901,
+      "step": 1068
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3730989499581585,
+      "learning_rate": 9.881865139518387e-06,
+      "loss": 0.2823,
+      "step": 1069
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2380187995918095,
+      "learning_rate": 9.8815464770004e-06,
+      "loss": 0.2252,
+      "step": 1070
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9575063275415405,
+      "learning_rate": 9.881227390425404e-06,
+      "loss": 0.2126,
+      "step": 1071
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.367039932864783,
+      "learning_rate": 9.880907879821115e-06,
+      "loss": 0.2389,
+      "step": 1072
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.403928509546694,
+      "learning_rate": 9.880587945215292e-06,
+      "loss": 0.3178,
+      "step": 1073
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.124601210349124,
+      "learning_rate": 9.880267586635726e-06,
+      "loss": 0.2979,
+      "step": 1074
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2124308545267084,
+      "learning_rate": 9.879946804110248e-06,
+      "loss": 0.2806,
+      "step": 1075
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2247892737126724,
+      "learning_rate": 9.879625597666721e-06,
+      "loss": 0.226,
+      "step": 1076
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.072346495670894,
+      "learning_rate": 9.879303967333053e-06,
+      "loss": 0.2681,
+      "step": 1077
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0390647081716544,
+      "learning_rate": 9.878981913137178e-06,
+      "loss": 0.2446,
+      "step": 1078
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8096996690715965,
+      "learning_rate": 9.878659435107078e-06,
+      "loss": 0.1933,
+      "step": 1079
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2478314529989514,
+      "learning_rate": 9.878336533270763e-06,
+      "loss": 0.2477,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8966634683247876,
+      "learning_rate": 9.878013207656285e-06,
+      "loss": 0.2187,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9379312569246974,
+      "learning_rate": 9.87768945829173e-06,
+      "loss": 0.2627,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.247422948684305,
+      "learning_rate": 9.87736528520522e-06,
+      "loss": 0.2769,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.145370943131256,
+      "learning_rate": 9.877040688424922e-06,
+      "loss": 0.2742,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9207543622331165,
+      "learning_rate": 9.876715667979026e-06,
+      "loss": 0.2456,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9626672840173347,
+      "learning_rate": 9.876390223895774e-06,
+      "loss": 0.2368,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0740793830612394,
+      "learning_rate": 9.87606435620343e-06,
+      "loss": 0.2419,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.202432612698249,
+      "learning_rate": 9.875738064930305e-06,
+      "loss": 0.2533,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.068348943939479,
+      "learning_rate": 9.875411350104745e-06,
+      "loss": 0.2431,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7994529309050415,
+      "learning_rate": 9.875084211755127e-06,
+      "loss": 0.273,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.094993820474425,
+      "learning_rate": 9.874756649909877e-06,
+      "loss": 0.2552,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.421990053116304,
+      "learning_rate": 9.874428664597444e-06,
+      "loss": 0.2584,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6811195491293667,
+      "learning_rate": 9.874100255846321e-06,
+      "loss": 0.2061,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5105282456220395,
+      "learning_rate": 9.873771423685037e-06,
+      "loss": 0.2912,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1829499290409946,
+      "learning_rate": 9.873442168142158e-06,
+      "loss": 0.2651,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8293048018507916,
+      "learning_rate": 9.873112489246286e-06,
+      "loss": 0.2112,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9748134768529795,
+      "learning_rate": 9.872782387026061e-06,
+      "loss": 0.2396,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0003203104271097,
+      "learning_rate": 9.872451861510157e-06,
+      "loss": 0.2598,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.328814951343707,
+      "learning_rate": 9.872120912727286e-06,
+      "loss": 0.2867,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6764398092307036,
+      "learning_rate": 9.8717895407062e-06,
+      "loss": 0.2308,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1102285081738597,
+      "learning_rate": 9.871457745475682e-06,
+      "loss": 0.2513,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.474952823633108,
+      "learning_rate": 9.871125527064559e-06,
+      "loss": 0.3014,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1945482815235877,
+      "learning_rate": 9.870792885501686e-06,
+      "loss": 0.2602,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7806017922969644,
+      "learning_rate": 9.87045982081596e-06,
+      "loss": 0.2249,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.889373640351625,
+      "learning_rate": 9.870126333036318e-06,
+      "loss": 0.2212,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.137399443034838,
+      "learning_rate": 9.869792422191727e-06,
+      "loss": 0.2845,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5451088105403756,
+      "learning_rate": 9.869458088311195e-06,
+      "loss": 0.2891,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8611552367886093,
+      "learning_rate": 9.869123331423763e-06,
+      "loss": 0.2412,
+      "step": 1108
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.139798328264714,
+      "learning_rate": 9.868788151558513e-06,
+      "loss": 0.2739,
+      "step": 1109
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.120058465279173,
+      "learning_rate": 9.868452548744563e-06,
+      "loss": 0.2506,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4306934130748865,
+      "learning_rate": 9.868116523011063e-06,
+      "loss": 0.2798,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0094312404190973,
+      "learning_rate": 9.867780074387207e-06,
+      "loss": 0.2247,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.121912789875263,
+      "learning_rate": 9.86744320290222e-06,
+      "loss": 0.2532,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8532544937814346,
+      "learning_rate": 9.867105908585366e-06,
+      "loss": 0.2292,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8893076443611174,
+      "learning_rate": 9.866768191465946e-06,
+      "loss": 0.1885,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3440291657838146,
+      "learning_rate": 9.866430051573296e-06,
+      "loss": 0.2732,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9496134563799656,
+      "learning_rate": 9.866091488936795e-06,
+      "loss": 0.2683,
+      "step": 1117
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0244362098680764,
+      "learning_rate": 9.865752503585848e-06,
+      "loss": 0.2496,
+      "step": 1118
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3000625365787393,
+      "learning_rate": 9.865413095549903e-06,
+      "loss": 0.2937,
+      "step": 1119
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2077587746808645,
+      "learning_rate": 9.865073264858447e-06,
+      "loss": 0.2792,
+      "step": 1120
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.880856128571587,
+      "learning_rate": 9.864733011541e-06,
+      "loss": 0.2744,
+      "step": 1121
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4214701914593606,
+      "learning_rate": 9.864392335627118e-06,
+      "loss": 0.2405,
+      "step": 1122
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7103203951537567,
+      "learning_rate": 9.864051237146395e-06,
+      "loss": 0.2235,
+      "step": 1123
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.946075951602688,
+      "learning_rate": 9.863709716128465e-06,
+      "loss": 0.2324,
+      "step": 1124
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.937384906212223,
+      "learning_rate": 9.863367772602994e-06,
+      "loss": 0.2448,
+      "step": 1125
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.173508622376695,
+      "learning_rate": 9.863025406599686e-06,
+      "loss": 0.2401,
+      "step": 1126
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.74269730960683,
+      "learning_rate": 9.862682618148286e-06,
+      "loss": 0.2584,
+      "step": 1127
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.993651776792849,
+      "learning_rate": 9.862339407278564e-06,
+      "loss": 0.2663,
+      "step": 1128
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2336638887277633,
+      "learning_rate": 9.861995774020341e-06,
+      "loss": 0.2485,
+      "step": 1129
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.873204241018341,
+      "learning_rate": 9.861651718403466e-06,
+      "loss": 0.2262,
+      "step": 1130
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1366089316912515,
+      "learning_rate": 9.861307240457828e-06,
+      "loss": 0.2602,
+      "step": 1131
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8291904007710147,
+      "learning_rate": 9.86096234021335e-06,
+      "loss": 0.2749,
+      "step": 1132
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2039991518725994,
+      "learning_rate": 9.860617017699993e-06,
+      "loss": 0.2952,
+      "step": 1133
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.72618487498062,
+      "learning_rate": 9.86027127294776e-06,
+      "loss": 0.227,
+      "step": 1134
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.941846000529975,
+      "learning_rate": 9.859925105986677e-06,
+      "loss": 0.2487,
+      "step": 1135
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5160975179268226,
+      "learning_rate": 9.859578516846822e-06,
+      "loss": 0.2442,
+      "step": 1136
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9299841378485487,
+      "learning_rate": 9.859231505558301e-06,
+      "loss": 0.2535,
+      "step": 1137
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.124378700379731,
+      "learning_rate": 9.858884072151258e-06,
+      "loss": 0.2476,
+      "step": 1138
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1642850268511196,
+      "learning_rate": 9.858536216655875e-06,
+      "loss": 0.2761,
+      "step": 1139
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.354830281022868,
+      "learning_rate": 9.85818793910237e-06,
+      "loss": 0.2496,
+      "step": 1140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9455152347080973,
+      "learning_rate": 9.857839239521e-06,
+      "loss": 0.2222,
+      "step": 1141
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0777993632048695,
+      "learning_rate": 9.85749011794205e-06,
+      "loss": 0.2382,
+      "step": 1142
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.288589623666548,
+      "learning_rate": 9.857140574395854e-06,
+      "loss": 0.2385,
+      "step": 1143
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0879987681864836,
+      "learning_rate": 9.856790608912775e-06,
+      "loss": 0.24,
+      "step": 1144
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7891813505355767,
+      "learning_rate": 9.856440221523211e-06,
+      "loss": 0.2225,
+      "step": 1145
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.930758829384716,
+      "learning_rate": 9.856089412257605e-06,
+      "loss": 0.2209,
+      "step": 1146
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9075032519743202,
+      "learning_rate": 9.855738181146427e-06,
+      "loss": 0.221,
+      "step": 1147
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.799038534038176,
+      "learning_rate": 9.855386528220194e-06,
+      "loss": 0.2453,
+      "step": 1148
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.031666778266753,
+      "learning_rate": 9.855034453509449e-06,
+      "loss": 0.2664,
+      "step": 1149
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.101983093152087,
+      "learning_rate": 9.854681957044779e-06,
+      "loss": 0.2506,
+      "step": 1150
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.538405884548164,
+      "learning_rate": 9.854329038856802e-06,
+      "loss": 0.295,
+      "step": 1151
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9999293062537107,
+      "learning_rate": 9.85397569897618e-06,
+      "loss": 0.2355,
+      "step": 1152
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0846498570996985,
+      "learning_rate": 9.853621937433603e-06,
+      "loss": 0.2286,
+      "step": 1153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0189315640607655,
+      "learning_rate": 9.853267754259808e-06,
+      "loss": 0.259,
+      "step": 1154
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.186436806117984,
+      "learning_rate": 9.852913149485556e-06,
+      "loss": 0.2426,
+      "step": 1155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9655259808621697,
+      "learning_rate": 9.852558123141656e-06,
+      "loss": 0.2479,
+      "step": 1156
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3247150840206916,
+      "learning_rate": 9.852202675258946e-06,
+      "loss": 0.2721,
+      "step": 1157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2023669493726996,
+      "learning_rate": 9.851846805868307e-06,
+      "loss": 0.2846,
+      "step": 1158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.972721170761225,
+      "learning_rate": 9.851490515000648e-06,
+      "loss": 0.2294,
+      "step": 1159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.157474039139291,
+      "learning_rate": 9.851133802686925e-06,
+      "loss": 0.2699,
+      "step": 1160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2081711511612174,
+      "learning_rate": 9.850776668958122e-06,
+      "loss": 0.2648,
+      "step": 1161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.87312505864339,
+      "learning_rate": 9.850419113845265e-06,
+      "loss": 0.2186,
+      "step": 1162
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.641740307924866,
+      "learning_rate": 9.850061137379414e-06,
+      "loss": 0.2877,
+      "step": 1163
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.399733383315742,
+      "learning_rate": 9.849702739591665e-06,
+      "loss": 0.293,
+      "step": 1164
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3365076997579006,
+      "learning_rate": 9.849343920513152e-06,
+      "loss": 0.2486,
+      "step": 1165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1083814329872936,
+      "learning_rate": 9.848984680175049e-06,
+      "loss": 0.235,
+      "step": 1166
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4197356076260603,
+      "learning_rate": 9.848625018608558e-06,
+      "loss": 0.3014,
+      "step": 1167
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1363940715215177,
+      "learning_rate": 9.848264935844924e-06,
+      "loss": 0.2725,
+      "step": 1168
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.010372502048954,
+      "learning_rate": 9.84790443191543e-06,
+      "loss": 0.2624,
+      "step": 1169
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.035017450320761,
+      "learning_rate": 9.84754350685139e-06,
+      "loss": 0.242,
+      "step": 1170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.973624503013562,
+      "learning_rate": 9.847182160684158e-06,
+      "loss": 0.2598,
+      "step": 1171
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2661895519406405,
+      "learning_rate": 9.846820393445125e-06,
+      "loss": 0.2409,
+      "step": 1172
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.378157190162606,
+      "learning_rate": 9.846458205165715e-06,
+      "loss": 0.2622,
+      "step": 1173
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3415247347195054,
+      "learning_rate": 9.846095595877392e-06,
+      "loss": 0.2468,
+      "step": 1174
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.267267724576772,
+      "learning_rate": 9.845732565611657e-06,
+      "loss": 0.2569,
+      "step": 1175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2069423372117574,
+      "learning_rate": 9.845369114400045e-06,
+      "loss": 0.2472,
+      "step": 1176
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1745211744508137,
+      "learning_rate": 9.84500524227413e-06,
+      "loss": 0.2757,
+      "step": 1177
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9301821375250134,
+      "learning_rate": 9.844640949265521e-06,
+      "loss": 0.2447,
+      "step": 1178
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.270593293447514,
+      "learning_rate": 9.844276235405861e-06,
+      "loss": 0.2501,
+      "step": 1179
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8292365517932136,
+      "learning_rate": 9.843911100726838e-06,
+      "loss": 0.2451,
+      "step": 1180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.24458190791579,
+      "learning_rate": 9.843545545260166e-06,
+      "loss": 0.255,
+      "step": 1181
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9921071614153956,
+      "learning_rate": 9.843179569037601e-06,
+      "loss": 0.2558,
+      "step": 1182
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.120135504640428,
+      "learning_rate": 9.84281317209094e-06,
+      "loss": 0.2527,
+      "step": 1183
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9574435847374505,
+      "learning_rate": 9.842446354452007e-06,
+      "loss": 0.2607,
+      "step": 1184
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.00655728891264,
+      "learning_rate": 9.84207911615267e-06,
+      "loss": 0.2348,
+      "step": 1185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.564259275203512,
+      "learning_rate": 9.841711457224827e-06,
+      "loss": 0.2315,
+      "step": 1186
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.178516775839786,
+      "learning_rate": 9.84134337770042e-06,
+      "loss": 0.2604,
+      "step": 1187
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9118185271293666,
+      "learning_rate": 9.840974877611423e-06,
+      "loss": 0.2391,
+      "step": 1188
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.999488208569082,
+      "learning_rate": 9.840605956989846e-06,
+      "loss": 0.2473,
+      "step": 1189
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.913629322338597,
+      "learning_rate": 9.840236615867738e-06,
+      "loss": 0.2154,
+      "step": 1190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.884895425217116,
+      "learning_rate": 9.839866854277182e-06,
+      "loss": 0.2496,
+      "step": 1191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0894615152157927,
+      "learning_rate": 9.839496672250301e-06,
+      "loss": 0.2567,
+      "step": 1192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.068759534760437,
+      "learning_rate": 9.839126069819254e-06,
+      "loss": 0.2507,
+      "step": 1193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.187900604153824,
+      "learning_rate": 9.838755047016229e-06,
+      "loss": 0.2576,
+      "step": 1194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9566174388486464,
+      "learning_rate": 9.838383603873463e-06,
+      "loss": 0.2433,
+      "step": 1195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9759192642662096,
+      "learning_rate": 9.838011740423219e-06,
+      "loss": 0.239,
+      "step": 1196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.972831661343069,
+      "learning_rate": 9.837639456697802e-06,
+      "loss": 0.232,
+      "step": 1197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7750806336622,
+      "learning_rate": 9.837266752729552e-06,
+      "loss": 0.2367,
+      "step": 1198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.827691379197103,
+      "learning_rate": 9.836893628550846e-06,
+      "loss": 0.2408,
+      "step": 1199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.815658170141337,
+      "learning_rate": 9.836520084194097e-06,
+      "loss": 0.2312,
+      "step": 1200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.057464773493854,
+      "learning_rate": 9.836146119691752e-06,
+      "loss": 0.2744,
+      "step": 1201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8656367413290393,
+      "learning_rate": 9.8357717350763e-06,
+      "loss": 0.2372,
+      "step": 1202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2514347865106568,
+      "learning_rate": 9.835396930380264e-06,
+      "loss": 0.2597,
+      "step": 1203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.955441368222323,
+      "learning_rate": 9.835021705636201e-06,
+      "loss": 0.2633,
+      "step": 1204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0960740638895206,
+      "learning_rate": 9.834646060876707e-06,
+      "loss": 0.2399,
+      "step": 1205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1750070887087336,
+      "learning_rate": 9.834269996134416e-06,
+      "loss": 0.2956,
+      "step": 1206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.628820775880156,
+      "learning_rate": 9.833893511441993e-06,
+      "loss": 0.2576,
+      "step": 1207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1801194195762683,
+      "learning_rate": 9.833516606832146e-06,
+      "loss": 0.29,
+      "step": 1208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.053432387455168,
+      "learning_rate": 9.833139282337615e-06,
+      "loss": 0.2735,
+      "step": 1209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0218155010339225,
+      "learning_rate": 9.832761537991177e-06,
+      "loss": 0.269,
+      "step": 1210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.163316769954804,
+      "learning_rate": 9.83238337382565e-06,
+      "loss": 0.2359,
+      "step": 1211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.341562744210851,
+      "learning_rate": 9.832004789873883e-06,
+      "loss": 0.2668,
+      "step": 1212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0137565479094057,
+      "learning_rate": 9.831625786168762e-06,
+      "loss": 0.2183,
+      "step": 1213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.134347588938876,
+      "learning_rate": 9.83124636274321e-06,
+      "loss": 0.2778,
+      "step": 1214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3810173760985975,
+      "learning_rate": 9.830866519630191e-06,
+      "loss": 0.2521,
+      "step": 1215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0092605989494436,
+      "learning_rate": 9.8304862568627e-06,
+      "loss": 0.2383,
+      "step": 1216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7881943397728652,
+      "learning_rate": 9.83010557447377e-06,
+      "loss": 0.2317,
+      "step": 1217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7706960576828523,
+      "learning_rate": 9.829724472496471e-06,
+      "loss": 0.2408,
+      "step": 1218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.931715255855614,
+      "learning_rate": 9.829342950963908e-06,
+      "loss": 0.2207,
+      "step": 1219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.164116215641434,
+      "learning_rate": 9.828961009909225e-06,
+      "loss": 0.2732,
+      "step": 1220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.843276605833962,
+      "learning_rate": 9.8285786493656e-06,
+      "loss": 0.2543,
+      "step": 1221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0044667189776066,
+      "learning_rate": 9.82819586936625e-06,
+      "loss": 0.2552,
+      "step": 1222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.415275420698183,
+      "learning_rate": 9.827812669944423e-06,
+      "loss": 0.2491,
+      "step": 1223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.848663639697217,
+      "learning_rate": 9.827429051133412e-06,
+      "loss": 0.2186,
+      "step": 1224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1255228590451396,
+      "learning_rate": 9.82704501296654e-06,
+      "loss": 0.2589,
+      "step": 1225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1109261829573573,
+      "learning_rate": 9.826660555477167e-06,
+      "loss": 0.2406,
+      "step": 1226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.130609667597135,
+      "learning_rate": 9.82627567869869e-06,
+      "loss": 0.2716,
+      "step": 1227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8743900904881006,
+      "learning_rate": 9.825890382664547e-06,
+      "loss": 0.2128,
+      "step": 1228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.134640376995852,
+      "learning_rate": 9.825504667408205e-06,
+      "loss": 0.2535,
+      "step": 1229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0572276927796738,
+      "learning_rate": 9.825118532963172e-06,
+      "loss": 0.2477,
+      "step": 1230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9615852301376675,
+      "learning_rate": 9.824731979362991e-06,
+      "loss": 0.2404,
+      "step": 1231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.195671642915052,
+      "learning_rate": 9.824345006641243e-06,
+      "loss": 0.2843,
+      "step": 1232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.933194017097645,
+      "learning_rate": 9.82395761483154e-06,
+      "loss": 0.2337,
+      "step": 1233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9125056938085208,
+      "learning_rate": 9.823569803967538e-06,
+      "loss": 0.2115,
+      "step": 1234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.588040278512406,
+      "learning_rate": 9.823181574082927e-06,
+      "loss": 0.2162,
+      "step": 1235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9811209941656434,
+      "learning_rate": 9.822792925211429e-06,
+      "loss": 0.2504,
+      "step": 1236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0058377718173963,
+      "learning_rate": 9.822403857386808e-06,
+      "loss": 0.2644,
+      "step": 1237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1204325390320045,
+      "learning_rate": 9.822014370642861e-06,
+      "loss": 0.2271,
+      "step": 1238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0256261042478316,
+      "learning_rate": 9.821624465013422e-06,
+      "loss": 0.2624,
+      "step": 1239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0178108233260814,
+      "learning_rate": 9.821234140532363e-06,
+      "loss": 0.2617,
+      "step": 1240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8272765368690966,
+      "learning_rate": 9.82084339723359e-06,
+      "loss": 0.259,
+      "step": 1241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8937466377850645,
+      "learning_rate": 9.82045223515105e-06,
+      "loss": 0.244,
+      "step": 1242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.969029717164541,
+      "learning_rate": 9.820060654318718e-06,
+      "loss": 0.2304,
+      "step": 1243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9931941186219126,
+      "learning_rate": 9.819668654770613e-06,
+      "loss": 0.2354,
+      "step": 1244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.389394950149006,
+      "learning_rate": 9.81927623654079e-06,
+      "loss": 0.2686,
+      "step": 1245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6485410148762205,
+      "learning_rate": 9.818883399663333e-06,
+      "loss": 0.2106,
+      "step": 1246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9073598792710653,
+      "learning_rate": 9.818490144172372e-06,
+      "loss": 0.2443,
+      "step": 1247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.364372678645247,
+      "learning_rate": 9.818096470102067e-06,
+      "loss": 0.2809,
+      "step": 1248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.551128418696761,
+      "learning_rate": 9.817702377486616e-06,
+      "loss": 0.2588,
+      "step": 1249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1391068765915704,
+      "learning_rate": 9.817307866360255e-06,
+      "loss": 0.2502,
+      "step": 1250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9030071165063744,
+      "learning_rate": 9.816912936757252e-06,
+      "loss": 0.2261,
+      "step": 1251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0904473213877157,
+      "learning_rate": 9.816517588711918e-06,
+      "loss": 0.2609,
+      "step": 1252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.141150929340751,
+      "learning_rate": 9.816121822258595e-06,
+      "loss": 0.2798,
+      "step": 1253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.645682764054394,
+      "learning_rate": 9.815725637431663e-06,
+      "loss": 0.2083,
+      "step": 1254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.771041108203385,
+      "learning_rate": 9.815329034265537e-06,
+      "loss": 0.2562,
+      "step": 1255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0609998682706734,
+      "learning_rate": 9.81493201279467e-06,
+      "loss": 0.2606,
+      "step": 1256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1785830974871714,
+      "learning_rate": 9.814534573053554e-06,
+      "loss": 0.2162,
+      "step": 1257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.781839154828549,
+      "learning_rate": 9.814136715076712e-06,
+      "loss": 0.2392,
+      "step": 1258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.732361163820835,
+      "learning_rate": 9.813738438898705e-06,
+      "loss": 0.217,
+      "step": 1259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.368933313189031,
+      "learning_rate": 9.813339744554134e-06,
+      "loss": 0.288,
+      "step": 1260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.622675303832583,
+      "learning_rate": 9.812940632077629e-06,
+      "loss": 0.2519,
+      "step": 1261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4208772684634035,
+      "learning_rate": 9.812541101503863e-06,
+      "loss": 0.2604,
+      "step": 1262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.113738877387231,
+      "learning_rate": 9.812141152867545e-06,
+      "loss": 0.2442,
+      "step": 1263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.139036950616636,
+      "learning_rate": 9.811740786203414e-06,
+      "loss": 0.2316,
+      "step": 1264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.419167717116625,
+      "learning_rate": 9.811340001546252e-06,
+      "loss": 0.2289,
+      "step": 1265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2143647000980833,
+      "learning_rate": 9.810938798930876e-06,
+      "loss": 0.2416,
+      "step": 1266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9808804309636843,
+      "learning_rate": 9.810537178392137e-06,
+      "loss": 0.2415,
+      "step": 1267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9939559005988365,
+      "learning_rate": 9.810135139964922e-06,
+      "loss": 0.2214,
+      "step": 1268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7757786289149005,
+      "learning_rate": 9.809732683684159e-06,
+      "loss": 0.2633,
+      "step": 1269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8083520955299472,
+      "learning_rate": 9.809329809584808e-06,
+      "loss": 0.2725,
+      "step": 1270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8358255488788093,
+      "learning_rate": 9.808926517701865e-06,
+      "loss": 0.213,
+      "step": 1271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.902968770899238,
+      "learning_rate": 9.808522808070365e-06,
+      "loss": 0.2064,
+      "step": 1272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.972331018988225,
+      "learning_rate": 9.808118680725376e-06,
+      "loss": 0.2501,
+      "step": 1273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.48357951576396,
+      "learning_rate": 9.807714135702008e-06,
+      "loss": 0.2162,
+      "step": 1274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.872521740272126,
+      "learning_rate": 9.8073091730354e-06,
+      "loss": 0.2495,
+      "step": 1275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7941610101784353,
+      "learning_rate": 9.806903792760733e-06,
+      "loss": 0.2341,
+      "step": 1276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.920416484684805,
+      "learning_rate": 9.806497994913223e-06,
+      "loss": 0.2274,
+      "step": 1277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.036458782351314,
+      "learning_rate": 9.806091779528119e-06,
+      "loss": 0.2396,
+      "step": 1278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0625407075135818,
+      "learning_rate": 9.80568514664071e-06,
+      "loss": 0.2218,
+      "step": 1279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.096712644925063,
+      "learning_rate": 9.805278096286318e-06,
+      "loss": 0.2557,
+      "step": 1280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7830752639758587,
+      "learning_rate": 9.804870628500306e-06,
+      "loss": 0.201,
+      "step": 1281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7330453381241333,
+      "learning_rate": 9.80446274331807e-06,
+      "loss": 0.2338,
+      "step": 1282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.309180277266247,
+      "learning_rate": 9.80405444077504e-06,
+      "loss": 0.292,
+      "step": 1283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.935178279963719,
+      "learning_rate": 9.803645720906689e-06,
+      "loss": 0.2319,
+      "step": 1284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.942633588779663,
+      "learning_rate": 9.80323658374852e-06,
+      "loss": 0.2355,
+      "step": 1285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7224101923256874,
+      "learning_rate": 9.802827029336076e-06,
+      "loss": 0.2756,
+      "step": 1286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3284034316773035,
+      "learning_rate": 9.80241705770493e-06,
+      "loss": 0.2635,
+      "step": 1287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.199097902914095,
+      "learning_rate": 9.802006668890702e-06,
+      "loss": 0.2569,
+      "step": 1288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.969066947405044,
+      "learning_rate": 9.80159586292904e-06,
+      "loss": 0.2515,
+      "step": 1289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.995045862758655,
+      "learning_rate": 9.80118463985563e-06,
+      "loss": 0.266,
+      "step": 1290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.146356230662008,
+      "learning_rate": 9.800772999706194e-06,
+      "loss": 0.2127,
+      "step": 1291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.84808712157093,
+      "learning_rate": 9.800360942516492e-06,
+      "loss": 0.2042,
+      "step": 1292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1417815974339587,
+      "learning_rate": 9.79994846832232e-06,
+      "loss": 0.2511,
+      "step": 1293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.846892211007102,
+      "learning_rate": 9.799535577159508e-06,
+      "loss": 0.2333,
+      "step": 1294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9185759963950715,
+      "learning_rate": 9.799122269063923e-06,
+      "loss": 0.2611,
+      "step": 1295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3653031968848564,
+      "learning_rate": 9.798708544071471e-06,
+      "loss": 0.2626,
+      "step": 1296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8032348067385886,
+      "learning_rate": 9.798294402218092e-06,
+      "loss": 0.208,
+      "step": 1297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.381782583910323,
+      "learning_rate": 9.797879843539759e-06,
+      "loss": 0.268,
+      "step": 1298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.956812105563334,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 0.2545,
+      "step": 1299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.291638112071125,
+      "learning_rate": 9.797049475852326e-06,
+      "loss": 0.2421,
+      "step": 1300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9288565883410804,
+      "learning_rate": 9.79663366691536e-06,
+      "loss": 0.2177,
+      "step": 1301
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0594606247297653,
+      "learning_rate": 9.796217441297704e-06,
+      "loss": 0.229,
+      "step": 1302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0042457235324296,
+      "learning_rate": 9.795800799035524e-06,
+      "loss": 0.2296,
+      "step": 1303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8891347001244454,
+      "learning_rate": 9.79538374016501e-06,
+      "loss": 0.2425,
+      "step": 1304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2146910253200183,
+      "learning_rate": 9.794966264722393e-06,
+      "loss": 0.2712,
+      "step": 1305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.854955435082528,
+      "learning_rate": 9.794548372743933e-06,
+      "loss": 0.2385,
+      "step": 1306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.713036302965482,
+      "learning_rate": 9.79413006426594e-06,
+      "loss": 0.3128,
+      "step": 1307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.010742657640395,
+      "learning_rate": 9.793711339324747e-06,
+      "loss": 0.2378,
+      "step": 1308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7513893304966746,
+      "learning_rate": 9.793292197956732e-06,
+      "loss": 0.2316,
+      "step": 1309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.632643880424819,
+      "learning_rate": 9.792872640198304e-06,
+      "loss": 0.2244,
+      "step": 1310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.121034875998835,
+      "learning_rate": 9.792452666085907e-06,
+      "loss": 0.2065,
+      "step": 1311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.916013310519197,
+      "learning_rate": 9.792032275656027e-06,
+      "loss": 0.2388,
+      "step": 1312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6940096650580627,
+      "learning_rate": 9.791611468945183e-06,
+      "loss": 0.2112,
+      "step": 1313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2924772744135278,
+      "learning_rate": 9.791190245989928e-06,
+      "loss": 0.2328,
+      "step": 1314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.312972018158128,
+      "learning_rate": 9.790768606826857e-06,
+      "loss": 0.2883,
+      "step": 1315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8283500515112956,
+      "learning_rate": 9.790346551492594e-06,
+      "loss": 0.2352,
+      "step": 1316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3789689431948906,
+      "learning_rate": 9.789924080023805e-06,
+      "loss": 0.2539,
+      "step": 1317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.025163570085979,
+      "learning_rate": 9.789501192457188e-06,
+      "loss": 0.2665,
+      "step": 1318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7201646541561666,
+      "learning_rate": 9.789077888829481e-06,
+      "loss": 0.2263,
+      "step": 1319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.105134247206127,
+      "learning_rate": 9.788654169177454e-06,
+      "loss": 0.239,
+      "step": 1320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9937715566232086,
+      "learning_rate": 9.788230033537918e-06,
+      "loss": 0.2326,
+      "step": 1321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8930068684455095,
+      "learning_rate": 9.787805481947715e-06,
+      "loss": 0.2618,
+      "step": 1322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8532096304885988,
+      "learning_rate": 9.787380514443727e-06,
+      "loss": 0.2379,
+      "step": 1323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.306251307084509,
+      "learning_rate": 9.78695513106287e-06,
+      "loss": 0.2648,
+      "step": 1324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.88144555529565,
+      "learning_rate": 9.786529331842096e-06,
+      "loss": 0.2249,
+      "step": 1325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8465201053620137,
+      "learning_rate": 9.786103116818394e-06,
+      "loss": 0.2161,
+      "step": 1326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6155688024809303,
+      "learning_rate": 9.78567648602879e-06,
+      "loss": 0.2117,
+      "step": 1327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2467769788155234,
+      "learning_rate": 9.785249439510348e-06,
+      "loss": 0.2793,
+      "step": 1328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.330731261149586,
+      "learning_rate": 9.784821977300159e-06,
+      "loss": 0.2838,
+      "step": 1329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9462581289474934,
+      "learning_rate": 9.78439409943536e-06,
+      "loss": 0.2699,
+      "step": 1330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.501323978467036,
+      "learning_rate": 9.78396580595312e-06,
+      "loss": 0.1978,
+      "step": 1331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.880850440291398,
+      "learning_rate": 9.783537096890647e-06,
+      "loss": 0.2732,
+      "step": 1332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.846927591391255,
+      "learning_rate": 9.783107972285177e-06,
+      "loss": 0.2391,
+      "step": 1333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.051728788517442,
+      "learning_rate": 9.782678432173992e-06,
+      "loss": 0.2244,
+      "step": 1334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.933609749329675,
+      "learning_rate": 9.782248476594408e-06,
+      "loss": 0.2356,
+      "step": 1335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.773405606029093,
+      "learning_rate": 9.781818105583771e-06,
+      "loss": 0.2244,
+      "step": 1336
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.846571843676018,
+      "learning_rate": 9.781387319179465e-06,
+      "loss": 0.232,
+      "step": 1337
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.011532741673929,
+      "learning_rate": 9.780956117418919e-06,
+      "loss": 0.2533,
+      "step": 1338
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.760565626444453,
+      "learning_rate": 9.780524500339585e-06,
+      "loss": 0.2161,
+      "step": 1339
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.92150953272804,
+      "learning_rate": 9.78009246797896e-06,
+      "loss": 0.2704,
+      "step": 1340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0816136912354355,
+      "learning_rate": 9.779660020374577e-06,
+      "loss": 0.2523,
+      "step": 1341
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9256642668861224,
+      "learning_rate": 9.779227157563998e-06,
+      "loss": 0.2247,
+      "step": 1342
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7818828946933163,
+      "learning_rate": 9.778793879584828e-06,
+      "loss": 0.2412,
+      "step": 1343
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6182256789978062,
+      "learning_rate": 9.778360186474703e-06,
+      "loss": 0.2242,
+      "step": 1344
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.780293571580932,
+      "learning_rate": 9.7779260782713e-06,
+      "loss": 0.2386,
+      "step": 1345
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7916525125110274,
+      "learning_rate": 9.777491555012331e-06,
+      "loss": 0.2148,
+      "step": 1346
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8414558705044697,
+      "learning_rate": 9.777056616735539e-06,
+      "loss": 0.2277,
+      "step": 1347
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7084149370794495,
+      "learning_rate": 9.77662126347871e-06,
+      "loss": 0.2204,
+      "step": 1348
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1534858821209246,
+      "learning_rate": 9.776185495279662e-06,
+      "loss": 0.2169,
+      "step": 1349
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.887329227267067,
+      "learning_rate": 9.775749312176249e-06,
+      "loss": 0.2392,
+      "step": 1350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0635723562757406,
+      "learning_rate": 9.77531271420636e-06,
+      "loss": 0.2363,
+      "step": 1351
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.723383931741273,
+      "learning_rate": 9.774875701407928e-06,
+      "loss": 0.2119,
+      "step": 1352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.101023708212512,
+      "learning_rate": 9.77443827381891e-06,
+      "loss": 0.2357,
+      "step": 1353
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1334044743601526,
+      "learning_rate": 9.774000431477311e-06,
+      "loss": 0.2711,
+      "step": 1354
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.963598890267245,
+      "learning_rate": 9.77356217442116e-06,
+      "loss": 0.2795,
+      "step": 1355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.4607461488362254,
+      "learning_rate": 9.773123502688532e-06,
+      "loss": 0.2721,
+      "step": 1356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1199559666763776,
+      "learning_rate": 9.772684416317534e-06,
+      "loss": 0.2609,
+      "step": 1357
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.881933361469555,
+      "learning_rate": 9.772244915346307e-06,
+      "loss": 0.2529,
+      "step": 1358
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.884999408927426,
+      "learning_rate": 9.771804999813033e-06,
+      "loss": 0.2291,
+      "step": 1359
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1996217454945173,
+      "learning_rate": 9.771364669755923e-06,
+      "loss": 0.2423,
+      "step": 1360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9066144554302924,
+      "learning_rate": 9.770923925213232e-06,
+      "loss": 0.2452,
+      "step": 1361
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.831117241446732,
+      "learning_rate": 9.770482766223246e-06,
+      "loss": 0.2457,
+      "step": 1362
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9745718789362137,
+      "learning_rate": 9.77004119282429e-06,
+      "loss": 0.239,
+      "step": 1363
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9381154286601836,
+      "learning_rate": 9.76959920505472e-06,
+      "loss": 0.2322,
+      "step": 1364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.207748380942559,
+      "learning_rate": 9.769156802952932e-06,
+      "loss": 0.251,
+      "step": 1365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.82081950408619,
+      "learning_rate": 9.768713986557359e-06,
+      "loss": 0.2391,
+      "step": 1366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.809456415586235,
+      "learning_rate": 9.768270755906467e-06,
+      "loss": 0.2528,
+      "step": 1367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3635368577684384,
+      "learning_rate": 9.767827111038757e-06,
+      "loss": 0.2137,
+      "step": 1368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.745610774592366,
+      "learning_rate": 9.767383051992774e-06,
+      "loss": 0.2456,
+      "step": 1369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5318381433366692,
+      "learning_rate": 9.766938578807088e-06,
+      "loss": 0.2175,
+      "step": 1370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6889640603088245,
+      "learning_rate": 9.766493691520312e-06,
+      "loss": 0.1818,
+      "step": 1371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5748940285815154,
+      "learning_rate": 9.766048390171091e-06,
+      "loss": 0.226,
+      "step": 1372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.658609980502939,
+      "learning_rate": 9.765602674798112e-06,
+      "loss": 0.2314,
+      "step": 1373
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7808521318830186,
+      "learning_rate": 9.76515654544009e-06,
+      "loss": 0.2225,
+      "step": 1374
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1063757115858017,
+      "learning_rate": 9.764710002135784e-06,
+      "loss": 0.242,
+      "step": 1375
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.979320561193384,
+      "learning_rate": 9.764263044923983e-06,
+      "loss": 0.2046,
+      "step": 1376
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.726814486021948,
+      "learning_rate": 9.763815673843511e-06,
+      "loss": 0.2154,
+      "step": 1377
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.08736527989591,
+      "learning_rate": 9.763367888933235e-06,
+      "loss": 0.2973,
+      "step": 1378
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.378369187913458,
+      "learning_rate": 9.762919690232053e-06,
+      "loss": 0.2429,
+      "step": 1379
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6106792037935933,
+      "learning_rate": 9.762471077778898e-06,
+      "loss": 0.2022,
+      "step": 1380
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.028391364636959,
+      "learning_rate": 9.762022051612742e-06,
+      "loss": 0.2316,
+      "step": 1381
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.64027256655154,
+      "learning_rate": 9.761572611772592e-06,
+      "loss": 0.1984,
+      "step": 1382
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3860530705961587,
+      "learning_rate": 9.76112275829749e-06,
+      "loss": 0.2753,
+      "step": 1383
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3926531943143763,
+      "learning_rate": 9.760672491226515e-06,
+      "loss": 0.1956,
+      "step": 1384
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6780198406056708,
+      "learning_rate": 9.76022181059878e-06,
+      "loss": 0.2218,
+      "step": 1385
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.014662980555568,
+      "learning_rate": 9.759770716453436e-06,
+      "loss": 0.2635,
+      "step": 1386
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8401462366944754,
+      "learning_rate": 9.759319208829671e-06,
+      "loss": 0.2184,
+      "step": 1387
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.60835863599508,
+      "learning_rate": 9.758867287766705e-06,
+      "loss": 0.2568,
+      "step": 1388
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.441044084131762,
+      "learning_rate": 9.758414953303796e-06,
+      "loss": 0.2675,
+      "step": 1389
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.880216438282025,
+      "learning_rate": 9.75796220548024e-06,
+      "loss": 0.2523,
+      "step": 1390
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.762354414228669,
+      "learning_rate": 9.757509044335367e-06,
+      "loss": 0.2574,
+      "step": 1391
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.10827643677901,
+      "learning_rate": 9.757055469908541e-06,
+      "loss": 0.2631,
+      "step": 1392
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.45109620298898,
+      "learning_rate": 9.756601482239162e-06,
+      "loss": 0.2915,
+      "step": 1393
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.814712553859058,
+      "learning_rate": 9.756147081366673e-06,
+      "loss": 0.251,
+      "step": 1394
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9052448870008205,
+      "learning_rate": 9.755692267330542e-06,
+      "loss": 0.237,
+      "step": 1395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.041018731645638,
+      "learning_rate": 9.755237040170284e-06,
+      "loss": 0.2328,
+      "step": 1396
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.988361672076705,
+      "learning_rate": 9.754781399925439e-06,
+      "loss": 0.2374,
+      "step": 1397
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6507589502827287,
+      "learning_rate": 9.754325346635592e-06,
+      "loss": 0.2248,
+      "step": 1398
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.881351900232617,
+      "learning_rate": 9.753868880340359e-06,
+      "loss": 0.2717,
+      "step": 1399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1008929889260797,
+      "learning_rate": 9.75341200107939e-06,
+      "loss": 0.2652,
+      "step": 1400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.627201150191938,
+      "learning_rate": 9.752954708892379e-06,
+      "loss": 0.2109,
+      "step": 1401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.789514054087141,
+      "learning_rate": 9.752497003819047e-06,
+      "loss": 0.221,
+      "step": 1402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7638005097773566,
+      "learning_rate": 9.752038885899154e-06,
+      "loss": 0.2154,
+      "step": 1403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.682996603419557,
+      "learning_rate": 9.7515803551725e-06,
+      "loss": 0.2129,
+      "step": 1404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9456725535516584,
+      "learning_rate": 9.751121411678915e-06,
+      "loss": 0.2246,
+      "step": 1405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7442855938194004,
+      "learning_rate": 9.750662055458268e-06,
+      "loss": 0.2121,
+      "step": 1406
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0050591184089117,
+      "learning_rate": 9.75020228655046e-06,
+      "loss": 0.241,
+      "step": 1407
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2350317194153346,
+      "learning_rate": 9.749742104995437e-06,
+      "loss": 0.2804,
+      "step": 1408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.970326638213501,
+      "learning_rate": 9.74928151083317e-06,
+      "loss": 0.2476,
+      "step": 1409
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1555676716786967,
+      "learning_rate": 9.748820504103671e-06,
+      "loss": 0.241,
+      "step": 1410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1348716369369978,
+      "learning_rate": 9.748359084846988e-06,
+      "loss": 0.2449,
+      "step": 1411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8001187749348473,
+      "learning_rate": 9.747897253103203e-06,
+      "loss": 0.238,
+      "step": 1412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7498967043995846,
+      "learning_rate": 9.747435008912438e-06,
+      "loss": 0.246,
+      "step": 1413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9037105384409476,
+      "learning_rate": 9.746972352314845e-06,
+      "loss": 0.1931,
+      "step": 1414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0233533254225695,
+      "learning_rate": 9.746509283350615e-06,
+      "loss": 0.2494,
+      "step": 1415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.709996444827517,
+      "learning_rate": 9.746045802059978e-06,
+      "loss": 0.2177,
+      "step": 1416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7811529903648444,
+      "learning_rate": 9.745581908483192e-06,
+      "loss": 0.2255,
+      "step": 1417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0215441716866382,
+      "learning_rate": 9.745117602660556e-06,
+      "loss": 0.237,
+      "step": 1418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.796389923691804,
+      "learning_rate": 9.744652884632406e-06,
+      "loss": 0.2196,
+      "step": 1419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.941517062979519,
+      "learning_rate": 9.74418775443911e-06,
+      "loss": 0.2382,
+      "step": 1420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8072558205273435,
+      "learning_rate": 9.743722212121075e-06,
+      "loss": 0.25,
+      "step": 1421
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.900209166241114,
+      "learning_rate": 9.743256257718741e-06,
+      "loss": 0.2139,
+      "step": 1422
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.912636878828581,
+      "learning_rate": 9.742789891272586e-06,
+      "loss": 0.2279,
+      "step": 1423
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8156612477239147,
+      "learning_rate": 9.742323112823123e-06,
+      "loss": 0.2057,
+      "step": 1424
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7964001395071136,
+      "learning_rate": 9.741855922410898e-06,
+      "loss": 0.2417,
+      "step": 1425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9162231197120745,
+      "learning_rate": 9.741388320076502e-06,
+      "loss": 0.2419,
+      "step": 1426
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7809086022401193,
+      "learning_rate": 9.740920305860548e-06,
+      "loss": 0.2265,
+      "step": 1427
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.983886010953564,
+      "learning_rate": 9.740451879803697e-06,
+      "loss": 0.224,
+      "step": 1428
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0570074553240496,
+      "learning_rate": 9.739983041946639e-06,
+      "loss": 0.2387,
+      "step": 1429
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6133874096467746,
+      "learning_rate": 9.7395137923301e-06,
+      "loss": 0.1918,
+      "step": 1430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.764566808220084,
+      "learning_rate": 9.739044130994848e-06,
+      "loss": 0.2353,
+      "step": 1431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6752715070348048,
+      "learning_rate": 9.73857405798168e-06,
+      "loss": 0.2515,
+      "step": 1432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5198857225780302,
+      "learning_rate": 9.738103573331427e-06,
+      "loss": 0.1994,
+      "step": 1433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0083749373139623,
+      "learning_rate": 9.737632677084967e-06,
+      "loss": 0.2537,
+      "step": 1434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8177570725757586,
+      "learning_rate": 9.737161369283201e-06,
+      "loss": 0.2399,
+      "step": 1435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8526563281320363,
+      "learning_rate": 9.736689649967074e-06,
+      "loss": 0.2477,
+      "step": 1436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1594934516407074,
+      "learning_rate": 9.736217519177562e-06,
+      "loss": 0.2089,
+      "step": 1437
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9541129715645886,
+      "learning_rate": 9.735744976955681e-06,
+      "loss": 0.237,
+      "step": 1438
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7483571705215533,
+      "learning_rate": 9.735272023342476e-06,
+      "loss": 0.247,
+      "step": 1439
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6214678485011973,
+      "learning_rate": 9.734798658379038e-06,
+      "loss": 0.1828,
+      "step": 1440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.874750182668853,
+      "learning_rate": 9.734324882106486e-06,
+      "loss": 0.247,
+      "step": 1441
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.602984845339882,
+      "learning_rate": 9.733850694565975e-06,
+      "loss": 0.2388,
+      "step": 1442
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8550399989129245,
+      "learning_rate": 9.733376095798698e-06,
+      "loss": 0.2194,
+      "step": 1443
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.874783969070111,
+      "learning_rate": 9.732901085845884e-06,
+      "loss": 0.2125,
+      "step": 1444
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4286741135106733,
+      "learning_rate": 9.732425664748794e-06,
+      "loss": 0.2127,
+      "step": 1445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9287319477749034,
+      "learning_rate": 9.731949832548733e-06,
+      "loss": 0.25,
+      "step": 1446
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.056696971915079,
+      "learning_rate": 9.731473589287031e-06,
+      "loss": 0.2447,
+      "step": 1447
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.625770808300889,
+      "learning_rate": 9.730996935005062e-06,
+      "loss": 0.2301,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.713978943811164,
+      "learning_rate": 9.730519869744231e-06,
+      "loss": 0.1856,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.911788171947011,
+      "learning_rate": 9.730042393545981e-06,
+      "loss": 0.2099,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.685149597225699,
+      "learning_rate": 9.729564506451791e-06,
+      "loss": 0.2496,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1868346876719644,
+      "learning_rate": 9.729086208503174e-06,
+      "loss": 0.2398,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9297875193130842,
+      "learning_rate": 9.72860749974168e-06,
+      "loss": 0.2114,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.027276424904841,
+      "learning_rate": 9.728128380208893e-06,
+      "loss": 0.253,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.775986367481444,
+      "learning_rate": 9.727648849946432e-06,
+      "loss": 0.2359,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3398402406789036,
+      "learning_rate": 9.727168908995958e-06,
+      "loss": 0.2268,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1048207044183993,
+      "learning_rate": 9.72668855739916e-06,
+      "loss": 0.2414,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9130027237547287,
+      "learning_rate": 9.726207795197768e-06,
+      "loss": 0.244,
+      "step": 1458
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.737956355562966,
+      "learning_rate": 9.725726622433544e-06,
+      "loss": 0.2005,
+      "step": 1459
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7139957498767293,
+      "learning_rate": 9.725245039148287e-06,
+      "loss": 0.218,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8155470025174916,
+      "learning_rate": 9.724763045383833e-06,
+      "loss": 0.2219,
+      "step": 1461
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.80076173339565,
+      "learning_rate": 9.724280641182052e-06,
+      "loss": 0.225,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9626463090711974,
+      "learning_rate": 9.723797826584849e-06,
+      "loss": 0.2399,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1460763243447225,
+      "learning_rate": 9.723314601634169e-06,
+      "loss": 0.2455,
+      "step": 1464
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.704317440787788,
+      "learning_rate": 9.722830966371985e-06,
+      "loss": 0.2313,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8420013040949295,
+      "learning_rate": 9.722346920840313e-06,
+      "loss": 0.2502,
+      "step": 1466
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.991844196300371,
+      "learning_rate": 9.721862465081202e-06,
+      "loss": 0.2233,
+      "step": 1467
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8066055210309897,
+      "learning_rate": 9.721377599136736e-06,
+      "loss": 0.2261,
+      "step": 1468
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.912754556760992,
+      "learning_rate": 9.720892323049034e-06,
+      "loss": 0.232,
+      "step": 1469
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5641644611215395,
+      "learning_rate": 9.720406636860252e-06,
+      "loss": 0.2125,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.574205658915224,
+      "learning_rate": 9.719920540612581e-06,
+      "loss": 0.235,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0153345293696905,
+      "learning_rate": 9.71943403434825e-06,
+      "loss": 0.2584,
+      "step": 1472
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7547468869911698,
+      "learning_rate": 9.71894711810952e-06,
+      "loss": 0.2331,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7033246964211477,
+      "learning_rate": 9.718459791938688e-06,
+      "loss": 0.2171,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.631225181848083,
+      "learning_rate": 9.717972055878088e-06,
+      "loss": 0.2364,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.764094585563155,
+      "learning_rate": 9.717483909970094e-06,
+      "loss": 0.2474,
+      "step": 1476
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.075218439107802,
+      "learning_rate": 9.716995354257103e-06,
+      "loss": 0.2607,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3546477061169537,
+      "learning_rate": 9.71650638878156e-06,
+      "loss": 0.25,
+      "step": 1478
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.633538440229838,
+      "learning_rate": 9.716017013585942e-06,
+      "loss": 0.243,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.871066080941459,
+      "learning_rate": 9.71552722871276e-06,
+      "loss": 0.2226,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.141714034012026,
+      "learning_rate": 9.71503703420456e-06,
+      "loss": 0.2607,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6676819877574576,
+      "learning_rate": 9.714546430103924e-06,
+      "loss": 0.2089,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9014722408892615,
+      "learning_rate": 9.714055416453473e-06,
+      "loss": 0.257,
+      "step": 1483
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9449452523977913,
+      "learning_rate": 9.713563993295862e-06,
+      "loss": 0.2273,
+      "step": 1484
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.823252338309778,
+      "learning_rate": 9.713072160673778e-06,
+      "loss": 0.2413,
+      "step": 1485
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4554738033481835,
+      "learning_rate": 9.712579918629947e-06,
+      "loss": 0.2045,
+      "step": 1486
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9483760562896486,
+      "learning_rate": 9.71208726720713e-06,
+      "loss": 0.2386,
+      "step": 1487
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9064400159585806,
+      "learning_rate": 9.711594206448123e-06,
+      "loss": 0.2531,
+      "step": 1488
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.458087638656289,
+      "learning_rate": 9.711100736395758e-06,
+      "loss": 0.2024,
+      "step": 1489
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3667108872172418,
+      "learning_rate": 9.710606857092903e-06,
+      "loss": 0.2365,
+      "step": 1490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5168877621793206,
+      "learning_rate": 9.71011256858246e-06,
+      "loss": 0.2007,
+      "step": 1491
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7584043646100014,
+      "learning_rate": 9.709617870907368e-06,
+      "loss": 0.2229,
+      "step": 1492
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6134092135665807,
+      "learning_rate": 9.7091227641106e-06,
+      "loss": 0.1967,
+      "step": 1493
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.907023087104539,
+      "learning_rate": 9.70862724823517e-06,
+      "loss": 0.2718,
+      "step": 1494
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7535465911622943,
+      "learning_rate": 9.708131323324117e-06,
+      "loss": 0.2009,
+      "step": 1495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.790231369678342,
+      "learning_rate": 9.707634989420525e-06,
+      "loss": 0.2307,
+      "step": 1496
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8397414552758313,
+      "learning_rate": 9.707138246567511e-06,
+      "loss": 0.201,
+      "step": 1497
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8221899219787123,
+      "learning_rate": 9.706641094808225e-06,
+      "loss": 0.2221,
+      "step": 1498
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6437144048989616,
+      "learning_rate": 9.706143534185854e-06,
+      "loss": 0.2189,
+      "step": 1499
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9433347949629662,
+      "learning_rate": 9.705645564743624e-06,
+      "loss": 0.223,
+      "step": 1500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0510375045265796,
+      "learning_rate": 9.70514718652479e-06,
+      "loss": 0.2274,
+      "step": 1501
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6224486945689525,
+      "learning_rate": 9.704648399572646e-06,
+      "loss": 0.203,
+      "step": 1502
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.417391792541818,
+      "learning_rate": 9.704149203930522e-06,
+      "loss": 0.2481,
+      "step": 1503
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4776124613410446,
+      "learning_rate": 9.703649599641785e-06,
+      "loss": 0.2551,
+      "step": 1504
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8031745721433508,
+      "learning_rate": 9.703149586749832e-06,
+      "loss": 0.2453,
+      "step": 1505
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5799424036544383,
+      "learning_rate": 9.702649165298098e-06,
+      "loss": 0.2357,
+      "step": 1506
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.479118377704076,
+      "learning_rate": 9.702148335330059e-06,
+      "loss": 0.2113,
+      "step": 1507
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.957523446181792,
+      "learning_rate": 9.70164709688922e-06,
+      "loss": 0.2212,
+      "step": 1508
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9003605433388135,
+      "learning_rate": 9.70114545001912e-06,
+      "loss": 0.2639,
+      "step": 1509
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.713050325946634,
+      "learning_rate": 9.70064339476334e-06,
+      "loss": 0.2176,
+      "step": 1510
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7310746005294435,
+      "learning_rate": 9.700140931165494e-06,
+      "loss": 0.2402,
+      "step": 1511
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6983111111860842,
+      "learning_rate": 9.699638059269228e-06,
+      "loss": 0.2153,
+      "step": 1512
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9712006245722553,
+      "learning_rate": 9.699134779118226e-06,
+      "loss": 0.2361,
+      "step": 1513
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0300691635280703,
+      "learning_rate": 9.698631090756211e-06,
+      "loss": 0.2609,
+      "step": 1514
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9187837623461137,
+      "learning_rate": 9.698126994226937e-06,
+      "loss": 0.2418,
+      "step": 1515
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7405932958273946,
+      "learning_rate": 9.697622489574192e-06,
+      "loss": 0.2039,
+      "step": 1516
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1415524299115414,
+      "learning_rate": 9.697117576841804e-06,
+      "loss": 0.2364,
+      "step": 1517
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9937275186872827,
+      "learning_rate": 9.696612256073634e-06,
+      "loss": 0.2328,
+      "step": 1518
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5861772953906312,
+      "learning_rate": 9.69610652731358e-06,
+      "loss": 0.2163,
+      "step": 1519
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.00855388669083,
+      "learning_rate": 9.695600390605573e-06,
+      "loss": 0.2655,
+      "step": 1520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7628333744866693,
+      "learning_rate": 9.69509384599358e-06,
+      "loss": 0.2055,
+      "step": 1521
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.843548450669024,
+      "learning_rate": 9.694586893521607e-06,
+      "loss": 0.2172,
+      "step": 1522
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8752611462885613,
+      "learning_rate": 9.694079533233692e-06,
+      "loss": 0.2449,
+      "step": 1523
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.036470017417834,
+      "learning_rate": 9.693571765173907e-06,
+      "loss": 0.2448,
+      "step": 1524
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.607871130670095,
+      "learning_rate": 9.693063589386361e-06,
+      "loss": 0.2293,
+      "step": 1525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.70990746902915,
+      "learning_rate": 9.692555005915203e-06,
+      "loss": 0.1998,
+      "step": 1526
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7762196484054082,
+      "learning_rate": 9.69204601480461e-06,
+      "loss": 0.234,
+      "step": 1527
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.027966924331917,
+      "learning_rate": 9.6915366160988e-06,
+      "loss": 0.2416,
+      "step": 1528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.210748036034756,
+      "learning_rate": 9.691026809842021e-06,
+      "loss": 0.2393,
+      "step": 1529
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.714439055952504,
+      "learning_rate": 9.690516596078563e-06,
+      "loss": 0.2186,
+      "step": 1530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.915272765772438,
+      "learning_rate": 9.690005974852746e-06,
+      "loss": 0.2209,
+      "step": 1531
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.854124562944047,
+      "learning_rate": 9.689494946208926e-06,
+      "loss": 0.2073,
+      "step": 1532
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.169506978735728,
+      "learning_rate": 9.688983510191498e-06,
+      "loss": 0.2587,
+      "step": 1533
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7947211413714785,
+      "learning_rate": 9.688471666844892e-06,
+      "loss": 0.2265,
+      "step": 1534
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7473053702464667,
+      "learning_rate": 9.687959416213568e-06,
+      "loss": 0.2014,
+      "step": 1535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.88756032851917,
+      "learning_rate": 9.687446758342025e-06,
+      "loss": 0.2372,
+      "step": 1536
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.730201404766415,
+      "learning_rate": 9.686933693274801e-06,
+      "loss": 0.2281,
+      "step": 1537
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.856505415518159,
+      "learning_rate": 9.68642022105646e-06,
+      "loss": 0.2426,
+      "step": 1538
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.439966690248097,
+      "learning_rate": 9.685906341731612e-06,
+      "loss": 0.1989,
+      "step": 1539
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.534229191798588,
+      "learning_rate": 9.685392055344894e-06,
+      "loss": 0.2107,
+      "step": 1540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5937486719178553,
+      "learning_rate": 9.684877361940985e-06,
+      "loss": 0.2458,
+      "step": 1541
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9766471585050844,
+      "learning_rate": 9.684362261564593e-06,
+      "loss": 0.2561,
+      "step": 1542
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7285978301285367,
+      "learning_rate": 9.683846754260467e-06,
+      "loss": 0.2318,
+      "step": 1543
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6490859759036534,
+      "learning_rate": 9.683330840073385e-06,
+      "loss": 0.1993,
+      "step": 1544
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6304394020217807,
+      "learning_rate": 9.68281451904817e-06,
+      "loss": 0.2009,
+      "step": 1545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.82355891124952,
+      "learning_rate": 9.682297791229668e-06,
+      "loss": 0.2356,
+      "step": 1546
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8447876460376453,
+      "learning_rate": 9.681780656662773e-06,
+      "loss": 0.2159,
+      "step": 1547
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8854951639768993,
+      "learning_rate": 9.681263115392403e-06,
+      "loss": 0.2066,
+      "step": 1548
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.874817457093106,
+      "learning_rate": 9.68074516746352e-06,
+      "loss": 0.2154,
+      "step": 1549
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6554675075569456,
+      "learning_rate": 9.680226812921117e-06,
+      "loss": 0.2369,
+      "step": 1550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8269212168665376,
+      "learning_rate": 9.679708051810222e-06,
+      "loss": 0.2609,
+      "step": 1551
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.425471588617914,
+      "learning_rate": 9.679188884175899e-06,
+      "loss": 0.1827,
+      "step": 1552
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.043539366403406,
+      "learning_rate": 9.67866931006325e-06,
+      "loss": 0.2295,
+      "step": 1553
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.944256704520571,
+      "learning_rate": 9.67814932951741e-06,
+      "loss": 0.2426,
+      "step": 1554
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.756378070320021,
+      "learning_rate": 9.677628942583546e-06,
+      "loss": 0.2047,
+      "step": 1555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4898495542831416,
+      "learning_rate": 9.677108149306868e-06,
+      "loss": 0.2083,
+      "step": 1556
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.992807003069182,
+      "learning_rate": 9.676586949732616e-06,
+      "loss": 0.2384,
+      "step": 1557
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.668642306173486,
+      "learning_rate": 9.676065343906062e-06,
+      "loss": 0.2189,
+      "step": 1558
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.631895639127102,
+      "learning_rate": 9.675543331872525e-06,
+      "loss": 0.2016,
+      "step": 1559
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0890160569099017,
+      "learning_rate": 9.675020913677345e-06,
+      "loss": 0.2293,
+      "step": 1560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9820345126258188,
+      "learning_rate": 9.67449808936591e-06,
+      "loss": 0.2375,
+      "step": 1561
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6081883274574142,
+      "learning_rate": 9.673974858983632e-06,
+      "loss": 0.2214,
+      "step": 1562
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.013177033536362,
+      "learning_rate": 9.673451222575966e-06,
+      "loss": 0.2515,
+      "step": 1563
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9825613183054993,
+      "learning_rate": 9.672927180188401e-06,
+      "loss": 0.2605,
+      "step": 1564
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1700277369011536,
+      "learning_rate": 9.67240273186646e-06,
+      "loss": 0.2524,
+      "step": 1565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5923557566562927,
+      "learning_rate": 9.6718778776557e-06,
+      "loss": 0.2158,
+      "step": 1566
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.759605462143463,
+      "learning_rate": 9.671352617601714e-06,
+      "loss": 0.2533,
+      "step": 1567
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5169542079163154,
+      "learning_rate": 9.670826951750136e-06,
+      "loss": 0.2398,
+      "step": 1568
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7234993794414883,
+      "learning_rate": 9.670300880146626e-06,
+      "loss": 0.2121,
+      "step": 1569
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9397000374259643,
+      "learning_rate": 9.669774402836883e-06,
+      "loss": 0.2187,
+      "step": 1570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7198101335816003,
+      "learning_rate": 9.669247519866645e-06,
+      "loss": 0.2316,
+      "step": 1571
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.008983223363173,
+      "learning_rate": 9.66872023128168e-06,
+      "loss": 0.2436,
+      "step": 1572
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9451381597613864,
+      "learning_rate": 9.668192537127793e-06,
+      "loss": 0.2249,
+      "step": 1573
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.955902351616932,
+      "learning_rate": 9.667664437450825e-06,
+      "loss": 0.2404,
+      "step": 1574
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.903320825647778,
+      "learning_rate": 9.667135932296653e-06,
+      "loss": 0.228,
+      "step": 1575
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4170988950285763,
+      "learning_rate": 9.666607021711185e-06,
+      "loss": 0.2205,
+      "step": 1576
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9882596573129945,
+      "learning_rate": 9.666077705740368e-06,
+      "loss": 0.2564,
+      "step": 1577
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5911391720617365,
+      "learning_rate": 9.665547984430186e-06,
+      "loss": 0.2126,
+      "step": 1578
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8208683827079954,
+      "learning_rate": 9.665017857826654e-06,
+      "loss": 0.2428,
+      "step": 1579
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0019933775974867,
+      "learning_rate": 9.664487325975822e-06,
+      "loss": 0.2649,
+      "step": 1580
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.814295091275241,
+      "learning_rate": 9.663956388923779e-06,
+      "loss": 0.2386,
+      "step": 1581
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3918956147036896,
+      "learning_rate": 9.663425046716648e-06,
+      "loss": 0.1696,
+      "step": 1582
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.671806418716493,
+      "learning_rate": 9.662893299400585e-06,
+      "loss": 0.2073,
+      "step": 1583
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6733044625154556,
+      "learning_rate": 9.66236114702178e-06,
+      "loss": 0.2077,
+      "step": 1584
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.212690294082634,
+      "learning_rate": 9.661828589626465e-06,
+      "loss": 0.2163,
+      "step": 1585
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9582787369390062,
+      "learning_rate": 9.661295627260901e-06,
+      "loss": 0.236,
+      "step": 1586
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7211921678156132,
+      "learning_rate": 9.660762259971386e-06,
+      "loss": 0.2061,
+      "step": 1587
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.045993606123908,
+      "learning_rate": 9.660228487804254e-06,
+      "loss": 0.2402,
+      "step": 1588
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6107671378707376,
+      "learning_rate": 9.659694310805874e-06,
+      "loss": 0.2166,
+      "step": 1589
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.854306288419431,
+      "learning_rate": 9.659159729022649e-06,
+      "loss": 0.2114,
+      "step": 1590
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1905254612518243,
+      "learning_rate": 9.658624742501018e-06,
+      "loss": 0.2499,
+      "step": 1591
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7649113408904396,
+      "learning_rate": 9.658089351287452e-06,
+      "loss": 0.205,
+      "step": 1592
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.80746686082814,
+      "learning_rate": 9.657553555428464e-06,
+      "loss": 0.1937,
+      "step": 1593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.808417049743047,
+      "learning_rate": 9.657017354970597e-06,
+      "loss": 0.2642,
+      "step": 1594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7248858319197664,
+      "learning_rate": 9.65648074996043e-06,
+      "loss": 0.232,
+      "step": 1595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5268189226850692,
+      "learning_rate": 9.655943740444579e-06,
+      "loss": 0.213,
+      "step": 1596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6470000281531716,
+      "learning_rate": 9.655406326469692e-06,
+      "loss": 0.2374,
+      "step": 1597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.728736640772866,
+      "learning_rate": 9.654868508082455e-06,
+      "loss": 0.2287,
+      "step": 1598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7662517925685073,
+      "learning_rate": 9.654330285329586e-06,
+      "loss": 0.2324,
+      "step": 1599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8220906486926056,
+      "learning_rate": 9.653791658257843e-06,
+      "loss": 0.2254,
+      "step": 1600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5777244323724395,
+      "learning_rate": 9.653252626914014e-06,
+      "loss": 0.23,
+      "step": 1601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8844888343714232,
+      "learning_rate": 9.652713191344925e-06,
+      "loss": 0.2502,
+      "step": 1602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.807909610431765,
+      "learning_rate": 9.652173351597435e-06,
+      "loss": 0.2084,
+      "step": 1603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3886962429509206,
+      "learning_rate": 9.651633107718443e-06,
+      "loss": 0.2113,
+      "step": 1604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6827822906980847,
+      "learning_rate": 9.651092459754879e-06,
+      "loss": 0.208,
+      "step": 1605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7765037544936737,
+      "learning_rate": 9.650551407753705e-06,
+      "loss": 0.227,
+      "step": 1606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.858755730947863,
+      "learning_rate": 9.650009951761926e-06,
+      "loss": 0.2541,
+      "step": 1607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5849049792674657,
+      "learning_rate": 9.649468091826575e-06,
+      "loss": 0.2027,
+      "step": 1608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.930380377264086,
+      "learning_rate": 9.648925827994725e-06,
+      "loss": 0.2327,
+      "step": 1609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3570562933030295,
+      "learning_rate": 9.64838316031348e-06,
+      "loss": 0.1835,
+      "step": 1610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6838342972952565,
+      "learning_rate": 9.647840088829984e-06,
+      "loss": 0.1922,
+      "step": 1611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.932592011487064,
+      "learning_rate": 9.647296613591411e-06,
+      "loss": 0.2225,
+      "step": 1612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7003931785238957,
+      "learning_rate": 9.646752734644974e-06,
+      "loss": 0.1956,
+      "step": 1613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.656601331116525,
+      "learning_rate": 9.646208452037919e-06,
+      "loss": 0.223,
+      "step": 1614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.805961420413957,
+      "learning_rate": 9.645663765817528e-06,
+      "loss": 0.1947,
+      "step": 1615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.718423148939623,
+      "learning_rate": 9.645118676031115e-06,
+      "loss": 0.2278,
+      "step": 1616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.023230417850136,
+      "learning_rate": 9.644573182726035e-06,
+      "loss": 0.2083,
+      "step": 1617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.648343553392819,
+      "learning_rate": 9.644027285949673e-06,
+      "loss": 0.2267,
+      "step": 1618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.896398420833112,
+      "learning_rate": 9.64348098574945e-06,
+      "loss": 0.2563,
+      "step": 1619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5017012795932616,
+      "learning_rate": 9.642934282172824e-06,
+      "loss": 0.1827,
+      "step": 1620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7858196091171687,
+      "learning_rate": 9.642387175267285e-06,
+      "loss": 0.2329,
+      "step": 1621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8165940319032696,
+      "learning_rate": 9.641839665080363e-06,
+      "loss": 0.2383,
+      "step": 1622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.806771898505537,
+      "learning_rate": 9.641291751659618e-06,
+      "loss": 0.1852,
+      "step": 1623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7910379629398294,
+      "learning_rate": 9.640743435052647e-06,
+      "loss": 0.2341,
+      "step": 1624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9556565949101397,
+      "learning_rate": 9.640194715307083e-06,
+      "loss": 0.2312,
+      "step": 1625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7559178244254405,
+      "learning_rate": 9.63964559247059e-06,
+      "loss": 0.2431,
+      "step": 1626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8777072599656943,
+      "learning_rate": 9.639096066590874e-06,
+      "loss": 0.2141,
+      "step": 1627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.886249274960757,
+      "learning_rate": 9.638546137715668e-06,
+      "loss": 0.2404,
+      "step": 1628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7941362488188304,
+      "learning_rate": 9.637995805892746e-06,
+      "loss": 0.2312,
+      "step": 1629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4425025204901574,
+      "learning_rate": 9.637445071169917e-06,
+      "loss": 0.2023,
+      "step": 1630
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6747070071904537,
+      "learning_rate": 9.63689393359502e-06,
+      "loss": 0.2143,
+      "step": 1631
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7787332078860385,
+      "learning_rate": 9.636342393215931e-06,
+      "loss": 0.2214,
+      "step": 1632
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7512638548427333,
+      "learning_rate": 9.635790450080566e-06,
+      "loss": 0.24,
+      "step": 1633
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9045180141424494,
+      "learning_rate": 9.63523810423687e-06,
+      "loss": 0.2277,
+      "step": 1634
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6887185550995873,
+      "learning_rate": 9.634685355732823e-06,
+      "loss": 0.1834,
+      "step": 1635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7723511221897974,
+      "learning_rate": 9.634132204616444e-06,
+      "loss": 0.2401,
+      "step": 1636
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.811919194203948,
+      "learning_rate": 9.633578650935786e-06,
+      "loss": 0.2275,
+      "step": 1637
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.728784944560048,
+      "learning_rate": 9.633024694738934e-06,
+      "loss": 0.228,
+      "step": 1638
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.59291729228635,
+      "learning_rate": 9.632470336074009e-06,
+      "loss": 0.2012,
+      "step": 1639
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.77904935255569,
+      "learning_rate": 9.631915574989171e-06,
+      "loss": 0.2443,
+      "step": 1640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2218525147178343,
+      "learning_rate": 9.631360411532609e-06,
+      "loss": 0.2696,
+      "step": 1641
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.920292090498311,
+      "learning_rate": 9.63080484575255e-06,
+      "loss": 0.2459,
+      "step": 1642
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5625603959113703,
+      "learning_rate": 9.630248877697259e-06,
+      "loss": 0.2307,
+      "step": 1643
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5111587966174493,
+      "learning_rate": 9.629692507415028e-06,
+      "loss": 0.1924,
+      "step": 1644
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9221839124698574,
+      "learning_rate": 9.62913573495419e-06,
+      "loss": 0.239,
+      "step": 1645
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.865562667929123,
+      "learning_rate": 9.628578560363113e-06,
+      "loss": 0.233,
+      "step": 1646
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8832287396785765,
+      "learning_rate": 9.628020983690197e-06,
+      "loss": 0.2277,
+      "step": 1647
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6577319695219286,
+      "learning_rate": 9.627463004983877e-06,
+      "loss": 0.2019,
+      "step": 1648
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6324502101358576,
+      "learning_rate": 9.626904624292629e-06,
+      "loss": 0.2289,
+      "step": 1649
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7031058190857595,
+      "learning_rate": 9.626345841664953e-06,
+      "loss": 0.2187,
+      "step": 1650
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.319627220530469,
+      "learning_rate": 9.625786657149396e-06,
+      "loss": 0.1619,
+      "step": 1651
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9793494221380317,
+      "learning_rate": 9.625227070794529e-06,
+      "loss": 0.2328,
+      "step": 1652
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0654668294341554,
+      "learning_rate": 9.624667082648966e-06,
+      "loss": 0.2422,
+      "step": 1653
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.816926895241324,
+      "learning_rate": 9.624106692761354e-06,
+      "loss": 0.238,
+      "step": 1654
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7852834344810513,
+      "learning_rate": 9.62354590118037e-06,
+      "loss": 0.242,
+      "step": 1655
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9036212936149406,
+      "learning_rate": 9.622984707954732e-06,
+      "loss": 0.2352,
+      "step": 1656
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.44481206178112,
+      "learning_rate": 9.62242311313319e-06,
+      "loss": 0.2284,
+      "step": 1657
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.365521546707559,
+      "learning_rate": 9.621861116764529e-06,
+      "loss": 0.1952,
+      "step": 1658
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5175533394484084,
+      "learning_rate": 9.621298718897569e-06,
+      "loss": 0.2103,
+      "step": 1659
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.103193049363998,
+      "learning_rate": 9.620735919581168e-06,
+      "loss": 0.2683,
+      "step": 1660
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.714802773138769,
+      "learning_rate": 9.620172718864213e-06,
+      "loss": 0.2482,
+      "step": 1661
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5411590433396687,
+      "learning_rate": 9.619609116795628e-06,
+      "loss": 0.2071,
+      "step": 1662
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6126050466901454,
+      "learning_rate": 9.619045113424376e-06,
+      "loss": 0.2201,
+      "step": 1663
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.368055393964109,
+      "learning_rate": 9.61848070879945e-06,
+      "loss": 0.1888,
+      "step": 1664
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7815542197529353,
+      "learning_rate": 9.617915902969879e-06,
+      "loss": 0.2612,
+      "step": 1665
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.543535657927365,
+      "learning_rate": 9.61735069598473e-06,
+      "loss": 0.2546,
+      "step": 1666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7180204939689463,
+      "learning_rate": 9.616785087893099e-06,
+      "loss": 0.2204,
+      "step": 1667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.101084274457579,
+      "learning_rate": 9.61621907874412e-06,
+      "loss": 0.281,
+      "step": 1668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.607747626404295,
+      "learning_rate": 9.615652668586965e-06,
+      "loss": 0.2186,
+      "step": 1669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.213761805253613,
+      "learning_rate": 9.615085857470835e-06,
+      "loss": 0.2085,
+      "step": 1670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6923825329172315,
+      "learning_rate": 9.61451864544497e-06,
+      "loss": 0.2249,
+      "step": 1671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.762454031205449,
+      "learning_rate": 9.613951032558641e-06,
+      "loss": 0.2218,
+      "step": 1672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5951883320325795,
+      "learning_rate": 9.613383018861159e-06,
+      "loss": 0.2274,
+      "step": 1673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.736871570936625,
+      "learning_rate": 9.612814604401868e-06,
+      "loss": 0.2099,
+      "step": 1674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5091108176230206,
+      "learning_rate": 9.61224578923014e-06,
+      "loss": 0.2109,
+      "step": 1675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.608786554200097,
+      "learning_rate": 9.611676573395394e-06,
+      "loss": 0.1975,
+      "step": 1676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6353856472027637,
+      "learning_rate": 9.611106956947073e-06,
+      "loss": 0.2314,
+      "step": 1677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1613038342922284,
+      "learning_rate": 9.610536939934663e-06,
+      "loss": 0.2529,
+      "step": 1678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.555757455711037,
+      "learning_rate": 9.609966522407678e-06,
+      "loss": 0.1815,
+      "step": 1679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2152204753078175,
+      "learning_rate": 9.609395704415672e-06,
+      "loss": 0.2457,
+      "step": 1680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.827473896740029,
+      "learning_rate": 9.608824486008228e-06,
+      "loss": 0.2327,
+      "step": 1681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6376256633618986,
+      "learning_rate": 9.608252867234972e-06,
+      "loss": 0.2351,
+      "step": 1682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6236339696338247,
+      "learning_rate": 9.607680848145557e-06,
+      "loss": 0.2151,
+      "step": 1683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8409205946611094,
+      "learning_rate": 9.607108428789677e-06,
+      "loss": 0.2346,
+      "step": 1684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.717881716079471,
+      "learning_rate": 9.606535609217054e-06,
+      "loss": 0.2186,
+      "step": 1685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7396456134689036,
+      "learning_rate": 9.60596238947745e-06,
+      "loss": 0.2328,
+      "step": 1686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.100764757098075,
+      "learning_rate": 9.605388769620663e-06,
+      "loss": 0.2555,
+      "step": 1687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4749698347744378,
+      "learning_rate": 9.604814749696519e-06,
+      "loss": 0.1782,
+      "step": 1688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.770567729618977,
+      "learning_rate": 9.604240329754883e-06,
+      "loss": 0.2484,
+      "step": 1689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.42417640332674,
+      "learning_rate": 9.603665509845657e-06,
+      "loss": 0.1926,
+      "step": 1690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.519824351316186,
+      "learning_rate": 9.603090290018774e-06,
+      "loss": 0.2179,
+      "step": 1691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2505042582116315,
+      "learning_rate": 9.602514670324204e-06,
+      "loss": 0.2687,
+      "step": 1692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.849312214312834,
+      "learning_rate": 9.601938650811949e-06,
+      "loss": 0.2332,
+      "step": 1693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.872267799669223,
+      "learning_rate": 9.601362231532047e-06,
+      "loss": 0.2251,
+      "step": 1694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5632923872826976,
+      "learning_rate": 9.600785412534575e-06,
+      "loss": 0.18,
+      "step": 1695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1272359230488287,
+      "learning_rate": 9.600208193869638e-06,
+      "loss": 0.2634,
+      "step": 1696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.720416841863935,
+      "learning_rate": 9.599630575587378e-06,
+      "loss": 0.2307,
+      "step": 1697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4673292796001958,
+      "learning_rate": 9.599052557737973e-06,
+      "loss": 0.2184,
+      "step": 1698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8346451288445516,
+      "learning_rate": 9.598474140371637e-06,
+      "loss": 0.2562,
+      "step": 1699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.548638751653133,
+      "learning_rate": 9.597895323538615e-06,
+      "loss": 0.1979,
+      "step": 1700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.742428112530759,
+      "learning_rate": 9.597316107289187e-06,
+      "loss": 0.1939,
+      "step": 1701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6337741204738876,
+      "learning_rate": 9.596736491673674e-06,
+      "loss": 0.1961,
+      "step": 1702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6800166273471153,
+      "learning_rate": 9.596156476742419e-06,
+      "loss": 0.2497,
+      "step": 1703
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.868938335305424,
+      "learning_rate": 9.595576062545817e-06,
+      "loss": 0.2305,
+      "step": 1704
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.600370957828557,
+      "learning_rate": 9.59499524913428e-06,
+      "loss": 0.235,
+      "step": 1705
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.263633374493889,
+      "learning_rate": 9.594414036558268e-06,
+      "loss": 0.2204,
+      "step": 1706
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7651879039745113,
+      "learning_rate": 9.593832424868271e-06,
+      "loss": 0.224,
+      "step": 1707
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.009188856854406,
+      "learning_rate": 9.59325041411481e-06,
+      "loss": 0.2195,
+      "step": 1708
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8061359410692273,
+      "learning_rate": 9.592668004348443e-06,
+      "loss": 0.2083,
+      "step": 1709
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6936061605824193,
+      "learning_rate": 9.592085195619767e-06,
+      "loss": 0.2064,
+      "step": 1710
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.930877583884321,
+      "learning_rate": 9.59150198797941e-06,
+      "loss": 0.2395,
+      "step": 1711
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6239550359661443,
+      "learning_rate": 9.590918381478034e-06,
+      "loss": 0.2228,
+      "step": 1712
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7394403666584504,
+      "learning_rate": 9.590334376166334e-06,
+      "loss": 0.235,
+      "step": 1713
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1226436039606473,
+      "learning_rate": 9.589749972095048e-06,
+      "loss": 0.2177,
+      "step": 1714
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.549593904820042,
+      "learning_rate": 9.589165169314938e-06,
+      "loss": 0.207,
+      "step": 1715
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8047872809286365,
+      "learning_rate": 9.588579967876806e-06,
+      "loss": 0.2455,
+      "step": 1716
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.152412334473281,
+      "learning_rate": 9.58799436783149e-06,
+      "loss": 0.2338,
+      "step": 1717
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6383426670233634,
+      "learning_rate": 9.58740836922986e-06,
+      "loss": 0.201,
+      "step": 1718
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5664408059758856,
+      "learning_rate": 9.586821972122822e-06,
+      "loss": 0.1984,
+      "step": 1719
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.616884817499784,
+      "learning_rate": 9.586235176561313e-06,
+      "loss": 0.2146,
+      "step": 1720
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7720308313210884,
+      "learning_rate": 9.58564798259631e-06,
+      "loss": 0.2363,
+      "step": 1721
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.519023353478564,
+      "learning_rate": 9.585060390278824e-06,
+      "loss": 0.1855,
+      "step": 1722
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1360082248795536,
+      "learning_rate": 9.584472399659895e-06,
+      "loss": 0.2539,
+      "step": 1723
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6233907807963837,
+      "learning_rate": 9.583884010790605e-06,
+      "loss": 0.2038,
+      "step": 1724
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.799187463421666,
+      "learning_rate": 9.583295223722062e-06,
+      "loss": 0.2399,
+      "step": 1725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7401234009096167,
+      "learning_rate": 9.582706038505418e-06,
+      "loss": 0.2465,
+      "step": 1726
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.402983256992335,
+      "learning_rate": 9.582116455191855e-06,
+      "loss": 0.2172,
+      "step": 1727
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6216783934659604,
+      "learning_rate": 9.581526473832585e-06,
+      "loss": 0.247,
+      "step": 1728
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0130415521406606,
+      "learning_rate": 9.580936094478865e-06,
+      "loss": 0.22,
+      "step": 1729
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6757854479097527,
+      "learning_rate": 9.58034531718198e-06,
+      "loss": 0.2159,
+      "step": 1730
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.719381500605708,
+      "learning_rate": 9.579754141993247e-06,
+      "loss": 0.2212,
+      "step": 1731
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9003634265410105,
+      "learning_rate": 9.579162568964025e-06,
+      "loss": 0.2039,
+      "step": 1732
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.818688851805939,
+      "learning_rate": 9.578570598145702e-06,
+      "loss": 0.2317,
+      "step": 1733
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.798901368085829,
+      "learning_rate": 9.577978229589702e-06,
+      "loss": 0.2015,
+      "step": 1734
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.759259072866002,
+      "learning_rate": 9.577385463347481e-06,
+      "loss": 0.2292,
+      "step": 1735
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0502353983735446,
+      "learning_rate": 9.576792299470537e-06,
+      "loss": 0.2408,
+      "step": 1736
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.756578316101421,
+      "learning_rate": 9.576198738010396e-06,
+      "loss": 0.1682,
+      "step": 1737
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.011501895584955,
+      "learning_rate": 9.57560477901862e-06,
+      "loss": 0.2237,
+      "step": 1738
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.735995439685285,
+      "learning_rate": 9.575010422546805e-06,
+      "loss": 0.2332,
+      "step": 1739
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.810411185932834,
+      "learning_rate": 9.574415668646584e-06,
+      "loss": 0.2217,
+      "step": 1740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6521338537732992,
+      "learning_rate": 9.573820517369623e-06,
+      "loss": 0.2205,
+      "step": 1741
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6750502634585613,
+      "learning_rate": 9.57322496876762e-06,
+      "loss": 0.2126,
+      "step": 1742
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.644547247288957,
+      "learning_rate": 9.572629022892312e-06,
+      "loss": 0.2061,
+      "step": 1743
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5504708251215007,
+      "learning_rate": 9.572032679795469e-06,
+      "loss": 0.2204,
+      "step": 1744
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.419006548996466,
+      "learning_rate": 9.571435939528893e-06,
+      "loss": 0.2386,
+      "step": 1745
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6413667336199613,
+      "learning_rate": 9.570838802144425e-06,
+      "loss": 0.1994,
+      "step": 1746
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0519880627966605,
+      "learning_rate": 9.570241267693935e-06,
+      "loss": 0.2604,
+      "step": 1747
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7924615919702345,
+      "learning_rate": 9.569643336229334e-06,
+      "loss": 0.2232,
+      "step": 1748
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.507699391405858,
+      "learning_rate": 9.569045007802558e-06,
+      "loss": 0.2198,
+      "step": 1749
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.868233561057278,
+      "learning_rate": 9.568446282465592e-06,
+      "loss": 0.2359,
+      "step": 1750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3797216668233028,
+      "learning_rate": 9.567847160270438e-06,
+      "loss": 0.2217,
+      "step": 1751
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.637712540644408,
+      "learning_rate": 9.567247641269148e-06,
+      "loss": 0.1944,
+      "step": 1752
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8500124457445506,
+      "learning_rate": 9.566647725513799e-06,
+      "loss": 0.2313,
+      "step": 1753
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.69311795556285,
+      "learning_rate": 9.566047413056506e-06,
+      "loss": 0.1994,
+      "step": 1754
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.663547303732506,
+      "learning_rate": 9.565446703949417e-06,
+      "loss": 0.2259,
+      "step": 1755
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.705919407798083,
+      "learning_rate": 9.564845598244717e-06,
+      "loss": 0.2496,
+      "step": 1756
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4229237870561717,
+      "learning_rate": 9.564244095994621e-06,
+      "loss": 0.1889,
+      "step": 1757
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3953847492794917,
+      "learning_rate": 9.563642197251382e-06,
+      "loss": 0.1879,
+      "step": 1758
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5483154714727347,
+      "learning_rate": 9.563039902067288e-06,
+      "loss": 0.1781,
+      "step": 1759
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6083217487177626,
+      "learning_rate": 9.56243721049466e-06,
+      "loss": 0.2401,
+      "step": 1760
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4330218004324977,
+      "learning_rate": 9.561834122585854e-06,
+      "loss": 0.2033,
+      "step": 1761
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.647433185175119,
+      "learning_rate": 9.561230638393255e-06,
+      "loss": 0.1895,
+      "step": 1762
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.351384241310298,
+      "learning_rate": 9.560626757969294e-06,
+      "loss": 0.1909,
+      "step": 1763
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7308059895075685,
+      "learning_rate": 9.560022481366424e-06,
+      "loss": 0.2479,
+      "step": 1764
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.099206116735547,
+      "learning_rate": 9.559417808637144e-06,
+      "loss": 0.2058,
+      "step": 1765
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7411520102567613,
+      "learning_rate": 9.558812739833976e-06,
+      "loss": 0.2017,
+      "step": 1766
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6059707581758467,
+      "learning_rate": 9.558207275009484e-06,
+      "loss": 0.224,
+      "step": 1767
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.699417109489021,
+      "learning_rate": 9.557601414216266e-06,
+      "loss": 0.2174,
+      "step": 1768
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2704084779433273,
+      "learning_rate": 9.55699515750695e-06,
+      "loss": 0.2148,
+      "step": 1769
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5127217709894043,
+      "learning_rate": 9.556388504934205e-06,
+      "loss": 0.1815,
+      "step": 1770
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9608885098882243,
+      "learning_rate": 9.555781456550725e-06,
+      "loss": 0.2227,
+      "step": 1771
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.742090598088731,
+      "learning_rate": 9.55517401240925e-06,
+      "loss": 0.2221,
+      "step": 1772
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.545091765276031,
+      "learning_rate": 9.554566172562543e-06,
+      "loss": 0.2099,
+      "step": 1773
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.982152951649236,
+      "learning_rate": 9.55395793706341e-06,
+      "loss": 0.277,
+      "step": 1774
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5354385601712255,
+      "learning_rate": 9.553349305964687e-06,
+      "loss": 0.225,
+      "step": 1775
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.79908848828513,
+      "learning_rate": 9.552740279319245e-06,
+      "loss": 0.2381,
+      "step": 1776
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.553719888278797,
+      "learning_rate": 9.55213085717999e-06,
+      "loss": 0.2207,
+      "step": 1777
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.605392576690462,
+      "learning_rate": 9.551521039599863e-06,
+      "loss": 0.2204,
+      "step": 1778
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5240178610291344,
+      "learning_rate": 9.550910826631838e-06,
+      "loss": 0.1899,
+      "step": 1779
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.846364174813725,
+      "learning_rate": 9.550300218328925e-06,
+      "loss": 0.219,
+      "step": 1780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8963255830032133,
+      "learning_rate": 9.549689214744164e-06,
+      "loss": 0.243,
+      "step": 1781
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.795185008332951,
+      "learning_rate": 9.549077815930636e-06,
+      "loss": 0.2629,
+      "step": 1782
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8229099089245624,
+      "learning_rate": 9.548466021941449e-06,
+      "loss": 0.2513,
+      "step": 1783
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2482969658252094,
+      "learning_rate": 9.547853832829755e-06,
+      "loss": 0.2531,
+      "step": 1784
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5229222999636374,
+      "learning_rate": 9.547241248648727e-06,
+      "loss": 0.2142,
+      "step": 1785
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3979692335565073,
+      "learning_rate": 9.546628269451585e-06,
+      "loss": 0.1935,
+      "step": 1786
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.754020006290599,
+      "learning_rate": 9.546014895291578e-06,
+      "loss": 0.2118,
+      "step": 1787
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7784641802939127,
+      "learning_rate": 9.54540112622199e-06,
+      "loss": 0.2547,
+      "step": 1788
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.731180462953888,
+      "learning_rate": 9.544786962296132e-06,
+      "loss": 0.2347,
+      "step": 1789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.040592126685657,
+      "learning_rate": 9.544172403567365e-06,
+      "loss": 0.1901,
+      "step": 1790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5571708814237177,
+      "learning_rate": 9.543557450089071e-06,
+      "loss": 0.2163,
+      "step": 1791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.484015609928097,
+      "learning_rate": 9.54294210191467e-06,
+      "loss": 0.2885,
+      "step": 1792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.324195827359162,
+      "learning_rate": 9.542326359097619e-06,
+      "loss": 0.1611,
+      "step": 1793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.590716916693087,
+      "learning_rate": 9.541710221691407e-06,
+      "loss": 0.2202,
+      "step": 1794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.923240766576748,
+      "learning_rate": 9.541093689749554e-06,
+      "loss": 0.2251,
+      "step": 1795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.496219520395009,
+      "learning_rate": 9.540476763325623e-06,
+      "loss": 0.25,
+      "step": 1796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.896660954214899,
+      "learning_rate": 9.539859442473203e-06,
+      "loss": 0.2337,
+      "step": 1797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7928843265518504,
+      "learning_rate": 9.539241727245921e-06,
+      "loss": 0.2324,
+      "step": 1798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7880172478124217,
+      "learning_rate": 9.538623617697437e-06,
+      "loss": 0.2095,
+      "step": 1799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.472749054959661,
+      "learning_rate": 9.538005113881445e-06,
+      "loss": 0.2659,
+      "step": 1800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.670830396455639,
+      "learning_rate": 9.537386215851677e-06,
+      "loss": 0.2332,
+      "step": 1801
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.080030042416917,
+      "learning_rate": 9.536766923661894e-06,
+      "loss": 0.2329,
+      "step": 1802
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7649874399804313,
+      "learning_rate": 9.536147237365895e-06,
+      "loss": 0.2226,
+      "step": 1803
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3235678329701828,
+      "learning_rate": 9.53552715701751e-06,
+      "loss": 0.2343,
+      "step": 1804
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.721784115519037,
+      "learning_rate": 9.534906682670606e-06,
+      "loss": 0.2197,
+      "step": 1805
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5320710013883065,
+      "learning_rate": 9.534285814379084e-06,
+      "loss": 0.1815,
+      "step": 1806
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.845195811264802,
+      "learning_rate": 9.533664552196875e-06,
+      "loss": 0.2107,
+      "step": 1807
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8196917814835873,
+      "learning_rate": 9.533042896177951e-06,
+      "loss": 0.2174,
+      "step": 1808
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2058179430262976,
+      "learning_rate": 9.532420846376316e-06,
+      "loss": 0.2779,
+      "step": 1809
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8220935091568498,
+      "learning_rate": 9.531798402846004e-06,
+      "loss": 0.2182,
+      "step": 1810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9622485496711426,
+      "learning_rate": 9.531175565641087e-06,
+      "loss": 0.2485,
+      "step": 1811
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.742569359962187,
+      "learning_rate": 9.530552334815672e-06,
+      "loss": 0.2205,
+      "step": 1812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.560874594311351,
+      "learning_rate": 9.529928710423897e-06,
+      "loss": 0.2021,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8660690670715376,
+      "learning_rate": 9.529304692519936e-06,
+      "loss": 0.2197,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.653512013513791,
+      "learning_rate": 9.528680281157999e-06,
+      "loss": 0.1918,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.358691075439933,
+      "learning_rate": 9.528055476392325e-06,
+      "loss": 0.2312,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.930750185807348,
+      "learning_rate": 9.527430278277194e-06,
+      "loss": 0.1845,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5720787870654163,
+      "learning_rate": 9.526804686866916e-06,
+      "loss": 0.1811,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5755857865455787,
+      "learning_rate": 9.526178702215833e-06,
+      "loss": 0.221,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.974370888850911,
+      "learning_rate": 9.525552324378324e-06,
+      "loss": 0.2462,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.656199892069847,
+      "learning_rate": 9.524925553408806e-06,
+      "loss": 0.2142,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4448411260276273,
+      "learning_rate": 9.524298389361724e-06,
+      "loss": 0.193,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.247907757639739,
+      "learning_rate": 9.523670832291556e-06,
+      "loss": 0.2239,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6984256328987652,
+      "learning_rate": 9.523042882252825e-06,
+      "loss": 0.224,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.664496157380522,
+      "learning_rate": 9.522414539300074e-06,
+      "loss": 0.2161,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.168184124135893,
+      "learning_rate": 9.521785803487888e-06,
+      "loss": 0.2655,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6135886449829258,
+      "learning_rate": 9.521156674870888e-06,
+      "loss": 0.2011,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4198571718264494,
+      "learning_rate": 9.520527153503722e-06,
+      "loss": 0.275,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5649898725678835,
+      "learning_rate": 9.51989723944108e-06,
+      "loss": 0.2152,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5715593733642272,
+      "learning_rate": 9.51926693273768e-06,
+      "loss": 0.2343,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.05875050110746,
+      "learning_rate": 9.518636233448276e-06,
+      "loss": 0.1952,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9192309259881792,
+      "learning_rate": 9.518005141627659e-06,
+      "loss": 0.1911,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7598320923834168,
+      "learning_rate": 9.517373657330648e-06,
+      "loss": 0.2255,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3713356821438447,
+      "learning_rate": 9.516741780612102e-06,
+      "loss": 0.2256,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0775933021261324,
+      "learning_rate": 9.516109511526912e-06,
+      "loss": 0.2287,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.92150254626363,
+      "learning_rate": 9.515476850130001e-06,
+      "loss": 0.2179,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1386410386127683,
+      "learning_rate": 9.514843796476329e-06,
+      "loss": 0.2004,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7784503953436968,
+      "learning_rate": 9.51421035062089e-06,
+      "loss": 0.2161,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6419659813585032,
+      "learning_rate": 9.51357651261871e-06,
+      "loss": 0.2222,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4760906899052904,
+      "learning_rate": 9.512942282524848e-06,
+      "loss": 0.2457,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5684668295644104,
+      "learning_rate": 9.512307660394404e-06,
+      "loss": 0.2055,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.964621561473322,
+      "learning_rate": 9.511672646282502e-06,
+      "loss": 0.204,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0596750014645933,
+      "learning_rate": 9.51103724024431e-06,
+      "loss": 0.2181,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5677103627021687,
+      "learning_rate": 9.510401442335022e-06,
+      "loss": 0.2149,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7876347799956704,
+      "learning_rate": 9.509765252609873e-06,
+      "loss": 0.2536,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8113449582417394,
+      "learning_rate": 9.509128671124123e-06,
+      "loss": 0.204,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3997174306890647,
+      "learning_rate": 9.508491697933076e-06,
+      "loss": 0.201,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6722596209316203,
+      "learning_rate": 9.507854333092064e-06,
+      "loss": 0.1983,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8171963877872015,
+      "learning_rate": 9.507216576656454e-06,
+      "loss": 0.2501,
+      "step": 1849
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.136077220446606,
+      "learning_rate": 9.506578428681648e-06,
+      "loss": 0.224,
+      "step": 1850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6603871708849227,
+      "learning_rate": 9.50593988922308e-06,
+      "loss": 0.2223,
+      "step": 1851
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.478401052836361,
+      "learning_rate": 9.505300958336224e-06,
+      "loss": 0.2123,
+      "step": 1852
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7983938569349647,
+      "learning_rate": 9.50466163607658e-06,
+      "loss": 0.2426,
+      "step": 1853
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1826528677294137,
+      "learning_rate": 9.504021922499685e-06,
+      "loss": 0.2295,
+      "step": 1854
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4609413388220713,
+      "learning_rate": 9.503381817661113e-06,
+      "loss": 0.1714,
+      "step": 1855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4388144063481425,
+      "learning_rate": 9.502741321616467e-06,
+      "loss": 0.1852,
+      "step": 1856
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5701719511300603,
+      "learning_rate": 9.502100434421388e-06,
+      "loss": 0.2182,
+      "step": 1857
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.630080586035678,
+      "learning_rate": 9.501459156131549e-06,
+      "loss": 0.1901,
+      "step": 1858
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5367991126218006,
+      "learning_rate": 9.500817486802658e-06,
+      "loss": 0.2091,
+      "step": 1859
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.977441082365511,
+      "learning_rate": 9.500175426490455e-06,
+      "loss": 0.2697,
+      "step": 1860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.694120040608933,
+      "learning_rate": 9.499532975250719e-06,
+      "loss": 0.2597,
+      "step": 1861
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8699728652688603,
+      "learning_rate": 9.498890133139253e-06,
+      "loss": 0.2311,
+      "step": 1862
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.592795339691541,
+      "learning_rate": 9.498246900211908e-06,
+      "loss": 0.2263,
+      "step": 1863
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8263425890119076,
+      "learning_rate": 9.497603276524555e-06,
+      "loss": 0.2416,
+      "step": 1864
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0721887535385144,
+      "learning_rate": 9.496959262133108e-06,
+      "loss": 0.2283,
+      "step": 1865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4867602124739987,
+      "learning_rate": 9.49631485709351e-06,
+      "loss": 0.2151,
+      "step": 1866
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5663160110048344,
+      "learning_rate": 9.495670061461747e-06,
+      "loss": 0.1792,
+      "step": 1867
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4583693021586006,
+      "learning_rate": 9.49502487529382e-06,
+      "loss": 0.1901,
+      "step": 1868
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.450647294077089,
+      "learning_rate": 9.494379298645788e-06,
+      "loss": 0.215,
+      "step": 1869
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1213740419409457,
+      "learning_rate": 9.493733331573724e-06,
+      "loss": 0.2618,
+      "step": 1870
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.919033858671396,
+      "learning_rate": 9.493086974133747e-06,
+      "loss": 0.2365,
+      "step": 1871
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1208642135231974,
+      "learning_rate": 9.492440226382003e-06,
+      "loss": 0.2599,
+      "step": 1872
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9299860824809874,
+      "learning_rate": 9.491793088374676e-06,
+      "loss": 0.2265,
+      "step": 1873
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5396018602628585,
+      "learning_rate": 9.491145560167983e-06,
+      "loss": 0.1909,
+      "step": 1874
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.733492147565226,
+      "learning_rate": 9.490497641818172e-06,
+      "loss": 0.2257,
+      "step": 1875
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5504400398440796,
+      "learning_rate": 9.489849333381529e-06,
+      "loss": 0.2288,
+      "step": 1876
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9391174795578725,
+      "learning_rate": 9.489200634914373e-06,
+      "loss": 0.2223,
+      "step": 1877
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.778446023649222,
+      "learning_rate": 9.488551546473055e-06,
+      "loss": 0.2553,
+      "step": 1878
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.973405428311063,
+      "learning_rate": 9.48790206811396e-06,
+      "loss": 0.2556,
+      "step": 1879
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.441651143798424,
+      "learning_rate": 9.48725219989351e-06,
+      "loss": 0.2079,
+      "step": 1880
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.694029475248282,
+      "learning_rate": 9.486601941868155e-06,
+      "loss": 0.1883,
+      "step": 1881
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4020601837291995,
+      "learning_rate": 9.485951294094386e-06,
+      "loss": 0.2231,
+      "step": 1882
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0210456620414887,
+      "learning_rate": 9.485300256628725e-06,
+      "loss": 0.2005,
+      "step": 1883
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.524550792070919,
+      "learning_rate": 9.484648829527722e-06,
+      "loss": 0.227,
+      "step": 1884
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1144946771860025,
+      "learning_rate": 9.483997012847971e-06,
+      "loss": 0.2383,
+      "step": 1885
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.485604120410346,
+      "learning_rate": 9.483344806646096e-06,
+      "loss": 0.1896,
+      "step": 1886
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4957293833751244,
+      "learning_rate": 9.48269221097875e-06,
+      "loss": 0.2012,
+      "step": 1887
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1301637551102504,
+      "learning_rate": 9.482039225902623e-06,
+      "loss": 0.2419,
+      "step": 1888
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2278398899690575,
+      "learning_rate": 9.481385851474443e-06,
+      "loss": 0.1573,
+      "step": 1889
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8055241281176797,
+      "learning_rate": 9.480732087750968e-06,
+      "loss": 0.2213,
+      "step": 1890
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4564968562722993,
+      "learning_rate": 9.480077934788987e-06,
+      "loss": 0.199,
+      "step": 1891
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.729906298416512,
+      "learning_rate": 9.479423392645327e-06,
+      "loss": 0.2137,
+      "step": 1892
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3677899548039942,
+      "learning_rate": 9.478768461376848e-06,
+      "loss": 0.1871,
+      "step": 1893
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6075451599195234,
+      "learning_rate": 9.478113141040444e-06,
+      "loss": 0.2289,
+      "step": 1894
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.483149257014688,
+      "learning_rate": 9.477457431693043e-06,
+      "loss": 0.1905,
+      "step": 1895
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9003563154187777,
+      "learning_rate": 9.476801333391604e-06,
+      "loss": 0.2148,
+      "step": 1896
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5408719970376263,
+      "learning_rate": 9.476144846193124e-06,
+      "loss": 0.2171,
+      "step": 1897
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8334286283044765,
+      "learning_rate": 9.475487970154628e-06,
+      "loss": 0.2191,
+      "step": 1898
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.749922657204987,
+      "learning_rate": 9.474830705333185e-06,
+      "loss": 0.2214,
+      "step": 1899
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4796788277965858,
+      "learning_rate": 9.474173051785884e-06,
+      "loss": 0.1978,
+      "step": 1900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8908276775330957,
+      "learning_rate": 9.473515009569857e-06,
+      "loss": 0.2654,
+      "step": 1901
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.568959426905677,
+      "learning_rate": 9.472856578742273e-06,
+      "loss": 0.2314,
+      "step": 1902
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.014301621071135,
+      "learning_rate": 9.472197759360322e-06,
+      "loss": 0.2192,
+      "step": 1903
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.851670442117005,
+      "learning_rate": 9.47153855148124e-06,
+      "loss": 0.2138,
+      "step": 1904
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4907931690439624,
+      "learning_rate": 9.470878955162291e-06,
+      "loss": 0.2079,
+      "step": 1905
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6247607760777676,
+      "learning_rate": 9.470218970460771e-06,
+      "loss": 0.2186,
+      "step": 1906
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.806282643848966,
+      "learning_rate": 9.469558597434018e-06,
+      "loss": 0.2217,
+      "step": 1907
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6951927795887323,
+      "learning_rate": 9.468897836139392e-06,
+      "loss": 0.1841,
+      "step": 1908
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7122144265711903,
+      "learning_rate": 9.468236686634298e-06,
+      "loss": 0.2002,
+      "step": 1909
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8418876411726717,
+      "learning_rate": 9.467575148976167e-06,
+      "loss": 0.2201,
+      "step": 1910
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7964638164990028,
+      "learning_rate": 9.466913223222467e-06,
+      "loss": 0.2007,
+      "step": 1911
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.714529518802107,
+      "learning_rate": 9.4662509094307e-06,
+      "loss": 0.2213,
+      "step": 1912
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5755466153761963,
+      "learning_rate": 9.465588207658398e-06,
+      "loss": 0.2126,
+      "step": 1913
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0776784341867045,
+      "learning_rate": 9.464925117963133e-06,
+      "loss": 0.2405,
+      "step": 1914
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.692709742743847,
+      "learning_rate": 9.464261640402504e-06,
+      "loss": 0.1977,
+      "step": 1915
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.61925345076181,
+      "learning_rate": 9.46359777503415e-06,
+      "loss": 0.1991,
+      "step": 1916
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9490438106366628,
+      "learning_rate": 9.462933521915738e-06,
+      "loss": 0.259,
+      "step": 1917
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8131085439040238,
+      "learning_rate": 9.462268881104973e-06,
+      "loss": 0.213,
+      "step": 1918
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6930230106023165,
+      "learning_rate": 9.461603852659592e-06,
+      "loss": 0.2219,
+      "step": 1919
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6248142589618286,
+      "learning_rate": 9.460938436637363e-06,
+      "loss": 0.1881,
+      "step": 1920
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.519783729717937,
+      "learning_rate": 9.460272633096093e-06,
+      "loss": 0.1977,
+      "step": 1921
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.44566254946816,
+      "learning_rate": 9.45960644209362e-06,
+      "loss": 0.1617,
+      "step": 1922
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.080743755461343,
+      "learning_rate": 9.458939863687814e-06,
+      "loss": 0.2351,
+      "step": 1923
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.372196875741983,
+      "learning_rate": 9.45827289793658e-06,
+      "loss": 0.1855,
+      "step": 1924
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8978361298701696,
+      "learning_rate": 9.45760554489786e-06,
+      "loss": 0.2039,
+      "step": 1925
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.938487337888618,
+      "learning_rate": 9.456937804629623e-06,
+      "loss": 0.2349,
+      "step": 1926
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5242025735422233,
+      "learning_rate": 9.456269677189878e-06,
+      "loss": 0.23,
+      "step": 1927
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9640757522941916,
+      "learning_rate": 9.455601162636662e-06,
+      "loss": 0.2345,
+      "step": 1928
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4592916506365734,
+      "learning_rate": 9.454932261028052e-06,
+      "loss": 0.1875,
+      "step": 1929
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.676826726363711,
+      "learning_rate": 9.45426297242215e-06,
+      "loss": 0.2357,
+      "step": 1930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.603045167142856,
+      "learning_rate": 9.4535932968771e-06,
+      "loss": 0.2186,
+      "step": 1931
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6179002499197743,
+      "learning_rate": 9.45292323445108e-06,
+      "loss": 0.2329,
+      "step": 1932
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4951787850459533,
+      "learning_rate": 9.452252785202291e-06,
+      "loss": 0.2556,
+      "step": 1933
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8161851093295622,
+      "learning_rate": 9.451581949188979e-06,
+      "loss": 0.2358,
+      "step": 1934
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8604351151213536,
+      "learning_rate": 9.450910726469415e-06,
+      "loss": 0.2146,
+      "step": 1935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7247453215479185,
+      "learning_rate": 9.450239117101913e-06,
+      "loss": 0.2197,
+      "step": 1936
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5932164082196785,
+      "learning_rate": 9.449567121144812e-06,
+      "loss": 0.1759,
+      "step": 1937
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.459127860060185,
+      "learning_rate": 9.448894738656488e-06,
+      "loss": 0.2054,
+      "step": 1938
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5567594832179057,
+      "learning_rate": 9.448221969695352e-06,
+      "loss": 0.2053,
+      "step": 1939
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.063139519828835,
+      "learning_rate": 9.447548814319844e-06,
+      "loss": 0.2648,
+      "step": 1940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7667141965766424,
+      "learning_rate": 9.446875272588444e-06,
+      "loss": 0.2322,
+      "step": 1941
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.65012761158475,
+      "learning_rate": 9.446201344559663e-06,
+      "loss": 0.2388,
+      "step": 1942
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.485013286242795,
+      "learning_rate": 9.445527030292038e-06,
+      "loss": 0.1689,
+      "step": 1943
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1674008356907786,
+      "learning_rate": 9.444852329844154e-06,
+      "loss": 0.1947,
+      "step": 1944
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5041850599188082,
+      "learning_rate": 9.444177243274619e-06,
+      "loss": 0.2318,
+      "step": 1945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.543262982507882,
+      "learning_rate": 9.443501770642074e-06,
+      "loss": 0.2224,
+      "step": 1946
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.552510002975641,
+      "learning_rate": 9.442825912005203e-06,
+      "loss": 0.1832,
+      "step": 1947
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6483771953343695,
+      "learning_rate": 9.442149667422712e-06,
+      "loss": 0.2122,
+      "step": 1948
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.116171727223382,
+      "learning_rate": 9.441473036953351e-06,
+      "loss": 0.2657,
+      "step": 1949
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4444632077571824,
+      "learning_rate": 9.440796020655893e-06,
+      "loss": 0.2112,
+      "step": 1950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8485890635983155,
+      "learning_rate": 9.440118618589153e-06,
+      "loss": 0.2249,
+      "step": 1951
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.857300519261187,
+      "learning_rate": 9.439440830811978e-06,
+      "loss": 0.2342,
+      "step": 1952
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9023746750587613,
+      "learning_rate": 9.438762657383244e-06,
+      "loss": 0.2244,
+      "step": 1953
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.660431395455733,
+      "learning_rate": 9.438084098361865e-06,
+      "loss": 0.1791,
+      "step": 1954
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1628614706220777,
+      "learning_rate": 9.437405153806786e-06,
+      "loss": 0.2179,
+      "step": 1955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6193031981195296,
+      "learning_rate": 9.43672582377699e-06,
+      "loss": 0.2186,
+      "step": 1956
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.637730547352104,
+      "learning_rate": 9.436046108331485e-06,
+      "loss": 0.1905,
+      "step": 1957
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.643787526420855,
+      "learning_rate": 9.435366007529321e-06,
+      "loss": 0.2487,
+      "step": 1958
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.806439694154534,
+      "learning_rate": 9.434685521429576e-06,
+      "loss": 0.256,
+      "step": 1959
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.426073384191156,
+      "learning_rate": 9.434004650091364e-06,
+      "loss": 0.2006,
+      "step": 1960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.848982888429369,
+      "learning_rate": 9.433323393573831e-06,
+      "loss": 0.2241,
+      "step": 1961
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4436292578971477,
+      "learning_rate": 9.432641751936162e-06,
+      "loss": 0.2039,
+      "step": 1962
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.690399285702892,
+      "learning_rate": 9.431959725237565e-06,
+      "loss": 0.2256,
+      "step": 1963
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3640975529400783,
+      "learning_rate": 9.43127731353729e-06,
+      "loss": 0.1991,
+      "step": 1964
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.664381998689749,
+      "learning_rate": 9.430594516894615e-06,
+      "loss": 0.2355,
+      "step": 1965
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.927538974930774,
+      "learning_rate": 9.42991133536886e-06,
+      "loss": 0.1872,
+      "step": 1966
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4192325030106843,
+      "learning_rate": 9.429227769019366e-06,
+      "loss": 0.1966,
+      "step": 1967
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5772244231544703,
+      "learning_rate": 9.428543817905518e-06,
+      "loss": 0.1878,
+      "step": 1968
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.945273229593765,
+      "learning_rate": 9.427859482086728e-06,
+      "loss": 0.2205,
+      "step": 1969
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4021082584549,
+      "learning_rate": 9.427174761622447e-06,
+      "loss": 0.178,
+      "step": 1970
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.723789379531242,
+      "learning_rate": 9.426489656572151e-06,
+      "loss": 0.224,
+      "step": 1971
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4069067741902193,
+      "learning_rate": 9.42580416699536e-06,
+      "loss": 0.185,
+      "step": 1972
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.61808192969402,
+      "learning_rate": 9.425118292951622e-06,
+      "loss": 0.1783,
+      "step": 1973
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6068471187754807,
+      "learning_rate": 9.424432034500514e-06,
+      "loss": 0.1997,
+      "step": 1974
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.505352164760058,
+      "learning_rate": 9.423745391701656e-06,
+      "loss": 0.1755,
+      "step": 1975
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.905321972183192,
+      "learning_rate": 9.423058364614692e-06,
+      "loss": 0.193,
+      "step": 1976
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.456636097164985,
+      "learning_rate": 9.422370953299305e-06,
+      "loss": 0.1936,
+      "step": 1977
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.651484055642835,
+      "learning_rate": 9.42168315781521e-06,
+      "loss": 0.2174,
+      "step": 1978
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2236332192989554,
+      "learning_rate": 9.420994978222156e-06,
+      "loss": 0.2565,
+      "step": 1979
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.662959400608087,
+      "learning_rate": 9.420306414579925e-06,
+      "loss": 0.2546,
+      "step": 1980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7209935987260008,
+      "learning_rate": 9.419617466948332e-06,
+      "loss": 0.1898,
+      "step": 1981
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.608906130950204,
+      "learning_rate": 9.418928135387224e-06,
+      "loss": 0.1959,
+      "step": 1982
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.892016124423646,
+      "learning_rate": 9.418238419956484e-06,
+      "loss": 0.2413,
+      "step": 1983
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5214807013393394,
+      "learning_rate": 9.417548320716027e-06,
+      "loss": 0.1986,
+      "step": 1984
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6225224108105416,
+      "learning_rate": 9.416857837725802e-06,
+      "loss": 0.1945,
+      "step": 1985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6927849854121133,
+      "learning_rate": 9.41616697104579e-06,
+      "loss": 0.234,
+      "step": 1986
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.885252230827505,
+      "learning_rate": 9.415475720736005e-06,
+      "loss": 0.193,
+      "step": 1987
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.406536944160158,
+      "learning_rate": 9.4147840868565e-06,
+      "loss": 0.2084,
+      "step": 1988
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3689466693689845,
+      "learning_rate": 9.41409206946735e-06,
+      "loss": 0.1754,
+      "step": 1989
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.732759605468346,
+      "learning_rate": 9.413399668628678e-06,
+      "loss": 0.234,
+      "step": 1990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8465027053557437,
+      "learning_rate": 9.412706884400626e-06,
+      "loss": 0.2314,
+      "step": 1991
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6859551471287766,
+      "learning_rate": 9.41201371684338e-06,
+      "loss": 0.2069,
+      "step": 1992
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4543680190490726,
+      "learning_rate": 9.41132016601715e-06,
+      "loss": 0.1891,
+      "step": 1993
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.543081842279785,
+      "learning_rate": 9.41062623198219e-06,
+      "loss": 0.2007,
+      "step": 1994
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.643504644548493,
+      "learning_rate": 9.40993191479878e-06,
+      "loss": 0.2005,
+      "step": 1995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6386626453698856,
+      "learning_rate": 9.40923721452723e-06,
+      "loss": 0.1965,
+      "step": 1996
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7883578537046163,
+      "learning_rate": 9.408542131227899e-06,
+      "loss": 0.2405,
+      "step": 1997
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5918221423216385,
+      "learning_rate": 9.407846664961156e-06,
+      "loss": 0.2374,
+      "step": 1998
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6790357210419495,
+      "learning_rate": 9.407150815787423e-06,
+      "loss": 0.2215,
+      "step": 1999
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.672340650914591,
+      "learning_rate": 9.406454583767148e-06,
+      "loss": 0.1919,
+      "step": 2000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.199263124042724,
+      "learning_rate": 9.405757968960809e-06,
+      "loss": 0.1702,
+      "step": 2001
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.826228979012623,
+      "learning_rate": 9.405060971428924e-06,
+      "loss": 0.1735,
+      "step": 2002
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8056742959283962,
+      "learning_rate": 9.404363591232038e-06,
+      "loss": 0.2377,
+      "step": 2003
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2464665235776438,
+      "learning_rate": 9.403665828430732e-06,
+      "loss": 0.1711,
+      "step": 2004
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.684128882037673,
+      "learning_rate": 9.402967683085622e-06,
+      "loss": 0.194,
+      "step": 2005
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.55592372971314,
+      "learning_rate": 9.402269155257355e-06,
+      "loss": 0.1856,
+      "step": 2006
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.810726533867702,
+      "learning_rate": 9.401570245006612e-06,
+      "loss": 0.1993,
+      "step": 2007
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.673834535407018,
+      "learning_rate": 9.400870952394105e-06,
+      "loss": 0.2109,
+      "step": 2008
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.554515647138085,
+      "learning_rate": 9.400171277480583e-06,
+      "loss": 0.2084,
+      "step": 2009
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2341086820966525,
+      "learning_rate": 9.399471220326827e-06,
+      "loss": 0.2471,
+      "step": 2010
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6214640591966494,
+      "learning_rate": 9.39877078099365e-06,
+      "loss": 0.2083,
+      "step": 2011
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.765259651089772,
+      "learning_rate": 9.398069959541895e-06,
+      "loss": 0.2277,
+      "step": 2012
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.426256706911028,
+      "learning_rate": 9.397368756032445e-06,
+      "loss": 0.1914,
+      "step": 2013
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.734865562586326,
+      "learning_rate": 9.396667170526215e-06,
+      "loss": 0.1872,
+      "step": 2014
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5323656842840556,
+      "learning_rate": 9.395965203084149e-06,
+      "loss": 0.1916,
+      "step": 2015
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.691272539878832,
+      "learning_rate": 9.39526285376723e-06,
+      "loss": 0.1791,
+      "step": 2016
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.290756305497156,
+      "learning_rate": 9.394560122636463e-06,
+      "loss": 0.1892,
+      "step": 2017
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.695272436574397,
+      "learning_rate": 9.3938570097529e-06,
+      "loss": 0.2303,
+      "step": 2018
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.856504204295195,
+      "learning_rate": 9.393153515177617e-06,
+      "loss": 0.2034,
+      "step": 2019
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6351374699877446,
+      "learning_rate": 9.39244963897173e-06,
+      "loss": 0.2342,
+      "step": 2020
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5280491061253145,
+      "learning_rate": 9.391745381196382e-06,
+      "loss": 0.2085,
+      "step": 2021
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.218726658355057,
+      "learning_rate": 9.39104074191275e-06,
+      "loss": 0.1772,
+      "step": 2022
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.569107509479353,
+      "learning_rate": 9.390335721182047e-06,
+      "loss": 0.2057,
+      "step": 2023
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.425718166092449,
+      "learning_rate": 9.389630319065518e-06,
+      "loss": 0.2033,
+      "step": 2024
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5682746905234297,
+      "learning_rate": 9.38892453562444e-06,
+      "loss": 0.1902,
+      "step": 2025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5803657481503133,
+      "learning_rate": 9.388218370920126e-06,
+      "loss": 0.1877,
+      "step": 2026
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.502568274778471,
+      "learning_rate": 9.387511825013917e-06,
+      "loss": 0.1952,
+      "step": 2027
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.367587850951012,
+      "learning_rate": 9.386804897967192e-06,
+      "loss": 0.215,
+      "step": 2028
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.483290930090042,
+      "learning_rate": 9.386097589841362e-06,
+      "loss": 0.192,
+      "step": 2029
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9069360870111547,
+      "learning_rate": 9.38538990069787e-06,
+      "loss": 0.2559,
+      "step": 2030
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4572354590028342,
+      "learning_rate": 9.384681830598192e-06,
+      "loss": 0.1854,
+      "step": 2031
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4336479586620023,
+      "learning_rate": 9.383973379603837e-06,
+      "loss": 0.2043,
+      "step": 2032
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.936384170713211,
+      "learning_rate": 9.383264547776348e-06,
+      "loss": 0.215,
+      "step": 2033
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0111136718837406,
+      "learning_rate": 9.382555335177301e-06,
+      "loss": 0.2232,
+      "step": 2034
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4228679706758345,
+      "learning_rate": 9.381845741868307e-06,
+      "loss": 0.2205,
+      "step": 2035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.350889180501579,
+      "learning_rate": 9.381135767911005e-06,
+      "loss": 0.1651,
+      "step": 2036
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.833836487250353,
+      "learning_rate": 9.380425413367072e-06,
+      "loss": 0.1983,
+      "step": 2037
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0301445337781305,
+      "learning_rate": 9.379714678298213e-06,
+      "loss": 0.2363,
+      "step": 2038
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6564261981617183,
+      "learning_rate": 9.379003562766172e-06,
+      "loss": 0.2102,
+      "step": 2039
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.605630163024274,
+      "learning_rate": 9.378292066832723e-06,
+      "loss": 0.2156,
+      "step": 2040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3876344503267166,
+      "learning_rate": 9.377580190559674e-06,
+      "loss": 0.1789,
+      "step": 2041
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.407973641103197,
+      "learning_rate": 9.376867934008862e-06,
+      "loss": 0.1637,
+      "step": 2042
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.255162426339039,
+      "learning_rate": 9.376155297242163e-06,
+      "loss": 0.1744,
+      "step": 2043
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5777906248589373,
+      "learning_rate": 9.375442280321483e-06,
+      "loss": 0.2254,
+      "step": 2044
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.560905524212858,
+      "learning_rate": 9.37472888330876e-06,
+      "loss": 0.1884,
+      "step": 2045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3351811571020478,
+      "learning_rate": 9.374015106265968e-06,
+      "loss": 0.1899,
+      "step": 2046
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.667532213283011,
+      "learning_rate": 9.373300949255112e-06,
+      "loss": 0.2127,
+      "step": 2047
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4224288583375313,
+      "learning_rate": 9.372586412338228e-06,
+      "loss": 0.2134,
+      "step": 2048
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.362493648754961,
+      "learning_rate": 9.371871495577391e-06,
+      "loss": 0.1995,
+      "step": 2049
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.930840530202458,
+      "learning_rate": 9.371156199034703e-06,
+      "loss": 0.2192,
+      "step": 2050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.486308103735155,
+      "learning_rate": 9.370440522772305e-06,
+      "loss": 0.217,
+      "step": 2051
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.778524384283164,
+      "learning_rate": 9.369724466852361e-06,
+      "loss": 0.2159,
+      "step": 2052
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.483776678284865,
+      "learning_rate": 9.36900803133708e-06,
+      "loss": 0.2255,
+      "step": 2053
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.490734908559257,
+      "learning_rate": 9.368291216288696e-06,
+      "loss": 0.2032,
+      "step": 2054
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.624057152426544,
+      "learning_rate": 9.367574021769477e-06,
+      "loss": 0.2028,
+      "step": 2055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.790844457952432,
+      "learning_rate": 9.36685644784173e-06,
+      "loss": 0.2242,
+      "step": 2056
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5409723039215666,
+      "learning_rate": 9.366138494567785e-06,
+      "loss": 0.2074,
+      "step": 2057
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5905844037475796,
+      "learning_rate": 9.365420162010011e-06,
+      "loss": 0.1891,
+      "step": 2058
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5827029992598973,
+      "learning_rate": 9.364701450230813e-06,
+      "loss": 0.2509,
+      "step": 2059
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5821468141160038,
+      "learning_rate": 9.36398235929262e-06,
+      "loss": 0.193,
+      "step": 2060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7476511990821915,
+      "learning_rate": 9.363262889257902e-06,
+      "loss": 0.2534,
+      "step": 2061
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5992986961210045,
+      "learning_rate": 9.36254304018916e-06,
+      "loss": 0.205,
+      "step": 2062
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5964621959352434,
+      "learning_rate": 9.361822812148925e-06,
+      "loss": 0.2411,
+      "step": 2063
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.348374146086673,
+      "learning_rate": 9.361102205199762e-06,
+      "loss": 0.1912,
+      "step": 2064
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.389593786680925,
+      "learning_rate": 9.360381219404268e-06,
+      "loss": 0.158,
+      "step": 2065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.342012078792294,
+      "learning_rate": 9.35965985482508e-06,
+      "loss": 0.1978,
+      "step": 2066
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.44827652113443,
+      "learning_rate": 9.35893811152486e-06,
+      "loss": 0.1982,
+      "step": 2067
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6125108406311552,
+      "learning_rate": 9.358215989566304e-06,
+      "loss": 0.2215,
+      "step": 2068
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6864063546168135,
+      "learning_rate": 9.357493489012147e-06,
+      "loss": 0.2506,
+      "step": 2069
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5296836077942304,
+      "learning_rate": 9.356770609925143e-06,
+      "loss": 0.1973,
+      "step": 2070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1646610855312853,
+      "learning_rate": 9.356047352368096e-06,
+      "loss": 0.2246,
+      "step": 2071
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.586692582512955,
+      "learning_rate": 9.355323716403834e-06,
+      "loss": 0.1868,
+      "step": 2072
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.697741948964955,
+      "learning_rate": 9.354599702095218e-06,
+      "loss": 0.2085,
+      "step": 2073
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5813862809412713,
+      "learning_rate": 9.353875309505141e-06,
+      "loss": 0.201,
+      "step": 2074
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6241177565790523,
+      "learning_rate": 9.353150538696531e-06,
+      "loss": 0.2017,
+      "step": 2075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3519312083482364,
+      "learning_rate": 9.35242538973235e-06,
+      "loss": 0.1834,
+      "step": 2076
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.443726068958758,
+      "learning_rate": 9.351699862675589e-06,
+      "loss": 0.1464,
+      "step": 2077
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.662134078970126,
+      "learning_rate": 9.350973957589278e-06,
+      "loss": 0.208,
+      "step": 2078
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.633001440059104,
+      "learning_rate": 9.35024767453647e-06,
+      "loss": 0.2104,
+      "step": 2079
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.504696823995845,
+      "learning_rate": 9.349521013580262e-06,
+      "loss": 0.2071,
+      "step": 2080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.85331303646671,
+      "learning_rate": 9.348793974783778e-06,
+      "loss": 0.2379,
+      "step": 2081
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5933256336894783,
+      "learning_rate": 9.348066558210174e-06,
+      "loss": 0.209,
+      "step": 2082
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.418385976699233,
+      "learning_rate": 9.34733876392264e-06,
+      "loss": 0.217,
+      "step": 2083
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.592403604363241,
+      "learning_rate": 9.346610591984398e-06,
+      "loss": 0.2376,
+      "step": 2084
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8134062455861577,
+      "learning_rate": 9.345882042458708e-06,
+      "loss": 0.248,
+      "step": 2085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8066052617759984,
+      "learning_rate": 9.345153115408854e-06,
+      "loss": 0.2222,
+      "step": 2086
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.526065099128459,
+      "learning_rate": 9.34442381089816e-06,
+      "loss": 0.2034,
+      "step": 2087
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7833584958751167,
+      "learning_rate": 9.343694128989979e-06,
+      "loss": 0.233,
+      "step": 2088
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7512879606674825,
+      "learning_rate": 9.3429640697477e-06,
+      "loss": 0.2513,
+      "step": 2089
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3346064071511567,
+      "learning_rate": 9.34223363323474e-06,
+      "loss": 0.1983,
+      "step": 2090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4612795853630147,
+      "learning_rate": 9.341502819514555e-06,
+      "loss": 0.2146,
+      "step": 2091
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5919535815213033,
+      "learning_rate": 9.340771628650628e-06,
+      "loss": 0.2189,
+      "step": 2092
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.845928305763761,
+      "learning_rate": 9.340040060706477e-06,
+      "loss": 0.2308,
+      "step": 2093
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7728025885276697,
+      "learning_rate": 9.339308115745654e-06,
+      "loss": 0.2416,
+      "step": 2094
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.756575008360787,
+      "learning_rate": 9.338575793831742e-06,
+      "loss": 0.197,
+      "step": 2095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5862909447618496,
+      "learning_rate": 9.337843095028357e-06,
+      "loss": 0.2095,
+      "step": 2096
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.49508953339067,
+      "learning_rate": 9.33711001939915e-06,
+      "loss": 0.1963,
+      "step": 2097
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2600581822721124,
+      "learning_rate": 9.336376567007799e-06,
+      "loss": 0.1723,
+      "step": 2098
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5511537118576566,
+      "learning_rate": 9.335642737918023e-06,
+      "loss": 0.225,
+      "step": 2099
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3327355427776744,
+      "learning_rate": 9.334908532193567e-06,
+      "loss": 0.1715,
+      "step": 2100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.920818904690184,
+      "learning_rate": 9.334173949898211e-06,
+      "loss": 0.2482,
+      "step": 2101
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7062295513631067,
+      "learning_rate": 9.333438991095767e-06,
+      "loss": 0.2147,
+      "step": 2102
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3807795559711673,
+      "learning_rate": 9.332703655850082e-06,
+      "loss": 0.2088,
+      "step": 2103
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3289614045043945,
+      "learning_rate": 9.331967944225034e-06,
+      "loss": 0.1775,
+      "step": 2104
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1916492647709345,
+      "learning_rate": 9.331231856284532e-06,
+      "loss": 0.2639,
+      "step": 2105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1850514504843503,
+      "learning_rate": 9.330495392092525e-06,
+      "loss": 0.2554,
+      "step": 2106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.567207117696942,
+      "learning_rate": 9.32975855171298e-06,
+      "loss": 0.1999,
+      "step": 2107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7436969616040896,
+      "learning_rate": 9.329021335209913e-06,
+      "loss": 0.2368,
+      "step": 2108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4937594081460155,
+      "learning_rate": 9.328283742647365e-06,
+      "loss": 0.1952,
+      "step": 2109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.565330792675296,
+      "learning_rate": 9.327545774089407e-06,
+      "loss": 0.2215,
+      "step": 2110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0092278156350196,
+      "learning_rate": 9.326807429600148e-06,
+      "loss": 0.2218,
+      "step": 2111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6764990407114406,
+      "learning_rate": 9.326068709243727e-06,
+      "loss": 0.209,
+      "step": 2112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.657926874049776,
+      "learning_rate": 9.325329613084317e-06,
+      "loss": 0.2355,
+      "step": 2113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6904005041241925,
+      "learning_rate": 9.324590141186123e-06,
+      "loss": 0.2072,
+      "step": 2114
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6118869426331783,
+      "learning_rate": 9.32385029361338e-06,
+      "loss": 0.2122,
+      "step": 2115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.451889794297275,
+      "learning_rate": 9.32311007043036e-06,
+      "loss": 0.2131,
+      "step": 2116
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4043279314029027,
+      "learning_rate": 9.322369471701367e-06,
+      "loss": 0.1735,
+      "step": 2117
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.413460952984375,
+      "learning_rate": 9.321628497490733e-06,
+      "loss": 0.1775,
+      "step": 2118
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8278178212703065,
+      "learning_rate": 9.32088714786283e-06,
+      "loss": 0.2253,
+      "step": 2119
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6485185472846444,
+      "learning_rate": 9.320145422882055e-06,
+      "loss": 0.2233,
+      "step": 2120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5247783694907433,
+      "learning_rate": 9.319403322612843e-06,
+      "loss": 0.2247,
+      "step": 2121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.220936374091986,
+      "learning_rate": 9.31866084711966e-06,
+      "loss": 0.1679,
+      "step": 2122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6856875359969132,
+      "learning_rate": 9.317917996467004e-06,
+      "loss": 0.2068,
+      "step": 2123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.448446404171035,
+      "learning_rate": 9.317174770719404e-06,
+      "loss": 0.2104,
+      "step": 2124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6013910956835167,
+      "learning_rate": 9.316431169941427e-06,
+      "loss": 0.2031,
+      "step": 2125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5404512092092766,
+      "learning_rate": 9.315687194197667e-06,
+      "loss": 0.1859,
+      "step": 2126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.536744435897603,
+      "learning_rate": 9.314942843552754e-06,
+      "loss": 0.196,
+      "step": 2127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.329118337890104,
+      "learning_rate": 9.314198118071349e-06,
+      "loss": 0.1949,
+      "step": 2128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3806052355601723,
+      "learning_rate": 9.313453017818144e-06,
+      "loss": 0.203,
+      "step": 2129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1550408684846882,
+      "learning_rate": 9.312707542857868e-06,
+      "loss": 0.1966,
+      "step": 2130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.510214207833926,
+      "learning_rate": 9.311961693255281e-06,
+      "loss": 0.2383,
+      "step": 2131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4929614422526685,
+      "learning_rate": 9.311215469075168e-06,
+      "loss": 0.2009,
+      "step": 2132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.460049342735931,
+      "learning_rate": 9.310468870382362e-06,
+      "loss": 0.176,
+      "step": 2133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.20905485079759,
+      "learning_rate": 9.309721897241712e-06,
+      "loss": 0.1701,
+      "step": 2134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6165381946184447,
+      "learning_rate": 9.30897454971811e-06,
+      "loss": 0.2305,
+      "step": 2135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.393248267509264,
+      "learning_rate": 9.308226827876478e-06,
+      "loss": 0.2181,
+      "step": 2136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7512517490500645,
+      "learning_rate": 9.307478731781772e-06,
+      "loss": 0.2472,
+      "step": 2137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2858045216344975,
+      "learning_rate": 9.306730261498973e-06,
+      "loss": 0.2112,
+      "step": 2138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5539745999920105,
+      "learning_rate": 9.305981417093106e-06,
+      "loss": 0.2101,
+      "step": 2139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4273099077392057,
+      "learning_rate": 9.30523219862922e-06,
+      "loss": 0.2299,
+      "step": 2140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5049969085942556,
+      "learning_rate": 9.304482606172401e-06,
+      "loss": 0.2145,
+      "step": 2141
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.699400857466148,
+      "learning_rate": 9.303732639787761e-06,
+      "loss": 0.2082,
+      "step": 2142
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7153707592171217,
+      "learning_rate": 9.302982299540455e-06,
+      "loss": 0.2369,
+      "step": 2143
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3292310329969435,
+      "learning_rate": 9.30223158549566e-06,
+      "loss": 0.1644,
+      "step": 2144
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5661604655312886,
+      "learning_rate": 9.301480497718594e-06,
+      "loss": 0.2138,
+      "step": 2145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.684056971993484,
+      "learning_rate": 9.300729036274501e-06,
+      "loss": 0.1889,
+      "step": 2146
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7374630586178834,
+      "learning_rate": 9.29997720122866e-06,
+      "loss": 0.1848,
+      "step": 2147
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6696952199659463,
+      "learning_rate": 9.299224992646383e-06,
+      "loss": 0.2356,
+      "step": 2148
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.192027484973054,
+      "learning_rate": 9.298472410593013e-06,
+      "loss": 0.1835,
+      "step": 2149
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2341993225528602,
+      "learning_rate": 9.29771945513393e-06,
+      "loss": 0.1701,
+      "step": 2150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7376584481437916,
+      "learning_rate": 9.296966126334538e-06,
+      "loss": 0.2127,
+      "step": 2151
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4134134284880453,
+      "learning_rate": 9.29621242426028e-06,
+      "loss": 0.2281,
+      "step": 2152
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3872216621139684,
+      "learning_rate": 9.295458348976632e-06,
+      "loss": 0.1927,
+      "step": 2153
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3646011573688273,
+      "learning_rate": 9.294703900549096e-06,
+      "loss": 0.1949,
+      "step": 2154
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6320640584006716,
+      "learning_rate": 9.293949079043212e-06,
+      "loss": 0.2274,
+      "step": 2155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8060272356137985,
+      "learning_rate": 9.293193884524554e-06,
+      "loss": 0.2024,
+      "step": 2156
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.521337567867084,
+      "learning_rate": 9.29243831705872e-06,
+      "loss": 0.219,
+      "step": 2157
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.569372889537845,
+      "learning_rate": 9.29168237671135e-06,
+      "loss": 0.1965,
+      "step": 2158
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9235388025968154,
+      "learning_rate": 9.290926063548109e-06,
+      "loss": 0.2257,
+      "step": 2159
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3844047758467566,
+      "learning_rate": 9.2901693776347e-06,
+      "loss": 0.2107,
+      "step": 2160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4737124488256255,
+      "learning_rate": 9.289412319036854e-06,
+      "loss": 0.1858,
+      "step": 2161
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1918016217053644,
+      "learning_rate": 9.288654887820337e-06,
+      "loss": 0.2093,
+      "step": 2162
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.60183909879754,
+      "learning_rate": 9.287897084050947e-06,
+      "loss": 0.2014,
+      "step": 2163
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2452810178972658,
+      "learning_rate": 9.287138907794514e-06,
+      "loss": 0.2155,
+      "step": 2164
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2047539569481898,
+      "learning_rate": 9.2863803591169e-06,
+      "loss": 0.1951,
+      "step": 2165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5579799380737205,
+      "learning_rate": 9.285621438083997e-06,
+      "loss": 0.2227,
+      "step": 2166
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.566045913231194,
+      "learning_rate": 9.284862144761736e-06,
+      "loss": 0.2266,
+      "step": 2167
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.288937732848183,
+      "learning_rate": 9.284102479216076e-06,
+      "loss": 0.1782,
+      "step": 2168
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5182564029899623,
+      "learning_rate": 9.283342441513008e-06,
+      "loss": 0.2119,
+      "step": 2169
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.650736729461245,
+      "learning_rate": 9.282582031718554e-06,
+      "loss": 0.2321,
+      "step": 2170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4123694904963933,
+      "learning_rate": 9.281821249898772e-06,
+      "loss": 0.1949,
+      "step": 2171
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.521404272246224,
+      "learning_rate": 9.281060096119751e-06,
+      "loss": 0.1934,
+      "step": 2172
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7190636854264847,
+      "learning_rate": 9.280298570447612e-06,
+      "loss": 0.2518,
+      "step": 2173
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8348195168176717,
+      "learning_rate": 9.279536672948508e-06,
+      "loss": 0.237,
+      "step": 2174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.568239341563141,
+      "learning_rate": 9.278774403688624e-06,
+      "loss": 0.1879,
+      "step": 2175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4247706948958436,
+      "learning_rate": 9.278011762734179e-06,
+      "loss": 0.1856,
+      "step": 2176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5458975973234548,
+      "learning_rate": 9.277248750151419e-06,
+      "loss": 0.2187,
+      "step": 2177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8079111609954728,
+      "learning_rate": 9.276485366006634e-06,
+      "loss": 0.2132,
+      "step": 2178
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6980350408333362,
+      "learning_rate": 9.275721610366134e-06,
+      "loss": 0.2008,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.449053889216658,
+      "learning_rate": 9.274957483296263e-06,
+      "loss": 0.1802,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.420252226896554,
+      "learning_rate": 9.274192984863409e-06,
+      "loss": 0.1927,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.720295193423951,
+      "learning_rate": 9.273428115133975e-06,
+      "loss": 0.217,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5627759383174817,
+      "learning_rate": 9.27266287417441e-06,
+      "loss": 0.1894,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.391417106873159,
+      "learning_rate": 9.271897262051186e-06,
+      "loss": 0.2124,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6370947993455442,
+      "learning_rate": 9.271131278830815e-06,
+      "loss": 0.19,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3735696682383924,
+      "learning_rate": 9.270364924579835e-06,
+      "loss": 0.1989,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4382396671086357,
+      "learning_rate": 9.269598199364821e-06,
+      "loss": 0.1997,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4815437912767506,
+      "learning_rate": 9.268831103252376e-06,
+      "loss": 0.2058,
+      "step": 2188
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6260524860068184,
+      "learning_rate": 9.268063636309138e-06,
+      "loss": 0.2074,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6209845059794405,
+      "learning_rate": 9.267295798601777e-06,
+      "loss": 0.2057,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.373248069102358,
+      "learning_rate": 9.266527590196992e-06,
+      "loss": 0.1927,
+      "step": 2191
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4207634492170342,
+      "learning_rate": 9.265759011161519e-06,
+      "loss": 0.2248,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.161131946062331,
+      "learning_rate": 9.264990061562125e-06,
+      "loss": 0.2892,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7128839987119275,
+      "learning_rate": 9.264220741465606e-06,
+      "loss": 0.2507,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9127508809489773,
+      "learning_rate": 9.263451050938792e-06,
+      "loss": 0.2079,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.602295486422228,
+      "learning_rate": 9.262680990048549e-06,
+      "loss": 0.2015,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.426353271983918,
+      "learning_rate": 9.261910558861767e-06,
+      "loss": 0.1721,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6281971888874023,
+      "learning_rate": 9.261139757445378e-06,
+      "loss": 0.2446,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6750678521031217,
+      "learning_rate": 9.260368585866338e-06,
+      "loss": 0.1887,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.439498290770255,
+      "learning_rate": 9.259597044191635e-06,
+      "loss": 0.1609,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.57565689154324,
+      "learning_rate": 9.258825132488301e-06,
+      "loss": 0.2046,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.469957936835051,
+      "learning_rate": 9.258052850823383e-06,
+      "loss": 0.1868,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.555981929385473,
+      "learning_rate": 9.257280199263975e-06,
+      "loss": 0.2084,
+      "step": 2203
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6755242032001685,
+      "learning_rate": 9.256507177877191e-06,
+      "loss": 0.2144,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9644514387571057,
+      "learning_rate": 9.255733786730187e-06,
+      "loss": 0.2233,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4233369565054117,
+      "learning_rate": 9.254960025890146e-06,
+      "loss": 0.1705,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.786675936072335,
+      "learning_rate": 9.254185895424284e-06,
+      "loss": 0.2475,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.583731197981944,
+      "learning_rate": 9.253411395399849e-06,
+      "loss": 0.2112,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.555591835235902,
+      "learning_rate": 9.25263652588412e-06,
+      "loss": 0.2283,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7282927973842215,
+      "learning_rate": 9.251861286944415e-06,
+      "loss": 0.183,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.60227318895484,
+      "learning_rate": 9.251085678648072e-06,
+      "loss": 0.2134,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5177181433857774,
+      "learning_rate": 9.25030970106247e-06,
+      "loss": 0.2091,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2608998775136233,
+      "learning_rate": 9.249533354255019e-06,
+      "loss": 0.1868,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4753627699685024,
+      "learning_rate": 9.248756638293156e-06,
+      "loss": 0.2058,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.103874205400295,
+      "learning_rate": 9.24797955324436e-06,
+      "loss": 0.1633,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3019823158672947,
+      "learning_rate": 9.24720209917613e-06,
+      "loss": 0.1926,
+      "step": 2216
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4833454095303154,
+      "learning_rate": 9.246424276156008e-06,
+      "loss": 0.1709,
+      "step": 2217
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6954470571054947,
+      "learning_rate": 9.245646084251558e-06,
+      "loss": 0.2511,
+      "step": 2218
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5725226450068868,
+      "learning_rate": 9.244867523530385e-06,
+      "loss": 0.1748,
+      "step": 2219
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4366714626084462,
+      "learning_rate": 9.24408859406012e-06,
+      "loss": 0.1892,
+      "step": 2220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9003770944177596,
+      "learning_rate": 9.243309295908429e-06,
+      "loss": 0.2223,
+      "step": 2221
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.373805078966787,
+      "learning_rate": 9.24252962914301e-06,
+      "loss": 0.1882,
+      "step": 2222
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3643264728169773,
+      "learning_rate": 9.241749593831588e-06,
+      "loss": 0.2111,
+      "step": 2223
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4892635326650145,
+      "learning_rate": 9.24096919004193e-06,
+      "loss": 0.2004,
+      "step": 2224
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2260087973239893,
+      "learning_rate": 9.240188417841824e-06,
+      "loss": 0.1972,
+      "step": 2225
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4923252320051366,
+      "learning_rate": 9.239407277299101e-06,
+      "loss": 0.1983,
+      "step": 2226
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.38887599003807,
+      "learning_rate": 9.238625768481612e-06,
+      "loss": 0.2028,
+      "step": 2227
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.702123916649951,
+      "learning_rate": 9.23784389145725e-06,
+      "loss": 0.2068,
+      "step": 2228
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.524056496186482,
+      "learning_rate": 9.237061646293937e-06,
+      "loss": 0.2283,
+      "step": 2229
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6790704610160523,
+      "learning_rate": 9.236279033059622e-06,
+      "loss": 0.2152,
+      "step": 2230
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.531486803326059,
+      "learning_rate": 9.235496051822293e-06,
+      "loss": 0.1936,
+      "step": 2231
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4294877192058992,
+      "learning_rate": 9.234712702649969e-06,
+      "loss": 0.2114,
+      "step": 2232
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.941051943504525,
+      "learning_rate": 9.233928985610693e-06,
+      "loss": 0.23,
+      "step": 2233
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4661294728495418,
+      "learning_rate": 9.233144900772553e-06,
+      "loss": 0.1802,
+      "step": 2234
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.562296352742423,
+      "learning_rate": 9.232360448203658e-06,
+      "loss": 0.1911,
+      "step": 2235
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3751514612177744,
+      "learning_rate": 9.231575627972153e-06,
+      "loss": 0.2145,
+      "step": 2236
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6790040889535183,
+      "learning_rate": 9.230790440146216e-06,
+      "loss": 0.2122,
+      "step": 2237
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6196786702959645,
+      "learning_rate": 9.230004884794056e-06,
+      "loss": 0.2186,
+      "step": 2238
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.507917127533199,
+      "learning_rate": 9.229218961983913e-06,
+      "loss": 0.1737,
+      "step": 2239
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.571893225282125,
+      "learning_rate": 9.228432671784057e-06,
+      "loss": 0.18,
+      "step": 2240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6517320268659414,
+      "learning_rate": 9.227646014262799e-06,
+      "loss": 0.2132,
+      "step": 2241
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4226470358598933,
+      "learning_rate": 9.22685898948847e-06,
+      "loss": 0.1812,
+      "step": 2242
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.691860727993957,
+      "learning_rate": 9.22607159752944e-06,
+      "loss": 0.2385,
+      "step": 2243
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.610457567357634,
+      "learning_rate": 9.225283838454111e-06,
+      "loss": 0.2069,
+      "step": 2244
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.287272287222549,
+      "learning_rate": 9.224495712330911e-06,
+      "loss": 0.1833,
+      "step": 2245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6967522817328424,
+      "learning_rate": 9.223707219228309e-06,
+      "loss": 0.2606,
+      "step": 2246
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5677464355327273,
+      "learning_rate": 9.222918359214798e-06,
+      "loss": 0.2088,
+      "step": 2247
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.228138654674976,
+      "learning_rate": 9.222129132358905e-06,
+      "loss": 0.1955,
+      "step": 2248
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7524106732804516,
+      "learning_rate": 9.221339538729191e-06,
+      "loss": 0.2326,
+      "step": 2249
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8469162027419754,
+      "learning_rate": 9.220549578394249e-06,
+      "loss": 0.2475,
+      "step": 2250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5654107410252474,
+      "learning_rate": 9.2197592514227e-06,
+      "loss": 0.1996,
+      "step": 2251
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.85828058348352,
+      "learning_rate": 9.2189685578832e-06,
+      "loss": 0.2241,
+      "step": 2252
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.6227627857354525,
+      "learning_rate": 9.218177497844438e-06,
+      "loss": 0.2305,
+      "step": 2253
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.489952855966967,
+      "learning_rate": 9.217386071375129e-06,
+      "loss": 0.2072,
+      "step": 2254
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3319205365781035,
+      "learning_rate": 9.216594278544026e-06,
+      "loss": 0.1919,
+      "step": 2255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3848896374278543,
+      "learning_rate": 9.215802119419912e-06,
+      "loss": 0.2,
+      "step": 2256
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7626309489641967,
+      "learning_rate": 9.2150095940716e-06,
+      "loss": 0.2021,
+      "step": 2257
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3814894832398634,
+      "learning_rate": 9.214216702567937e-06,
+      "loss": 0.1596,
+      "step": 2258
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4858932665283797,
+      "learning_rate": 9.213423444977802e-06,
+      "loss": 0.1887,
+      "step": 2259
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8814946682273184,
+      "learning_rate": 9.212629821370104e-06,
+      "loss": 0.2191,
+      "step": 2260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3771866338389738,
+      "learning_rate": 9.211835831813782e-06,
+      "loss": 0.1945,
+      "step": 2261
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5824248953859152,
+      "learning_rate": 9.211041476377815e-06,
+      "loss": 0.1795,
+      "step": 2262
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6829547605660515,
+      "learning_rate": 9.210246755131204e-06,
+      "loss": 0.2167,
+      "step": 2263
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8895884976101955,
+      "learning_rate": 9.209451668142985e-06,
+      "loss": 0.2107,
+      "step": 2264
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3545112337814396,
+      "learning_rate": 9.20865621548223e-06,
+      "loss": 0.1627,
+      "step": 2265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.643819032093641,
+      "learning_rate": 9.20786039721804e-06,
+      "loss": 0.2269,
+      "step": 2266
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5771608728719277,
+      "learning_rate": 9.207064213419543e-06,
+      "loss": 0.1847,
+      "step": 2267
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4209425767600563,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 0.2118,
+      "step": 2268
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5796833818822664,
+      "learning_rate": 9.205470749496326e-06,
+      "loss": 0.226,
+      "step": 2269
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.842896162356135,
+      "learning_rate": 9.204673469510025e-06,
+      "loss": 0.1982,
+      "step": 2270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.529868123364016,
+      "learning_rate": 9.203875824266269e-06,
+      "loss": 0.1942,
+      "step": 2271
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.302580492244617,
+      "learning_rate": 9.203077813834345e-06,
+      "loss": 0.1705,
+      "step": 2272
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1416593207256684,
+      "learning_rate": 9.202279438283577e-06,
+      "loss": 0.1855,
+      "step": 2273
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.492684078471484,
+      "learning_rate": 9.201480697683319e-06,
+      "loss": 0.1913,
+      "step": 2274
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6180141685541685,
+      "learning_rate": 9.200681592102955e-06,
+      "loss": 0.2137,
+      "step": 2275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.630713575189787,
+      "learning_rate": 9.199882121611907e-06,
+      "loss": 0.2268,
+      "step": 2276
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.499227011095003,
+      "learning_rate": 9.199082286279622e-06,
+      "loss": 0.2072,
+      "step": 2277
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2589567657768264,
+      "learning_rate": 9.198282086175582e-06,
+      "loss": 0.1974,
+      "step": 2278
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.30818219670286,
+      "learning_rate": 9.197481521369299e-06,
+      "loss": 0.1922,
+      "step": 2279
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6818294975956243,
+      "learning_rate": 9.196680591930318e-06,
+      "loss": 0.1994,
+      "step": 2280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.581348724363768,
+      "learning_rate": 9.195879297928217e-06,
+      "loss": 0.1948,
+      "step": 2281
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.373275487807223,
+      "learning_rate": 9.195077639432599e-06,
+      "loss": 0.181,
+      "step": 2282
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.789135756288639,
+      "learning_rate": 9.19427561651311e-06,
+      "loss": 0.2037,
+      "step": 2283
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4823239168037237,
+      "learning_rate": 9.193473229239417e-06,
+      "loss": 0.21,
+      "step": 2284
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.657455426959781,
+      "learning_rate": 9.192670477681224e-06,
+      "loss": 0.2134,
+      "step": 2285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8368552341607973,
+      "learning_rate": 9.191867361908265e-06,
+      "loss": 0.2164,
+      "step": 2286
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.372927718917402,
+      "learning_rate": 9.191063881990308e-06,
+      "loss": 0.1829,
+      "step": 2287
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4615733313114796,
+      "learning_rate": 9.190260037997149e-06,
+      "loss": 0.1821,
+      "step": 2288
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.955995801986834,
+      "learning_rate": 9.18945582999862e-06,
+      "loss": 0.1968,
+      "step": 2289
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.99586152877953,
+      "learning_rate": 9.188651258064578e-06,
+      "loss": 0.2663,
+      "step": 2290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3867822819661932,
+      "learning_rate": 9.187846322264918e-06,
+      "loss": 0.2123,
+      "step": 2291
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4441363284071396,
+      "learning_rate": 9.187041022669562e-06,
+      "loss": 0.1639,
+      "step": 2292
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2085418968273136,
+      "learning_rate": 9.186235359348472e-06,
+      "loss": 0.2219,
+      "step": 2293
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5316935222268997,
+      "learning_rate": 9.18542933237163e-06,
+      "loss": 0.2067,
+      "step": 2294
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4250991928934607,
+      "learning_rate": 9.184622941809056e-06,
+      "loss": 0.2069,
+      "step": 2295
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.451692458129849,
+      "learning_rate": 9.183816187730801e-06,
+      "loss": 0.2226,
+      "step": 2296
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.819762800542777,
+      "learning_rate": 9.183009070206947e-06,
+      "loss": 0.1914,
+      "step": 2297
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1409874568250555,
+      "learning_rate": 9.18220158930761e-06,
+      "loss": 0.1817,
+      "step": 2298
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7705844408200377,
+      "learning_rate": 9.181393745102933e-06,
+      "loss": 0.2391,
+      "step": 2299
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3212913216712927,
+      "learning_rate": 9.180585537663093e-06,
+      "loss": 0.211,
+      "step": 2300
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.490546854112704,
+      "learning_rate": 9.179776967058301e-06,
+      "loss": 0.2167,
+      "step": 2301
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.785336218043496,
+      "learning_rate": 9.178968033358792e-06,
+      "loss": 0.2155,
+      "step": 2302
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.529619917330041,
+      "learning_rate": 9.178158736634843e-06,
+      "loss": 0.2219,
+      "step": 2303
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.141144489191175,
+      "learning_rate": 9.177349076956755e-06,
+      "loss": 0.2405,
+      "step": 2304
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390693943568701,
+      "learning_rate": 9.176539054394861e-06,
+      "loss": 0.1634,
+      "step": 2305
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4829239723812413,
+      "learning_rate": 9.17572866901953e-06,
+      "loss": 0.2161,
+      "step": 2306
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3760852544012767,
+      "learning_rate": 9.174917920901156e-06,
+      "loss": 0.1572,
+      "step": 2307
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2753702872193706,
+      "learning_rate": 9.174106810110173e-06,
+      "loss": 0.1803,
+      "step": 2308
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.424409432888883,
+      "learning_rate": 9.173295336717039e-06,
+      "loss": 0.2061,
+      "step": 2309
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.380403302759444,
+      "learning_rate": 9.172483500792246e-06,
+      "loss": 0.2143,
+      "step": 2310
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.366735171800615,
+      "learning_rate": 9.171671302406317e-06,
+      "loss": 0.2072,
+      "step": 2311
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5404571786951338,
+      "learning_rate": 9.17085874162981e-06,
+      "loss": 0.1761,
+      "step": 2312
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5021129312986186,
+      "learning_rate": 9.17004581853331e-06,
+      "loss": 0.23,
+      "step": 2313
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.423851749270581,
+      "learning_rate": 9.169232533187434e-06,
+      "loss": 0.1916,
+      "step": 2314
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.277235982766273,
+      "learning_rate": 9.168418885662833e-06,
+      "loss": 0.1872,
+      "step": 2315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.524266286636938,
+      "learning_rate": 9.16760487603019e-06,
+      "loss": 0.2128,
+      "step": 2316
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4324222436109526,
+      "learning_rate": 9.166790504360213e-06,
+      "loss": 0.1722,
+      "step": 2317
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.775058622059398,
+      "learning_rate": 9.165975770723649e-06,
+      "loss": 0.2067,
+      "step": 2318
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7675433299186416,
+      "learning_rate": 9.165160675191272e-06,
+      "loss": 0.2304,
+      "step": 2319
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4508904010052195,
+      "learning_rate": 9.164345217833892e-06,
+      "loss": 0.1771,
+      "step": 2320
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.928070351068871,
+      "learning_rate": 9.163529398722341e-06,
+      "loss": 0.1783,
+      "step": 2321
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.532753074523733,
+      "learning_rate": 9.162713217927496e-06,
+      "loss": 0.1929,
+      "step": 2322
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4023598033593556,
+      "learning_rate": 9.161896675520255e-06,
+      "loss": 0.178,
+      "step": 2323
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.521448453266078,
+      "learning_rate": 9.161079771571548e-06,
+      "loss": 0.1823,
+      "step": 2324
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2797182925069497,
+      "learning_rate": 9.160262506152343e-06,
+      "loss": 0.2003,
+      "step": 2325
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4963420733253887,
+      "learning_rate": 9.159444879333632e-06,
+      "loss": 0.1956,
+      "step": 2326
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5921579676631157,
+      "learning_rate": 9.158626891186444e-06,
+      "loss": 0.2148,
+      "step": 2327
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5903548975508017,
+      "learning_rate": 9.157808541781837e-06,
+      "loss": 0.2428,
+      "step": 2328
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.688871673826555,
+      "learning_rate": 9.1569898311909e-06,
+      "loss": 0.1799,
+      "step": 2329
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6669491093957283,
+      "learning_rate": 9.156170759484754e-06,
+      "loss": 0.2222,
+      "step": 2330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2530311279808366,
+      "learning_rate": 9.15535132673455e-06,
+      "loss": 0.1704,
+      "step": 2331
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.525118043745158,
+      "learning_rate": 9.154531533011474e-06,
+      "loss": 0.2325,
+      "step": 2332
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.611461519542473,
+      "learning_rate": 9.15371137838674e-06,
+      "loss": 0.2184,
+      "step": 2333
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5116887195049347,
+      "learning_rate": 9.152890862931594e-06,
+      "loss": 0.1927,
+      "step": 2334
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6384672348798244,
+      "learning_rate": 9.152069986717313e-06,
+      "loss": 0.2143,
+      "step": 2335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.602262864379223,
+      "learning_rate": 9.151248749815208e-06,
+      "loss": 0.2304,
+      "step": 2336
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4878049048128283,
+      "learning_rate": 9.150427152296617e-06,
+      "loss": 0.2165,
+      "step": 2337
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.449314616898853,
+      "learning_rate": 9.149605194232915e-06,
+      "loss": 0.2058,
+      "step": 2338
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.616224754079725,
+      "learning_rate": 9.1487828756955e-06,
+      "loss": 0.204,
+      "step": 2339
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.826569344632306,
+      "learning_rate": 9.147960196755811e-06,
+      "loss": 0.2051,
+      "step": 2340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.632979306591604,
+      "learning_rate": 9.147137157485313e-06,
+      "loss": 0.2354,
+      "step": 2341
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.614383356719092,
+      "learning_rate": 9.146313757955501e-06,
+      "loss": 0.2203,
+      "step": 2342
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6829268903903087,
+      "learning_rate": 9.145489998237902e-06,
+      "loss": 0.2123,
+      "step": 2343
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7470750383820572,
+      "learning_rate": 9.14466587840408e-06,
+      "loss": 0.2152,
+      "step": 2344
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4731689934132466,
+      "learning_rate": 9.143841398525621e-06,
+      "loss": 0.1678,
+      "step": 2345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7265545322283593,
+      "learning_rate": 9.14301655867415e-06,
+      "loss": 0.2235,
+      "step": 2346
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4005540914942793,
+      "learning_rate": 9.14219135892132e-06,
+      "loss": 0.1817,
+      "step": 2347
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.63075932459022,
+      "learning_rate": 9.141365799338817e-06,
+      "loss": 0.222,
+      "step": 2348
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.637870873628337,
+      "learning_rate": 9.140539879998353e-06,
+      "loss": 0.2368,
+      "step": 2349
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2160145430440275,
+      "learning_rate": 9.139713600971677e-06,
+      "loss": 0.185,
+      "step": 2350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.705252575292958,
+      "learning_rate": 9.13888696233057e-06,
+      "loss": 0.2223,
+      "step": 2351
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.529558226859568,
+      "learning_rate": 9.138059964146839e-06,
+      "loss": 0.1875,
+      "step": 2352
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.364523498187168,
+      "learning_rate": 9.137232606492323e-06,
+      "loss": 0.2073,
+      "step": 2353
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5583423616556535,
+      "learning_rate": 9.136404889438898e-06,
+      "loss": 0.2287,
+      "step": 2354
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4761735879009543,
+      "learning_rate": 9.135576813058465e-06,
+      "loss": 0.1977,
+      "step": 2355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7569134643346818,
+      "learning_rate": 9.134748377422959e-06,
+      "loss": 0.2077,
+      "step": 2356
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.29017605428271,
+      "learning_rate": 9.133919582604344e-06,
+      "loss": 0.1863,
+      "step": 2357
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7071280771654864,
+      "learning_rate": 9.133090428674621e-06,
+      "loss": 0.1938,
+      "step": 2358
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6404789634689285,
+      "learning_rate": 9.132260915705814e-06,
+      "loss": 0.1861,
+      "step": 2359
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1560569466912836,
+      "learning_rate": 9.131431043769986e-06,
+      "loss": 0.2064,
+      "step": 2360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1571917601712345,
+      "learning_rate": 9.130600812939223e-06,
+      "loss": 0.1698,
+      "step": 2361
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.517039789499306,
+      "learning_rate": 9.12977022328565e-06,
+      "loss": 0.201,
+      "step": 2362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6096427349526934,
+      "learning_rate": 9.12893927488142e-06,
+      "loss": 0.2066,
+      "step": 2363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.35847259995961,
+      "learning_rate": 9.128107967798716e-06,
+      "loss": 0.1938,
+      "step": 2364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5139641439215072,
+      "learning_rate": 9.127276302109751e-06,
+      "loss": 0.1883,
+      "step": 2365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5282524066097487,
+      "learning_rate": 9.126444277886775e-06,
+      "loss": 0.2252,
+      "step": 2366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2701435754023134,
+      "learning_rate": 9.125611895202062e-06,
+      "loss": 0.1505,
+      "step": 2367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.31553176464601,
+      "learning_rate": 9.124779154127925e-06,
+      "loss": 0.1861,
+      "step": 2368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5582724634325844,
+      "learning_rate": 9.123946054736699e-06,
+      "loss": 0.192,
+      "step": 2369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4666701265703606,
+      "learning_rate": 9.123112597100759e-06,
+      "loss": 0.1873,
+      "step": 2370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4398544974358582,
+      "learning_rate": 9.122278781292502e-06,
+      "loss": 0.1956,
+      "step": 2371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5212396981436584,
+      "learning_rate": 9.121444607384366e-06,
+      "loss": 0.2083,
+      "step": 2372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4531087112740484,
+      "learning_rate": 9.120610075448812e-06,
+      "loss": 0.2159,
+      "step": 2373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.521625924341954,
+      "learning_rate": 9.119775185558337e-06,
+      "loss": 0.205,
+      "step": 2374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9055400242238525,
+      "learning_rate": 9.118939937785468e-06,
+      "loss": 0.2391,
+      "step": 2375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.127400798654293,
+      "learning_rate": 9.11810433220276e-06,
+      "loss": 0.2125,
+      "step": 2376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.712235334496407,
+      "learning_rate": 9.117268368882804e-06,
+      "loss": 0.2485,
+      "step": 2377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.244793115850796,
+      "learning_rate": 9.116432047898218e-06,
+      "loss": 0.1939,
+      "step": 2378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.547615944695285,
+      "learning_rate": 9.115595369321653e-06,
+      "loss": 0.2319,
+      "step": 2379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5217760753167404,
+      "learning_rate": 9.11475833322579e-06,
+      "loss": 0.1851,
+      "step": 2380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.662425007047709,
+      "learning_rate": 9.113920939683343e-06,
+      "loss": 0.2018,
+      "step": 2381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.450000356913756,
+      "learning_rate": 9.113083188767057e-06,
+      "loss": 0.2029,
+      "step": 2382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.707311504512976,
+      "learning_rate": 9.112245080549705e-06,
+      "loss": 0.2436,
+      "step": 2383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.327568324823103,
+      "learning_rate": 9.111406615104093e-06,
+      "loss": 0.1746,
+      "step": 2384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6303578016025506,
+      "learning_rate": 9.11056779250306e-06,
+      "loss": 0.2035,
+      "step": 2385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.473639109241671,
+      "learning_rate": 9.10972861281947e-06,
+      "loss": 0.1922,
+      "step": 2386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3177880383312552,
+      "learning_rate": 9.108889076126226e-06,
+      "loss": 0.1689,
+      "step": 2387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.357540115084761,
+      "learning_rate": 9.108049182496258e-06,
+      "loss": 0.1958,
+      "step": 2388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4387198546760693,
+      "learning_rate": 9.107208932002524e-06,
+      "loss": 0.2026,
+      "step": 2389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4623057030615767,
+      "learning_rate": 9.106368324718018e-06,
+      "loss": 0.1777,
+      "step": 2390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4349492491255837,
+      "learning_rate": 9.105527360715762e-06,
+      "loss": 0.2273,
+      "step": 2391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5841649533417765,
+      "learning_rate": 9.104686040068813e-06,
+      "loss": 0.225,
+      "step": 2392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.473556789946725,
+      "learning_rate": 9.103844362850252e-06,
+      "loss": 0.1945,
+      "step": 2393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2122557317828595,
+      "learning_rate": 9.103002329133198e-06,
+      "loss": 0.1831,
+      "step": 2394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2404121280900147,
+      "learning_rate": 9.102159938990795e-06,
+      "loss": 0.1823,
+      "step": 2395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4931415163162076,
+      "learning_rate": 9.101317192496223e-06,
+      "loss": 0.1873,
+      "step": 2396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5390114877106584,
+      "learning_rate": 9.100474089722693e-06,
+      "loss": 0.2256,
+      "step": 2397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.33274111323077,
+      "learning_rate": 9.09963063074344e-06,
+      "loss": 0.1917,
+      "step": 2398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.533292050495369,
+      "learning_rate": 9.09878681563174e-06,
+      "loss": 0.2359,
+      "step": 2399
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5700281243605305,
+      "learning_rate": 9.097942644460889e-06,
+      "loss": 0.2128,
+      "step": 2400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4012519604445712,
+      "learning_rate": 9.097098117304223e-06,
+      "loss": 0.1974,
+      "step": 2401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.344381550523882,
+      "learning_rate": 9.096253234235106e-06,
+      "loss": 0.1738,
+      "step": 2402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.489285497686882,
+      "learning_rate": 9.095407995326932e-06,
+      "loss": 0.1938,
+      "step": 2403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.095489186777263,
+      "learning_rate": 9.094562400653127e-06,
+      "loss": 0.1659,
+      "step": 2404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.556714026206172,
+      "learning_rate": 9.093716450287144e-06,
+      "loss": 0.1799,
+      "step": 2405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3237891630283176,
+      "learning_rate": 9.092870144302473e-06,
+      "loss": 0.1819,
+      "step": 2406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5541341537198963,
+      "learning_rate": 9.092023482772632e-06,
+      "loss": 0.2073,
+      "step": 2407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3328547608535075,
+      "learning_rate": 9.09117646577117e-06,
+      "loss": 0.1799,
+      "step": 2408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.859401939505249,
+      "learning_rate": 9.090329093371667e-06,
+      "loss": 0.2167,
+      "step": 2409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.818978506251009,
+      "learning_rate": 9.089481365647731e-06,
+      "loss": 0.2378,
+      "step": 2410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3793785038393134,
+      "learning_rate": 9.088633282673007e-06,
+      "loss": 0.2031,
+      "step": 2411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.551510400384724,
+      "learning_rate": 9.087784844521165e-06,
+      "loss": 0.1998,
+      "step": 2412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.575366290435541,
+      "learning_rate": 9.086936051265911e-06,
+      "loss": 0.1912,
+      "step": 2413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.388430625458781,
+      "learning_rate": 9.086086902980977e-06,
+      "loss": 0.1829,
+      "step": 2414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3011581260158582,
+      "learning_rate": 9.08523739974013e-06,
+      "loss": 0.1944,
+      "step": 2415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.688993983604584,
+      "learning_rate": 9.084387541617163e-06,
+      "loss": 0.2373,
+      "step": 2416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.280031351020771,
+      "learning_rate": 9.083537328685905e-06,
+      "loss": 0.186,
+      "step": 2417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.278659189003653,
+      "learning_rate": 9.082686761020213e-06,
+      "loss": 0.1908,
+      "step": 2418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.358950756818462,
+      "learning_rate": 9.081835838693975e-06,
+      "loss": 0.1911,
+      "step": 2419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.127556016921645,
+      "learning_rate": 9.08098456178111e-06,
+      "loss": 0.1829,
+      "step": 2420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5318188618630315,
+      "learning_rate": 9.080132930355567e-06,
+      "loss": 0.2198,
+      "step": 2421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.79596352908327,
+      "learning_rate": 9.079280944491328e-06,
+      "loss": 0.2098,
+      "step": 2422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3481607225525325,
+      "learning_rate": 9.078428604262404e-06,
+      "loss": 0.1626,
+      "step": 2423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1612195744560623,
+      "learning_rate": 9.07757590974284e-06,
+      "loss": 0.1624,
+      "step": 2424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5113771269556318,
+      "learning_rate": 9.076722861006703e-06,
+      "loss": 0.2046,
+      "step": 2425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.599015448839853,
+      "learning_rate": 9.075869458128104e-06,
+      "loss": 0.2052,
+      "step": 2426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.61695649889905,
+      "learning_rate": 9.075015701181171e-06,
+      "loss": 0.2105,
+      "step": 2427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1843430611268624,
+      "learning_rate": 9.074161590240073e-06,
+      "loss": 0.2002,
+      "step": 2428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6209019914824756,
+      "learning_rate": 9.073307125379007e-06,
+      "loss": 0.2016,
+      "step": 2429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7647355309744546,
+      "learning_rate": 9.072452306672197e-06,
+      "loss": 0.188,
+      "step": 2430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.377126561331576,
+      "learning_rate": 9.071597134193902e-06,
+      "loss": 0.2037,
+      "step": 2431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4947093898491084,
+      "learning_rate": 9.070741608018412e-06,
+      "loss": 0.2051,
+      "step": 2432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2788322416194218,
+      "learning_rate": 9.06988572822004e-06,
+      "loss": 0.1962,
+      "step": 2433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0697105314086937,
+      "learning_rate": 9.069029494873143e-06,
+      "loss": 0.1682,
+      "step": 2434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7096674897716584,
+      "learning_rate": 9.0681729080521e-06,
+      "loss": 0.1893,
+      "step": 2435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6083361680593873,
+      "learning_rate": 9.067315967831318e-06,
+      "loss": 0.2095,
+      "step": 2436
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4684044310368867,
+      "learning_rate": 9.066458674285244e-06,
+      "loss": 0.1899,
+      "step": 2437
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3594764790592024,
+      "learning_rate": 9.065601027488345e-06,
+      "loss": 0.1698,
+      "step": 2438
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6024058463927062,
+      "learning_rate": 9.064743027515127e-06,
+      "loss": 0.2011,
+      "step": 2439
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.482605069921783,
+      "learning_rate": 9.06388467444013e-06,
+      "loss": 0.1787,
+      "step": 2440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0751819433097958,
+      "learning_rate": 9.063025968337909e-06,
+      "loss": 0.2365,
+      "step": 2441
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4685563561276096,
+      "learning_rate": 9.062166909283062e-06,
+      "loss": 0.2171,
+      "step": 2442
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5874773352701843,
+      "learning_rate": 9.061307497350218e-06,
+      "loss": 0.2034,
+      "step": 2443
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4994176068440694,
+      "learning_rate": 9.060447732614032e-06,
+      "loss": 0.1794,
+      "step": 2444
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.747185862872241,
+      "learning_rate": 9.05958761514919e-06,
+      "loss": 0.1889,
+      "step": 2445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.312992293654448,
+      "learning_rate": 9.058727145030412e-06,
+      "loss": 0.1939,
+      "step": 2446
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.284478786525657,
+      "learning_rate": 9.057866322332444e-06,
+      "loss": 0.1795,
+      "step": 2447
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.505819551045151,
+      "learning_rate": 9.057005147130069e-06,
+      "loss": 0.2145,
+      "step": 2448
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.644106449204228,
+      "learning_rate": 9.056143619498092e-06,
+      "loss": 0.2226,
+      "step": 2449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5437561761628147,
+      "learning_rate": 9.055281739511357e-06,
+      "loss": 0.2118,
+      "step": 2450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3994728772589786,
+      "learning_rate": 9.054419507244733e-06,
+      "loss": 0.2066,
+      "step": 2451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6583797683100525,
+      "learning_rate": 9.053556922773123e-06,
+      "loss": 0.243,
+      "step": 2452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.24078021872897,
+      "learning_rate": 9.052693986171458e-06,
+      "loss": 0.2034,
+      "step": 2453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8615102174164577,
+      "learning_rate": 9.0518306975147e-06,
+      "loss": 0.2117,
+      "step": 2454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6016578727282496,
+      "learning_rate": 9.050967056877846e-06,
+      "loss": 0.184,
+      "step": 2455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6024994900214606,
+      "learning_rate": 9.050103064335918e-06,
+      "loss": 0.2334,
+      "step": 2456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.565712321449078,
+      "learning_rate": 9.049238719963968e-06,
+      "loss": 0.2069,
+      "step": 2457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3959642164561785,
+      "learning_rate": 9.048374023837086e-06,
+      "loss": 0.1892,
+      "step": 2458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4830850896312624,
+      "learning_rate": 9.047508976030383e-06,
+      "loss": 0.1815,
+      "step": 2459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3611934810103556,
+      "learning_rate": 9.046643576619007e-06,
+      "loss": 0.2017,
+      "step": 2460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.419255193019906,
+      "learning_rate": 9.045777825678135e-06,
+      "loss": 0.1934,
+      "step": 2461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.759285035517681,
+      "learning_rate": 9.044911723282974e-06,
+      "loss": 0.2286,
+      "step": 2462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2771914060996736,
+      "learning_rate": 9.044045269508762e-06,
+      "loss": 0.1905,
+      "step": 2463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1274783459043265,
+      "learning_rate": 9.043178464430767e-06,
+      "loss": 0.1773,
+      "step": 2464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.045621177965386,
+      "learning_rate": 9.042311308124287e-06,
+      "loss": 0.1873,
+      "step": 2465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.319864296319806,
+      "learning_rate": 9.041443800664653e-06,
+      "loss": 0.1906,
+      "step": 2466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4311581086363736,
+      "learning_rate": 9.040575942127225e-06,
+      "loss": 0.2106,
+      "step": 2467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.549959377963562,
+      "learning_rate": 9.039707732587393e-06,
+      "loss": 0.2248,
+      "step": 2468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.440690134697464,
+      "learning_rate": 9.038839172120575e-06,
+      "loss": 0.2064,
+      "step": 2469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.524740913080514,
+      "learning_rate": 9.037970260802227e-06,
+      "loss": 0.2381,
+      "step": 2470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5354858092078034,
+      "learning_rate": 9.037100998707829e-06,
+      "loss": 0.2091,
+      "step": 2471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9763266121004284,
+      "learning_rate": 9.03623138591289e-06,
+      "loss": 0.2129,
+      "step": 2472
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.565062421558977,
+      "learning_rate": 9.035361422492956e-06,
+      "loss": 0.2157,
+      "step": 2473
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3279172146755873,
+      "learning_rate": 9.034491108523603e-06,
+      "loss": 0.2095,
+      "step": 2474
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7564536824755943,
+      "learning_rate": 9.033620444080427e-06,
+      "loss": 0.2123,
+      "step": 2475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.192012728465349,
+      "learning_rate": 9.032749429239069e-06,
+      "loss": 0.1692,
+      "step": 2476
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1868698557802526,
+      "learning_rate": 9.03187806407519e-06,
+      "loss": 0.1572,
+      "step": 2477
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5350585643309613,
+      "learning_rate": 9.031006348664487e-06,
+      "loss": 0.1739,
+      "step": 2478
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3235306470499935,
+      "learning_rate": 9.030134283082683e-06,
+      "loss": 0.2116,
+      "step": 2479
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3368059019653877,
+      "learning_rate": 9.029261867405536e-06,
+      "loss": 0.1881,
+      "step": 2480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.983056906125631,
+      "learning_rate": 9.028389101708833e-06,
+      "loss": 0.1533,
+      "step": 2481
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.510747102752466,
+      "learning_rate": 9.027515986068387e-06,
+      "loss": 0.1907,
+      "step": 2482
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3864579964373323,
+      "learning_rate": 9.026642520560047e-06,
+      "loss": 0.229,
+      "step": 2483
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.472303629375403,
+      "learning_rate": 9.025768705259691e-06,
+      "loss": 0.1719,
+      "step": 2484
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.514599661066843,
+      "learning_rate": 9.024894540243227e-06,
+      "loss": 0.1832,
+      "step": 2485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6526308781800343,
+      "learning_rate": 9.024020025586592e-06,
+      "loss": 0.1982,
+      "step": 2486
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.529140028023631,
+      "learning_rate": 9.023145161365755e-06,
+      "loss": 0.1875,
+      "step": 2487
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.219666914906933,
+      "learning_rate": 9.022269947656714e-06,
+      "loss": 0.172,
+      "step": 2488
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2115965038405214,
+      "learning_rate": 9.0213943845355e-06,
+      "loss": 0.1639,
+      "step": 2489
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5239739531750796,
+      "learning_rate": 9.020518472078172e-06,
+      "loss": 0.1766,
+      "step": 2490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7168355638313866,
+      "learning_rate": 9.019642210360821e-06,
+      "loss": 0.2226,
+      "step": 2491
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6850842491003104,
+      "learning_rate": 9.018765599459564e-06,
+      "loss": 0.1593,
+      "step": 2492
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.157432193569933,
+      "learning_rate": 9.017888639450557e-06,
+      "loss": 0.1671,
+      "step": 2493
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1735623248198594,
+      "learning_rate": 9.017011330409975e-06,
+      "loss": 0.1777,
+      "step": 2494
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.531257473174138,
+      "learning_rate": 9.016133672414034e-06,
+      "loss": 0.2359,
+      "step": 2495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6663386506280053,
+      "learning_rate": 9.015255665538972e-06,
+      "loss": 0.2259,
+      "step": 2496
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46173077673191,
+      "learning_rate": 9.014377309861064e-06,
+      "loss": 0.2247,
+      "step": 2497
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5898726981211895,
+      "learning_rate": 9.01349860545661e-06,
+      "loss": 0.1998,
+      "step": 2498
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.55733901522693,
+      "learning_rate": 9.012619552401945e-06,
+      "loss": 0.1852,
+      "step": 2499
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5922058456675705,
+      "learning_rate": 9.01174015077343e-06,
+      "loss": 0.2169,
+      "step": 2500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.432134587314923,
+      "learning_rate": 9.010860400647457e-06,
+      "loss": 0.179,
+      "step": 2501
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7066590187228057,
+      "learning_rate": 9.009980302100452e-06,
+      "loss": 0.2221,
+      "step": 2502
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1703036328767773,
+      "learning_rate": 9.009099855208867e-06,
+      "loss": 0.1749,
+      "step": 2503
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3696854242886443,
+      "learning_rate": 9.008219060049188e-06,
+      "loss": 0.2019,
+      "step": 2504
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6393038590768536,
+      "learning_rate": 9.007337916697925e-06,
+      "loss": 0.1896,
+      "step": 2505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2458042418158373,
+      "learning_rate": 9.006456425231624e-06,
+      "loss": 0.171,
+      "step": 2506
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4766936086919618,
+      "learning_rate": 9.005574585726864e-06,
+      "loss": 0.1838,
+      "step": 2507
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4339046291804265,
+      "learning_rate": 9.004692398260243e-06,
+      "loss": 0.2033,
+      "step": 2508
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1070770820711737,
+      "learning_rate": 9.003809862908401e-06,
+      "loss": 0.165,
+      "step": 2509
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3495901576145233,
+      "learning_rate": 9.002926979748003e-06,
+      "loss": 0.1534,
+      "step": 2510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.679900667081155,
+      "learning_rate": 9.002043748855742e-06,
+      "loss": 0.1875,
+      "step": 2511
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.278490696235511,
+      "learning_rate": 9.001160170308346e-06,
+      "loss": 0.1882,
+      "step": 2512
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2114416143034665,
+      "learning_rate": 9.000276244182567e-06,
+      "loss": 0.1469,
+      "step": 2513
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.14285991408518,
+      "learning_rate": 8.999391970555197e-06,
+      "loss": 0.1727,
+      "step": 2514
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.516106427016791,
+      "learning_rate": 8.998507349503048e-06,
+      "loss": 0.1956,
+      "step": 2515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3448306359367037,
+      "learning_rate": 8.997622381102968e-06,
+      "loss": 0.1837,
+      "step": 2516
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3617013676775653,
+      "learning_rate": 8.996737065431834e-06,
+      "loss": 0.1828,
+      "step": 2517
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.760758996202912,
+      "learning_rate": 8.995851402566553e-06,
+      "loss": 0.2136,
+      "step": 2518
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6919894177214885,
+      "learning_rate": 8.99496539258406e-06,
+      "loss": 0.2138,
+      "step": 2519
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.578201382299165,
+      "learning_rate": 8.994079035561325e-06,
+      "loss": 0.1909,
+      "step": 2520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.384298749070819,
+      "learning_rate": 8.993192331575342e-06,
+      "loss": 0.212,
+      "step": 2521
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.567503057419424,
+      "learning_rate": 8.992305280703141e-06,
+      "loss": 0.195,
+      "step": 2522
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1983066858698934,
+      "learning_rate": 8.99141788302178e-06,
+      "loss": 0.1862,
+      "step": 2523
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.648225708757083,
+      "learning_rate": 8.990530138608346e-06,
+      "loss": 0.1958,
+      "step": 2524
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.75944415053918,
+      "learning_rate": 8.989642047539956e-06,
+      "loss": 0.2242,
+      "step": 2525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.270021089247332,
+      "learning_rate": 8.988753609893757e-06,
+      "loss": 0.1976,
+      "step": 2526
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.220451854767584,
+      "learning_rate": 8.987864825746929e-06,
+      "loss": 0.175,
+      "step": 2527
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1935898661889714,
+      "learning_rate": 8.986975695176678e-06,
+      "loss": 0.1887,
+      "step": 2528
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3983034854133365,
+      "learning_rate": 8.986086218260247e-06,
+      "loss": 0.1734,
+      "step": 2529
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5217345647835656,
+      "learning_rate": 8.985196395074899e-06,
+      "loss": 0.2166,
+      "step": 2530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3942090099163704,
+      "learning_rate": 8.984306225697935e-06,
+      "loss": 0.1987,
+      "step": 2531
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.357183499918745,
+      "learning_rate": 8.983415710206683e-06,
+      "loss": 0.1847,
+      "step": 2532
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4056164139368894,
+      "learning_rate": 8.982524848678502e-06,
+      "loss": 0.1946,
+      "step": 2533
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.273062218632727,
+      "learning_rate": 8.981633641190779e-06,
+      "loss": 0.1757,
+      "step": 2534
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3669871095189774,
+      "learning_rate": 8.980742087820935e-06,
+      "loss": 0.164,
+      "step": 2535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7304213900609895,
+      "learning_rate": 8.979850188646418e-06,
+      "loss": 0.2254,
+      "step": 2536
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.669376962879526,
+      "learning_rate": 8.978957943744703e-06,
+      "loss": 0.168,
+      "step": 2537
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.586408129218207,
+      "learning_rate": 8.978065353193305e-06,
+      "loss": 0.2103,
+      "step": 2538
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9451901423347504,
+      "learning_rate": 8.97717241706976e-06,
+      "loss": 0.1771,
+      "step": 2539
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.367741511715015,
+      "learning_rate": 8.976279135451636e-06,
+      "loss": 0.2065,
+      "step": 2540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.539884020529156,
+      "learning_rate": 8.975385508416532e-06,
+      "loss": 0.2096,
+      "step": 2541
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.298470116584281,
+      "learning_rate": 8.974491536042079e-06,
+      "loss": 0.1823,
+      "step": 2542
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5577680136098957,
+      "learning_rate": 8.973597218405931e-06,
+      "loss": 0.2171,
+      "step": 2543
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3682258359476225,
+      "learning_rate": 8.972702555585783e-06,
+      "loss": 0.1696,
+      "step": 2544
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.45072881335195,
+      "learning_rate": 8.971807547659349e-06,
+      "loss": 0.2182,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.225956724766786,
+      "learning_rate": 8.970912194704379e-06,
+      "loss": 0.1949,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2924213374139786,
+      "learning_rate": 8.970016496798655e-06,
+      "loss": 0.1555,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3598353685315243,
+      "learning_rate": 8.969120454019983e-06,
+      "loss": 0.1968,
+      "step": 2548
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140298079840243,
+      "learning_rate": 8.9682240664462e-06,
+      "loss": 0.1939,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5966158436583457,
+      "learning_rate": 8.967327334155179e-06,
+      "loss": 0.1976,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.110009483547678,
+      "learning_rate": 8.966430257224814e-06,
+      "loss": 0.1868,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.469090312716522,
+      "learning_rate": 8.965532835733035e-06,
+      "loss": 0.2249,
+      "step": 2552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.370113641352872,
+      "learning_rate": 8.964635069757803e-06,
+      "loss": 0.2061,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.546990219077739,
+      "learning_rate": 8.963736959377103e-06,
+      "loss": 0.2428,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3509801057604114,
+      "learning_rate": 8.962838504668956e-06,
+      "loss": 0.1919,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5454718768347426,
+      "learning_rate": 8.961939705711407e-06,
+      "loss": 0.1909,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3623847434195118,
+      "learning_rate": 8.96104056258254e-06,
+      "loss": 0.2182,
+      "step": 2557
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4994084402990775,
+      "learning_rate": 8.960141075360455e-06,
+      "loss": 0.2169,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7713601720157772,
+      "learning_rate": 8.959241244123296e-06,
+      "loss": 0.2162,
+      "step": 2559
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.510208846466004,
+      "learning_rate": 8.95834106894923e-06,
+      "loss": 0.1802,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.683542890183487,
+      "learning_rate": 8.95744054991645e-06,
+      "loss": 0.22,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2349075239697282,
+      "learning_rate": 8.95653968710319e-06,
+      "loss": 0.1904,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4025046067993463,
+      "learning_rate": 8.955638480587705e-06,
+      "loss": 0.1888,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3291027823376704,
+      "learning_rate": 8.95473693044828e-06,
+      "loss": 0.1932,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3294861762512635,
+      "learning_rate": 8.953835036763234e-06,
+      "loss": 0.1829,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2631303033466508,
+      "learning_rate": 8.952932799610916e-06,
+      "loss": 0.1981,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6347584544532427,
+      "learning_rate": 8.952030219069699e-06,
+      "loss": 0.1978,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3729548502433104,
+      "learning_rate": 8.951127295217991e-06,
+      "loss": 0.1992,
+      "step": 2568
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3796328147454724,
+      "learning_rate": 8.950224028134228e-06,
+      "loss": 0.1749,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2032308613495553,
+      "learning_rate": 8.949320417896878e-06,
+      "loss": 0.1623,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.349496398032231,
+      "learning_rate": 8.948416464584437e-06,
+      "loss": 0.1738,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4801402860656125,
+      "learning_rate": 8.94751216827543e-06,
+      "loss": 0.2164,
+      "step": 2572
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.616846553003231,
+      "learning_rate": 8.946607529048413e-06,
+      "loss": 0.2054,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5664926276413067,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 0.1987,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.595175724344372,
+      "learning_rate": 8.944797222154717e-06,
+      "loss": 0.1782,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4764064047660423,
+      "learning_rate": 8.943891554645298e-06,
+      "loss": 0.2036,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.50359201275801,
+      "learning_rate": 8.942985544532392e-06,
+      "loss": 0.1792,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2853237458671276,
+      "learning_rate": 8.942079191894699e-06,
+      "loss": 0.1869,
+      "step": 2578
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4175671303708506,
+      "learning_rate": 8.941172496810956e-06,
+      "loss": 0.2111,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4312945112777866,
+      "learning_rate": 8.940265459359927e-06,
+      "loss": 0.2159,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3866714982493313,
+      "learning_rate": 8.939358079620404e-06,
+      "loss": 0.2222,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.507593823712348,
+      "learning_rate": 8.938450357671211e-06,
+      "loss": 0.1952,
+      "step": 2582
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4490264204042433,
+      "learning_rate": 8.937542293591201e-06,
+      "loss": 0.2059,
+      "step": 2583
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5427577795874403,
+      "learning_rate": 8.936633887459259e-06,
+      "loss": 0.2066,
+      "step": 2584
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0981010453887894,
+      "learning_rate": 8.935725139354296e-06,
+      "loss": 0.1692,
+      "step": 2585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4256197886371558,
+      "learning_rate": 8.934816049355255e-06,
+      "loss": 0.1774,
+      "step": 2586
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.561659906225474,
+      "learning_rate": 8.933906617541107e-06,
+      "loss": 0.1897,
+      "step": 2587
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.455147206768468,
+      "learning_rate": 8.932996843990855e-06,
+      "loss": 0.2136,
+      "step": 2588
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.671623848703675,
+      "learning_rate": 8.932086728783531e-06,
+      "loss": 0.2089,
+      "step": 2589
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4538046001696907,
+      "learning_rate": 8.931176271998195e-06,
+      "loss": 0.181,
+      "step": 2590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1588708352543704,
+      "learning_rate": 8.930265473713939e-06,
+      "loss": 0.1803,
+      "step": 2591
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4808194384474196,
+      "learning_rate": 8.92935433400988e-06,
+      "loss": 0.2281,
+      "step": 2592
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.342660061878623,
+      "learning_rate": 8.928442852965174e-06,
+      "loss": 0.1996,
+      "step": 2593
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3213882097449123,
+      "learning_rate": 8.927531030658996e-06,
+      "loss": 0.1862,
+      "step": 2594
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.279203330837764,
+      "learning_rate": 8.926618867170555e-06,
+      "loss": 0.2044,
+      "step": 2595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6501150126803514,
+      "learning_rate": 8.925706362579097e-06,
+      "loss": 0.2396,
+      "step": 2596
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3008470660473077,
+      "learning_rate": 8.924793516963881e-06,
+      "loss": 0.1902,
+      "step": 2597
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.525054516359642,
+      "learning_rate": 8.923880330404213e-06,
+      "loss": 0.1973,
+      "step": 2598
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.985375270909967,
+      "learning_rate": 8.922966802979419e-06,
+      "loss": 0.1781,
+      "step": 2599
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3014604694785024,
+      "learning_rate": 8.922052934768853e-06,
+      "loss": 0.1848,
+      "step": 2600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3950569056470057,
+      "learning_rate": 8.921138725851905e-06,
+      "loss": 0.1597,
+      "step": 2601
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3100562560786737,
+      "learning_rate": 8.920224176307994e-06,
+      "loss": 0.1987,
+      "step": 2602
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6765741555939098,
+      "learning_rate": 8.919309286216564e-06,
+      "loss": 0.2271,
+      "step": 2603
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4074240365292305,
+      "learning_rate": 8.918394055657091e-06,
+      "loss": 0.2097,
+      "step": 2604
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8491943599881737,
+      "learning_rate": 8.917478484709078e-06,
+      "loss": 0.2319,
+      "step": 2605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.074805592357438,
+      "learning_rate": 8.916562573452066e-06,
+      "loss": 0.201,
+      "step": 2606
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.349371380185813,
+      "learning_rate": 8.915646321965615e-06,
+      "loss": 0.2053,
+      "step": 2607
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.244421993615951,
+      "learning_rate": 8.914729730329321e-06,
+      "loss": 0.1849,
+      "step": 2608
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.707783615326083,
+      "learning_rate": 8.913812798622806e-06,
+      "loss": 0.2418,
+      "step": 2609
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4143826250045017,
+      "learning_rate": 8.912895526925726e-06,
+      "loss": 0.2264,
+      "step": 2610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.75017070650157,
+      "learning_rate": 8.911977915317763e-06,
+      "loss": 0.212,
+      "step": 2611
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4221582789612217,
+      "learning_rate": 8.911059963878628e-06,
+      "loss": 0.1865,
+      "step": 2612
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.465579895374271,
+      "learning_rate": 8.910141672688063e-06,
+      "loss": 0.2352,
+      "step": 2613
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2840294665344527,
+      "learning_rate": 8.90922304182584e-06,
+      "loss": 0.1946,
+      "step": 2614
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.509777791455074,
+      "learning_rate": 8.90830407137176e-06,
+      "loss": 0.2016,
+      "step": 2615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.304029709912657,
+      "learning_rate": 8.907384761405655e-06,
+      "loss": 0.1898,
+      "step": 2616
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3494404301626983,
+      "learning_rate": 8.906465112007383e-06,
+      "loss": 0.1888,
+      "step": 2617
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2915500246689295,
+      "learning_rate": 8.905545123256834e-06,
+      "loss": 0.1903,
+      "step": 2618
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4862369114630454,
+      "learning_rate": 8.904624795233926e-06,
+      "loss": 0.1719,
+      "step": 2619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4562799641917175,
+      "learning_rate": 8.903704128018608e-06,
+      "loss": 0.1868,
+      "step": 2620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.396967219162005,
+      "learning_rate": 8.90278312169086e-06,
+      "loss": 0.2036,
+      "step": 2621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5542097177272853,
+      "learning_rate": 8.901861776330682e-06,
+      "loss": 0.1857,
+      "step": 2622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.640828208504622,
+      "learning_rate": 8.90094009201812e-06,
+      "loss": 0.2163,
+      "step": 2623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9035881114351754,
+      "learning_rate": 8.900018068833233e-06,
+      "loss": 0.2161,
+      "step": 2624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1263250135172362,
+      "learning_rate": 8.899095706856122e-06,
+      "loss": 0.172,
+      "step": 2625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4189926210944415,
+      "learning_rate": 8.89817300616691e-06,
+      "loss": 0.2049,
+      "step": 2626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5532193452171397,
+      "learning_rate": 8.897249966845748e-06,
+      "loss": 0.1889,
+      "step": 2627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.272106966068934,
+      "learning_rate": 8.896326588972826e-06,
+      "loss": 0.1815,
+      "step": 2628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3596544155483374,
+      "learning_rate": 8.895402872628352e-06,
+      "loss": 0.1928,
+      "step": 2629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4576055375391777,
+      "learning_rate": 8.894478817892574e-06,
+      "loss": 0.1976,
+      "step": 2630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3334390278615667,
+      "learning_rate": 8.893554424845758e-06,
+      "loss": 0.1906,
+      "step": 2631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3957573839948045,
+      "learning_rate": 8.892629693568209e-06,
+      "loss": 0.2194,
+      "step": 2632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.593040381054902,
+      "learning_rate": 8.891704624140257e-06,
+      "loss": 0.2009,
+      "step": 2633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.366141529724085,
+      "learning_rate": 8.890779216642263e-06,
+      "loss": 0.1875,
+      "step": 2634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0176872815599713,
+      "learning_rate": 8.889853471154615e-06,
+      "loss": 0.1905,
+      "step": 2635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2938314093045014,
+      "learning_rate": 8.888927387757735e-06,
+      "loss": 0.1832,
+      "step": 2636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5337541768176055,
+      "learning_rate": 8.88800096653207e-06,
+      "loss": 0.1771,
+      "step": 2637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5706883467893515,
+      "learning_rate": 8.887074207558092e-06,
+      "loss": 0.1959,
+      "step": 2638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.613025290018213,
+      "learning_rate": 8.886147110916316e-06,
+      "loss": 0.2099,
+      "step": 2639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5296313291740886,
+      "learning_rate": 8.885219676687277e-06,
+      "loss": 0.1964,
+      "step": 2640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6274165654596917,
+      "learning_rate": 8.884291904951538e-06,
+      "loss": 0.1955,
+      "step": 2641
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.322922156189818,
+      "learning_rate": 8.883363795789694e-06,
+      "loss": 0.1902,
+      "step": 2642
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7197900251218736,
+      "learning_rate": 8.882435349282371e-06,
+      "loss": 0.1966,
+      "step": 2643
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5146984303852613,
+      "learning_rate": 8.88150656551022e-06,
+      "loss": 0.2029,
+      "step": 2644
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.471498318081182,
+      "learning_rate": 8.880577444553929e-06,
+      "loss": 0.1632,
+      "step": 2645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.382739921323251,
+      "learning_rate": 8.879647986494205e-06,
+      "loss": 0.1877,
+      "step": 2646
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.353491698934403,
+      "learning_rate": 8.878718191411792e-06,
+      "loss": 0.1942,
+      "step": 2647
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0638506623885307,
+      "learning_rate": 8.87778805938746e-06,
+      "loss": 0.1796,
+      "step": 2648
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.398036634202962,
+      "learning_rate": 8.876857590502008e-06,
+      "loss": 0.2193,
+      "step": 2649
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.016808796523212,
+      "learning_rate": 8.875926784836267e-06,
+      "loss": 0.1839,
+      "step": 2650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.372394166875836,
+      "learning_rate": 8.874995642471094e-06,
+      "loss": 0.2073,
+      "step": 2651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.64394057706979,
+      "learning_rate": 8.87406416348738e-06,
+      "loss": 0.2165,
+      "step": 2652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2874222473065564,
+      "learning_rate": 8.873132347966038e-06,
+      "loss": 0.169,
+      "step": 2653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5881586221816844,
+      "learning_rate": 8.872200195988016e-06,
+      "loss": 0.2303,
+      "step": 2654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.360985619004908,
+      "learning_rate": 8.87126770763429e-06,
+      "loss": 0.2072,
+      "step": 2655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.379383349603289,
+      "learning_rate": 8.870334882985866e-06,
+      "loss": 0.2116,
+      "step": 2656
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3236815237475192,
+      "learning_rate": 8.869401722123771e-06,
+      "loss": 0.1735,
+      "step": 2657
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.381407411566154,
+      "learning_rate": 8.868468225129078e-06,
+      "loss": 0.1991,
+      "step": 2658
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5332292689228026,
+      "learning_rate": 8.867534392082873e-06,
+      "loss": 0.2303,
+      "step": 2659
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0766636850425058,
+      "learning_rate": 8.866600223066277e-06,
+      "loss": 0.212,
+      "step": 2660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.230793299068313,
+      "learning_rate": 8.865665718160445e-06,
+      "loss": 0.1952,
+      "step": 2661
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28309000536618,
+      "learning_rate": 8.864730877446555e-06,
+      "loss": 0.1981,
+      "step": 2662
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.275711567135502,
+      "learning_rate": 8.863795701005813e-06,
+      "loss": 0.1946,
+      "step": 2663
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28430198968044,
+      "learning_rate": 8.862860188919462e-06,
+      "loss": 0.2033,
+      "step": 2664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.368544603110747,
+      "learning_rate": 8.861924341268768e-06,
+      "loss": 0.2011,
+      "step": 2665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4294646823414103,
+      "learning_rate": 8.860988158135025e-06,
+      "loss": 0.1739,
+      "step": 2666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.221545826217993,
+      "learning_rate": 8.86005163959956e-06,
+      "loss": 0.168,
+      "step": 2667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1390555800485322,
+      "learning_rate": 8.85911478574373e-06,
+      "loss": 0.1711,
+      "step": 2668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.772727096983433,
+      "learning_rate": 8.858177596648915e-06,
+      "loss": 0.26,
+      "step": 2669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.558961118855876,
+      "learning_rate": 8.857240072396533e-06,
+      "loss": 0.1786,
+      "step": 2670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.386775291108183,
+      "learning_rate": 8.856302213068022e-06,
+      "loss": 0.2214,
+      "step": 2671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3310011330571294,
+      "learning_rate": 8.855364018744854e-06,
+      "loss": 0.1684,
+      "step": 2672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.273939990036212,
+      "learning_rate": 8.85442548950853e-06,
+      "loss": 0.1911,
+      "step": 2673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4815458541950965,
+      "learning_rate": 8.853486625440581e-06,
+      "loss": 0.2182,
+      "step": 2674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0978484759341267,
+      "learning_rate": 8.852547426622563e-06,
+      "loss": 0.1582,
+      "step": 2675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.169755746431044,
+      "learning_rate": 8.851607893136065e-06,
+      "loss": 0.1461,
+      "step": 2676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.444823710297576,
+      "learning_rate": 8.850668025062704e-06,
+      "loss": 0.2153,
+      "step": 2677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.349062825390895,
+      "learning_rate": 8.849727822484125e-06,
+      "loss": 0.1737,
+      "step": 2678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.334075504860098,
+      "learning_rate": 8.848787285482003e-06,
+      "loss": 0.1639,
+      "step": 2679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2785551599374436,
+      "learning_rate": 8.847846414138041e-06,
+      "loss": 0.1872,
+      "step": 2680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.353327556189428,
+      "learning_rate": 8.846905208533974e-06,
+      "loss": 0.2066,
+      "step": 2681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7052969734121,
+      "learning_rate": 8.84596366875156e-06,
+      "loss": 0.2187,
+      "step": 2682
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4316913495022687,
+      "learning_rate": 8.845021794872597e-06,
+      "loss": 0.1951,
+      "step": 2683
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.530526962112897,
+      "learning_rate": 8.844079586978897e-06,
+      "loss": 0.1605,
+      "step": 2684
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.025354715696344,
+      "learning_rate": 8.843137045152314e-06,
+      "loss": 0.1467,
+      "step": 2685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5048048944435073,
+      "learning_rate": 8.842194169474727e-06,
+      "loss": 0.2003,
+      "step": 2686
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.055699239640041,
+      "learning_rate": 8.84125096002804e-06,
+      "loss": 0.1944,
+      "step": 2687
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.508366549060347,
+      "learning_rate": 8.840307416894189e-06,
+      "loss": 0.2225,
+      "step": 2688
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1510090795677193,
+      "learning_rate": 8.83936354015514e-06,
+      "loss": 0.1931,
+      "step": 2689
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.007995593537831,
+      "learning_rate": 8.838419329892887e-06,
+      "loss": 0.2572,
+      "step": 2690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.431869073316424,
+      "learning_rate": 8.837474786189454e-06,
+      "loss": 0.1721,
+      "step": 2691
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.144026062533291,
+      "learning_rate": 8.836529909126891e-06,
+      "loss": 0.1978,
+      "step": 2692
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4982088085582665,
+      "learning_rate": 8.83558469878728e-06,
+      "loss": 0.2134,
+      "step": 2693
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40689261859619,
+      "learning_rate": 8.834639155252732e-06,
+      "loss": 0.2002,
+      "step": 2694
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3185364909248567,
+      "learning_rate": 8.833693278605381e-06,
+      "loss": 0.1995,
+      "step": 2695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3452714908893015,
+      "learning_rate": 8.832747068927404e-06,
+      "loss": 0.1777,
+      "step": 2696
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4296731297586613,
+      "learning_rate": 8.831800526300987e-06,
+      "loss": 0.1996,
+      "step": 2697
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1262934381815453,
+      "learning_rate": 8.830853650808361e-06,
+      "loss": 0.172,
+      "step": 2698
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.387946308830362,
+      "learning_rate": 8.829906442531782e-06,
+      "loss": 0.1819,
+      "step": 2699
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5480912583075277,
+      "learning_rate": 8.828958901553529e-06,
+      "loss": 0.1925,
+      "step": 2700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2675911521147873,
+      "learning_rate": 8.828011027955918e-06,
+      "loss": 0.1949,
+      "step": 2701
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.235832415913824,
+      "learning_rate": 8.82706282182129e-06,
+      "loss": 0.1804,
+      "step": 2702
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.405518129795668,
+      "learning_rate": 8.826114283232012e-06,
+      "loss": 0.1918,
+      "step": 2703
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1489028022114343,
+      "learning_rate": 8.825165412270487e-06,
+      "loss": 0.183,
+      "step": 2704
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4898813225517396,
+      "learning_rate": 8.824216209019139e-06,
+      "loss": 0.2136,
+      "step": 2705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4290380574607413,
+      "learning_rate": 8.823266673560426e-06,
+      "loss": 0.1997,
+      "step": 2706
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0547336409764037,
+      "learning_rate": 8.822316805976836e-06,
+      "loss": 0.1964,
+      "step": 2707
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4995177184981383,
+      "learning_rate": 8.821366606350882e-06,
+      "loss": 0.204,
+      "step": 2708
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.496877932327667,
+      "learning_rate": 8.820416074765106e-06,
+      "loss": 0.1846,
+      "step": 2709
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.811538995486479,
+      "learning_rate": 8.819465211302081e-06,
+      "loss": 0.2538,
+      "step": 2710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3483936457928305,
+      "learning_rate": 8.818514016044405e-06,
+      "loss": 0.2007,
+      "step": 2711
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.459001267028272,
+      "learning_rate": 8.817562489074714e-06,
+      "loss": 0.1892,
+      "step": 2712
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.330527072438074,
+      "learning_rate": 8.816610630475664e-06,
+      "loss": 0.1985,
+      "step": 2713
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.927556743327112,
+      "learning_rate": 8.81565844032994e-06,
+      "loss": 0.1527,
+      "step": 2714
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5477202344241388,
+      "learning_rate": 8.814705918720259e-06,
+      "loss": 0.205,
+      "step": 2715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4638115812631862,
+      "learning_rate": 8.813753065729369e-06,
+      "loss": 0.1914,
+      "step": 2716
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1382782141388263,
+      "learning_rate": 8.812799881440039e-06,
+      "loss": 0.1869,
+      "step": 2717
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2137579397035516,
+      "learning_rate": 8.811846365935076e-06,
+      "loss": 0.1826,
+      "step": 2718
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3893327699204407,
+      "learning_rate": 8.810892519297308e-06,
+      "loss": 0.1719,
+      "step": 2719
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2884005897257405,
+      "learning_rate": 8.809938341609596e-06,
+      "loss": 0.17,
+      "step": 2720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7329048226931616,
+      "learning_rate": 8.808983832954831e-06,
+      "loss": 0.256,
+      "step": 2721
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6693921189089678,
+      "learning_rate": 8.808028993415929e-06,
+      "loss": 0.206,
+      "step": 2722
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4346854883180216,
+      "learning_rate": 8.807073823075835e-06,
+      "loss": 0.1957,
+      "step": 2723
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.129420663344039,
+      "learning_rate": 8.806118322017525e-06,
+      "loss": 0.1711,
+      "step": 2724
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0543657210581103,
+      "learning_rate": 8.805162490324005e-06,
+      "loss": 0.1643,
+      "step": 2725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2182602963716076,
+      "learning_rate": 8.804206328078304e-06,
+      "loss": 0.2095,
+      "step": 2726
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4027833208632905,
+      "learning_rate": 8.803249835363486e-06,
+      "loss": 0.2085,
+      "step": 2727
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.16120960638039,
+      "learning_rate": 8.80229301226264e-06,
+      "loss": 0.1855,
+      "step": 2728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.108758597089169,
+      "learning_rate": 8.801335858858883e-06,
+      "loss": 0.1567,
+      "step": 2729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2403412709481745,
+      "learning_rate": 8.800378375235365e-06,
+      "loss": 0.1895,
+      "step": 2730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.338011618501962,
+      "learning_rate": 8.79942056147526e-06,
+      "loss": 0.1889,
+      "step": 2731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3050329311885926,
+      "learning_rate": 8.798462417661775e-06,
+      "loss": 0.1801,
+      "step": 2732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.309140102025805,
+      "learning_rate": 8.79750394387814e-06,
+      "loss": 0.1691,
+      "step": 2733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3946601334039133,
+      "learning_rate": 8.796545140207622e-06,
+      "loss": 0.1955,
+      "step": 2734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9992360363500568,
+      "learning_rate": 8.795586006733505e-06,
+      "loss": 0.1381,
+      "step": 2735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.309804412797823,
+      "learning_rate": 8.794626543539114e-06,
+      "loss": 0.1834,
+      "step": 2736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.157161474511113,
+      "learning_rate": 8.793666750707795e-06,
+      "loss": 0.1843,
+      "step": 2737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4118610090961554,
+      "learning_rate": 8.792706628322924e-06,
+      "loss": 0.2072,
+      "step": 2738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.153840754720639,
+      "learning_rate": 8.791746176467908e-06,
+      "loss": 0.1753,
+      "step": 2739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2432703281122897,
+      "learning_rate": 8.79078539522618e-06,
+      "loss": 0.2047,
+      "step": 2740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.199319382192434,
+      "learning_rate": 8.789824284681201e-06,
+      "loss": 0.1813,
+      "step": 2741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4394308727285465,
+      "learning_rate": 8.788862844916464e-06,
+      "loss": 0.1931,
+      "step": 2742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3116635056711408,
+      "learning_rate": 8.787901076015487e-06,
+      "loss": 0.1791,
+      "step": 2743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078360072119509,
+      "learning_rate": 8.78693897806182e-06,
+      "loss": 0.1734,
+      "step": 2744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6638077309788724,
+      "learning_rate": 8.78597655113904e-06,
+      "loss": 0.207,
+      "step": 2745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.381023662094938,
+      "learning_rate": 8.78501379533075e-06,
+      "loss": 0.2085,
+      "step": 2746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1840748034821784,
+      "learning_rate": 8.784050710720587e-06,
+      "loss": 0.1615,
+      "step": 2747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2008103336316793,
+      "learning_rate": 8.783087297392212e-06,
+      "loss": 0.1742,
+      "step": 2748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.232934320641409,
+      "learning_rate": 8.782123555429315e-06,
+      "loss": 0.1971,
+      "step": 2749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375914138269291,
+      "learning_rate": 8.78115948491562e-06,
+      "loss": 0.2037,
+      "step": 2750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.345778592442171,
+      "learning_rate": 8.780195085934871e-06,
+      "loss": 0.1819,
+      "step": 2751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7953484726190094,
+      "learning_rate": 8.779230358570845e-06,
+      "loss": 0.2017,
+      "step": 2752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5488411936839976,
+      "learning_rate": 8.77826530290735e-06,
+      "loss": 0.2177,
+      "step": 2753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.45947770505814,
+      "learning_rate": 8.777299919028217e-06,
+      "loss": 0.1952,
+      "step": 2754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6336880197170376,
+      "learning_rate": 8.77633420701731e-06,
+      "loss": 0.2218,
+      "step": 2755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.282826229800347,
+      "learning_rate": 8.775368166958518e-06,
+      "loss": 0.1649,
+      "step": 2756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3111710868660253,
+      "learning_rate": 8.774401798935763e-06,
+      "loss": 0.209,
+      "step": 2757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4752426681147495,
+      "learning_rate": 8.773435103032992e-06,
+      "loss": 0.1833,
+      "step": 2758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2713680512389063,
+      "learning_rate": 8.77246807933418e-06,
+      "loss": 0.1962,
+      "step": 2759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.117424035034632,
+      "learning_rate": 8.771500727923332e-06,
+      "loss": 0.1544,
+      "step": 2760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4703833500846306,
+      "learning_rate": 8.770533048884483e-06,
+      "loss": 0.1948,
+      "step": 2761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.163878655433132,
+      "learning_rate": 8.769565042301692e-06,
+      "loss": 0.1916,
+      "step": 2762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4381167007851654,
+      "learning_rate": 8.768596708259052e-06,
+      "loss": 0.2201,
+      "step": 2763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2743277856091626,
+      "learning_rate": 8.767628046840677e-06,
+      "loss": 0.161,
+      "step": 2764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5351153775040127,
+      "learning_rate": 8.766659058130719e-06,
+      "loss": 0.2141,
+      "step": 2765
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4758388521291677,
+      "learning_rate": 8.765689742213353e-06,
+      "loss": 0.1988,
+      "step": 2766
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5028856392124017,
+      "learning_rate": 8.764720099172781e-06,
+      "loss": 0.2122,
+      "step": 2767
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0529987408669244,
+      "learning_rate": 8.763750129093236e-06,
+      "loss": 0.2422,
+      "step": 2768
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.500728904048456,
+      "learning_rate": 8.762779832058978e-06,
+      "loss": 0.1867,
+      "step": 2769
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.355871069715481,
+      "learning_rate": 8.761809208154297e-06,
+      "loss": 0.1948,
+      "step": 2770
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.586134685319701,
+      "learning_rate": 8.760838257463511e-06,
+      "loss": 0.2041,
+      "step": 2771
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.099813190305629,
+      "learning_rate": 8.759866980070963e-06,
+      "loss": 0.1795,
+      "step": 2772
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.399160429165581,
+      "learning_rate": 8.758895376061032e-06,
+      "loss": 0.1882,
+      "step": 2773
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.220740916464358,
+      "learning_rate": 8.757923445518116e-06,
+      "loss": 0.1793,
+      "step": 2774
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.456157289890256,
+      "learning_rate": 8.75695118852665e-06,
+      "loss": 0.2157,
+      "step": 2775
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.523721127510485,
+      "learning_rate": 8.755978605171089e-06,
+      "loss": 0.1935,
+      "step": 2776
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.47234853171708,
+      "learning_rate": 8.755005695535925e-06,
+      "loss": 0.1925,
+      "step": 2777
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1301879411098334,
+      "learning_rate": 8.754032459705672e-06,
+      "loss": 0.1751,
+      "step": 2778
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4418392813033867,
+      "learning_rate": 8.753058897764874e-06,
+      "loss": 0.1967,
+      "step": 2779
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6905797584173454,
+      "learning_rate": 8.752085009798106e-06,
+      "loss": 0.1807,
+      "step": 2780
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.494328923529313,
+      "learning_rate": 8.751110795889966e-06,
+      "loss": 0.1697,
+      "step": 2781
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3040684749027185,
+      "learning_rate": 8.750136256125085e-06,
+      "loss": 0.172,
+      "step": 2782
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4668805819280553,
+      "learning_rate": 8.749161390588121e-06,
+      "loss": 0.1678,
+      "step": 2783
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.836425295505188,
+      "learning_rate": 8.74818619936376e-06,
+      "loss": 0.2205,
+      "step": 2784
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.550242630675199,
+      "learning_rate": 8.747210682536715e-06,
+      "loss": 0.1905,
+      "step": 2785
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.449602648461097,
+      "learning_rate": 8.746234840191729e-06,
+      "loss": 0.2055,
+      "step": 2786
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5010618604504757,
+      "learning_rate": 8.745258672413574e-06,
+      "loss": 0.1836,
+      "step": 2787
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.474184030193847,
+      "learning_rate": 8.744282179287049e-06,
+      "loss": 0.237,
+      "step": 2788
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1031667080320315,
+      "learning_rate": 8.743305360896978e-06,
+      "loss": 0.1334,
+      "step": 2789
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.565738634157848,
+      "learning_rate": 8.742328217328221e-06,
+      "loss": 0.1951,
+      "step": 2790
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7330100746431305,
+      "learning_rate": 8.741350748665662e-06,
+      "loss": 0.202,
+      "step": 2791
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.30154469479652,
+      "learning_rate": 8.74037295499421e-06,
+      "loss": 0.2157,
+      "step": 2792
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.189408271568324,
+      "learning_rate": 8.739394836398806e-06,
+      "loss": 0.1714,
+      "step": 2793
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.224480666727112,
+      "learning_rate": 8.73841639296442e-06,
+      "loss": 0.1877,
+      "step": 2794
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2626532348442527,
+      "learning_rate": 8.737437624776047e-06,
+      "loss": 0.1926,
+      "step": 2795
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.628893566052792,
+      "learning_rate": 8.736458531918714e-06,
+      "loss": 0.218,
+      "step": 2796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2522327405947307,
+      "learning_rate": 8.735479114477472e-06,
+      "loss": 0.1874,
+      "step": 2797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.591948674583733,
+      "learning_rate": 8.734499372537406e-06,
+      "loss": 0.1992,
+      "step": 2798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.51752308255524,
+      "learning_rate": 8.73351930618362e-06,
+      "loss": 0.2157,
+      "step": 2799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3201063266190713,
+      "learning_rate": 8.732538915501257e-06,
+      "loss": 0.1637,
+      "step": 2800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4899502973650116,
+      "learning_rate": 8.73155820057548e-06,
+      "loss": 0.2145,
+      "step": 2801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.129743947493152,
+      "learning_rate": 8.730577161491486e-06,
+      "loss": 0.1636,
+      "step": 2802
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1384861431605815,
+      "learning_rate": 8.729595798334494e-06,
+      "loss": 0.1534,
+      "step": 2803
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7007994397624007,
+      "learning_rate": 8.728614111189756e-06,
+      "loss": 0.2017,
+      "step": 2804
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.780240542908485,
+      "learning_rate": 8.72763210014255e-06,
+      "loss": 0.2,
+      "step": 2805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1232296947977547,
+      "learning_rate": 8.726649765278184e-06,
+      "loss": 0.1977,
+      "step": 2806
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4397195719291567,
+      "learning_rate": 8.72566710668199e-06,
+      "loss": 0.201,
+      "step": 2807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6438181958017797,
+      "learning_rate": 8.724684124439336e-06,
+      "loss": 0.2249,
+      "step": 2808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3711227058858615,
+      "learning_rate": 8.723700818635608e-06,
+      "loss": 0.1919,
+      "step": 2809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.180779817131397,
+      "learning_rate": 8.722717189356226e-06,
+      "loss": 0.1603,
+      "step": 2810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.49436174495027,
+      "learning_rate": 8.72173323668664e-06,
+      "loss": 0.2141,
+      "step": 2811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2868530248417716,
+      "learning_rate": 8.720748960712323e-06,
+      "loss": 0.1736,
+      "step": 2812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.383150595717509,
+      "learning_rate": 8.71976436151878e-06,
+      "loss": 0.1723,
+      "step": 2813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.119588970630532,
+      "learning_rate": 8.718779439191543e-06,
+      "loss": 0.1602,
+      "step": 2814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3213459267634593,
+      "learning_rate": 8.717794193816166e-06,
+      "loss": 0.1608,
+      "step": 2815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4550755801244257,
+      "learning_rate": 8.716808625478245e-06,
+      "loss": 0.1963,
+      "step": 2816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4747411684013088,
+      "learning_rate": 8.715822734263391e-06,
+      "loss": 0.1783,
+      "step": 2817
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.572036209057203,
+      "learning_rate": 8.714836520257248e-06,
+      "loss": 0.2092,
+      "step": 2818
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7284523647952152,
+      "learning_rate": 8.71384998354549e-06,
+      "loss": 0.2255,
+      "step": 2819
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4605970978778173,
+      "learning_rate": 8.712863124213814e-06,
+      "loss": 0.1908,
+      "step": 2820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4768126847342353,
+      "learning_rate": 8.711875942347949e-06,
+      "loss": 0.2375,
+      "step": 2821
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4665160610714607,
+      "learning_rate": 8.710888438033651e-06,
+      "loss": 0.2214,
+      "step": 2822
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.282577791392892,
+      "learning_rate": 8.709900611356703e-06,
+      "loss": 0.162,
+      "step": 2823
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.026581680861299,
+      "learning_rate": 8.708912462402921e-06,
+      "loss": 0.1699,
+      "step": 2824
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4428791122302482,
+      "learning_rate": 8.70792399125814e-06,
+      "loss": 0.1952,
+      "step": 2825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3454189423426235,
+      "learning_rate": 8.706935198008228e-06,
+      "loss": 0.1852,
+      "step": 2826
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.89607961041875,
+      "learning_rate": 8.705946082739085e-06,
+      "loss": 0.2184,
+      "step": 2827
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4226062878841166,
+      "learning_rate": 8.70495664553663e-06,
+      "loss": 0.1992,
+      "step": 2828
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.573541371576731,
+      "learning_rate": 8.703966886486819e-06,
+      "loss": 0.2227,
+      "step": 2829
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4774018351164053,
+      "learning_rate": 8.702976805675629e-06,
+      "loss": 0.2,
+      "step": 2830
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0534267382939277,
+      "learning_rate": 8.70198640318907e-06,
+      "loss": 0.1667,
+      "step": 2831
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5087219925522373,
+      "learning_rate": 8.700995679113175e-06,
+      "loss": 0.2074,
+      "step": 2832
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9257746165405534,
+      "learning_rate": 8.70000463353401e-06,
+      "loss": 0.1813,
+      "step": 2833
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.273680968541367,
+      "learning_rate": 8.699013266537663e-06,
+      "loss": 0.185,
+      "step": 2834
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3327119556671003,
+      "learning_rate": 8.698021578210258e-06,
+      "loss": 0.2063,
+      "step": 2835
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1195830299768024,
+      "learning_rate": 8.697029568637942e-06,
+      "loss": 0.1717,
+      "step": 2836
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1554648168964543,
+      "learning_rate": 8.696037237906887e-06,
+      "loss": 0.1708,
+      "step": 2837
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.193543189385725,
+      "learning_rate": 8.695044586103297e-06,
+      "loss": 0.1869,
+      "step": 2838
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.330837839686286,
+      "learning_rate": 8.694051613313404e-06,
+      "loss": 0.1728,
+      "step": 2839
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5632516650720967,
+      "learning_rate": 8.693058319623466e-06,
+      "loss": 0.2162,
+      "step": 2840
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3568423874980344,
+      "learning_rate": 8.692064705119773e-06,
+      "loss": 0.1773,
+      "step": 2841
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2891638023206524,
+      "learning_rate": 8.691070769888637e-06,
+      "loss": 0.176,
+      "step": 2842
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.484993679095583,
+      "learning_rate": 8.690076514016399e-06,
+      "loss": 0.1897,
+      "step": 2843
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4749565428766696,
+      "learning_rate": 8.689081937589432e-06,
+      "loss": 0.2004,
+      "step": 2844
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6288869732455398,
+      "learning_rate": 8.688087040694133e-06,
+      "loss": 0.1926,
+      "step": 2845
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1442859947017343,
+      "learning_rate": 8.68709182341693e-06,
+      "loss": 0.1631,
+      "step": 2846
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3355028331418413,
+      "learning_rate": 8.686096285844274e-06,
+      "loss": 0.1781,
+      "step": 2847
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1889818725678225,
+      "learning_rate": 8.68510042806265e-06,
+      "loss": 0.1887,
+      "step": 2848
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.454479286130754,
+      "learning_rate": 8.684104250158565e-06,
+      "loss": 0.2128,
+      "step": 2849
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.26999384322895,
+      "learning_rate": 8.683107752218557e-06,
+      "loss": 0.158,
+      "step": 2850
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4794635176905886,
+      "learning_rate": 8.682110934329191e-06,
+      "loss": 0.1707,
+      "step": 2851
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6289178468210173,
+      "learning_rate": 8.681113796577063e-06,
+      "loss": 0.205,
+      "step": 2852
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3575225229142247,
+      "learning_rate": 8.680116339048787e-06,
+      "loss": 0.2212,
+      "step": 2853
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1829393498673357,
+      "learning_rate": 8.679118561831018e-06,
+      "loss": 0.1676,
+      "step": 2854
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1871362611364096,
+      "learning_rate": 8.678120465010431e-06,
+      "loss": 0.1775,
+      "step": 2855
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4979950234212405,
+      "learning_rate": 8.677122048673727e-06,
+      "loss": 0.2119,
+      "step": 2856
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.304731113369537,
+      "learning_rate": 8.676123312907641e-06,
+      "loss": 0.1798,
+      "step": 2857
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3456569745337896,
+      "learning_rate": 8.675124257798933e-06,
+      "loss": 0.1984,
+      "step": 2858
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.419655888175134,
+      "learning_rate": 8.674124883434386e-06,
+      "loss": 0.199,
+      "step": 2859
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6063851727841976,
+      "learning_rate": 8.67312518990082e-06,
+      "loss": 0.1833,
+      "step": 2860
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.351231335326036,
+      "learning_rate": 8.672125177285073e-06,
+      "loss": 0.1712,
+      "step": 2861
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5861423461475384,
+      "learning_rate": 8.67112484567402e-06,
+      "loss": 0.2243,
+      "step": 2862
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.277944174305463,
+      "learning_rate": 8.670124195154557e-06,
+      "loss": 0.1825,
+      "step": 2863
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.388511731547195,
+      "learning_rate": 8.669123225813611e-06,
+      "loss": 0.2155,
+      "step": 2864
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0744506189039975,
+      "learning_rate": 8.668121937738134e-06,
+      "loss": 0.1904,
+      "step": 2865
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.540421638416718,
+      "learning_rate": 8.667120331015107e-06,
+      "loss": 0.2151,
+      "step": 2866
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1769251559270395,
+      "learning_rate": 8.666118405731542e-06,
+      "loss": 0.2024,
+      "step": 2867
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2955279155522024,
+      "learning_rate": 8.665116161974473e-06,
+      "loss": 0.1532,
+      "step": 2868
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.252405629246522,
+      "learning_rate": 8.664113599830965e-06,
+      "loss": 0.1902,
+      "step": 2869
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.401060287230373,
+      "learning_rate": 8.66311071938811e-06,
+      "loss": 0.2054,
+      "step": 2870
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3898162386650443,
+      "learning_rate": 8.662107520733027e-06,
+      "loss": 0.1725,
+      "step": 2871
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2622532107999294,
+      "learning_rate": 8.661104003952866e-06,
+      "loss": 0.2061,
+      "step": 2872
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2400789912091432,
+      "learning_rate": 8.660100169134797e-06,
+      "loss": 0.1832,
+      "step": 2873
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0082931945123224,
+      "learning_rate": 8.659096016366027e-06,
+      "loss": 0.1442,
+      "step": 2874
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.149202520874365,
+      "learning_rate": 8.658091545733785e-06,
+      "loss": 0.1542,
+      "step": 2875
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4157046924148946,
+      "learning_rate": 8.657086757325328e-06,
+      "loss": 0.2143,
+      "step": 2876
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4375182585215973,
+      "learning_rate": 8.65608165122794e-06,
+      "loss": 0.1901,
+      "step": 2877
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6077155006900785,
+      "learning_rate": 8.655076227528937e-06,
+      "loss": 0.2247,
+      "step": 2878
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.44056015973225,
+      "learning_rate": 8.654070486315658e-06,
+      "loss": 0.1822,
+      "step": 2879
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0831534440376775,
+      "learning_rate": 8.65306442767547e-06,
+      "loss": 0.1228,
+      "step": 2880
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2978488650552795,
+      "learning_rate": 8.652058051695772e-06,
+      "loss": 0.2031,
+      "step": 2881
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.200378755762137,
+      "learning_rate": 8.651051358463984e-06,
+      "loss": 0.1636,
+      "step": 2882
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3073567230467473,
+      "learning_rate": 8.650044348067558e-06,
+      "loss": 0.1845,
+      "step": 2883
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.361397506047959,
+      "learning_rate": 8.649037020593974e-06,
+      "loss": 0.2009,
+      "step": 2884
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.513137026791609,
+      "learning_rate": 8.648029376130735e-06,
+      "loss": 0.1934,
+      "step": 2885
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.381683246996296,
+      "learning_rate": 8.647021414765376e-06,
+      "loss": 0.1841,
+      "step": 2886
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0265282393780493,
+      "learning_rate": 8.646013136585457e-06,
+      "loss": 0.1438,
+      "step": 2887
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.499891823240599,
+      "learning_rate": 8.64500454167857e-06,
+      "loss": 0.1949,
+      "step": 2888
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.424010980934348,
+      "learning_rate": 8.643995630132326e-06,
+      "loss": 0.213,
+      "step": 2889
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1534888939475394,
+      "learning_rate": 8.642986402034373e-06,
+      "loss": 0.1984,
+      "step": 2890
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6980642477388956,
+      "learning_rate": 8.641976857472378e-06,
+      "loss": 0.1634,
+      "step": 2891
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1505693614550183,
+      "learning_rate": 8.640966996534043e-06,
+      "loss": 0.1604,
+      "step": 2892
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.256234672206739,
+      "learning_rate": 8.639956819307092e-06,
+      "loss": 0.1914,
+      "step": 2893
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0887059862256607,
+      "learning_rate": 8.638946325879278e-06,
+      "loss": 0.1628,
+      "step": 2894
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3956840289056602,
+      "learning_rate": 8.637935516338384e-06,
+      "loss": 0.192,
+      "step": 2895
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.482079862747106,
+      "learning_rate": 8.636924390772217e-06,
+      "loss": 0.1735,
+      "step": 2896
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3747609164764185,
+      "learning_rate": 8.635912949268614e-06,
+      "loss": 0.1864,
+      "step": 2897
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2545481998290158,
+      "learning_rate": 8.634901191915438e-06,
+      "loss": 0.1375,
+      "step": 2898
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1432853337502666,
+      "learning_rate": 8.633889118800578e-06,
+      "loss": 0.1604,
+      "step": 2899
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.195843722058196,
+      "learning_rate": 8.632876730011955e-06,
+      "loss": 0.1844,
+      "step": 2900
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2080382931556977,
+      "learning_rate": 8.631864025637511e-06,
+      "loss": 0.1843,
+      "step": 2901
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5617323967118195,
+      "learning_rate": 8.630851005765223e-06,
+      "loss": 0.2174,
+      "step": 2902
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.330087877657432,
+      "learning_rate": 8.62983767048309e-06,
+      "loss": 0.1574,
+      "step": 2903
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6095905353837923,
+      "learning_rate": 8.628824019879137e-06,
+      "loss": 0.2094,
+      "step": 2904
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.113043926884788,
+      "learning_rate": 8.627810054041423e-06,
+      "loss": 0.1982,
+      "step": 2905
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.394773063638068,
+      "learning_rate": 8.62679577305803e-06,
+      "loss": 0.1764,
+      "step": 2906
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.237974930921654,
+      "learning_rate": 8.625781177017066e-06,
+      "loss": 0.1884,
+      "step": 2907
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.411786800496304,
+      "learning_rate": 8.62476626600667e-06,
+      "loss": 0.1936,
+      "step": 2908
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.676020172408539,
+      "learning_rate": 8.623751040115007e-06,
+      "loss": 0.1638,
+      "step": 2909
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2170382092354903,
+      "learning_rate": 8.622735499430267e-06,
+      "loss": 0.182,
+      "step": 2910
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5105724065421393,
+      "learning_rate": 8.62171964404067e-06,
+      "loss": 0.2039,
+      "step": 2911
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.371694386308627,
+      "learning_rate": 8.620703474034466e-06,
+      "loss": 0.1815,
+      "step": 2912
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3530216986637034,
+      "learning_rate": 8.619686989499926e-06,
+      "loss": 0.2152,
+      "step": 2913
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3072036955774147,
+      "learning_rate": 8.61867019052535e-06,
+      "loss": 0.1757,
+      "step": 2914
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1529166265273343,
+      "learning_rate": 8.617653077199073e-06,
+      "loss": 0.1941,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.364777274540656,
+      "learning_rate": 8.616635649609443e-06,
+      "loss": 0.1929,
+      "step": 2916
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2779472042187847,
+      "learning_rate": 8.615617907844848e-06,
+      "loss": 0.1985,
+      "step": 2917
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.214152739006492,
+      "learning_rate": 8.614599851993697e-06,
+      "loss": 0.1833,
+      "step": 2918
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1385745194723342,
+      "learning_rate": 8.613581482144428e-06,
+      "loss": 0.1666,
+      "step": 2919
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9988892369314604,
+      "learning_rate": 8.612562798385508e-06,
+      "loss": 0.1582,
+      "step": 2920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.145132150836671,
+      "learning_rate": 8.61154380080543e-06,
+      "loss": 0.1851,
+      "step": 2921
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.324306245473324,
+      "learning_rate": 8.610524489492709e-06,
+      "loss": 0.2049,
+      "step": 2922
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2495750454299692,
+      "learning_rate": 8.609504864535896e-06,
+      "loss": 0.1775,
+      "step": 2923
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3201180801364054,
+      "learning_rate": 8.608484926023564e-06,
+      "loss": 0.1865,
+      "step": 2924
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.192310554152911,
+      "learning_rate": 8.607464674044315e-06,
+      "loss": 0.1771,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5280041140581537,
+      "learning_rate": 8.606444108686775e-06,
+      "loss": 0.2193,
+      "step": 2926
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.121601675028039,
+      "learning_rate": 8.605423230039605e-06,
+      "loss": 0.1609,
+      "step": 2927
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.960615936056803,
+      "learning_rate": 8.604402038191483e-06,
+      "loss": 0.1492,
+      "step": 2928
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.234093324508485,
+      "learning_rate": 8.603380533231123e-06,
+      "loss": 0.1824,
+      "step": 2929
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.882866952366496,
+      "learning_rate": 8.60235871524726e-06,
+      "loss": 0.2241,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2962682529563767,
+      "learning_rate": 8.601336584328659e-06,
+      "loss": 0.1876,
+      "step": 2931
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3056479835593757,
+      "learning_rate": 8.600314140564114e-06,
+      "loss": 0.2271,
+      "step": 2932
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4034838788210604,
+      "learning_rate": 8.599291384042442e-06,
+      "loss": 0.2053,
+      "step": 2933
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.107982531790283,
+      "learning_rate": 8.598268314852492e-06,
+      "loss": 0.202,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.455988044209978,
+      "learning_rate": 8.597244933083133e-06,
+      "loss": 0.2168,
+      "step": 2935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.301990901965644,
+      "learning_rate": 8.596221238823269e-06,
+      "loss": 0.1734,
+      "step": 2936
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0704976389102803,
+      "learning_rate": 8.595197232161824e-06,
+      "loss": 0.1816,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1749843939542703,
+      "learning_rate": 8.594172913187759e-06,
+      "loss": 0.1621,
+      "step": 2938
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3930631179372077,
+      "learning_rate": 8.593148281990052e-06,
+      "loss": 0.1699,
+      "step": 2939
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3075746391802316,
+      "learning_rate": 8.592123338657713e-06,
+      "loss": 0.1931,
+      "step": 2940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2129723600778597,
+      "learning_rate": 8.591098083279774e-06,
+      "loss": 0.1624,
+      "step": 2941
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0610936259969437,
+      "learning_rate": 8.590072515945305e-06,
+      "loss": 0.1711,
+      "step": 2942
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.74212252456524,
+      "learning_rate": 8.589046636743394e-06,
+      "loss": 0.2182,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4806255070332215,
+      "learning_rate": 8.588020445763156e-06,
+      "loss": 0.2078,
+      "step": 2944
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.339736639385108,
+      "learning_rate": 8.58699394309374e-06,
+      "loss": 0.1504,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.393556450157724,
+      "learning_rate": 8.585967128824313e-06,
+      "loss": 0.1769,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2620493563224335,
+      "learning_rate": 8.584940003044078e-06,
+      "loss": 0.1972,
+      "step": 2947
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5604571786252897,
+      "learning_rate": 8.583912565842258e-06,
+      "loss": 0.2324,
+      "step": 2948
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2752450658443713,
+      "learning_rate": 8.582884817308106e-06,
+      "loss": 0.1481,
+      "step": 2949
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.394276666398064,
+      "learning_rate": 8.581856757530902e-06,
+      "loss": 0.1835,
+      "step": 2950
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.45337753520434,
+      "learning_rate": 8.580828386599955e-06,
+      "loss": 0.2006,
+      "step": 2951
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6284522610521486,
+      "learning_rate": 8.579799704604597e-06,
+      "loss": 0.1906,
+      "step": 2952
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0366292427105246,
+      "learning_rate": 8.57877071163419e-06,
+      "loss": 0.1434,
+      "step": 2953
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.249679144464604,
+      "learning_rate": 8.57774140777812e-06,
+      "loss": 0.1718,
+      "step": 2954
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.652569711235235,
+      "learning_rate": 8.576711793125804e-06,
+      "loss": 0.1909,
+      "step": 2955
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.725252432798393,
+      "learning_rate": 8.575681867766685e-06,
+      "loss": 0.2155,
+      "step": 2956
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.890890149773693,
+      "learning_rate": 8.574651631790229e-06,
+      "loss": 0.1701,
+      "step": 2957
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.296949025815334,
+      "learning_rate": 8.573621085285934e-06,
+      "loss": 0.2085,
+      "step": 2958
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.029623932412452,
+      "learning_rate": 8.572590228343322e-06,
+      "loss": 0.1562,
+      "step": 2959
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.521867740967801,
+      "learning_rate": 8.571559061051943e-06,
+      "loss": 0.2039,
+      "step": 2960
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3599661575880275,
+      "learning_rate": 8.570527583501374e-06,
+      "loss": 0.2269,
+      "step": 2961
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3909147968253603,
+      "learning_rate": 8.569495795781221e-06,
+      "loss": 0.2201,
+      "step": 2962
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4161298466611467,
+      "learning_rate": 8.568463697981112e-06,
+      "loss": 0.2188,
+      "step": 2963
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.240662591106111,
+      "learning_rate": 8.567431290190705e-06,
+      "loss": 0.1746,
+      "step": 2964
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.271007932073467,
+      "learning_rate": 8.566398572499685e-06,
+      "loss": 0.2024,
+      "step": 2965
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3613354212421203,
+      "learning_rate": 8.565365544997763e-06,
+      "loss": 0.1807,
+      "step": 2966
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.498127016720821,
+      "learning_rate": 8.56433220777468e-06,
+      "loss": 0.217,
+      "step": 2967
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2452359753478834,
+      "learning_rate": 8.563298560920198e-06,
+      "loss": 0.184,
+      "step": 2968
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1818094509200914,
+      "learning_rate": 8.562264604524112e-06,
+      "loss": 0.1907,
+      "step": 2969
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3948536569139094,
+      "learning_rate": 8.56123033867624e-06,
+      "loss": 0.2004,
+      "step": 2970
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6730079276605223,
+      "learning_rate": 8.560195763466428e-06,
+      "loss": 0.1843,
+      "step": 2971
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.341976726801267,
+      "learning_rate": 8.559160878984548e-06,
+      "loss": 0.1884,
+      "step": 2972
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1177567859730697,
+      "learning_rate": 8.558125685320502e-06,
+      "loss": 0.169,
+      "step": 2973
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0266131283623263,
+      "learning_rate": 8.557090182564215e-06,
+      "loss": 0.1782,
+      "step": 2974
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0774743600694725,
+      "learning_rate": 8.556054370805642e-06,
+      "loss": 0.167,
+      "step": 2975
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9627260224740766,
+      "learning_rate": 8.555018250134761e-06,
+      "loss": 0.1465,
+      "step": 2976
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2128573763105877,
+      "learning_rate": 8.553981820641582e-06,
+      "loss": 0.1614,
+      "step": 2977
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.44005337598289,
+      "learning_rate": 8.552945082416135e-06,
+      "loss": 0.1942,
+      "step": 2978
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5369090014104283,
+      "learning_rate": 8.551908035548486e-06,
+      "loss": 0.1994,
+      "step": 2979
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.249862410876498,
+      "learning_rate": 8.550870680128718e-06,
+      "loss": 0.1732,
+      "step": 2980
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4587537349780977,
+      "learning_rate": 8.549833016246948e-06,
+      "loss": 0.2018,
+      "step": 2981
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1910853260507452,
+      "learning_rate": 8.548795043993316e-06,
+      "loss": 0.174,
+      "step": 2982
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0283828584050654,
+      "learning_rate": 8.547756763457993e-06,
+      "loss": 0.1824,
+      "step": 2983
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1443991911136666,
+      "learning_rate": 8.54671817473117e-06,
+      "loss": 0.197,
+      "step": 2984
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.466367033374896,
+      "learning_rate": 8.54567927790307e-06,
+      "loss": 0.2043,
+      "step": 2985
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3139226927566536,
+      "learning_rate": 8.544640073063941e-06,
+      "loss": 0.2045,
+      "step": 2986
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0997027130755943,
+      "learning_rate": 8.543600560304059e-06,
+      "loss": 0.1866,
+      "step": 2987
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1389997681869377,
+      "learning_rate": 8.542560739713726e-06,
+      "loss": 0.167,
+      "step": 2988
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2853143360034194,
+      "learning_rate": 8.54152061138327e-06,
+      "loss": 0.1751,
+      "step": 2989
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1682988469979314,
+      "learning_rate": 8.540480175403045e-06,
+      "loss": 0.1736,
+      "step": 2990
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4002264765603365,
+      "learning_rate": 8.539439431863434e-06,
+      "loss": 0.1904,
+      "step": 2991
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.427987034433787,
+      "learning_rate": 8.538398380854848e-06,
+      "loss": 0.1823,
+      "step": 2992
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1709130880004173,
+      "learning_rate": 8.53735702246772e-06,
+      "loss": 0.1487,
+      "step": 2993
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1147905296149347,
+      "learning_rate": 8.536315356792513e-06,
+      "loss": 0.1483,
+      "step": 2994
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.512373886935549,
+      "learning_rate": 8.535273383919715e-06,
+      "loss": 0.2015,
+      "step": 2995
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8196023608468455,
+      "learning_rate": 8.534231103939842e-06,
+      "loss": 0.2452,
+      "step": 2996
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5105635412716727,
+      "learning_rate": 8.533188516943436e-06,
+      "loss": 0.212,
+      "step": 2997
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.27547473821853,
+      "learning_rate": 8.532145623021067e-06,
+      "loss": 0.1748,
+      "step": 2998
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.24749024744668,
+      "learning_rate": 8.53110242226333e-06,
+      "loss": 0.205,
+      "step": 2999
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.296213499562884,
+      "learning_rate": 8.530058914760846e-06,
+      "loss": 0.1995,
+      "step": 3000
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.191079970170801,
+      "learning_rate": 8.529015100604267e-06,
+      "loss": 0.1424,
+      "step": 3001
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3979563801602555,
+      "learning_rate": 8.527970979884266e-06,
+      "loss": 0.2067,
+      "step": 3002
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1809938735388568,
+      "learning_rate": 8.526926552691545e-06,
+      "loss": 0.184,
+      "step": 3003
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.306277374137795,
+      "learning_rate": 8.525881819116832e-06,
+      "loss": 0.1918,
+      "step": 3004
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.34026456006689,
+      "learning_rate": 8.524836779250886e-06,
+      "loss": 0.1494,
+      "step": 3005
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2721382838304565,
+      "learning_rate": 8.523791433184486e-06,
+      "loss": 0.2074,
+      "step": 3006
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.222874464450221,
+      "learning_rate": 8.522745781008442e-06,
+      "loss": 0.2302,
+      "step": 3007
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.325379369042873,
+      "learning_rate": 8.521699822813587e-06,
+      "loss": 0.1908,
+      "step": 3008
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4122651401773894,
+      "learning_rate": 8.520653558690785e-06,
+      "loss": 0.2084,
+      "step": 3009
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1546434756554875,
+      "learning_rate": 8.519606988730924e-06,
+      "loss": 0.1842,
+      "step": 3010
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.865708444249605,
+      "learning_rate": 8.518560113024918e-06,
+      "loss": 0.169,
+      "step": 3011
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2554067090977776,
+      "learning_rate": 8.51751293166371e-06,
+      "loss": 0.1703,
+      "step": 3012
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.357181088308605,
+      "learning_rate": 8.516465444738264e-06,
+      "loss": 0.1735,
+      "step": 3013
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.425931728761658,
+      "learning_rate": 8.51541765233958e-06,
+      "loss": 0.1971,
+      "step": 3014
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.365361162220478,
+      "learning_rate": 8.514369554558677e-06,
+      "loss": 0.2018,
+      "step": 3015
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0394891389537,
+      "learning_rate": 8.513321151486602e-06,
+      "loss": 0.1462,
+      "step": 3016
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6773440927342125,
+      "learning_rate": 8.512272443214428e-06,
+      "loss": 0.1901,
+      "step": 3017
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4491676839132874,
+      "learning_rate": 8.511223429833258e-06,
+      "loss": 0.1913,
+      "step": 3018
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0677356406947327,
+      "learning_rate": 8.510174111434219e-06,
+      "loss": 0.1713,
+      "step": 3019
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3485120970242437,
+      "learning_rate": 8.509124488108462e-06,
+      "loss": 0.1811,
+      "step": 3020
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.46341352234346,
+      "learning_rate": 8.508074559947172e-06,
+      "loss": 0.1709,
+      "step": 3021
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.685440054496452,
+      "learning_rate": 8.507024327041551e-06,
+      "loss": 0.1824,
+      "step": 3022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4808934965049363,
+      "learning_rate": 8.505973789482833e-06,
+      "loss": 0.2093,
+      "step": 3023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8556894455660642,
+      "learning_rate": 8.50492294736228e-06,
+      "loss": 0.1524,
+      "step": 3024
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4881506016204358,
+      "learning_rate": 8.503871800771175e-06,
+      "loss": 0.2049,
+      "step": 3025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.240489646190458,
+      "learning_rate": 8.502820349800832e-06,
+      "loss": 0.196,
+      "step": 3026
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.691420655024295,
+      "learning_rate": 8.50176859454259e-06,
+      "loss": 0.2091,
+      "step": 3027
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.053217633237131,
+      "learning_rate": 8.500716535087815e-06,
+      "loss": 0.1731,
+      "step": 3028
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2403328224929324,
+      "learning_rate": 8.499664171527895e-06,
+      "loss": 0.1428,
+      "step": 3029
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4978588685201775,
+      "learning_rate": 8.498611503954253e-06,
+      "loss": 0.2214,
+      "step": 3030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4065018938809706,
+      "learning_rate": 8.497558532458333e-06,
+      "loss": 0.2216,
+      "step": 3031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5129761814306266,
+      "learning_rate": 8.496505257131602e-06,
+      "loss": 0.2129,
+      "step": 3032
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.363418496015452,
+      "learning_rate": 8.495451678065563e-06,
+      "loss": 0.1981,
+      "step": 3033
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3725663050997725,
+      "learning_rate": 8.494397795351735e-06,
+      "loss": 0.1729,
+      "step": 3034
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6043223027279883,
+      "learning_rate": 8.49334360908167e-06,
+      "loss": 0.2415,
+      "step": 3035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.200892946502124,
+      "learning_rate": 8.492289119346944e-06,
+      "loss": 0.2071,
+      "step": 3036
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2410103218519954,
+      "learning_rate": 8.491234326239162e-06,
+      "loss": 0.1557,
+      "step": 3037
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4070035528104743,
+      "learning_rate": 8.49017922984995e-06,
+      "loss": 0.207,
+      "step": 3038
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.237562903347781,
+      "learning_rate": 8.489123830270966e-06,
+      "loss": 0.1919,
+      "step": 3039
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3614936089828213,
+      "learning_rate": 8.488068127593892e-06,
+      "loss": 0.2001,
+      "step": 3040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.175831430982278,
+      "learning_rate": 8.487012121910435e-06,
+      "loss": 0.1487,
+      "step": 3041
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1806127137435847,
+      "learning_rate": 8.485955813312328e-06,
+      "loss": 0.1924,
+      "step": 3042
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2384261705794195,
+      "learning_rate": 8.484899201891336e-06,
+      "loss": 0.1762,
+      "step": 3043
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.343339274052141,
+      "learning_rate": 8.483842287739244e-06,
+      "loss": 0.1853,
+      "step": 3044
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.910824238326714,
+      "learning_rate": 8.482785070947866e-06,
+      "loss": 0.2042,
+      "step": 3045
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7607282515390614,
+      "learning_rate": 8.48172755160904e-06,
+      "loss": 0.2072,
+      "step": 3046
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.146948793894683,
+      "learning_rate": 8.480669729814635e-06,
+      "loss": 0.1675,
+      "step": 3047
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.218717007846125,
+      "learning_rate": 8.479611605656541e-06,
+      "loss": 0.1702,
+      "step": 3048
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.325550412706671,
+      "learning_rate": 8.478553179226676e-06,
+      "loss": 0.1925,
+      "step": 3049
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.242273133710287,
+      "learning_rate": 8.477494450616988e-06,
+      "loss": 0.1816,
+      "step": 3050
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.94647079451854,
+      "learning_rate": 8.476435419919446e-06,
+      "loss": 0.2217,
+      "step": 3051
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1939567209115025,
+      "learning_rate": 8.475376087226048e-06,
+      "loss": 0.1751,
+      "step": 3052
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2597485431400823,
+      "learning_rate": 8.474316452628816e-06,
+      "loss": 0.2272,
+      "step": 3053
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.382621343745723,
+      "learning_rate": 8.473256516219803e-06,
+      "loss": 0.1747,
+      "step": 3054
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4858681339125206,
+      "learning_rate": 8.472196278091083e-06,
+      "loss": 0.1875,
+      "step": 3055
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.211265026347421,
+      "learning_rate": 8.471135738334758e-06,
+      "loss": 0.1826,
+      "step": 3056
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5829061069828247,
+      "learning_rate": 8.470074897042958e-06,
+      "loss": 0.1901,
+      "step": 3057
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4037883591247065,
+      "learning_rate": 8.469013754307834e-06,
+      "loss": 0.2005,
+      "step": 3058
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1486518031212425,
+      "learning_rate": 8.46795231022157e-06,
+      "loss": 0.1593,
+      "step": 3059
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3014271772524304,
+      "learning_rate": 8.466890564876374e-06,
+      "loss": 0.1904,
+      "step": 3060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3152604521832143,
+      "learning_rate": 8.465828518364476e-06,
+      "loss": 0.19,
+      "step": 3061
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.496576917370413,
+      "learning_rate": 8.464766170778138e-06,
+      "loss": 0.2097,
+      "step": 3062
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0387693913764022,
+      "learning_rate": 8.463703522209644e-06,
+      "loss": 0.1546,
+      "step": 3063
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.279854354405747,
+      "learning_rate": 8.462640572751306e-06,
+      "loss": 0.1831,
+      "step": 3064
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3758804873078194,
+      "learning_rate": 8.461577322495463e-06,
+      "loss": 0.1929,
+      "step": 3065
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.385642994484938,
+      "learning_rate": 8.460513771534475e-06,
+      "loss": 0.2021,
+      "step": 3066
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.700259176909793,
+      "learning_rate": 8.459449919960737e-06,
+      "loss": 0.2065,
+      "step": 3067
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.631203541207082,
+      "learning_rate": 8.458385767866662e-06,
+      "loss": 0.1705,
+      "step": 3068
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.447619822150682,
+      "learning_rate": 8.457321315344695e-06,
+      "loss": 0.1886,
+      "step": 3069
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4516625794127993,
+      "learning_rate": 8.456256562487301e-06,
+      "loss": 0.1905,
+      "step": 3070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.299592883000959,
+      "learning_rate": 8.455191509386975e-06,
+      "loss": 0.182,
+      "step": 3071
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2269770745840445,
+      "learning_rate": 8.45412615613624e-06,
+      "loss": 0.2016,
+      "step": 3072
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.085420537972154,
+      "learning_rate": 8.45306050282764e-06,
+      "loss": 0.1708,
+      "step": 3073
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3862694234581605,
+      "learning_rate": 8.45199454955375e-06,
+      "loss": 0.1878,
+      "step": 3074
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.353820514730244,
+      "learning_rate": 8.450928296407168e-06,
+      "loss": 0.1744,
+      "step": 3075
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3247469904138276,
+      "learning_rate": 8.449861743480517e-06,
+      "loss": 0.1847,
+      "step": 3076
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.074744225118243,
+      "learning_rate": 8.44879489086645e-06,
+      "loss": 0.1635,
+      "step": 3077
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8597912145251223,
+      "learning_rate": 8.44772773865764e-06,
+      "loss": 0.1338,
+      "step": 3078
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.125846950269848,
+      "learning_rate": 8.446660286946796e-06,
+      "loss": 0.1703,
+      "step": 3079
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0872555692773047,
+      "learning_rate": 8.445592535826643e-06,
+      "loss": 0.1761,
+      "step": 3080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1181738011558635,
+      "learning_rate": 8.444524485389936e-06,
+      "loss": 0.1456,
+      "step": 3081
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4989665979744675,
+      "learning_rate": 8.443456135729458e-06,
+      "loss": 0.188,
+      "step": 3082
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5422381304376507,
+      "learning_rate": 8.442387486938013e-06,
+      "loss": 0.1858,
+      "step": 3083
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.590302299506852,
+      "learning_rate": 8.441318539108433e-06,
+      "loss": 0.2033,
+      "step": 3084
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3115779754728045,
+      "learning_rate": 8.440249292333583e-06,
+      "loss": 0.1878,
+      "step": 3085
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3282731352024673,
+      "learning_rate": 8.439179746706343e-06,
+      "loss": 0.1708,
+      "step": 3086
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.75669156663873,
+      "learning_rate": 8.438109902319622e-06,
+      "loss": 0.2234,
+      "step": 3087
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1802476879487545,
+      "learning_rate": 8.437039759266364e-06,
+      "loss": 0.1879,
+      "step": 3088
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.412437474357647,
+      "learning_rate": 8.435969317639522e-06,
+      "loss": 0.1834,
+      "step": 3089
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2251027093776927,
+      "learning_rate": 8.434898577532094e-06,
+      "loss": 0.1928,
+      "step": 3090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5865316177069135,
+      "learning_rate": 8.433827539037088e-06,
+      "loss": 0.1917,
+      "step": 3091
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.578706726599439,
+      "learning_rate": 8.432756202247547e-06,
+      "loss": 0.2108,
+      "step": 3092
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8375642700619648,
+      "learning_rate": 8.431684567256537e-06,
+      "loss": 0.1851,
+      "step": 3093
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.259006862859739,
+      "learning_rate": 8.430612634157152e-06,
+      "loss": 0.187,
+      "step": 3094
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2474961826152158,
+      "learning_rate": 8.429540403042507e-06,
+      "loss": 0.2066,
+      "step": 3095
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.738646810269791,
+      "learning_rate": 8.42846787400575e-06,
+      "loss": 0.1719,
+      "step": 3096
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9865551837167141,
+      "learning_rate": 8.427395047140046e-06,
+      "loss": 0.1636,
+      "step": 3097
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9476311838629043,
+      "learning_rate": 8.426321922538594e-06,
+      "loss": 0.213,
+      "step": 3098
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.316636072549775,
+      "learning_rate": 8.425248500294616e-06,
+      "loss": 0.1726,
+      "step": 3099
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0406642084271187,
+      "learning_rate": 8.424174780501359e-06,
+      "loss": 0.1494,
+      "step": 3100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3455637830238585,
+      "learning_rate": 8.423100763252094e-06,
+      "loss": 0.1637,
+      "step": 3101
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.499855280905456,
+      "learning_rate": 8.422026448640124e-06,
+      "loss": 0.2014,
+      "step": 3102
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.331657645185503,
+      "learning_rate": 8.420951836758774e-06,
+      "loss": 0.1723,
+      "step": 3103
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3615940104180857,
+      "learning_rate": 8.41987692770139e-06,
+      "loss": 0.1931,
+      "step": 3104
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2436713661654646,
+      "learning_rate": 8.418801721561355e-06,
+      "loss": 0.1838,
+      "step": 3105
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.427231359923422,
+      "learning_rate": 8.417726218432065e-06,
+      "loss": 0.1825,
+      "step": 3106
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3052245007264798,
+      "learning_rate": 8.416650418406956e-06,
+      "loss": 0.1901,
+      "step": 3107
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2037188012982165,
+      "learning_rate": 8.415574321579474e-06,
+      "loss": 0.1778,
+      "step": 3108
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2248530678225924,
+      "learning_rate": 8.414497928043104e-06,
+      "loss": 0.191,
+      "step": 3109
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.489295525766973,
+      "learning_rate": 8.413421237891352e-06,
+      "loss": 0.2177,
+      "step": 3110
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1842004842443665,
+      "learning_rate": 8.412344251217746e-06,
+      "loss": 0.1683,
+      "step": 3111
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.270047635910235,
+      "learning_rate": 8.411266968115847e-06,
+      "loss": 0.2077,
+      "step": 3112
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1802499926960626,
+      "learning_rate": 8.410189388679234e-06,
+      "loss": 0.1844,
+      "step": 3113
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.623144868565555,
+      "learning_rate": 8.409111513001519e-06,
+      "loss": 0.2277,
+      "step": 3114
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3706240208993665,
+      "learning_rate": 8.408033341176333e-06,
+      "loss": 0.2081,
+      "step": 3115
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.239003050831606,
+      "learning_rate": 8.406954873297342e-06,
+      "loss": 0.1768,
+      "step": 3116
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.093780183056687,
+      "learning_rate": 8.405876109458225e-06,
+      "loss": 0.1933,
+      "step": 3117
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4175316107952383,
+      "learning_rate": 8.404797049752697e-06,
+      "loss": 0.2196,
+      "step": 3118
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.403451209470031,
+      "learning_rate": 8.403717694274498e-06,
+      "loss": 0.1985,
+      "step": 3119
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.165214059904794,
+      "learning_rate": 8.402638043117384e-06,
+      "loss": 0.1801,
+      "step": 3120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4554409846973364,
+      "learning_rate": 8.401558096375149e-06,
+      "loss": 0.2003,
+      "step": 3121
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4169821631854447,
+      "learning_rate": 8.400477854141606e-06,
+      "loss": 0.202,
+      "step": 3122
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.60438028892301,
+      "learning_rate": 8.399397316510596e-06,
+      "loss": 0.1805,
+      "step": 3123
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.617154542372659,
+      "learning_rate": 8.398316483575981e-06,
+      "loss": 0.1991,
+      "step": 3124
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.189831265922532,
+      "learning_rate": 8.397235355431656e-06,
+      "loss": 0.1776,
+      "step": 3125
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.533707995638101,
+      "learning_rate": 8.396153932171538e-06,
+      "loss": 0.2125,
+      "step": 3126
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2225113045189127,
+      "learning_rate": 8.395072213889567e-06,
+      "loss": 0.1739,
+      "step": 3127
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2039562971032574,
+      "learning_rate": 8.393990200679714e-06,
+      "loss": 0.1546,
+      "step": 3128
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5193649829042535,
+      "learning_rate": 8.39290789263597e-06,
+      "loss": 0.197,
+      "step": 3129
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.290784657932828,
+      "learning_rate": 8.391825289852355e-06,
+      "loss": 0.2148,
+      "step": 3130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.118159324819893,
+      "learning_rate": 8.390742392422916e-06,
+      "loss": 0.1718,
+      "step": 3131
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1319792493852017,
+      "learning_rate": 8.389659200441722e-06,
+      "loss": 0.1792,
+      "step": 3132
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.222421329953683,
+      "learning_rate": 8.388575714002872e-06,
+      "loss": 0.1979,
+      "step": 3133
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3475733503971337,
+      "learning_rate": 8.387491933200483e-06,
+      "loss": 0.154,
+      "step": 3134
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3936045453675576,
+      "learning_rate": 8.386407858128707e-06,
+      "loss": 0.1923,
+      "step": 3135
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.287446122286075,
+      "learning_rate": 8.385323488881714e-06,
+      "loss": 0.1614,
+      "step": 3136
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.621889330018262,
+      "learning_rate": 8.384238825553704e-06,
+      "loss": 0.207,
+      "step": 3137
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.238718143378392,
+      "learning_rate": 8.383153868238898e-06,
+      "loss": 0.1701,
+      "step": 3138
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1291662501656803,
+      "learning_rate": 8.382068617031552e-06,
+      "loss": 0.1736,
+      "step": 3139
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3343925441577413,
+      "learning_rate": 8.380983072025934e-06,
+      "loss": 0.1717,
+      "step": 3140
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2028058505377515,
+      "learning_rate": 8.37989723331635e-06,
+      "loss": 0.1625,
+      "step": 3141
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.342067928948073,
+      "learning_rate": 8.378811100997122e-06,
+      "loss": 0.2015,
+      "step": 3142
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.408933931685575,
+      "learning_rate": 8.377724675162607e-06,
+      "loss": 0.1793,
+      "step": 3143
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.572768283139493,
+      "learning_rate": 8.376637955907176e-06,
+      "loss": 0.2256,
+      "step": 3144
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1310873301182287,
+      "learning_rate": 8.375550943325235e-06,
+      "loss": 0.1638,
+      "step": 3145
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.463516843479315,
+      "learning_rate": 8.374463637511212e-06,
+      "loss": 0.2336,
+      "step": 3146
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.969188980344144,
+      "learning_rate": 8.37337603855956e-06,
+      "loss": 0.152,
+      "step": 3147
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.496477007147593,
+      "learning_rate": 8.372288146564757e-06,
+      "loss": 0.2056,
+      "step": 3148
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.135154155548586,
+      "learning_rate": 8.371199961621312e-06,
+      "loss": 0.1861,
+      "step": 3149
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3125600317055137,
+      "learning_rate": 8.370111483823749e-06,
+      "loss": 0.207,
+      "step": 3150
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.483728856043701,
+      "learning_rate": 8.369022713266629e-06,
+      "loss": 0.1885,
+      "step": 3151
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.185622752091959,
+      "learning_rate": 8.367933650044526e-06,
+      "loss": 0.1773,
+      "step": 3152
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4317937062140706,
+      "learning_rate": 8.366844294252054e-06,
+      "loss": 0.1909,
+      "step": 3153
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2057446814003066,
+      "learning_rate": 8.365754645983839e-06,
+      "loss": 0.1998,
+      "step": 3154
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2408568551905246,
+      "learning_rate": 8.36466470533454e-06,
+      "loss": 0.183,
+      "step": 3155
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.187634157667538,
+      "learning_rate": 8.363574472398841e-06,
+      "loss": 0.1812,
+      "step": 3156
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.428166441488641,
+      "learning_rate": 8.362483947271446e-06,
+      "loss": 0.1812,
+      "step": 3157
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.303233504588605,
+      "learning_rate": 8.361393130047093e-06,
+      "loss": 0.1894,
+      "step": 3158
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1228275158824457,
+      "learning_rate": 8.360302020820538e-06,
+      "loss": 0.1653,
+      "step": 3159
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6528927843667383,
+      "learning_rate": 8.359210619686565e-06,
+      "loss": 0.2106,
+      "step": 3160
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.059399745173855,
+      "learning_rate": 8.358118926739984e-06,
+      "loss": 0.1607,
+      "step": 3161
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1633688696032256,
+      "learning_rate": 8.35702694207563e-06,
+      "loss": 0.1848,
+      "step": 3162
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9429359947337626,
+      "learning_rate": 8.355934665788361e-06,
+      "loss": 0.1565,
+      "step": 3163
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.124194516661543,
+      "learning_rate": 8.354842097973065e-06,
+      "loss": 0.1796,
+      "step": 3164
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5037622521669856,
+      "learning_rate": 8.35374923872465e-06,
+      "loss": 0.1994,
+      "step": 3165
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.271565297502334,
+      "learning_rate": 8.352656088138056e-06,
+      "loss": 0.1799,
+      "step": 3166
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.997459133750597,
+      "learning_rate": 8.35156264630824e-06,
+      "loss": 0.1507,
+      "step": 3167
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3259655677813296,
+      "learning_rate": 8.350468913330192e-06,
+      "loss": 0.191,
+      "step": 3168
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3097531818512205,
+      "learning_rate": 8.349374889298923e-06,
+      "loss": 0.2023,
+      "step": 3169
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.627053274025339,
+      "learning_rate": 8.348280574309468e-06,
+      "loss": 0.181,
+      "step": 3170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.495673766202746,
+      "learning_rate": 8.347185968456891e-06,
+      "loss": 0.2025,
+      "step": 3171
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.278958573588023,
+      "learning_rate": 8.346091071836281e-06,
+      "loss": 0.1756,
+      "step": 3172
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0677608065011257,
+      "learning_rate": 8.34499588454275e-06,
+      "loss": 0.1665,
+      "step": 3173
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1306891930787017,
+      "learning_rate": 8.343900406671434e-06,
+      "loss": 0.151,
+      "step": 3174
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.178018017246907,
+      "learning_rate": 8.342804638317502e-06,
+      "loss": 0.1824,
+      "step": 3175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3232082669398086,
+      "learning_rate": 8.341708579576138e-06,
+      "loss": 0.2091,
+      "step": 3176
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1334689352633336,
+      "learning_rate": 8.340612230542557e-06,
+      "loss": 0.1791,
+      "step": 3177
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2534095598490946,
+      "learning_rate": 8.339515591312e-06,
+      "loss": 0.1651,
+      "step": 3178
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.265424416753896,
+      "learning_rate": 8.338418661979729e-06,
+      "loss": 0.1869,
+      "step": 3179
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.374421167726862,
+      "learning_rate": 8.337321442641036e-06,
+      "loss": 0.1945,
+      "step": 3180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3292860052044833,
+      "learning_rate": 8.336223933391232e-06,
+      "loss": 0.164,
+      "step": 3181
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.257221654444349,
+      "learning_rate": 8.33512613432566e-06,
+      "loss": 0.1412,
+      "step": 3182
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1604269318398046,
+      "learning_rate": 8.334028045539685e-06,
+      "loss": 0.161,
+      "step": 3183
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6048587704344675,
+      "learning_rate": 8.332929667128698e-06,
+      "loss": 0.2226,
+      "step": 3184
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.09938770425246,
+      "learning_rate": 8.33183099918811e-06,
+      "loss": 0.1715,
+      "step": 3185
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1683859617254693,
+      "learning_rate": 8.330732041813367e-06,
+      "loss": 0.1821,
+      "step": 3186
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4259187194768157,
+      "learning_rate": 8.329632795099934e-06,
+      "loss": 0.2045,
+      "step": 3187
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.218391832587804,
+      "learning_rate": 8.328533259143298e-06,
+      "loss": 0.1693,
+      "step": 3188
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2756406388422636,
+      "learning_rate": 8.327433434038979e-06,
+      "loss": 0.1927,
+      "step": 3189
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.456680147484425,
+      "learning_rate": 8.326333319882516e-06,
+      "loss": 0.2015,
+      "step": 3190
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.132591417501493,
+      "learning_rate": 8.325232916769477e-06,
+      "loss": 0.1811,
+      "step": 3191
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.155834474540361,
+      "learning_rate": 8.324132224795453e-06,
+      "loss": 0.2041,
+      "step": 3192
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2093705082676047,
+      "learning_rate": 8.323031244056058e-06,
+      "loss": 0.1858,
+      "step": 3193
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.470273888322643,
+      "learning_rate": 8.321929974646936e-06,
+      "loss": 0.1899,
+      "step": 3194
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7339549245730934,
+      "learning_rate": 8.320828416663753e-06,
+      "loss": 0.1525,
+      "step": 3195
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9586981936767238,
+      "learning_rate": 8.319726570202201e-06,
+      "loss": 0.151,
+      "step": 3196
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.363058366062218,
+      "learning_rate": 8.318624435357995e-06,
+      "loss": 0.1842,
+      "step": 3197
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.801335386185857,
+      "learning_rate": 8.31752201222688e-06,
+      "loss": 0.2212,
+      "step": 3198
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1806504636883726,
+      "learning_rate": 8.316419300904622e-06,
+      "loss": 0.1885,
+      "step": 3199
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.45678693859772,
+      "learning_rate": 8.315316301487009e-06,
+      "loss": 0.2254,
+      "step": 3200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3243264293220367,
+      "learning_rate": 8.31421301406986e-06,
+      "loss": 0.1747,
+      "step": 3201
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4487585774361067,
+      "learning_rate": 8.313109438749021e-06,
+      "loss": 0.1936,
+      "step": 3202
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.093960751141969,
+      "learning_rate": 8.312005575620355e-06,
+      "loss": 0.13,
+      "step": 3203
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4272617352385084,
+      "learning_rate": 8.310901424779752e-06,
+      "loss": 0.1932,
+      "step": 3204
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.201052640840292,
+      "learning_rate": 8.309796986323135e-06,
+      "loss": 0.1998,
+      "step": 3205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.394518667058749,
+      "learning_rate": 8.308692260346439e-06,
+      "loss": 0.1889,
+      "step": 3206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7775850170271346,
+      "learning_rate": 8.307587246945636e-06,
+      "loss": 0.1477,
+      "step": 3207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.134855455428797,
+      "learning_rate": 8.306481946216716e-06,
+      "loss": 0.1662,
+      "step": 3208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4727849088191625,
+      "learning_rate": 8.305376358255695e-06,
+      "loss": 0.2219,
+      "step": 3209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1528447216291586,
+      "learning_rate": 8.304270483158617e-06,
+      "loss": 0.1931,
+      "step": 3210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3480119030798785,
+      "learning_rate": 8.303164321021547e-06,
+      "loss": 0.1877,
+      "step": 3211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0817444970144234,
+      "learning_rate": 8.302057871940577e-06,
+      "loss": 0.1733,
+      "step": 3212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2849706467352564,
+      "learning_rate": 8.300951136011824e-06,
+      "loss": 0.2033,
+      "step": 3213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0795125906451535,
+      "learning_rate": 8.299844113331428e-06,
+      "loss": 0.1777,
+      "step": 3214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9746048251693091,
+      "learning_rate": 8.298736803995558e-06,
+      "loss": 0.1653,
+      "step": 3215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.32948345963972,
+      "learning_rate": 8.297629208100402e-06,
+      "loss": 0.1799,
+      "step": 3216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1454475713257284,
+      "learning_rate": 8.296521325742178e-06,
+      "loss": 0.1797,
+      "step": 3217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.29301092899836,
+      "learning_rate": 8.295413157017127e-06,
+      "loss": 0.195,
+      "step": 3218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.150827013458793,
+      "learning_rate": 8.294304702021515e-06,
+      "loss": 0.1692,
+      "step": 3219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1776073753888765,
+      "learning_rate": 8.293195960851634e-06,
+      "loss": 0.1796,
+      "step": 3220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2560544368660844,
+      "learning_rate": 8.292086933603799e-06,
+      "loss": 0.2095,
+      "step": 3221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.187506780002228,
+      "learning_rate": 8.290977620374348e-06,
+      "loss": 0.1972,
+      "step": 3222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.357901176192247,
+      "learning_rate": 8.28986802125965e-06,
+      "loss": 0.1953,
+      "step": 3223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2398928673990413,
+      "learning_rate": 8.288758136356093e-06,
+      "loss": 0.1966,
+      "step": 3224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8708969413786636,
+      "learning_rate": 8.287647965760092e-06,
+      "loss": 0.2051,
+      "step": 3225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.076145330091136,
+      "learning_rate": 8.28653750956809e-06,
+      "loss": 0.1505,
+      "step": 3226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4300211893441044,
+      "learning_rate": 8.285426767876546e-06,
+      "loss": 0.1803,
+      "step": 3227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.533166184322002,
+      "learning_rate": 8.284315740781953e-06,
+      "loss": 0.2261,
+      "step": 3228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0927025673180455,
+      "learning_rate": 8.283204428380826e-06,
+      "loss": 0.1569,
+      "step": 3229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1302617408923785,
+      "learning_rate": 8.282092830769703e-06,
+      "loss": 0.1704,
+      "step": 3230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2232777757208675,
+      "learning_rate": 8.280980948045146e-06,
+      "loss": 0.1766,
+      "step": 3231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3813699646542235,
+      "learning_rate": 8.279868780303745e-06,
+      "loss": 0.192,
+      "step": 3232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2768997991403905,
+      "learning_rate": 8.278756327642116e-06,
+      "loss": 0.1874,
+      "step": 3233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2832862280711472,
+      "learning_rate": 8.277643590156893e-06,
+      "loss": 0.2008,
+      "step": 3234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.064851399007841,
+      "learning_rate": 8.276530567944742e-06,
+      "loss": 0.172,
+      "step": 3235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.410946951271658,
+      "learning_rate": 8.27541726110235e-06,
+      "loss": 0.1931,
+      "step": 3236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4255657343587296,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 0.2167,
+      "step": 3237
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.148485397334108,
+      "learning_rate": 8.273189793913711e-06,
+      "loss": 0.1732,
+      "step": 3238
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3862713143435936,
+      "learning_rate": 8.272075633760966e-06,
+      "loss": 0.2084,
+      "step": 3239
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5697572088737193,
+      "learning_rate": 8.270961189364974e-06,
+      "loss": 0.1892,
+      "step": 3240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.223355889643228,
+      "learning_rate": 8.26984646082255e-06,
+      "loss": 0.1754,
+      "step": 3241
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1495131546296182,
+      "learning_rate": 8.268731448230527e-06,
+      "loss": 0.1737,
+      "step": 3242
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.13242767275266,
+      "learning_rate": 8.267616151685768e-06,
+      "loss": 0.1817,
+      "step": 3243
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2890027917155145,
+      "learning_rate": 8.266500571285159e-06,
+      "loss": 0.1949,
+      "step": 3244
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.500480168591124,
+      "learning_rate": 8.265384707125607e-06,
+      "loss": 0.219,
+      "step": 3245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.085684452421729,
+      "learning_rate": 8.264268559304046e-06,
+      "loss": 0.1745,
+      "step": 3246
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.136115020302101,
+      "learning_rate": 8.263152127917438e-06,
+      "loss": 0.1718,
+      "step": 3247
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.063594587049693,
+      "learning_rate": 8.262035413062763e-06,
+      "loss": 0.1604,
+      "step": 3248
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3298035804113724,
+      "learning_rate": 8.260918414837034e-06,
+      "loss": 0.1806,
+      "step": 3249
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.088763092076401,
+      "learning_rate": 8.25980113333728e-06,
+      "loss": 0.1824,
+      "step": 3250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.338366015080485,
+      "learning_rate": 8.258683568660561e-06,
+      "loss": 0.1972,
+      "step": 3251
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.462788790077738,
+      "learning_rate": 8.257565720903957e-06,
+      "loss": 0.1773,
+      "step": 3252
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1223564596571878,
+      "learning_rate": 8.256447590164576e-06,
+      "loss": 0.1964,
+      "step": 3253
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.391580740908043,
+      "learning_rate": 8.255329176539552e-06,
+      "loss": 0.1953,
+      "step": 3254
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.210927125449662,
+      "learning_rate": 8.254210480126036e-06,
+      "loss": 0.1922,
+      "step": 3255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.053793442990861,
+      "learning_rate": 8.25309150102121e-06,
+      "loss": 0.1543,
+      "step": 3256
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0110743721123066,
+      "learning_rate": 8.251972239322283e-06,
+      "loss": 0.159,
+      "step": 3257
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.064272753404483,
+      "learning_rate": 8.250852695126478e-06,
+      "loss": 0.16,
+      "step": 3258
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.36416064058127,
+      "learning_rate": 8.249732868531056e-06,
+      "loss": 0.1903,
+      "step": 3259
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2913187130041472,
+      "learning_rate": 8.24861275963329e-06,
+      "loss": 0.2064,
+      "step": 3260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3229313929607476,
+      "learning_rate": 8.247492368530485e-06,
+      "loss": 0.186,
+      "step": 3261
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2712649861955145,
+      "learning_rate": 8.246371695319968e-06,
+      "loss": 0.1759,
+      "step": 3262
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2843930624863416,
+      "learning_rate": 8.245250740099095e-06,
+      "loss": 0.18,
+      "step": 3263
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2399401489642865,
+      "learning_rate": 8.244129502965239e-06,
+      "loss": 0.1816,
+      "step": 3264
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.358125578004363,
+      "learning_rate": 8.243007984015801e-06,
+      "loss": 0.2038,
+      "step": 3265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.540874782184986,
+      "learning_rate": 8.24188618334821e-06,
+      "loss": 0.2072,
+      "step": 3266
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5915999466287727,
+      "learning_rate": 8.240764101059913e-06,
+      "loss": 0.2011,
+      "step": 3267
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.140998716900764,
+      "learning_rate": 8.239641737248386e-06,
+      "loss": 0.1814,
+      "step": 3268
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1999473038819537,
+      "learning_rate": 8.238519092011125e-06,
+      "loss": 0.1835,
+      "step": 3269
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0586290044788584,
+      "learning_rate": 8.237396165445661e-06,
+      "loss": 0.1557,
+      "step": 3270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.140028167761064,
+      "learning_rate": 8.236272957649534e-06,
+      "loss": 0.1545,
+      "step": 3271
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.474330908193338,
+      "learning_rate": 8.23514946872032e-06,
+      "loss": 0.1623,
+      "step": 3272
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.435467950221756,
+      "learning_rate": 8.234025698755616e-06,
+      "loss": 0.1792,
+      "step": 3273
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2334372698809437,
+      "learning_rate": 8.232901647853043e-06,
+      "loss": 0.1605,
+      "step": 3274
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1493466075789187,
+      "learning_rate": 8.231777316110245e-06,
+      "loss": 0.1803,
+      "step": 3275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.003315284589848,
+      "learning_rate": 8.230652703624893e-06,
+      "loss": 0.1557,
+      "step": 3276
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.043228461711188,
+      "learning_rate": 8.229527810494682e-06,
+      "loss": 0.1599,
+      "step": 3277
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9786669354200614,
+      "learning_rate": 8.228402636817331e-06,
+      "loss": 0.1461,
+      "step": 3278
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2086720767807426,
+      "learning_rate": 8.227277182690582e-06,
+      "loss": 0.1667,
+      "step": 3279
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3602797058448477,
+      "learning_rate": 8.226151448212202e-06,
+      "loss": 0.1797,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.225265777224894,
+      "learning_rate": 8.225025433479987e-06,
+      "loss": 0.1833,
+      "step": 3281
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2059809692911934,
+      "learning_rate": 8.22389913859175e-06,
+      "loss": 0.1687,
+      "step": 3282
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.159812376875455,
+      "learning_rate": 8.222772563645329e-06,
+      "loss": 0.184,
+      "step": 3283
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.228120479440467,
+      "learning_rate": 8.221645708738594e-06,
+      "loss": 0.1806,
+      "step": 3284
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4583856589863826,
+      "learning_rate": 8.220518573969432e-06,
+      "loss": 0.1798,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.191022395410361,
+      "learning_rate": 8.219391159435755e-06,
+      "loss": 0.1828,
+      "step": 3286
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.356795918174125,
+      "learning_rate": 8.218263465235502e-06,
+      "loss": 0.2138,
+      "step": 3287
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4955424270318827,
+      "learning_rate": 8.217135491466636e-06,
+      "loss": 0.1955,
+      "step": 3288
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.156001927303675,
+      "learning_rate": 8.216007238227142e-06,
+      "loss": 0.1834,
+      "step": 3289
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1896506011684567,
+      "learning_rate": 8.214878705615033e-06,
+      "loss": 0.1682,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.382573553593165,
+      "learning_rate": 8.213749893728342e-06,
+      "loss": 0.1453,
+      "step": 3291
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.148992685201197,
+      "learning_rate": 8.212620802665127e-06,
+      "loss": 0.1795,
+      "step": 3292
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.251030436578612,
+      "learning_rate": 8.211491432523474e-06,
+      "loss": 0.1511,
+      "step": 3293
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2823839686944583,
+      "learning_rate": 8.210361783401491e-06,
+      "loss": 0.187,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2377797315918566,
+      "learning_rate": 8.209231855397309e-06,
+      "loss": 0.1606,
+      "step": 3295
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3382576365493137,
+      "learning_rate": 8.208101648609082e-06,
+      "loss": 0.1841,
+      "step": 3296
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.235197394522637,
+      "learning_rate": 8.206971163134992e-06,
+      "loss": 0.1892,
+      "step": 3297
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5282837116080947,
+      "learning_rate": 8.205840399073245e-06,
+      "loss": 0.183,
+      "step": 3298
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4616697894816446,
+      "learning_rate": 8.204709356522069e-06,
+      "loss": 0.1895,
+      "step": 3299
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.434362636902406,
+      "learning_rate": 8.203578035579716e-06,
+      "loss": 0.1662,
+      "step": 3300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3371972078139636,
+      "learning_rate": 8.202446436344463e-06,
+      "loss": 0.1983,
+      "step": 3301
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.435275274808842,
+      "learning_rate": 8.201314558914613e-06,
+      "loss": 0.1979,
+      "step": 3302
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4478602119627815,
+      "learning_rate": 8.20018240338849e-06,
+      "loss": 0.2396,
+      "step": 3303
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.625698385241079,
+      "learning_rate": 8.199049969864445e-06,
+      "loss": 0.2171,
+      "step": 3304
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.505762837331745,
+      "learning_rate": 8.197917258440851e-06,
+      "loss": 0.1549,
+      "step": 3305
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0453110101659964,
+      "learning_rate": 8.196784269216107e-06,
+      "loss": 0.1789,
+      "step": 3306
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2866626290225827,
+      "learning_rate": 8.195651002288633e-06,
+      "loss": 0.1898,
+      "step": 3307
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.283789599413208,
+      "learning_rate": 8.194517457756877e-06,
+      "loss": 0.1957,
+      "step": 3308
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4412990804505323,
+      "learning_rate": 8.193383635719308e-06,
+      "loss": 0.1838,
+      "step": 3309
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.138029238240916,
+      "learning_rate": 8.192249536274421e-06,
+      "loss": 0.1817,
+      "step": 3310
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.324398080938735,
+      "learning_rate": 8.191115159520735e-06,
+      "loss": 0.211,
+      "step": 3311
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2155668721038135,
+      "learning_rate": 8.189980505556793e-06,
+      "loss": 0.1935,
+      "step": 3312
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4183581811832786,
+      "learning_rate": 8.188845574481162e-06,
+      "loss": 0.1902,
+      "step": 3313
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.38784633428729,
+      "learning_rate": 8.187710366392431e-06,
+      "loss": 0.1862,
+      "step": 3314
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.426471563918452,
+      "learning_rate": 8.186574881389216e-06,
+      "loss": 0.2169,
+      "step": 3315
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6638006352140082,
+      "learning_rate": 8.185439119570154e-06,
+      "loss": 0.2402,
+      "step": 3316
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.112580427314221,
+      "learning_rate": 8.184303081033911e-06,
+      "loss": 0.1782,
+      "step": 3317
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.156329305632411,
+      "learning_rate": 8.183166765879171e-06,
+      "loss": 0.1624,
+      "step": 3318
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.425715856479946,
+      "learning_rate": 8.182030174204648e-06,
+      "loss": 0.222,
+      "step": 3319
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2116814000944442,
+      "learning_rate": 8.180893306109075e-06,
+      "loss": 0.1789,
+      "step": 3320
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.359940770493828,
+      "learning_rate": 8.179756161691212e-06,
+      "loss": 0.1748,
+      "step": 3321
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0130173846981254,
+      "learning_rate": 8.178618741049841e-06,
+      "loss": 0.1541,
+      "step": 3322
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.28094935573533,
+      "learning_rate": 8.17748104428377e-06,
+      "loss": 0.1834,
+      "step": 3323
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3078207516293694,
+      "learning_rate": 8.17634307149183e-06,
+      "loss": 0.1969,
+      "step": 3324
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2269701877737553,
+      "learning_rate": 8.175204822772875e-06,
+      "loss": 0.1831,
+      "step": 3325
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2811932793283645,
+      "learning_rate": 8.174066298225785e-06,
+      "loss": 0.1894,
+      "step": 3326
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.157994459926913,
+      "learning_rate": 8.172927497949463e-06,
+      "loss": 0.1875,
+      "step": 3327
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0926735948127244,
+      "learning_rate": 8.171788422042837e-06,
+      "loss": 0.1723,
+      "step": 3328
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1976119524902717,
+      "learning_rate": 8.170649070604855e-06,
+      "loss": 0.1886,
+      "step": 3329
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0687673978670937,
+      "learning_rate": 8.169509443734493e-06,
+      "loss": 0.1873,
+      "step": 3330
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1869805702470875,
+      "learning_rate": 8.16836954153075e-06,
+      "loss": 0.1958,
+      "step": 3331
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.303212673810355,
+      "learning_rate": 8.167229364092648e-06,
+      "loss": 0.1681,
+      "step": 3332
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2794315472409097,
+      "learning_rate": 8.166088911519236e-06,
+      "loss": 0.1804,
+      "step": 3333
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.182459423048527,
+      "learning_rate": 8.16494818390958e-06,
+      "loss": 0.1834,
+      "step": 3334
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1105704925939714,
+      "learning_rate": 8.163807181362778e-06,
+      "loss": 0.1731,
+      "step": 3335
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3344840179572337,
+      "learning_rate": 8.162665903977947e-06,
+      "loss": 0.2257,
+      "step": 3336
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.178533704416329,
+      "learning_rate": 8.161524351854229e-06,
+      "loss": 0.1764,
+      "step": 3337
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2327794298956363,
+      "learning_rate": 8.16038252509079e-06,
+      "loss": 0.1864,
+      "step": 3338
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.22333040777405,
+      "learning_rate": 8.15924042378682e-06,
+      "loss": 0.1708,
+      "step": 3339
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.325766604148024,
+      "learning_rate": 8.158098048041534e-06,
+      "loss": 0.196,
+      "step": 3340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.127296045212197,
+      "learning_rate": 8.156955397954166e-06,
+      "loss": 0.1844,
+      "step": 3341
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2199497615834316,
+      "learning_rate": 8.15581247362398e-06,
+      "loss": 0.203,
+      "step": 3342
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3653662688952943,
+      "learning_rate": 8.154669275150259e-06,
+      "loss": 0.1887,
+      "step": 3343
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8745965289880373,
+      "learning_rate": 8.153525802632314e-06,
+      "loss": 0.1612,
+      "step": 3344
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.117495794730123,
+      "learning_rate": 8.15238205616948e-06,
+      "loss": 0.153,
+      "step": 3345
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.037321887315161,
+      "learning_rate": 8.151238035861108e-06,
+      "loss": 0.1584,
+      "step": 3346
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.196284899438368,
+      "learning_rate": 8.150093741806582e-06,
+      "loss": 0.1971,
+      "step": 3347
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.204382918961807,
+      "learning_rate": 8.148949174105305e-06,
+      "loss": 0.1787,
+      "step": 3348
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.092713372572776,
+      "learning_rate": 8.147804332856705e-06,
+      "loss": 0.162,
+      "step": 3349
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3956935239184567,
+      "learning_rate": 8.146659218160233e-06,
+      "loss": 0.1915,
+      "step": 3350
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.046048622108676,
+      "learning_rate": 8.145513830115367e-06,
+      "loss": 0.194,
+      "step": 3351
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.298114634723322,
+      "learning_rate": 8.144368168821603e-06,
+      "loss": 0.1856,
+      "step": 3352
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.17841290441522,
+      "learning_rate": 8.143222234378467e-06,
+      "loss": 0.1747,
+      "step": 3353
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.191878202826198,
+      "learning_rate": 8.142076026885504e-06,
+      "loss": 0.1712,
+      "step": 3354
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.377293506076023,
+      "learning_rate": 8.140929546442282e-06,
+      "loss": 0.2064,
+      "step": 3355
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1353715101449913,
+      "learning_rate": 8.1397827931484e-06,
+      "loss": 0.1638,
+      "step": 3356
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.33039694183237,
+      "learning_rate": 8.13863576710347e-06,
+      "loss": 0.1741,
+      "step": 3357
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3258499626196913,
+      "learning_rate": 8.13748846840714e-06,
+      "loss": 0.2203,
+      "step": 3358
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3207823728598234,
+      "learning_rate": 8.136340897159071e-06,
+      "loss": 0.1852,
+      "step": 3359
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2941575772141247,
+      "learning_rate": 8.135193053458952e-06,
+      "loss": 0.1817,
+      "step": 3360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0603519800074728,
+      "learning_rate": 8.134044937406496e-06,
+      "loss": 0.1469,
+      "step": 3361
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5186651452807243,
+      "learning_rate": 8.13289654910144e-06,
+      "loss": 0.1841,
+      "step": 3362
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.162276033773308,
+      "learning_rate": 8.131747888643541e-06,
+      "loss": 0.2057,
+      "step": 3363
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9764896077994298,
+      "learning_rate": 8.130598956132587e-06,
+      "loss": 0.1523,
+      "step": 3364
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.330984831901305,
+      "learning_rate": 8.129449751668382e-06,
+      "loss": 0.1965,
+      "step": 3365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1995998810972517,
+      "learning_rate": 8.128300275350756e-06,
+      "loss": 0.1759,
+      "step": 3366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0550206928753614,
+      "learning_rate": 8.127150527279565e-06,
+      "loss": 0.1602,
+      "step": 3367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.200589011616952,
+      "learning_rate": 8.126000507554688e-06,
+      "loss": 0.1838,
+      "step": 3368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1442422115987174,
+      "learning_rate": 8.124850216276023e-06,
+      "loss": 0.2015,
+      "step": 3369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3846688878144655,
+      "learning_rate": 8.1236996535435e-06,
+      "loss": 0.2376,
+      "step": 3370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.262346544704719,
+      "learning_rate": 8.122548819457063e-06,
+      "loss": 0.1841,
+      "step": 3371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1317126817830725,
+      "learning_rate": 8.121397714116686e-06,
+      "loss": 0.1769,
+      "step": 3372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1110246032851303,
+      "learning_rate": 8.120246337622364e-06,
+      "loss": 0.1797,
+      "step": 3373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0968224383971763,
+      "learning_rate": 8.119094690074119e-06,
+      "loss": 0.1717,
+      "step": 3374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0633835379639307,
+      "learning_rate": 8.117942771571992e-06,
+      "loss": 0.1776,
+      "step": 3375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.343492438496204,
+      "learning_rate": 8.11679058221605e-06,
+      "loss": 0.1941,
+      "step": 3376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1212228457088727,
+      "learning_rate": 8.115638122106382e-06,
+      "loss": 0.2008,
+      "step": 3377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1254516867786184,
+      "learning_rate": 8.114485391343102e-06,
+      "loss": 0.1267,
+      "step": 3378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2565782996683383,
+      "learning_rate": 8.113332390026348e-06,
+      "loss": 0.2063,
+      "step": 3379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.08004835438089,
+      "learning_rate": 8.11217911825628e-06,
+      "loss": 0.1941,
+      "step": 3380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0275372993617298,
+      "learning_rate": 8.11102557613308e-06,
+      "loss": 0.166,
+      "step": 3381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3199502608216522,
+      "learning_rate": 8.10987176375696e-06,
+      "loss": 0.1903,
+      "step": 3382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0210505290996137,
+      "learning_rate": 8.108717681228146e-06,
+      "loss": 0.1744,
+      "step": 3383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0014804385007214,
+      "learning_rate": 8.107563328646897e-06,
+      "loss": 0.1987,
+      "step": 3384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2547966612495665,
+      "learning_rate": 8.106408706113486e-06,
+      "loss": 0.2141,
+      "step": 3385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1260768948327278,
+      "learning_rate": 8.10525381372822e-06,
+      "loss": 0.1909,
+      "step": 3386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9969525668783519,
+      "learning_rate": 8.104098651591418e-06,
+      "loss": 0.1608,
+      "step": 3387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.231321357827387,
+      "learning_rate": 8.102943219803433e-06,
+      "loss": 0.1844,
+      "step": 3388
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9586615143030417,
+      "learning_rate": 8.101787518464634e-06,
+      "loss": 0.1747,
+      "step": 3389
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3204957236827197,
+      "learning_rate": 8.100631547675417e-06,
+      "loss": 0.198,
+      "step": 3390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.053834824641304,
+      "learning_rate": 8.0994753075362e-06,
+      "loss": 0.168,
+      "step": 3391
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1464494303607586,
+      "learning_rate": 8.098318798147426e-06,
+      "loss": 0.149,
+      "step": 3392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4694733791375962,
+      "learning_rate": 8.097162019609562e-06,
+      "loss": 0.2131,
+      "step": 3393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2703502085464873,
+      "learning_rate": 8.09600497202309e-06,
+      "loss": 0.1768,
+      "step": 3394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2156428920251434,
+      "learning_rate": 8.094847655488528e-06,
+      "loss": 0.2003,
+      "step": 3395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1656901559941146,
+      "learning_rate": 8.09369007010641e-06,
+      "loss": 0.1946,
+      "step": 3396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.625584483631351,
+      "learning_rate": 8.092532215977293e-06,
+      "loss": 0.1912,
+      "step": 3397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.58355268290637,
+      "learning_rate": 8.09137409320176e-06,
+      "loss": 0.2121,
+      "step": 3398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.361937405251314,
+      "learning_rate": 8.090215701880418e-06,
+      "loss": 0.1655,
+      "step": 3399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9789854426305598,
+      "learning_rate": 8.089057042113895e-06,
+      "loss": 0.1502,
+      "step": 3400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2216580271250836,
+      "learning_rate": 8.087898114002842e-06,
+      "loss": 0.1795,
+      "step": 3401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0856964605391375,
+      "learning_rate": 8.086738917647937e-06,
+      "loss": 0.1378,
+      "step": 3402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.020846775830208,
+      "learning_rate": 8.085579453149874e-06,
+      "loss": 0.1549,
+      "step": 3403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2020348895434685,
+      "learning_rate": 8.084419720609377e-06,
+      "loss": 0.1748,
+      "step": 3404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1278798565663997,
+      "learning_rate": 8.083259720127195e-06,
+      "loss": 0.1756,
+      "step": 3405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5077072419121653,
+      "learning_rate": 8.082099451804093e-06,
+      "loss": 0.1605,
+      "step": 3406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2672746401792714,
+      "learning_rate": 8.080938915740863e-06,
+      "loss": 0.1736,
+      "step": 3407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2788171992700734,
+      "learning_rate": 8.079778112038318e-06,
+      "loss": 0.1598,
+      "step": 3408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2866941428106684,
+      "learning_rate": 8.078617040797304e-06,
+      "loss": 0.1981,
+      "step": 3409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.13636483216342,
+      "learning_rate": 8.077455702118673e-06,
+      "loss": 0.1578,
+      "step": 3410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2036606064180004,
+      "learning_rate": 8.076294096103316e-06,
+      "loss": 0.1739,
+      "step": 3411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.043876503178284,
+      "learning_rate": 8.075132222852138e-06,
+      "loss": 0.1709,
+      "step": 3412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1230623903068997,
+      "learning_rate": 8.073970082466071e-06,
+      "loss": 0.1859,
+      "step": 3413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1414403624526486,
+      "learning_rate": 8.072807675046073e-06,
+      "loss": 0.1707,
+      "step": 3414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0688715663520507,
+      "learning_rate": 8.071645000693116e-06,
+      "loss": 0.1678,
+      "step": 3415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1794382570736275,
+      "learning_rate": 8.070482059508202e-06,
+      "loss": 0.2066,
+      "step": 3416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9796256375435533,
+      "learning_rate": 8.069318851592358e-06,
+      "loss": 0.1688,
+      "step": 3417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1099252518864406,
+      "learning_rate": 8.068155377046629e-06,
+      "loss": 0.1614,
+      "step": 3418
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.178794028016435,
+      "learning_rate": 8.066991635972087e-06,
+      "loss": 0.172,
+      "step": 3419
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0753428201336597,
+      "learning_rate": 8.065827628469823e-06,
+      "loss": 0.1825,
+      "step": 3420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0694085117241268,
+      "learning_rate": 8.064663354640956e-06,
+      "loss": 0.159,
+      "step": 3421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.328959460484489,
+      "learning_rate": 8.063498814586623e-06,
+      "loss": 0.1977,
+      "step": 3422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9669738818523599,
+      "learning_rate": 8.06233400840799e-06,
+      "loss": 0.159,
+      "step": 3423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1442275564750948,
+      "learning_rate": 8.06116893620624e-06,
+      "loss": 0.1666,
+      "step": 3424
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2924947982463078,
+      "learning_rate": 8.060003598082587e-06,
+      "loss": 0.166,
+      "step": 3425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.134976702211855,
+      "learning_rate": 8.058837994138256e-06,
+      "loss": 0.1493,
+      "step": 3426
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2141414172383787,
+      "learning_rate": 8.057672124474508e-06,
+      "loss": 0.1676,
+      "step": 3427
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.212568054359841,
+      "learning_rate": 8.05650598919262e-06,
+      "loss": 0.1552,
+      "step": 3428
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2876868730832722,
+      "learning_rate": 8.055339588393892e-06,
+      "loss": 0.1708,
+      "step": 3429
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.239595571033942,
+      "learning_rate": 8.05417292217965e-06,
+      "loss": 0.1888,
+      "step": 3430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.070562564412396,
+      "learning_rate": 8.053005990651242e-06,
+      "loss": 0.1446,
+      "step": 3431
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.16333935420736,
+      "learning_rate": 8.051838793910038e-06,
+      "loss": 0.1709,
+      "step": 3432
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1235719033835774,
+      "learning_rate": 8.05067133205743e-06,
+      "loss": 0.1644,
+      "step": 3433
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2918465911043255,
+      "learning_rate": 8.049503605194837e-06,
+      "loss": 0.1693,
+      "step": 3434
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2404753228145404,
+      "learning_rate": 8.0483356134237e-06,
+      "loss": 0.1726,
+      "step": 3435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0528987096696114,
+      "learning_rate": 8.047167356845475e-06,
+      "loss": 0.1615,
+      "step": 3436
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0699111281915896,
+      "learning_rate": 8.045998835561656e-06,
+      "loss": 0.1408,
+      "step": 3437
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2646101158294734,
+      "learning_rate": 8.04483004967375e-06,
+      "loss": 0.1653,
+      "step": 3438
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.211999081533399,
+      "learning_rate": 8.043660999283282e-06,
+      "loss": 0.1941,
+      "step": 3439
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2064024177081976,
+      "learning_rate": 8.042491684491816e-06,
+      "loss": 0.1997,
+      "step": 3440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0289132652750563,
+      "learning_rate": 8.041322105400923e-06,
+      "loss": 0.1471,
+      "step": 3441
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1208623405231006,
+      "learning_rate": 8.040152262112206e-06,
+      "loss": 0.1798,
+      "step": 3442
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.219163433886025,
+      "learning_rate": 8.038982154727288e-06,
+      "loss": 0.1723,
+      "step": 3443
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2545709877848497,
+      "learning_rate": 8.03781178334782e-06,
+      "loss": 0.1752,
+      "step": 3444
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.314027216226458,
+      "learning_rate": 8.036641148075463e-06,
+      "loss": 0.2001,
+      "step": 3445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1919804614253366,
+      "learning_rate": 8.035470249011916e-06,
+      "loss": 0.1817,
+      "step": 3446
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1350167097930974,
+      "learning_rate": 8.034299086258892e-06,
+      "loss": 0.1505,
+      "step": 3447
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1082956861360125,
+      "learning_rate": 8.03312765991813e-06,
+      "loss": 0.1474,
+      "step": 3448
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.469382675126546,
+      "learning_rate": 8.031955970091389e-06,
+      "loss": 0.1539,
+      "step": 3449
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.347055738968163,
+      "learning_rate": 8.030784016880456e-06,
+      "loss": 0.1952,
+      "step": 3450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.292347892682746,
+      "learning_rate": 8.029611800387134e-06,
+      "loss": 0.1701,
+      "step": 3451
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.36423604990879,
+      "learning_rate": 8.028439320713256e-06,
+      "loss": 0.1809,
+      "step": 3452
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1421634728899854,
+      "learning_rate": 8.027266577960676e-06,
+      "loss": 0.1847,
+      "step": 3453
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9370081267153783,
+      "learning_rate": 8.026093572231266e-06,
+      "loss": 0.1421,
+      "step": 3454
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1827786032604166,
+      "learning_rate": 8.024920303626925e-06,
+      "loss": 0.1753,
+      "step": 3455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.333584098516215,
+      "learning_rate": 8.023746772249574e-06,
+      "loss": 0.1992,
+      "step": 3456
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0367031019980093,
+      "learning_rate": 8.02257297820116e-06,
+      "loss": 0.1793,
+      "step": 3457
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.26262519575056,
+      "learning_rate": 8.021398921583644e-06,
+      "loss": 0.1834,
+      "step": 3458
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.108617365234379,
+      "learning_rate": 8.020224602499024e-06,
+      "loss": 0.1538,
+      "step": 3459
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4657179891321555,
+      "learning_rate": 8.019050021049303e-06,
+      "loss": 0.1867,
+      "step": 3460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9461986777864206,
+      "learning_rate": 8.017875177336522e-06,
+      "loss": 0.1787,
+      "step": 3461
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9401714096142866,
+      "learning_rate": 8.016700071462736e-06,
+      "loss": 0.1564,
+      "step": 3462
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1614582313937976,
+      "learning_rate": 8.015524703530028e-06,
+      "loss": 0.1725,
+      "step": 3463
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2348289930688177,
+      "learning_rate": 8.014349073640504e-06,
+      "loss": 0.2069,
+      "step": 3464
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0380167889262735,
+      "learning_rate": 8.013173181896283e-06,
+      "loss": 0.1764,
+      "step": 3465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9548494264634033,
+      "learning_rate": 8.011997028399518e-06,
+      "loss": 0.1635,
+      "step": 3466
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.183734264974981,
+      "learning_rate": 8.010820613252383e-06,
+      "loss": 0.1786,
+      "step": 3467
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0667665607313888,
+      "learning_rate": 8.00964393655707e-06,
+      "loss": 0.1738,
+      "step": 3468
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2817955488929207,
+      "learning_rate": 8.008466998415795e-06,
+      "loss": 0.1822,
+      "step": 3469
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1144768277086348,
+      "learning_rate": 8.0072897989308e-06,
+      "loss": 0.1683,
+      "step": 3470
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.283772571639778,
+      "learning_rate": 8.006112338204348e-06,
+      "loss": 0.2007,
+      "step": 3471
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0776714690143367,
+      "learning_rate": 8.004934616338721e-06,
+      "loss": 0.1766,
+      "step": 3472
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3441820787104968,
+      "learning_rate": 8.003756633436233e-06,
+      "loss": 0.2139,
+      "step": 3473
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.106388984871136,
+      "learning_rate": 8.002578389599208e-06,
+      "loss": 0.1665,
+      "step": 3474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.155528158851314,
+      "learning_rate": 8.001399884930004e-06,
+      "loss": 0.1392,
+      "step": 3475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4168729617208884,
+      "learning_rate": 8.000221119530993e-06,
+      "loss": 0.2105,
+      "step": 3476
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7080133395427586,
+      "learning_rate": 7.999042093504578e-06,
+      "loss": 0.1258,
+      "step": 3477
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9973846088719347,
+      "learning_rate": 7.997862806953177e-06,
+      "loss": 0.1594,
+      "step": 3478
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.198990238001898,
+      "learning_rate": 7.996683259979237e-06,
+      "loss": 0.1759,
+      "step": 3479
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1789223569845575,
+      "learning_rate": 7.99550345268522e-06,
+      "loss": 0.1672,
+      "step": 3480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6053157771372377,
+      "learning_rate": 7.994323385173618e-06,
+      "loss": 0.1711,
+      "step": 3481
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1243318283349617,
+      "learning_rate": 7.993143057546943e-06,
+      "loss": 0.1467,
+      "step": 3482
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3651409279697218,
+      "learning_rate": 7.99196246990773e-06,
+      "loss": 0.1786,
+      "step": 3483
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1039675744831974,
+      "learning_rate": 7.990781622358535e-06,
+      "loss": 0.1657,
+      "step": 3484
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3422274905057185,
+      "learning_rate": 7.989600515001936e-06,
+      "loss": 0.1951,
+      "step": 3485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2320673052169613,
+      "learning_rate": 7.988419147940538e-06,
+      "loss": 0.1974,
+      "step": 3486
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.571505983897901,
+      "learning_rate": 7.987237521276962e-06,
+      "loss": 0.2343,
+      "step": 3487
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8619746740621699,
+      "learning_rate": 7.986055635113859e-06,
+      "loss": 0.1595,
+      "step": 3488
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0890751586542735,
+      "learning_rate": 7.984873489553896e-06,
+      "loss": 0.2056,
+      "step": 3489
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3582593618879937,
+      "learning_rate": 7.983691084699768e-06,
+      "loss": 0.1986,
+      "step": 3490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2436576397191965,
+      "learning_rate": 7.982508420654187e-06,
+      "loss": 0.2073,
+      "step": 3491
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.079998338064337,
+      "learning_rate": 7.981325497519892e-06,
+      "loss": 0.1919,
+      "step": 3492
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.931521167618295,
+      "learning_rate": 7.980142315399641e-06,
+      "loss": 0.1518,
+      "step": 3493
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0978927386404465,
+      "learning_rate": 7.978958874396219e-06,
+      "loss": 0.1746,
+      "step": 3494
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2262282445467982,
+      "learning_rate": 7.977775174612427e-06,
+      "loss": 0.1991,
+      "step": 3495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1060791843223274,
+      "learning_rate": 7.976591216151097e-06,
+      "loss": 0.1791,
+      "step": 3496
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.449195918529177,
+      "learning_rate": 7.975406999115077e-06,
+      "loss": 0.1755,
+      "step": 3497
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.180690228956373,
+      "learning_rate": 7.974222523607236e-06,
+      "loss": 0.1876,
+      "step": 3498
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0838083420367175,
+      "learning_rate": 7.973037789730473e-06,
+      "loss": 0.1881,
+      "step": 3499
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1994620223233645,
+      "learning_rate": 7.971852797587703e-06,
+      "loss": 0.1827,
+      "step": 3500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1028586358959234,
+      "learning_rate": 7.970667547281864e-06,
+      "loss": 0.1656,
+      "step": 3501
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.001220486504472,
+      "learning_rate": 7.969482038915924e-06,
+      "loss": 0.1622,
+      "step": 3502
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.593041833952668,
+      "learning_rate": 7.968296272592862e-06,
+      "loss": 0.1961,
+      "step": 3503
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2221717220322943,
+      "learning_rate": 7.967110248415684e-06,
+      "loss": 0.1651,
+      "step": 3504
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0717810340002614,
+      "learning_rate": 7.965923966487423e-06,
+      "loss": 0.1623,
+      "step": 3505
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1646505978563892,
+      "learning_rate": 7.964737426911129e-06,
+      "loss": 0.203,
+      "step": 3506
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.079734266405713,
+      "learning_rate": 7.963550629789875e-06,
+      "loss": 0.1835,
+      "step": 3507
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2339854210790286,
+      "learning_rate": 7.962363575226762e-06,
+      "loss": 0.1775,
+      "step": 3508
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1768851108290446,
+      "learning_rate": 7.961176263324902e-06,
+      "loss": 0.153,
+      "step": 3509
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.299772264652932,
+      "learning_rate": 7.959988694187438e-06,
+      "loss": 0.1692,
+      "step": 3510
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.274121660579695,
+      "learning_rate": 7.958800867917536e-06,
+      "loss": 0.1562,
+      "step": 3511
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1210062952820934,
+      "learning_rate": 7.95761278461838e-06,
+      "loss": 0.1912,
+      "step": 3512
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3986542287361936,
+      "learning_rate": 7.956424444393179e-06,
+      "loss": 0.1834,
+      "step": 3513
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0478773258113043,
+      "learning_rate": 7.955235847345162e-06,
+      "loss": 0.1727,
+      "step": 3514
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.510867699445433,
+      "learning_rate": 7.954046993577585e-06,
+      "loss": 0.1494,
+      "step": 3515
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3300426179512552,
+      "learning_rate": 7.952857883193716e-06,
+      "loss": 0.1902,
+      "step": 3516
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3439448304895225,
+      "learning_rate": 7.95166851629686e-06,
+      "loss": 0.162,
+      "step": 3517
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8758526820004597,
+      "learning_rate": 7.950478892990334e-06,
+      "loss": 0.1461,
+      "step": 3518
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1069712167874437,
+      "learning_rate": 7.949289013377476e-06,
+      "loss": 0.1515,
+      "step": 3519
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2333934607562704,
+      "learning_rate": 7.948098877561657e-06,
+      "loss": 0.1502,
+      "step": 3520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.108495201748499,
+      "learning_rate": 7.946908485646256e-06,
+      "loss": 0.1855,
+      "step": 3521
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0846383761194605,
+      "learning_rate": 7.945717837734688e-06,
+      "loss": 0.1761,
+      "step": 3522
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.023161732682581,
+      "learning_rate": 7.94452693393038e-06,
+      "loss": 0.1369,
+      "step": 3523
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9717515278406028,
+      "learning_rate": 7.943335774336788e-06,
+      "loss": 0.1721,
+      "step": 3524
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0739791365240667,
+      "learning_rate": 7.942144359057385e-06,
+      "loss": 0.1798,
+      "step": 3525
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4664264572135206,
+      "learning_rate": 7.940952688195668e-06,
+      "loss": 0.135,
+      "step": 3526
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.218813882567977,
+      "learning_rate": 7.93976076185516e-06,
+      "loss": 0.1685,
+      "step": 3527
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1221146363517813,
+      "learning_rate": 7.9385685801394e-06,
+      "loss": 0.1751,
+      "step": 3528
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.246879673976531,
+      "learning_rate": 7.937376143151952e-06,
+      "loss": 0.1625,
+      "step": 3529
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8532174437228135,
+      "learning_rate": 7.936183450996402e-06,
+      "loss": 0.1496,
+      "step": 3530
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2308435776722457,
+      "learning_rate": 7.934990503776363e-06,
+      "loss": 0.1859,
+      "step": 3531
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0046124455969823,
+      "learning_rate": 7.933797301595461e-06,
+      "loss": 0.1747,
+      "step": 3532
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2308768703659925,
+      "learning_rate": 7.93260384455735e-06,
+      "loss": 0.2035,
+      "step": 3533
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2799449370538554,
+      "learning_rate": 7.931410132765705e-06,
+      "loss": 0.1943,
+      "step": 3534
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0629483766815184,
+      "learning_rate": 7.930216166324222e-06,
+      "loss": 0.1684,
+      "step": 3535
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.101585602817571,
+      "learning_rate": 7.929021945336622e-06,
+      "loss": 0.1586,
+      "step": 3536
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.395470517501492,
+      "learning_rate": 7.927827469906646e-06,
+      "loss": 0.2002,
+      "step": 3537
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0733325071509,
+      "learning_rate": 7.926632740138056e-06,
+      "loss": 0.1973,
+      "step": 3538
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.173906098320224,
+      "learning_rate": 7.925437756134638e-06,
+      "loss": 0.1849,
+      "step": 3539
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0835824339257663,
+      "learning_rate": 7.9242425180002e-06,
+      "loss": 0.1991,
+      "step": 3540
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.178607895657494,
+      "learning_rate": 7.923047025838573e-06,
+      "loss": 0.184,
+      "step": 3541
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8557686680474144,
+      "learning_rate": 7.921851279753606e-06,
+      "loss": 0.1478,
+      "step": 3542
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.226527726189626,
+      "learning_rate": 7.920655279849173e-06,
+      "loss": 0.1731,
+      "step": 3543
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.263842902523124,
+      "learning_rate": 7.91945902622917e-06,
+      "loss": 0.164,
+      "step": 3544
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.222812893933422,
+      "learning_rate": 7.918262518997517e-06,
+      "loss": 0.1633,
+      "step": 3545
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.345809616476078,
+      "learning_rate": 7.917065758258152e-06,
+      "loss": 0.1384,
+      "step": 3546
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1612431231219293,
+      "learning_rate": 7.915868744115036e-06,
+      "loss": 0.2035,
+      "step": 3547
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1440309679864504,
+      "learning_rate": 7.914671476672156e-06,
+      "loss": 0.1661,
+      "step": 3548
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1923003326283714,
+      "learning_rate": 7.913473956033515e-06,
+      "loss": 0.1897,
+      "step": 3549
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.373787317714028,
+      "learning_rate": 7.912276182303142e-06,
+      "loss": 0.1977,
+      "step": 3550
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.297594420104452,
+      "learning_rate": 7.911078155585086e-06,
+      "loss": 0.186,
+      "step": 3551
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.324221234412241,
+      "learning_rate": 7.909879875983422e-06,
+      "loss": 0.1742,
+      "step": 3552
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.891183263793428,
+      "learning_rate": 7.90868134360224e-06,
+      "loss": 0.1526,
+      "step": 3553
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1037199362594383,
+      "learning_rate": 7.907482558545656e-06,
+      "loss": 0.1878,
+      "step": 3554
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0096736546789993,
+      "learning_rate": 7.90628352091781e-06,
+      "loss": 0.1553,
+      "step": 3555
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.031641656027865,
+      "learning_rate": 7.90508423082286e-06,
+      "loss": 0.1565,
+      "step": 3556
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.185495313637928,
+      "learning_rate": 7.90388468836499e-06,
+      "loss": 0.1759,
+      "step": 3557
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.057603718256187,
+      "learning_rate": 7.9026848936484e-06,
+      "loss": 0.1663,
+      "step": 3558
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0721491998024595,
+      "learning_rate": 7.901484846777318e-06,
+      "loss": 0.1807,
+      "step": 3559
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.052862677469006,
+      "learning_rate": 7.900284547855992e-06,
+      "loss": 0.1592,
+      "step": 3560
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.308106480239381,
+      "learning_rate": 7.899083996988688e-06,
+      "loss": 0.185,
+      "step": 3561
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2031022737955976,
+      "learning_rate": 7.8978831942797e-06,
+      "loss": 0.1883,
+      "step": 3562
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0664145451726816,
+      "learning_rate": 7.89668213983334e-06,
+      "loss": 0.1854,
+      "step": 3563
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.252339097079376,
+      "learning_rate": 7.895480833753942e-06,
+      "loss": 0.1899,
+      "step": 3564
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2414139940806104,
+      "learning_rate": 7.894279276145864e-06,
+      "loss": 0.1687,
+      "step": 3565
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2418496338234584,
+      "learning_rate": 7.893077467113484e-06,
+      "loss": 0.2083,
+      "step": 3566
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0779303500377067,
+      "learning_rate": 7.891875406761203e-06,
+      "loss": 0.1646,
+      "step": 3567
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5537845178608345,
+      "learning_rate": 7.890673095193444e-06,
+      "loss": 0.2314,
+      "step": 3568
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1283538997559455,
+      "learning_rate": 7.889470532514648e-06,
+      "loss": 0.1805,
+      "step": 3569
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.085581684336599,
+      "learning_rate": 7.888267718829283e-06,
+      "loss": 0.1738,
+      "step": 3570
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2679729945442935,
+      "learning_rate": 7.887064654241837e-06,
+      "loss": 0.1628,
+      "step": 3571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0552760794668705,
+      "learning_rate": 7.88586133885682e-06,
+      "loss": 0.1581,
+      "step": 3572
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2094837558584857,
+      "learning_rate": 7.884657772778761e-06,
+      "loss": 0.1704,
+      "step": 3573
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7712533441270153,
+      "learning_rate": 7.883453956112215e-06,
+      "loss": 0.1429,
+      "step": 3574
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2793545454710697,
+      "learning_rate": 7.882249888961755e-06,
+      "loss": 0.183,
+      "step": 3575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.254728717853408,
+      "learning_rate": 7.881045571431982e-06,
+      "loss": 0.1751,
+      "step": 3576
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9783315134459758,
+      "learning_rate": 7.87984100362751e-06,
+      "loss": 0.146,
+      "step": 3577
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2016207543503645,
+      "learning_rate": 7.878636185652977e-06,
+      "loss": 0.1843,
+      "step": 3578
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.393689134300773,
+      "learning_rate": 7.87743111761305e-06,
+      "loss": 0.179,
+      "step": 3579
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4130954387645085,
+      "learning_rate": 7.876225799612413e-06,
+      "loss": 0.1879,
+      "step": 3580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.382150322873826,
+      "learning_rate": 7.875020231755766e-06,
+      "loss": 0.1866,
+      "step": 3581
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1816127050711116,
+      "learning_rate": 7.87381441414784e-06,
+      "loss": 0.1699,
+      "step": 3582
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0412915989688094,
+      "learning_rate": 7.872608346893384e-06,
+      "loss": 0.1451,
+      "step": 3583
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1345226653301457,
+      "learning_rate": 7.871402030097164e-06,
+      "loss": 0.1699,
+      "step": 3584
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2619096872164257,
+      "learning_rate": 7.870195463863976e-06,
+      "loss": 0.205,
+      "step": 3585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3620466324276643,
+      "learning_rate": 7.868988648298632e-06,
+      "loss": 0.2061,
+      "step": 3586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.195944428029587,
+      "learning_rate": 7.867781583505968e-06,
+      "loss": 0.2029,
+      "step": 3587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.085782909047239,
+      "learning_rate": 7.866574269590842e-06,
+      "loss": 0.1839,
+      "step": 3588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.114297515579239,
+      "learning_rate": 7.86536670665813e-06,
+      "loss": 0.1631,
+      "step": 3589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4204376026162704,
+      "learning_rate": 7.864158894812734e-06,
+      "loss": 0.1874,
+      "step": 3590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0932578100402917,
+      "learning_rate": 7.862950834159577e-06,
+      "loss": 0.1759,
+      "step": 3591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9257234083304198,
+      "learning_rate": 7.8617425248036e-06,
+      "loss": 0.1529,
+      "step": 3592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8362723308079476,
+      "learning_rate": 7.86053396684977e-06,
+      "loss": 0.1658,
+      "step": 3593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.092559604046123,
+      "learning_rate": 7.859325160403073e-06,
+      "loss": 0.1885,
+      "step": 3594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.08425924056734,
+      "learning_rate": 7.858116105568515e-06,
+      "loss": 0.1631,
+      "step": 3595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.115557562049383,
+      "learning_rate": 7.856906802451129e-06,
+      "loss": 0.1911,
+      "step": 3596
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0559289771974267,
+      "learning_rate": 7.855697251155967e-06,
+      "loss": 0.1761,
+      "step": 3597
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9382733345003595,
+      "learning_rate": 7.8544874517881e-06,
+      "loss": 0.1667,
+      "step": 3598
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.187024826977775,
+      "learning_rate": 7.853277404452622e-06,
+      "loss": 0.1533,
+      "step": 3599
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4844125425681365,
+      "learning_rate": 7.85206710925465e-06,
+      "loss": 0.2034,
+      "step": 3600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2583852699595086,
+      "learning_rate": 7.850856566299326e-06,
+      "loss": 0.187,
+      "step": 3601
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.240428346124938,
+      "learning_rate": 7.8496457756918e-06,
+      "loss": 0.1903,
+      "step": 3602
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.41276433076512,
+      "learning_rate": 7.848434737537258e-06,
+      "loss": 0.2157,
+      "step": 3603
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.23616051730908,
+      "learning_rate": 7.847223451940903e-06,
+      "loss": 0.2073,
+      "step": 3604
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3841639453354038,
+      "learning_rate": 7.846011919007958e-06,
+      "loss": 0.1879,
+      "step": 3605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.238312520591959,
+      "learning_rate": 7.844800138843667e-06,
+      "loss": 0.1883,
+      "step": 3606
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9405395143229405,
+      "learning_rate": 7.843588111553297e-06,
+      "loss": 0.1697,
+      "step": 3607
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9771168877083594,
+      "learning_rate": 7.842375837242135e-06,
+      "loss": 0.1653,
+      "step": 3608
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9357223547452334,
+      "learning_rate": 7.841163316015495e-06,
+      "loss": 0.1664,
+      "step": 3609
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9883664464605586,
+      "learning_rate": 7.839950547978701e-06,
+      "loss": 0.1646,
+      "step": 3610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1696297046698176,
+      "learning_rate": 7.838737533237111e-06,
+      "loss": 0.1842,
+      "step": 3611
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.084704948118195,
+      "learning_rate": 7.837524271896097e-06,
+      "loss": 0.1775,
+      "step": 3612
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.306153241843188,
+      "learning_rate": 7.836310764061054e-06,
+      "loss": 0.2025,
+      "step": 3613
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0841266268660177,
+      "learning_rate": 7.8350970098374e-06,
+      "loss": 0.1546,
+      "step": 3614
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.169456772883008,
+      "learning_rate": 7.833883009330573e-06,
+      "loss": 0.1683,
+      "step": 3615
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0801329041012435,
+      "learning_rate": 7.832668762646027e-06,
+      "loss": 0.1588,
+      "step": 3616
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4326142118595597,
+      "learning_rate": 7.831454269889251e-06,
+      "loss": 0.2003,
+      "step": 3617
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.068686739382579,
+      "learning_rate": 7.830239531165744e-06,
+      "loss": 0.1649,
+      "step": 3618
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2495280147914,
+      "learning_rate": 7.829024546581028e-06,
+      "loss": 0.1607,
+      "step": 3619
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.106744303449247,
+      "learning_rate": 7.82780931624065e-06,
+      "loss": 0.1745,
+      "step": 3620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.954901983198919,
+      "learning_rate": 7.826593840250175e-06,
+      "loss": 0.1496,
+      "step": 3621
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2596144770210866,
+      "learning_rate": 7.825378118715192e-06,
+      "loss": 0.2083,
+      "step": 3622
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2979179680004824,
+      "learning_rate": 7.824162151741309e-06,
+      "loss": 0.1792,
+      "step": 3623
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5728516987634054,
+      "learning_rate": 7.822945939434156e-06,
+      "loss": 0.1467,
+      "step": 3624
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.32380576821315,
+      "learning_rate": 7.821729481899388e-06,
+      "loss": 0.1801,
+      "step": 3625
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.341874849029923,
+      "learning_rate": 7.820512779242673e-06,
+      "loss": 0.1856,
+      "step": 3626
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9082342384228124,
+      "learning_rate": 7.819295831569708e-06,
+      "loss": 0.1423,
+      "step": 3627
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.057660440735419,
+      "learning_rate": 7.818078638986208e-06,
+      "loss": 0.1851,
+      "step": 3628
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9273872153792628,
+      "learning_rate": 7.81686120159791e-06,
+      "loss": 0.1606,
+      "step": 3629
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.275162468742561,
+      "learning_rate": 7.815643519510571e-06,
+      "loss": 0.1723,
+      "step": 3630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2205730258665675,
+      "learning_rate": 7.81442559282997e-06,
+      "loss": 0.1975,
+      "step": 3631
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3172581151458465,
+      "learning_rate": 7.813207421661911e-06,
+      "loss": 0.2031,
+      "step": 3632
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0986324998598818,
+      "learning_rate": 7.811989006112212e-06,
+      "loss": 0.1582,
+      "step": 3633
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.074398153701755,
+      "learning_rate": 7.81077034628672e-06,
+      "loss": 0.1678,
+      "step": 3634
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.288249927364208,
+      "learning_rate": 7.809551442291294e-06,
+      "loss": 0.1947,
+      "step": 3635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.079007515606117,
+      "learning_rate": 7.808332294231824e-06,
+      "loss": 0.1763,
+      "step": 3636
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7444197919171367,
+      "learning_rate": 7.807112902214213e-06,
+      "loss": 0.1334,
+      "step": 3637
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5377783164990193,
+      "learning_rate": 7.805893266344393e-06,
+      "loss": 0.1846,
+      "step": 3638
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9755284217690599,
+      "learning_rate": 7.80467338672831e-06,
+      "loss": 0.1755,
+      "step": 3639
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1564631769413003,
+      "learning_rate": 7.803453263471933e-06,
+      "loss": 0.1971,
+      "step": 3640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.166983344278988,
+      "learning_rate": 7.802232896681259e-06,
+      "loss": 0.1427,
+      "step": 3641
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0767136281066185,
+      "learning_rate": 7.801012286462294e-06,
+      "loss": 0.1546,
+      "step": 3642
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3473963684789525,
+      "learning_rate": 7.799791432921075e-06,
+      "loss": 0.1734,
+      "step": 3643
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.321964229409665,
+      "learning_rate": 7.798570336163658e-06,
+      "loss": 0.1905,
+      "step": 3644
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.269420383017954,
+      "learning_rate": 7.797348996296116e-06,
+      "loss": 0.1738,
+      "step": 3645
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2162877772274028,
+      "learning_rate": 7.796127413424547e-06,
+      "loss": 0.1634,
+      "step": 3646
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1415424951487454,
+      "learning_rate": 7.794905587655071e-06,
+      "loss": 0.1677,
+      "step": 3647
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2387061876461987,
+      "learning_rate": 7.793683519093825e-06,
+      "loss": 0.1937,
+      "step": 3648
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.595956527584737,
+      "learning_rate": 7.79246120784697e-06,
+      "loss": 0.1937,
+      "step": 3649
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.222507791529048,
+      "learning_rate": 7.791238654020686e-06,
+      "loss": 0.1689,
+      "step": 3650
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.040363762646887,
+      "learning_rate": 7.79001585772118e-06,
+      "loss": 0.1563,
+      "step": 3651
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1002846408100044,
+      "learning_rate": 7.788792819054672e-06,
+      "loss": 0.1717,
+      "step": 3652
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9722218800934836,
+      "learning_rate": 7.787569538127406e-06,
+      "loss": 0.1284,
+      "step": 3653
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.18165286904065,
+      "learning_rate": 7.78634601504565e-06,
+      "loss": 0.1586,
+      "step": 3654
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9558197607387153,
+      "learning_rate": 7.785122249915688e-06,
+      "loss": 0.1391,
+      "step": 3655
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.02086360664275,
+      "learning_rate": 7.783898242843832e-06,
+      "loss": 0.1724,
+      "step": 3656
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2701007232384374,
+      "learning_rate": 7.782673993936408e-06,
+      "loss": 0.1975,
+      "step": 3657
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.23036440140428,
+      "learning_rate": 7.781449503299764e-06,
+      "loss": 0.1932,
+      "step": 3658
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2008271452439723,
+      "learning_rate": 7.780224771040275e-06,
+      "loss": 0.1758,
+      "step": 3659
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0738143177534427,
+      "learning_rate": 7.77899979726433e-06,
+      "loss": 0.1504,
+      "step": 3660
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.315121437885548,
+      "learning_rate": 7.777774582078342e-06,
+      "loss": 0.212,
+      "step": 3661
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.262291693554671,
+      "learning_rate": 7.776549125588743e-06,
+      "loss": 0.1997,
+      "step": 3662
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0893907627131463,
+      "learning_rate": 7.775323427901993e-06,
+      "loss": 0.1618,
+      "step": 3663
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8423042655805577,
+      "learning_rate": 7.774097489124562e-06,
+      "loss": 0.1244,
+      "step": 3664
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8418254739445166,
+      "learning_rate": 7.77287130936295e-06,
+      "loss": 0.1253,
+      "step": 3665
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.657948802551241,
+      "learning_rate": 7.771644888723672e-06,
+      "loss": 0.1314,
+      "step": 3666
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9144428569459166,
+      "learning_rate": 7.77041822731327e-06,
+      "loss": 0.1279,
+      "step": 3667
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2205524232330776,
+      "learning_rate": 7.7691913252383e-06,
+      "loss": 0.1421,
+      "step": 3668
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9555913677668197,
+      "learning_rate": 7.767964182605344e-06,
+      "loss": 0.1336,
+      "step": 3669
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.13653645398747,
+      "learning_rate": 7.766736799521e-06,
+      "loss": 0.1051,
+      "step": 3670
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9341991943882766,
+      "learning_rate": 7.765509176091894e-06,
+      "loss": 0.1026,
+      "step": 3671
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.357264880011443,
+      "learning_rate": 7.764281312424668e-06,
+      "loss": 0.162,
+      "step": 3672
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9847222722541236,
+      "learning_rate": 7.763053208625985e-06,
+      "loss": 0.128,
+      "step": 3673
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.895491652290345,
+      "learning_rate": 7.76182486480253e-06,
+      "loss": 0.1211,
+      "step": 3674
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0884058412451094,
+      "learning_rate": 7.760596281061008e-06,
+      "loss": 0.1084,
+      "step": 3675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5606195645329892,
+      "learning_rate": 7.759367457508145e-06,
+      "loss": 0.152,
+      "step": 3676
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2232985036917103,
+      "learning_rate": 7.75813839425069e-06,
+      "loss": 0.1297,
+      "step": 3677
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1246733133536146,
+      "learning_rate": 7.756909091395409e-06,
+      "loss": 0.1489,
+      "step": 3678
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3298561093334205,
+      "learning_rate": 7.755679549049093e-06,
+      "loss": 0.1301,
+      "step": 3679
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.787679274015443,
+      "learning_rate": 7.754449767318548e-06,
+      "loss": 0.1325,
+      "step": 3680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2458370436488218,
+      "learning_rate": 7.753219746310607e-06,
+      "loss": 0.1342,
+      "step": 3681
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0569591421773312,
+      "learning_rate": 7.751989486132122e-06,
+      "loss": 0.1255,
+      "step": 3682
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9517107500015811,
+      "learning_rate": 7.750758986889963e-06,
+      "loss": 0.1225,
+      "step": 3683
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7270387776582103,
+      "learning_rate": 7.749528248691026e-06,
+      "loss": 0.1009,
+      "step": 3684
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9922838751071787,
+      "learning_rate": 7.748297271642218e-06,
+      "loss": 0.1153,
+      "step": 3685
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.022761341914585,
+      "learning_rate": 7.747066055850479e-06,
+      "loss": 0.1188,
+      "step": 3686
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.002268246115833,
+      "learning_rate": 7.745834601422762e-06,
+      "loss": 0.1089,
+      "step": 3687
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0257435845184664,
+      "learning_rate": 7.744602908466044e-06,
+      "loss": 0.1335,
+      "step": 3688
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.906314083171679,
+      "learning_rate": 7.743370977087318e-06,
+      "loss": 0.1185,
+      "step": 3689
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9820137401200066,
+      "learning_rate": 7.742138807393607e-06,
+      "loss": 0.1203,
+      "step": 3690
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7000351401395506,
+      "learning_rate": 7.740906399491941e-06,
+      "loss": 0.1156,
+      "step": 3691
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.853642092895378,
+      "learning_rate": 7.739673753489386e-06,
+      "loss": 0.1048,
+      "step": 3692
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.8665274999096684,
+      "learning_rate": 7.738440869493018e-06,
+      "loss": 0.1399,
+      "step": 3693
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8854289206162174,
+      "learning_rate": 7.737207747609936e-06,
+      "loss": 0.1197,
+      "step": 3694
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.982596994289207,
+      "learning_rate": 7.73597438794726e-06,
+      "loss": 0.1263,
+      "step": 3695
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.2178625175209574,
+      "learning_rate": 7.734740790612137e-06,
+      "loss": 0.1424,
+      "step": 3696
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.3352004148220162,
+      "learning_rate": 7.73350695571172e-06,
+      "loss": 0.1402,
+      "step": 3697
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8392551497793572,
+      "learning_rate": 7.732272883353197e-06,
+      "loss": 0.1006,
+      "step": 3698
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1135223526792193,
+      "learning_rate": 7.731038573643772e-06,
+      "loss": 0.1352,
+      "step": 3699
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.4588044373414597,
+      "learning_rate": 7.729804026690667e-06,
+      "loss": 0.1231,
+      "step": 3700
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1401973444446876,
+      "learning_rate": 7.728569242601125e-06,
+      "loss": 0.1308,
+      "step": 3701
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.933891212732119,
+      "learning_rate": 7.727334221482412e-06,
+      "loss": 0.0984,
+      "step": 3702
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1801745492849016,
+      "learning_rate": 7.726098963441815e-06,
+      "loss": 0.1443,
+      "step": 3703
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1297045046097427,
+      "learning_rate": 7.72486346858664e-06,
+      "loss": 0.114,
+      "step": 3704
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.057459886527118,
+      "learning_rate": 7.72362773702421e-06,
+      "loss": 0.1184,
+      "step": 3705
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1681983924518784,
+      "learning_rate": 7.722391768861875e-06,
+      "loss": 0.1256,
+      "step": 3706
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8668252852955303,
+      "learning_rate": 7.721155564207003e-06,
+      "loss": 0.1188,
+      "step": 3707
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.2069628875416116,
+      "learning_rate": 7.719919123166984e-06,
+      "loss": 0.1349,
+      "step": 3708
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0008419645704647,
+      "learning_rate": 7.718682445849224e-06,
+      "loss": 0.1316,
+      "step": 3709
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.2983687304425855,
+      "learning_rate": 7.717445532361152e-06,
+      "loss": 0.1345,
+      "step": 3710
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.05051866052114,
+      "learning_rate": 7.716208382810221e-06,
+      "loss": 0.1058,
+      "step": 3711
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0233261525054567,
+      "learning_rate": 7.714970997303898e-06,
+      "loss": 0.1222,
+      "step": 3712
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9752764858048422,
+      "learning_rate": 7.713733375949677e-06,
+      "loss": 0.1201,
+      "step": 3713
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.873273117532547,
+      "learning_rate": 7.712495518855067e-06,
+      "loss": 0.1262,
+      "step": 3714
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7002184049434432,
+      "learning_rate": 7.711257426127601e-06,
+      "loss": 0.1121,
+      "step": 3715
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9103501157364418,
+      "learning_rate": 7.710019097874833e-06,
+      "loss": 0.1208,
+      "step": 3716
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9079401513506784,
+      "learning_rate": 7.708780534204332e-06,
+      "loss": 0.1129,
+      "step": 3717
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.004981317480421,
+      "learning_rate": 7.707541735223696e-06,
+      "loss": 0.1248,
+      "step": 3718
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9873701586717412,
+      "learning_rate": 7.706302701040534e-06,
+      "loss": 0.1214,
+      "step": 3719
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.7659708433541366,
+      "learning_rate": 7.705063431762482e-06,
+      "loss": 0.1345,
+      "step": 3720
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0630262358471736,
+      "learning_rate": 7.703823927497196e-06,
+      "loss": 0.1267,
+      "step": 3721
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1202664639270123,
+      "learning_rate": 7.702584188352351e-06,
+      "loss": 0.1227,
+      "step": 3722
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9194476615333196,
+      "learning_rate": 7.701344214435639e-06,
+      "loss": 0.1093,
+      "step": 3723
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1041895925649134,
+      "learning_rate": 7.70010400585478e-06,
+      "loss": 0.1311,
+      "step": 3724
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.791171872663612,
+      "learning_rate": 7.69886356271751e-06,
+      "loss": 0.1044,
+      "step": 3725
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.3081099524846302,
+      "learning_rate": 7.697622885131579e-06,
+      "loss": 0.1394,
+      "step": 3726
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1810024380765274,
+      "learning_rate": 7.696381973204772e-06,
+      "loss": 0.1309,
+      "step": 3727
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1530724119101965,
+      "learning_rate": 7.695140827044882e-06,
+      "loss": 0.1143,
+      "step": 3728
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.027416485250483,
+      "learning_rate": 7.693899446759727e-06,
+      "loss": 0.1189,
+      "step": 3729
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.103253731855292,
+      "learning_rate": 7.692657832457146e-06,
+      "loss": 0.1221,
+      "step": 3730
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.4178987848133366,
+      "learning_rate": 7.691415984244998e-06,
+      "loss": 0.1243,
+      "step": 3731
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9307468950288047,
+      "learning_rate": 7.69017390223116e-06,
+      "loss": 0.1167,
+      "step": 3732
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2186650535101213,
+      "learning_rate": 7.688931586523531e-06,
+      "loss": 0.1327,
+      "step": 3733
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9361333348080088,
+      "learning_rate": 7.687689037230031e-06,
+      "loss": 0.117,
+      "step": 3734
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8704648712928205,
+      "learning_rate": 7.686446254458598e-06,
+      "loss": 0.1118,
+      "step": 3735
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0021402115150564,
+      "learning_rate": 7.685203238317194e-06,
+      "loss": 0.1282,
+      "step": 3736
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8389938289627499,
+      "learning_rate": 7.683959988913798e-06,
+      "loss": 0.1172,
+      "step": 3737
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2211323360832544,
+      "learning_rate": 7.68271650635641e-06,
+      "loss": 0.1363,
+      "step": 3738
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.38660748421882,
+      "learning_rate": 7.68147279075305e-06,
+      "loss": 0.1215,
+      "step": 3739
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.043336109098562,
+      "learning_rate": 7.680228842211762e-06,
+      "loss": 0.1292,
+      "step": 3740
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.077726491515614,
+      "learning_rate": 7.678984660840603e-06,
+      "loss": 0.1461,
+      "step": 3741
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9413258346644382,
+      "learning_rate": 7.677740246747657e-06,
+      "loss": 0.1257,
+      "step": 3742
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.3790580947663225,
+      "learning_rate": 7.676495600041025e-06,
+      "loss": 0.117,
+      "step": 3743
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.931216787036907,
+      "learning_rate": 7.675250720828827e-06,
+      "loss": 0.1127,
+      "step": 3744
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2817151131499283,
+      "learning_rate": 7.674005609219208e-06,
+      "loss": 0.1673,
+      "step": 3745
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.322409997825016,
+      "learning_rate": 7.672760265320326e-06,
+      "loss": 0.1399,
+      "step": 3746
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0328968578081628,
+      "learning_rate": 7.671514689240366e-06,
+      "loss": 0.1073,
+      "step": 3747
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.190794783708201,
+      "learning_rate": 7.670268881087532e-06,
+      "loss": 0.1163,
+      "step": 3748
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8360948468549523,
+      "learning_rate": 7.669022840970042e-06,
+      "loss": 0.0977,
+      "step": 3749
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.159420469519051,
+      "learning_rate": 7.667776568996143e-06,
+      "loss": 0.1101,
+      "step": 3750
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9715156209059748,
+      "learning_rate": 7.666530065274096e-06,
+      "loss": 0.1094,
+      "step": 3751
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0006232383871634,
+      "learning_rate": 7.665283329912183e-06,
+      "loss": 0.1159,
+      "step": 3752
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.891669338202314,
+      "learning_rate": 7.664036363018709e-06,
+      "loss": 0.0982,
+      "step": 3753
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.932178421119616,
+      "learning_rate": 7.662789164702e-06,
+      "loss": 0.1091,
+      "step": 3754
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9550871481428564,
+      "learning_rate": 7.661541735070392e-06,
+      "loss": 0.1096,
+      "step": 3755
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.309750163994795,
+      "learning_rate": 7.660294074232254e-06,
+      "loss": 0.1355,
+      "step": 3756
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9411718065366854,
+      "learning_rate": 7.659046182295968e-06,
+      "loss": 0.1159,
+      "step": 3757
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0663007125384802,
+      "learning_rate": 7.657798059369938e-06,
+      "loss": 0.1033,
+      "step": 3758
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.699637014114936,
+      "learning_rate": 7.656549705562588e-06,
+      "loss": 0.104,
+      "step": 3759
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0754942082614556,
+      "learning_rate": 7.655301120982362e-06,
+      "loss": 0.1247,
+      "step": 3760
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.884264305624918,
+      "learning_rate": 7.65405230573772e-06,
+      "loss": 0.1182,
+      "step": 3761
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9442666225404777,
+      "learning_rate": 7.65280325993715e-06,
+      "loss": 0.1159,
+      "step": 3762
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.259687442842502,
+      "learning_rate": 7.651553983689155e-06,
+      "loss": 0.1399,
+      "step": 3763
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9731502958964626,
+      "learning_rate": 7.650304477102258e-06,
+      "loss": 0.1169,
+      "step": 3764
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.8490187668665032,
+      "learning_rate": 7.649054740285005e-06,
+      "loss": 0.1338,
+      "step": 3765
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.093527942133517,
+      "learning_rate": 7.647804773345957e-06,
+      "loss": 0.1154,
+      "step": 3766
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9360838865901533,
+      "learning_rate": 7.6465545763937e-06,
+      "loss": 0.1463,
+      "step": 3767
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7420134041239128,
+      "learning_rate": 7.645304149536833e-06,
+      "loss": 0.1174,
+      "step": 3768
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.124621494746595,
+      "learning_rate": 7.64405349288399e-06,
+      "loss": 0.1416,
+      "step": 3769
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.032178427084541,
+      "learning_rate": 7.642802606543805e-06,
+      "loss": 0.1246,
+      "step": 3770
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0429903039752175,
+      "learning_rate": 7.641551490624945e-06,
+      "loss": 0.1211,
+      "step": 3771
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.4544332284707857,
+      "learning_rate": 7.640300145236096e-06,
+      "loss": 0.1405,
+      "step": 3772
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9425654583940424,
+      "learning_rate": 7.63904857048596e-06,
+      "loss": 0.1105,
+      "step": 3773
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9763481900564752,
+      "learning_rate": 7.637796766483259e-06,
+      "loss": 0.1108,
+      "step": 3774
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.213957311424813,
+      "learning_rate": 7.636544733336739e-06,
+      "loss": 0.1459,
+      "step": 3775
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9079870239827432,
+      "learning_rate": 7.63529247115516e-06,
+      "loss": 0.1262,
+      "step": 3776
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9417706053933135,
+      "learning_rate": 7.634039980047308e-06,
+      "loss": 0.1249,
+      "step": 3777
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8839413417862498,
+      "learning_rate": 7.632787260121987e-06,
+      "loss": 0.1223,
+      "step": 3778
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.730359986752707,
+      "learning_rate": 7.631534311488016e-06,
+      "loss": 0.1031,
+      "step": 3779
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.091028049090303,
+      "learning_rate": 7.630281134254243e-06,
+      "loss": 0.1119,
+      "step": 3780
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9783565560528857,
+      "learning_rate": 7.629027728529527e-06,
+      "loss": 0.1146,
+      "step": 3781
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.207020810025933,
+      "learning_rate": 7.627774094422751e-06,
+      "loss": 0.1354,
+      "step": 3782
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0780893017509325,
+      "learning_rate": 7.626520232042819e-06,
+      "loss": 0.1281,
+      "step": 3783
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.326454344067117,
+      "learning_rate": 7.625266141498653e-06,
+      "loss": 0.1508,
+      "step": 3784
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.2877337956174024,
+      "learning_rate": 7.624011822899193e-06,
+      "loss": 0.1338,
+      "step": 3785
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0433247797705896,
+      "learning_rate": 7.622757276353404e-06,
+      "loss": 0.1318,
+      "step": 3786
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.477176615181054,
+      "learning_rate": 7.621502501970266e-06,
+      "loss": 0.1198,
+      "step": 3787
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8899613382321498,
+      "learning_rate": 7.62024749985878e-06,
+      "loss": 0.1166,
+      "step": 3788
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.2823334619309605,
+      "learning_rate": 7.618992270127968e-06,
+      "loss": 0.1533,
+      "step": 3789
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.187582203800892,
+      "learning_rate": 7.617736812886873e-06,
+      "loss": 0.143,
+      "step": 3790
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.060816786485956,
+      "learning_rate": 7.616481128244552e-06,
+      "loss": 0.1247,
+      "step": 3791
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1473759311973075,
+      "learning_rate": 7.615225216310087e-06,
+      "loss": 0.1215,
+      "step": 3792
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.027390599332439,
+      "learning_rate": 7.61396907719258e-06,
+      "loss": 0.1036,
+      "step": 3793
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.2999710157707582,
+      "learning_rate": 7.612712711001149e-06,
+      "loss": 0.1579,
+      "step": 3794
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9420989573866498,
+      "learning_rate": 7.611456117844934e-06,
+      "loss": 0.1109,
+      "step": 3795
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9202408310627497,
+      "learning_rate": 7.610199297833097e-06,
+      "loss": 0.1099,
+      "step": 3796
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.4931245594801488,
+      "learning_rate": 7.6089422510748135e-06,
+      "loss": 0.1296,
+      "step": 3797
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.487569412261905,
+      "learning_rate": 7.607684977679284e-06,
+      "loss": 0.133,
+      "step": 3798
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.2629390578533757,
+      "learning_rate": 7.606427477755729e-06,
+      "loss": 0.1063,
+      "step": 3799
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.815007814451184,
+      "learning_rate": 7.605169751413382e-06,
+      "loss": 0.1061,
+      "step": 3800
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.78255486807673,
+      "learning_rate": 7.603911798761506e-06,
+      "loss": 0.0984,
+      "step": 3801
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9411296500169395,
+      "learning_rate": 7.602653619909377e-06,
+      "loss": 0.1046,
+      "step": 3802
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9040436529835725,
+      "learning_rate": 7.6013952149662905e-06,
+      "loss": 0.1162,
+      "step": 3803
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.2592443603321635,
+      "learning_rate": 7.600136584041564e-06,
+      "loss": 0.1322,
+      "step": 3804
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.18545426159128,
+      "learning_rate": 7.598877727244538e-06,
+      "loss": 0.1287,
+      "step": 3805
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.5845484981997244,
+      "learning_rate": 7.597618644684561e-06,
+      "loss": 0.1423,
+      "step": 3806
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.122224330553804,
+      "learning_rate": 7.596359336471015e-06,
+      "loss": 0.1461,
+      "step": 3807
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.400711901823705,
+      "learning_rate": 7.595099802713293e-06,
+      "loss": 0.1438,
+      "step": 3808
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1607933436257425,
+      "learning_rate": 7.593840043520811e-06,
+      "loss": 0.141,
+      "step": 3809
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.5193339257391743,
+      "learning_rate": 7.592580059003002e-06,
+      "loss": 0.1677,
+      "step": 3810
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1070766809372454,
+      "learning_rate": 7.591319849269322e-06,
+      "loss": 0.1285,
+      "step": 3811
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7893219394998408,
+      "learning_rate": 7.590059414429243e-06,
+      "loss": 0.1095,
+      "step": 3812
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9859167953025019,
+      "learning_rate": 7.588798754592258e-06,
+      "loss": 0.1109,
+      "step": 3813
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9754442486155872,
+      "learning_rate": 7.5875378698678825e-06,
+      "loss": 0.1204,
+      "step": 3814
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1919939021031456,
+      "learning_rate": 7.586276760365645e-06,
+      "loss": 0.1571,
+      "step": 3815
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8636472283709595,
+      "learning_rate": 7.585015426195099e-06,
+      "loss": 0.1072,
+      "step": 3816
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9258476176040322,
+      "learning_rate": 7.583753867465819e-06,
+      "loss": 0.116,
+      "step": 3817
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8914851751763961,
+      "learning_rate": 7.582492084287389e-06,
+      "loss": 0.1092,
+      "step": 3818
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.2381351920084693,
+      "learning_rate": 7.581230076769426e-06,
+      "loss": 0.1539,
+      "step": 3819
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0667574388644927,
+      "learning_rate": 7.5799678450215566e-06,
+      "loss": 0.124,
+      "step": 3820
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.2180434666246245,
+      "learning_rate": 7.57870538915343e-06,
+      "loss": 0.1375,
+      "step": 3821
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0791209671074786,
+      "learning_rate": 7.577442709274716e-06,
+      "loss": 0.1174,
+      "step": 3822
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8182884625637064,
+      "learning_rate": 7.576179805495102e-06,
+      "loss": 0.0924,
+      "step": 3823
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0908213312737605,
+      "learning_rate": 7.574916677924295e-06,
+      "loss": 0.1067,
+      "step": 3824
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0439814288268674,
+      "learning_rate": 7.573653326672026e-06,
+      "loss": 0.1147,
+      "step": 3825
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.234533068126777,
+      "learning_rate": 7.572389751848037e-06,
+      "loss": 0.1191,
+      "step": 3826
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9979746648423296,
+      "learning_rate": 7.571125953562095e-06,
+      "loss": 0.1113,
+      "step": 3827
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0435986277618454,
+      "learning_rate": 7.569861931923989e-06,
+      "loss": 0.1165,
+      "step": 3828
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9489731784843585,
+      "learning_rate": 7.5685976870435185e-06,
+      "loss": 0.1088,
+      "step": 3829
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0278297352860477,
+      "learning_rate": 7.567333219030511e-06,
+      "loss": 0.1234,
+      "step": 3830
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1953538524287923,
+      "learning_rate": 7.566068527994809e-06,
+      "loss": 0.1457,
+      "step": 3831
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7399577961527786,
+      "learning_rate": 7.564803614046276e-06,
+      "loss": 0.1027,
+      "step": 3832
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.100384229954873,
+      "learning_rate": 7.563538477294793e-06,
+      "loss": 0.1416,
+      "step": 3833
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.3432059152091114,
+      "learning_rate": 7.562273117850264e-06,
+      "loss": 0.1296,
+      "step": 3834
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.232665061085612,
+      "learning_rate": 7.561007535822608e-06,
+      "loss": 0.1246,
+      "step": 3835
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.03871537857279,
+      "learning_rate": 7.5597417313217655e-06,
+      "loss": 0.132,
+      "step": 3836
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1104292454682887,
+      "learning_rate": 7.558475704457698e-06,
+      "loss": 0.115,
+      "step": 3837
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.134435422165674,
+      "learning_rate": 7.557209455340382e-06,
+      "loss": 0.1225,
+      "step": 3838
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.623820967149173,
+      "learning_rate": 7.5559429840798185e-06,
+      "loss": 0.1598,
+      "step": 3839
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1751083310927912,
+      "learning_rate": 7.554676290786023e-06,
+      "loss": 0.1149,
+      "step": 3840
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0688845722642686,
+      "learning_rate": 7.553409375569032e-06,
+      "loss": 0.1351,
+      "step": 3841
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4061758756107574,
+      "learning_rate": 7.552142238538905e-06,
+      "loss": 0.1438,
+      "step": 3842
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.959849350279419,
+      "learning_rate": 7.550874879805713e-06,
+      "loss": 0.1201,
+      "step": 3843
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.195586633950834,
+      "learning_rate": 7.549607299479554e-06,
+      "loss": 0.1339,
+      "step": 3844
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.275838343695774,
+      "learning_rate": 7.548339497670538e-06,
+      "loss": 0.1185,
+      "step": 3845
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0138949984997723,
+      "learning_rate": 7.547071474488804e-06,
+      "loss": 0.1185,
+      "step": 3846
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.440760860070949,
+      "learning_rate": 7.5458032300445e-06,
+      "loss": 0.1235,
+      "step": 3847
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.148225878770096,
+      "learning_rate": 7.5445347644478e-06,
+      "loss": 0.1159,
+      "step": 3848
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0426812647668724,
+      "learning_rate": 7.543266077808893e-06,
+      "loss": 0.1301,
+      "step": 3849
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.040079652236069,
+      "learning_rate": 7.541997170237989e-06,
+      "loss": 0.1216,
+      "step": 3850
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8651198756252683,
+      "learning_rate": 7.540728041845319e-06,
+      "loss": 0.1149,
+      "step": 3851
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9001572836042329,
+      "learning_rate": 7.539458692741131e-06,
+      "loss": 0.1296,
+      "step": 3852
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0255828982058763,
+      "learning_rate": 7.538189123035691e-06,
+      "loss": 0.1347,
+      "step": 3853
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.379063105768789,
+      "learning_rate": 7.536919332839288e-06,
+      "loss": 0.1487,
+      "step": 3854
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.5585705811249126,
+      "learning_rate": 7.5356493222622265e-06,
+      "loss": 0.153,
+      "step": 3855
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.3292888958864677,
+      "learning_rate": 7.534379091414832e-06,
+      "loss": 0.1378,
+      "step": 3856
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9274956079730192,
+      "learning_rate": 7.533108640407447e-06,
+      "loss": 0.1258,
+      "step": 3857
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.042712402018402,
+      "learning_rate": 7.5318379693504375e-06,
+      "loss": 0.1077,
+      "step": 3858
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.3887910584551473,
+      "learning_rate": 7.530567078354185e-06,
+      "loss": 0.158,
+      "step": 3859
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.459217105232712,
+      "learning_rate": 7.52929596752909e-06,
+      "loss": 0.1375,
+      "step": 3860
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4524818431952595,
+      "learning_rate": 7.528024636985575e-06,
+      "loss": 0.1578,
+      "step": 3861
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.217551888440984,
+      "learning_rate": 7.5267530868340775e-06,
+      "loss": 0.1331,
+      "step": 3862
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.217046594652403,
+      "learning_rate": 7.525481317185057e-06,
+      "loss": 0.1461,
+      "step": 3863
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.1391646749176862,
+      "learning_rate": 7.524209328148995e-06,
+      "loss": 0.1448,
+      "step": 3864
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9448940713996767,
+      "learning_rate": 7.5229371198363824e-06,
+      "loss": 0.1051,
+      "step": 3865
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1864608556086482,
+      "learning_rate": 7.521664692357737e-06,
+      "loss": 0.128,
+      "step": 3866
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.22166140657511,
+      "learning_rate": 7.520392045823598e-06,
+      "loss": 0.1465,
+      "step": 3867
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9892002574237704,
+      "learning_rate": 7.519119180344514e-06,
+      "loss": 0.1391,
+      "step": 3868
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9478256230445294,
+      "learning_rate": 7.5178460960310605e-06,
+      "loss": 0.1148,
+      "step": 3869
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1832304119031076,
+      "learning_rate": 7.51657279299383e-06,
+      "loss": 0.1401,
+      "step": 3870
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6455865180261677,
+      "learning_rate": 7.515299271343434e-06,
+      "loss": 0.0897,
+      "step": 3871
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.3708326081904145,
+      "learning_rate": 7.514025531190499e-06,
+      "loss": 0.1542,
+      "step": 3872
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8935678478048354,
+      "learning_rate": 7.512751572645679e-06,
+      "loss": 0.1034,
+      "step": 3873
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1230092424681124,
+      "learning_rate": 7.5114773958196385e-06,
+      "loss": 0.1224,
+      "step": 3874
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.027034265766682,
+      "learning_rate": 7.510203000823066e-06,
+      "loss": 0.1114,
+      "step": 3875
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.222676575068051,
+      "learning_rate": 7.5089283877666664e-06,
+      "loss": 0.1477,
+      "step": 3876
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1273060136462276,
+      "learning_rate": 7.507653556761166e-06,
+      "loss": 0.1172,
+      "step": 3877
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.087109178883188,
+      "learning_rate": 7.506378507917306e-06,
+      "loss": 0.1227,
+      "step": 3878
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.143495203803203,
+      "learning_rate": 7.505103241345853e-06,
+      "loss": 0.1225,
+      "step": 3879
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.251255473238864,
+      "learning_rate": 7.503827757157584e-06,
+      "loss": 0.1318,
+      "step": 3880
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.165190009236253,
+      "learning_rate": 7.5025520554633035e-06,
+      "loss": 0.1035,
+      "step": 3881
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.176057028029448,
+      "learning_rate": 7.501276136373831e-06,
+      "loss": 0.1391,
+      "step": 3882
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0142652621974895,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.1153,
+      "step": 3883
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0047299306248467,
+      "learning_rate": 7.498723646452673e-06,
+      "loss": 0.1226,
+      "step": 3884
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1608921058753405,
+      "learning_rate": 7.497447075842725e-06,
+      "loss": 0.1073,
+      "step": 3885
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2977880579664305,
+      "learning_rate": 7.496170288281049e-06,
+      "loss": 0.1578,
+      "step": 3886
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.133512346587424,
+      "learning_rate": 7.494893283878559e-06,
+      "loss": 0.1007,
+      "step": 3887
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1753885488805498,
+      "learning_rate": 7.493616062746191e-06,
+      "loss": 0.1184,
+      "step": 3888
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1006548687026956,
+      "learning_rate": 7.492338624994892e-06,
+      "loss": 0.1376,
+      "step": 3889
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8971780208166353,
+      "learning_rate": 7.491060970735633e-06,
+      "loss": 0.1123,
+      "step": 3890
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.125157506528644,
+      "learning_rate": 7.489783100079407e-06,
+      "loss": 0.1423,
+      "step": 3891
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9354704301464662,
+      "learning_rate": 7.488505013137215e-06,
+      "loss": 0.1273,
+      "step": 3892
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0189909509120425,
+      "learning_rate": 7.4872267100200905e-06,
+      "loss": 0.1281,
+      "step": 3893
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9884918843137125,
+      "learning_rate": 7.485948190839076e-06,
+      "loss": 0.1123,
+      "step": 3894
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9607487448291196,
+      "learning_rate": 7.484669455705235e-06,
+      "loss": 0.1321,
+      "step": 3895
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8519986073858707,
+      "learning_rate": 7.483390504729651e-06,
+      "loss": 0.1116,
+      "step": 3896
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1195844912532094,
+      "learning_rate": 7.4821113380234266e-06,
+      "loss": 0.1325,
+      "step": 3897
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.329881985749329,
+      "learning_rate": 7.48083195569768e-06,
+      "loss": 0.1309,
+      "step": 3898
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.127672858599593,
+      "learning_rate": 7.479552357863553e-06,
+      "loss": 0.1207,
+      "step": 3899
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8440888140963385,
+      "learning_rate": 7.478272544632204e-06,
+      "loss": 0.1173,
+      "step": 3900
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.094780808136034,
+      "learning_rate": 7.476992516114805e-06,
+      "loss": 0.1201,
+      "step": 3901
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.3466968102559167,
+      "learning_rate": 7.4757122724225575e-06,
+      "loss": 0.1301,
+      "step": 3902
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.038753418217479,
+      "learning_rate": 7.474431813666669e-06,
+      "loss": 0.1226,
+      "step": 3903
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.197397302185069,
+      "learning_rate": 7.473151139958378e-06,
+      "loss": 0.1126,
+      "step": 3904
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0627679163946873,
+      "learning_rate": 7.4718702514089324e-06,
+      "loss": 0.1351,
+      "step": 3905
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0491556330215053,
+      "learning_rate": 7.470589148129603e-06,
+      "loss": 0.1371,
+      "step": 3906
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0169410858543992,
+      "learning_rate": 7.469307830231679e-06,
+      "loss": 0.1333,
+      "step": 3907
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.185654585940526,
+      "learning_rate": 7.468026297826468e-06,
+      "loss": 0.1582,
+      "step": 3908
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.356927378141032,
+      "learning_rate": 7.4667445510252945e-06,
+      "loss": 0.1439,
+      "step": 3909
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.742003771019665,
+      "learning_rate": 7.465462589939504e-06,
+      "loss": 0.1368,
+      "step": 3910
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7950421170417288,
+      "learning_rate": 7.4641804146804605e-06,
+      "loss": 0.1153,
+      "step": 3911
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.128224600591649,
+      "learning_rate": 7.462898025359544e-06,
+      "loss": 0.112,
+      "step": 3912
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1396326082975095,
+      "learning_rate": 7.461615422088155e-06,
+      "loss": 0.1272,
+      "step": 3913
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0025502724788504,
+      "learning_rate": 7.460332604977716e-06,
+      "loss": 0.1202,
+      "step": 3914
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2624811154948556,
+      "learning_rate": 7.4590495741396585e-06,
+      "loss": 0.1349,
+      "step": 3915
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9355326775586001,
+      "learning_rate": 7.457766329685444e-06,
+      "loss": 0.1259,
+      "step": 3916
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0140273416791428,
+      "learning_rate": 7.456482871726545e-06,
+      "loss": 0.1255,
+      "step": 3917
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9606837035385274,
+      "learning_rate": 7.4551992003744545e-06,
+      "loss": 0.1172,
+      "step": 3918
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0180270701257226,
+      "learning_rate": 7.4539153157406825e-06,
+      "loss": 0.1306,
+      "step": 3919
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.856103025005708,
+      "learning_rate": 7.4526312179367656e-06,
+      "loss": 0.1179,
+      "step": 3920
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1450723568870793,
+      "learning_rate": 7.451346907074245e-06,
+      "loss": 0.1231,
+      "step": 3921
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2612429769892226,
+      "learning_rate": 7.450062383264692e-06,
+      "loss": 0.1434,
+      "step": 3922
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.518162038010842,
+      "learning_rate": 7.448777646619693e-06,
+      "loss": 0.1241,
+      "step": 3923
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.067242591568339,
+      "learning_rate": 7.44749269725085e-06,
+      "loss": 0.1378,
+      "step": 3924
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0774798553528186,
+      "learning_rate": 7.4462075352697875e-06,
+      "loss": 0.1205,
+      "step": 3925
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.075024819502733,
+      "learning_rate": 7.444922160788146e-06,
+      "loss": 0.1303,
+      "step": 3926
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.926040180986208,
+      "learning_rate": 7.443636573917585e-06,
+      "loss": 0.113,
+      "step": 3927
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.384360386854931,
+      "learning_rate": 7.442350774769782e-06,
+      "loss": 0.1411,
+      "step": 3928
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.673943894788367,
+      "learning_rate": 7.441064763456437e-06,
+      "loss": 0.0857,
+      "step": 3929
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.521919680103763,
+      "learning_rate": 7.439778540089261e-06,
+      "loss": 0.1559,
+      "step": 3930
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.513935300832426,
+      "learning_rate": 7.438492104779989e-06,
+      "loss": 0.1274,
+      "step": 3931
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.176807780283664,
+      "learning_rate": 7.437205457640374e-06,
+      "loss": 0.1044,
+      "step": 3932
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1076513224284428,
+      "learning_rate": 7.435918598782183e-06,
+      "loss": 0.1228,
+      "step": 3933
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.279043079419693,
+      "learning_rate": 7.434631528317209e-06,
+      "loss": 0.1522,
+      "step": 3934
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.204071430556285,
+      "learning_rate": 7.433344246357257e-06,
+      "loss": 0.13,
+      "step": 3935
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2060846039932938,
+      "learning_rate": 7.432056753014152e-06,
+      "loss": 0.1438,
+      "step": 3936
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1204035655348754,
+      "learning_rate": 7.4307690483997365e-06,
+      "loss": 0.1246,
+      "step": 3937
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1186187320136267,
+      "learning_rate": 7.429481132625876e-06,
+      "loss": 0.1421,
+      "step": 3938
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9005706004579819,
+      "learning_rate": 7.428193005804449e-06,
+      "loss": 0.1079,
+      "step": 3939
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.083741061405799,
+      "learning_rate": 7.426904668047352e-06,
+      "loss": 0.1178,
+      "step": 3940
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.063866709633579,
+      "learning_rate": 7.425616119466508e-06,
+      "loss": 0.1286,
+      "step": 3941
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.817758440126498,
+      "learning_rate": 7.424327360173847e-06,
+      "loss": 0.1023,
+      "step": 3942
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.013458294290249,
+      "learning_rate": 7.4230383902813255e-06,
+      "loss": 0.1321,
+      "step": 3943
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2409224472735123,
+      "learning_rate": 7.421749209900916e-06,
+      "loss": 0.1172,
+      "step": 3944
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3202713111420383,
+      "learning_rate": 7.420459819144605e-06,
+      "loss": 0.1297,
+      "step": 3945
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9171512332914806,
+      "learning_rate": 7.419170218124405e-06,
+      "loss": 0.116,
+      "step": 3946
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.926502387883841,
+      "learning_rate": 7.417880406952343e-06,
+      "loss": 0.1188,
+      "step": 3947
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1402951453430905,
+      "learning_rate": 7.4165903857404606e-06,
+      "loss": 0.1121,
+      "step": 3948
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.146439520986459,
+      "learning_rate": 7.4153001546008245e-06,
+      "loss": 0.14,
+      "step": 3949
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9027406826044377,
+      "learning_rate": 7.414009713645516e-06,
+      "loss": 0.1156,
+      "step": 3950
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3179571518912288,
+      "learning_rate": 7.412719062986632e-06,
+      "loss": 0.1386,
+      "step": 3951
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1585023801670684,
+      "learning_rate": 7.411428202736293e-06,
+      "loss": 0.1269,
+      "step": 3952
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7807928987543278,
+      "learning_rate": 7.410137133006636e-06,
+      "loss": 0.1023,
+      "step": 3953
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2934532102517853,
+      "learning_rate": 7.408845853909813e-06,
+      "loss": 0.1208,
+      "step": 3954
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2432775515713814,
+      "learning_rate": 7.407554365557999e-06,
+      "loss": 0.1272,
+      "step": 3955
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.69488860170964,
+      "learning_rate": 7.406262668063383e-06,
+      "loss": 0.1003,
+      "step": 3956
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.082637074657893,
+      "learning_rate": 7.404970761538175e-06,
+      "loss": 0.1256,
+      "step": 3957
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.277575887253905,
+      "learning_rate": 7.403678646094602e-06,
+      "loss": 0.1449,
+      "step": 3958
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.257923864518211,
+      "learning_rate": 7.40238632184491e-06,
+      "loss": 0.1316,
+      "step": 3959
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0746344557535656,
+      "learning_rate": 7.40109378890136e-06,
+      "loss": 0.1217,
+      "step": 3960
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8769338531399027,
+      "learning_rate": 7.399801047376235e-06,
+      "loss": 0.111,
+      "step": 3961
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8318698550524606,
+      "learning_rate": 7.398508097381837e-06,
+      "loss": 0.111,
+      "step": 3962
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1451980967544824,
+      "learning_rate": 7.397214939030479e-06,
+      "loss": 0.1562,
+      "step": 3963
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.44948514918435,
+      "learning_rate": 7.395921572434501e-06,
+      "loss": 0.1159,
+      "step": 3964
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9230492125242697,
+      "learning_rate": 7.394627997706256e-06,
+      "loss": 0.1095,
+      "step": 3965
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7568446334639645,
+      "learning_rate": 7.393334214958114e-06,
+      "loss": 0.0903,
+      "step": 3966
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9733458104791217,
+      "learning_rate": 7.392040224302468e-06,
+      "loss": 0.1214,
+      "step": 3967
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.4210225728514954,
+      "learning_rate": 7.390746025851725e-06,
+      "loss": 0.1368,
+      "step": 3968
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6787517447564464,
+      "learning_rate": 7.389451619718311e-06,
+      "loss": 0.0943,
+      "step": 3969
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2614749821578637,
+      "learning_rate": 7.388157006014669e-06,
+      "loss": 0.1329,
+      "step": 3970
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.239093522756117,
+      "learning_rate": 7.386862184853264e-06,
+      "loss": 0.1227,
+      "step": 3971
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2995402571318095,
+      "learning_rate": 7.3855671563465745e-06,
+      "loss": 0.1351,
+      "step": 3972
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.968195674552456,
+      "learning_rate": 7.3842719206071e-06,
+      "loss": 0.1166,
+      "step": 3973
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.241333931537465,
+      "learning_rate": 7.382976477747357e-06,
+      "loss": 0.1467,
+      "step": 3974
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 20.921668138100443,
+      "learning_rate": 7.381680827879877e-06,
+      "loss": 0.1516,
+      "step": 3975
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9626431379187887,
+      "learning_rate": 7.380384971117215e-06,
+      "loss": 0.1199,
+      "step": 3976
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.273551048713562,
+      "learning_rate": 7.379088907571942e-06,
+      "loss": 0.1522,
+      "step": 3977
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 21.837456878779356,
+      "learning_rate": 7.377792637356644e-06,
+      "loss": 0.1428,
+      "step": 3978
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.833715634757583,
+      "learning_rate": 7.376496160583928e-06,
+      "loss": 0.1185,
+      "step": 3979
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.096116938590306,
+      "learning_rate": 7.3751994773664195e-06,
+      "loss": 0.1303,
+      "step": 3980
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.274181786895908,
+      "learning_rate": 7.373902587816758e-06,
+      "loss": 0.1305,
+      "step": 3981
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9311979803562314,
+      "learning_rate": 7.372605492047605e-06,
+      "loss": 0.114,
+      "step": 3982
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.129251198279499,
+      "learning_rate": 7.37130819017164e-06,
+      "loss": 0.1392,
+      "step": 3983
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.2371882114083426,
+      "learning_rate": 7.370010682301556e-06,
+      "loss": 0.136,
+      "step": 3984
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.9793736302005662,
+      "learning_rate": 7.368712968550068e-06,
+      "loss": 0.1156,
+      "step": 3985
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8591369501604358,
+      "learning_rate": 7.367415049029909e-06,
+      "loss": 0.1155,
+      "step": 3986
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.2948435519030217,
+      "learning_rate": 7.3661169238538255e-06,
+      "loss": 0.1251,
+      "step": 3987
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9748317862585145,
+      "learning_rate": 7.364818593134586e-06,
+      "loss": 0.1225,
+      "step": 3988
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1696537685157304,
+      "learning_rate": 7.363520056984977e-06,
+      "loss": 0.1463,
+      "step": 3989
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0301156159973863,
+      "learning_rate": 7.362221315517801e-06,
+      "loss": 0.1226,
+      "step": 3990
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.3661832290675964,
+      "learning_rate": 7.3609223688458775e-06,
+      "loss": 0.1212,
+      "step": 3991
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1250212628680742,
+      "learning_rate": 7.359623217082047e-06,
+      "loss": 0.1072,
+      "step": 3992
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0771167817979572,
+      "learning_rate": 7.358323860339165e-06,
+      "loss": 0.1062,
+      "step": 3993
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.121546310168771,
+      "learning_rate": 7.357024298730107e-06,
+      "loss": 0.13,
+      "step": 3994
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.2314031551528086,
+      "learning_rate": 7.355724532367763e-06,
+      "loss": 0.1374,
+      "step": 3995
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0199067337626078,
+      "learning_rate": 7.354424561365046e-06,
+      "loss": 0.1128,
+      "step": 3996
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.3121695047480118,
+      "learning_rate": 7.35312438583488e-06,
+      "loss": 0.1448,
+      "step": 3997
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8886272701778675,
+      "learning_rate": 7.351824005890213e-06,
+      "loss": 0.1146,
+      "step": 3998
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.250740809519039,
+      "learning_rate": 7.350523421644008e-06,
+      "loss": 0.1392,
+      "step": 3999
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1534237343358567,
+      "learning_rate": 7.349222633209246e-06,
+      "loss": 0.1277,
+      "step": 4000
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.079741345723657,
+      "learning_rate": 7.347921640698925e-06,
+      "loss": 0.124,
+      "step": 4001
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.4299492899779693,
+      "learning_rate": 7.3466204442260605e-06,
+      "loss": 0.1246,
+      "step": 4002
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1011338876661463,
+      "learning_rate": 7.345319043903689e-06,
+      "loss": 0.1111,
+      "step": 4003
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7390941763705936,
+      "learning_rate": 7.344017439844862e-06,
+      "loss": 0.1093,
+      "step": 4004
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.077205117707561,
+      "learning_rate": 7.342715632162647e-06,
+      "loss": 0.1172,
+      "step": 4005
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.153527928267279,
+      "learning_rate": 7.3414136209701335e-06,
+      "loss": 0.0958,
+      "step": 4006
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.2102528197082947,
+      "learning_rate": 7.340111406380425e-06,
+      "loss": 0.1179,
+      "step": 4007
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1591957922579565,
+      "learning_rate": 7.338808988506644e-06,
+      "loss": 0.1329,
+      "step": 4008
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9922621824714042,
+      "learning_rate": 7.337506367461933e-06,
+      "loss": 0.1247,
+      "step": 4009
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.2014104256536773,
+      "learning_rate": 7.336203543359446e-06,
+      "loss": 0.129,
+      "step": 4010
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8946157551562837,
+      "learning_rate": 7.3349005163123625e-06,
+      "loss": 0.1038,
+      "step": 4011
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0086623417674847,
+      "learning_rate": 7.333597286433873e-06,
+      "loss": 0.1173,
+      "step": 4012
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2230856327408817,
+      "learning_rate": 7.33229385383719e-06,
+      "loss": 0.1224,
+      "step": 4013
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5089897680310016,
+      "learning_rate": 7.330990218635541e-06,
+      "loss": 0.1147,
+      "step": 4014
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.14394160648444,
+      "learning_rate": 7.329686380942172e-06,
+      "loss": 0.1332,
+      "step": 4015
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2508440942690697,
+      "learning_rate": 7.3283823408703466e-06,
+      "loss": 0.1132,
+      "step": 4016
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1391547224960457,
+      "learning_rate": 7.327078098533347e-06,
+      "loss": 0.1365,
+      "step": 4017
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1416846784491317,
+      "learning_rate": 7.3257736540444715e-06,
+      "loss": 0.123,
+      "step": 4018
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.927397974785919,
+      "learning_rate": 7.324469007517035e-06,
+      "loss": 0.1172,
+      "step": 4019
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.801071893702841,
+      "learning_rate": 7.323164159064372e-06,
+      "loss": 0.1252,
+      "step": 4020
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.3330053919857847,
+      "learning_rate": 7.321859108799836e-06,
+      "loss": 0.1234,
+      "step": 4021
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.09285207012069,
+      "learning_rate": 7.320553856836792e-06,
+      "loss": 0.1045,
+      "step": 4022
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.84329052109295,
+      "learning_rate": 7.319248403288629e-06,
+      "loss": 0.1056,
+      "step": 4023
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.11669083553945,
+      "learning_rate": 7.317942748268753e-06,
+      "loss": 0.1342,
+      "step": 4024
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.015587895149446,
+      "learning_rate": 7.31663689189058e-06,
+      "loss": 0.1282,
+      "step": 4025
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.395268345813073,
+      "learning_rate": 7.315330834267553e-06,
+      "loss": 0.1415,
+      "step": 4026
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0938613208144936,
+      "learning_rate": 7.31402457551313e-06,
+      "loss": 0.123,
+      "step": 4027
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9680644762034556,
+      "learning_rate": 7.31271811574078e-06,
+      "loss": 0.1211,
+      "step": 4028
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.115640362741935,
+      "learning_rate": 7.311411455063997e-06,
+      "loss": 0.1285,
+      "step": 4029
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1288514908174747,
+      "learning_rate": 7.31010459359629e-06,
+      "loss": 0.1331,
+      "step": 4030
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1912549131546237,
+      "learning_rate": 7.308797531451185e-06,
+      "loss": 0.1472,
+      "step": 4031
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2871869773379374,
+      "learning_rate": 7.307490268742224e-06,
+      "loss": 0.1267,
+      "step": 4032
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0732219243656513,
+      "learning_rate": 7.306182805582972e-06,
+      "loss": 0.1304,
+      "step": 4033
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.120908086417422,
+      "learning_rate": 7.304875142087005e-06,
+      "loss": 0.1164,
+      "step": 4034
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9752969660723898,
+      "learning_rate": 7.303567278367918e-06,
+      "loss": 0.1249,
+      "step": 4035
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.305683229994794,
+      "learning_rate": 7.302259214539327e-06,
+      "loss": 0.14,
+      "step": 4036
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0519892133694357,
+      "learning_rate": 7.300950950714859e-06,
+      "loss": 0.1298,
+      "step": 4037
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.229621746050729,
+      "learning_rate": 7.299642487008166e-06,
+      "loss": 0.1185,
+      "step": 4038
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1076792448967523,
+      "learning_rate": 7.298333823532913e-06,
+      "loss": 0.1312,
+      "step": 4039
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8214709736410473,
+      "learning_rate": 7.297024960402779e-06,
+      "loss": 0.1308,
+      "step": 4040
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.099988719820571,
+      "learning_rate": 7.295715897731468e-06,
+      "loss": 0.1289,
+      "step": 4041
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9086126172107463,
+      "learning_rate": 7.294406635632696e-06,
+      "loss": 0.125,
+      "step": 4042
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9388755983126214,
+      "learning_rate": 7.293097174220199e-06,
+      "loss": 0.1373,
+      "step": 4043
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9252420528037315,
+      "learning_rate": 7.291787513607727e-06,
+      "loss": 0.1267,
+      "step": 4044
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.943952817642541,
+      "learning_rate": 7.290477653909052e-06,
+      "loss": 0.1158,
+      "step": 4045
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.186985218661055,
+      "learning_rate": 7.289167595237957e-06,
+      "loss": 0.139,
+      "step": 4046
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1620383014593028,
+      "learning_rate": 7.28785733770825e-06,
+      "loss": 0.1332,
+      "step": 4047
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.012440181711628,
+      "learning_rate": 7.28654688143375e-06,
+      "loss": 0.1339,
+      "step": 4048
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0694763750739615,
+      "learning_rate": 7.285236226528297e-06,
+      "loss": 0.1259,
+      "step": 4049
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.879491122272998,
+      "learning_rate": 7.283925373105745e-06,
+      "loss": 0.1021,
+      "step": 4050
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.065723860322678,
+      "learning_rate": 7.282614321279969e-06,
+      "loss": 0.1445,
+      "step": 4051
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3948578867793446,
+      "learning_rate": 7.281303071164858e-06,
+      "loss": 0.1073,
+      "step": 4052
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2230764969291266,
+      "learning_rate": 7.279991622874319e-06,
+      "loss": 0.1155,
+      "step": 4053
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3494294967172453,
+      "learning_rate": 7.278679976522279e-06,
+      "loss": 0.1298,
+      "step": 4054
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0682955922314874,
+      "learning_rate": 7.277368132222678e-06,
+      "loss": 0.122,
+      "step": 4055
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1057988091073554,
+      "learning_rate": 7.276056090089475e-06,
+      "loss": 0.1307,
+      "step": 4056
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.066580174324056,
+      "learning_rate": 7.274743850236649e-06,
+      "loss": 0.1206,
+      "step": 4057
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.059543126056794,
+      "learning_rate": 7.273431412778189e-06,
+      "loss": 0.1411,
+      "step": 4058
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8810898036468067,
+      "learning_rate": 7.272118777828109e-06,
+      "loss": 0.1037,
+      "step": 4059
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.05918874136541,
+      "learning_rate": 7.270805945500436e-06,
+      "loss": 0.1222,
+      "step": 4060
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9418198299457987,
+      "learning_rate": 7.269492915909214e-06,
+      "loss": 0.0895,
+      "step": 4061
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0836958443411864,
+      "learning_rate": 7.268179689168507e-06,
+      "loss": 0.1376,
+      "step": 4062
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.16593274620396,
+      "learning_rate": 7.266866265392394e-06,
+      "loss": 0.1376,
+      "step": 4063
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.022972715275929,
+      "learning_rate": 7.265552644694969e-06,
+      "loss": 0.1225,
+      "step": 4064
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0873612386337173,
+      "learning_rate": 7.264238827190346e-06,
+      "loss": 0.1178,
+      "step": 4065
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.985209312165002,
+      "learning_rate": 7.2629248129926576e-06,
+      "loss": 0.1114,
+      "step": 4066
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.04820684925789,
+      "learning_rate": 7.26161060221605e-06,
+      "loss": 0.1193,
+      "step": 4067
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.174628697363053,
+      "learning_rate": 7.2602961949746886e-06,
+      "loss": 0.1323,
+      "step": 4068
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8469254903009882,
+      "learning_rate": 7.258981591382756e-06,
+      "loss": 0.1181,
+      "step": 4069
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.190923853319206,
+      "learning_rate": 7.257666791554448e-06,
+      "loss": 0.1345,
+      "step": 4070
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2819848208376468,
+      "learning_rate": 7.256351795603982e-06,
+      "loss": 0.1449,
+      "step": 4071
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1012960945037884,
+      "learning_rate": 7.255036603645593e-06,
+      "loss": 0.1231,
+      "step": 4072
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9111178390492092,
+      "learning_rate": 7.253721215793528e-06,
+      "loss": 0.1205,
+      "step": 4073
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9868463170310164,
+      "learning_rate": 7.252405632162054e-06,
+      "loss": 0.1326,
+      "step": 4074
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8498058036585032,
+      "learning_rate": 7.251089852865458e-06,
+      "loss": 0.1111,
+      "step": 4075
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.130595669503501,
+      "learning_rate": 7.2497738780180375e-06,
+      "loss": 0.1496,
+      "step": 4076
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7243798034060782,
+      "learning_rate": 7.248457707734113e-06,
+      "loss": 0.1065,
+      "step": 4077
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.021871491947951,
+      "learning_rate": 7.247141342128017e-06,
+      "loss": 0.098,
+      "step": 4078
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8678339379041473,
+      "learning_rate": 7.245824781314104e-06,
+      "loss": 0.1021,
+      "step": 4079
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1691268925388405,
+      "learning_rate": 7.24450802540674e-06,
+      "loss": 0.1184,
+      "step": 4080
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0599849185264376,
+      "learning_rate": 7.243191074520314e-06,
+      "loss": 0.1128,
+      "step": 4081
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.397295444122207,
+      "learning_rate": 7.2418739287692266e-06,
+      "loss": 0.1494,
+      "step": 4082
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8887882863137924,
+      "learning_rate": 7.240556588267897e-06,
+      "loss": 0.1334,
+      "step": 4083
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0671605015297705,
+      "learning_rate": 7.2392390531307634e-06,
+      "loss": 0.1297,
+      "step": 4084
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9278132265550376,
+      "learning_rate": 7.237921323472279e-06,
+      "loss": 0.1235,
+      "step": 4085
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.126243458354478,
+      "learning_rate": 7.236603399406914e-06,
+      "loss": 0.1264,
+      "step": 4086
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.073927395193306,
+      "learning_rate": 7.235285281049154e-06,
+      "loss": 0.132,
+      "step": 4087
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.00195967613182,
+      "learning_rate": 7.233966968513506e-06,
+      "loss": 0.1181,
+      "step": 4088
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9167174467624488,
+      "learning_rate": 7.23264846191449e-06,
+      "loss": 0.1247,
+      "step": 4089
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8760304689855907,
+      "learning_rate": 7.231329761366642e-06,
+      "loss": 0.1191,
+      "step": 4090
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9159775973618893,
+      "learning_rate": 7.230010866984518e-06,
+      "loss": 0.1162,
+      "step": 4091
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9277798756667899,
+      "learning_rate": 7.2286917788826926e-06,
+      "loss": 0.0942,
+      "step": 4092
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.111950398625085,
+      "learning_rate": 7.2273724971757484e-06,
+      "loss": 0.1279,
+      "step": 4093
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9492437452475793,
+      "learning_rate": 7.226053021978295e-06,
+      "loss": 0.1159,
+      "step": 4094
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1127674985376426,
+      "learning_rate": 7.2247333534049536e-06,
+      "loss": 0.1219,
+      "step": 4095
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.337096983319986,
+      "learning_rate": 7.2234134915703616e-06,
+      "loss": 0.1232,
+      "step": 4096
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.191416578479189,
+      "learning_rate": 7.222093436589175e-06,
+      "loss": 0.1345,
+      "step": 4097
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9533956736046576,
+      "learning_rate": 7.220773188576068e-06,
+      "loss": 0.1088,
+      "step": 4098
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.905760269854261,
+      "learning_rate": 7.219452747645728e-06,
+      "loss": 0.1023,
+      "step": 4099
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3912067840711795,
+      "learning_rate": 7.218132113912859e-06,
+      "loss": 0.1539,
+      "step": 4100
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.210437236899646,
+      "learning_rate": 7.216811287492189e-06,
+      "loss": 0.137,
+      "step": 4101
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2701933019852327,
+      "learning_rate": 7.215490268498453e-06,
+      "loss": 0.1295,
+      "step": 4102
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2963509680678045,
+      "learning_rate": 7.2141690570464074e-06,
+      "loss": 0.1415,
+      "step": 4103
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.953002537439698,
+      "learning_rate": 7.212847653250828e-06,
+      "loss": 0.1015,
+      "step": 4104
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2133252702056136,
+      "learning_rate": 7.211526057226502e-06,
+      "loss": 0.1454,
+      "step": 4105
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7610991350213612,
+      "learning_rate": 7.2102042690882356e-06,
+      "loss": 0.0934,
+      "step": 4106
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.298450285512435,
+      "learning_rate": 7.208882288950854e-06,
+      "loss": 0.146,
+      "step": 4107
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.237744583653582,
+      "learning_rate": 7.207560116929192e-06,
+      "loss": 0.1247,
+      "step": 4108
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8205207035711697,
+      "learning_rate": 7.20623775313811e-06,
+      "loss": 0.1116,
+      "step": 4109
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.861896084213048,
+      "learning_rate": 7.204915197692481e-06,
+      "loss": 0.1191,
+      "step": 4110
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.107175259116252,
+      "learning_rate": 7.203592450707193e-06,
+      "loss": 0.1435,
+      "step": 4111
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.717745939394215,
+      "learning_rate": 7.202269512297153e-06,
+      "loss": 0.1128,
+      "step": 4112
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.291724950550943,
+      "learning_rate": 7.200946382577284e-06,
+      "loss": 0.1233,
+      "step": 4113
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0100059226373426,
+      "learning_rate": 7.199623061662524e-06,
+      "loss": 0.134,
+      "step": 4114
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.061720671735907,
+      "learning_rate": 7.1982995496678306e-06,
+      "loss": 0.1285,
+      "step": 4115
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9410259925562126,
+      "learning_rate": 7.196975846708176e-06,
+      "loss": 0.1054,
+      "step": 4116
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0518441549823327,
+      "learning_rate": 7.19565195289855e-06,
+      "loss": 0.1256,
+      "step": 4117
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7990715197265263,
+      "learning_rate": 7.194327868353958e-06,
+      "loss": 0.0977,
+      "step": 4118
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2207897239223384,
+      "learning_rate": 7.193003593189423e-06,
+      "loss": 0.122,
+      "step": 4119
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2162668701755543,
+      "learning_rate": 7.191679127519981e-06,
+      "loss": 0.13,
+      "step": 4120
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.871066645582895,
+      "learning_rate": 7.190354471460692e-06,
+      "loss": 0.1225,
+      "step": 4121
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.104715667461744,
+      "learning_rate": 7.189029625126627e-06,
+      "loss": 0.1234,
+      "step": 4122
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.05382869277293,
+      "learning_rate": 7.187704588632871e-06,
+      "loss": 0.14,
+      "step": 4123
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9002726339466374,
+      "learning_rate": 7.186379362094533e-06,
+      "loss": 0.0959,
+      "step": 4124
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8299457765103404,
+      "learning_rate": 7.185053945626734e-06,
+      "loss": 0.1054,
+      "step": 4125
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9894973271417298,
+      "learning_rate": 7.183728339344611e-06,
+      "loss": 0.1181,
+      "step": 4126
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9189738564018048,
+      "learning_rate": 7.182402543363319e-06,
+      "loss": 0.1159,
+      "step": 4127
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9614473738193228,
+      "learning_rate": 7.1810765577980305e-06,
+      "loss": 0.1084,
+      "step": 4128
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.102583468885272,
+      "learning_rate": 7.179750382763931e-06,
+      "loss": 0.1053,
+      "step": 4129
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0328218705613574,
+      "learning_rate": 7.178424018376224e-06,
+      "loss": 0.1301,
+      "step": 4130
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.066696744552882,
+      "learning_rate": 7.177097464750134e-06,
+      "loss": 0.119,
+      "step": 4131
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.071562043633715,
+      "learning_rate": 7.175770722000893e-06,
+      "loss": 0.1305,
+      "step": 4132
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.294928397577337,
+      "learning_rate": 7.174443790243758e-06,
+      "loss": 0.133,
+      "step": 4133
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0890502330090364,
+      "learning_rate": 7.173116669593997e-06,
+      "loss": 0.139,
+      "step": 4134
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1316413776312366,
+      "learning_rate": 7.171789360166896e-06,
+      "loss": 0.1233,
+      "step": 4135
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9121317906783983,
+      "learning_rate": 7.170461862077759e-06,
+      "loss": 0.1187,
+      "step": 4136
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1341586131665102,
+      "learning_rate": 7.169134175441904e-06,
+      "loss": 0.129,
+      "step": 4137
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.152178980078153,
+      "learning_rate": 7.167806300374665e-06,
+      "loss": 0.1355,
+      "step": 4138
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.260737540531499,
+      "learning_rate": 7.166478236991396e-06,
+      "loss": 0.144,
+      "step": 4139
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.3243095875135924,
+      "learning_rate": 7.165149985407465e-06,
+      "loss": 0.1214,
+      "step": 4140
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9983298131638916,
+      "learning_rate": 7.163821545738254e-06,
+      "loss": 0.097,
+      "step": 4141
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.14539639922315,
+      "learning_rate": 7.162492918099167e-06,
+      "loss": 0.1354,
+      "step": 4142
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.073087929254994,
+      "learning_rate": 7.16116410260562e-06,
+      "loss": 0.1284,
+      "step": 4143
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8304532592972178,
+      "learning_rate": 7.1598350993730435e-06,
+      "loss": 0.1099,
+      "step": 4144
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.290558617469023,
+      "learning_rate": 7.158505908516891e-06,
+      "loss": 0.1211,
+      "step": 4145
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.834973411605714,
+      "learning_rate": 7.157176530152628e-06,
+      "loss": 0.1192,
+      "step": 4146
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.069403813635896,
+      "learning_rate": 7.155846964395734e-06,
+      "loss": 0.1241,
+      "step": 4147
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9038946446698481,
+      "learning_rate": 7.154517211361709e-06,
+      "loss": 0.1285,
+      "step": 4148
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1139802214615795,
+      "learning_rate": 7.153187271166071e-06,
+      "loss": 0.1209,
+      "step": 4149
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.5077831189380517,
+      "learning_rate": 7.151857143924345e-06,
+      "loss": 0.1387,
+      "step": 4150
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2675835021650275,
+      "learning_rate": 7.150526829752085e-06,
+      "loss": 0.1404,
+      "step": 4151
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8652222644044676,
+      "learning_rate": 7.14919632876485e-06,
+      "loss": 0.1228,
+      "step": 4152
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8811022419661327,
+      "learning_rate": 7.147865641078221e-06,
+      "loss": 0.1173,
+      "step": 4153
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1861974372829143,
+      "learning_rate": 7.146534766807794e-06,
+      "loss": 0.1421,
+      "step": 4154
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0373985670142067,
+      "learning_rate": 7.145203706069183e-06,
+      "loss": 0.1448,
+      "step": 4155
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0403613502975824,
+      "learning_rate": 7.143872458978013e-06,
+      "loss": 0.1294,
+      "step": 4156
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0392816974201753,
+      "learning_rate": 7.142541025649932e-06,
+      "loss": 0.128,
+      "step": 4157
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9652661242245302,
+      "learning_rate": 7.1412094062005985e-06,
+      "loss": 0.1199,
+      "step": 4158
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.969055061144305,
+      "learning_rate": 7.139877600745691e-06,
+      "loss": 0.1145,
+      "step": 4159
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.90720565346627,
+      "learning_rate": 7.138545609400901e-06,
+      "loss": 0.1341,
+      "step": 4160
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.033805183744819,
+      "learning_rate": 7.13721343228194e-06,
+      "loss": 0.1222,
+      "step": 4161
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7017616871810028,
+      "learning_rate": 7.135881069504531e-06,
+      "loss": 0.0993,
+      "step": 4162
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.003142843982908,
+      "learning_rate": 7.134548521184417e-06,
+      "loss": 0.1227,
+      "step": 4163
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8475285773033214,
+      "learning_rate": 7.1332157874373565e-06,
+      "loss": 0.1044,
+      "step": 4164
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9241259271038214,
+      "learning_rate": 7.1318828683791205e-06,
+      "loss": 0.114,
+      "step": 4165
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.873873295806884,
+      "learning_rate": 7.130549764125502e-06,
+      "loss": 0.1159,
+      "step": 4166
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1965279420717434,
+      "learning_rate": 7.129216474792305e-06,
+      "loss": 0.1238,
+      "step": 4167
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0584641430052115,
+      "learning_rate": 7.127883000495353e-06,
+      "loss": 0.1282,
+      "step": 4168
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2082759397085514,
+      "learning_rate": 7.1265493413504815e-06,
+      "loss": 0.1227,
+      "step": 4169
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0151240493765266,
+      "learning_rate": 7.125215497473548e-06,
+      "loss": 0.1054,
+      "step": 4170
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.000606501275892,
+      "learning_rate": 7.123881468980419e-06,
+      "loss": 0.1216,
+      "step": 4171
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.4278567707426846,
+      "learning_rate": 7.122547255986985e-06,
+      "loss": 0.1339,
+      "step": 4172
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.128830850778144,
+      "learning_rate": 7.121212858609146e-06,
+      "loss": 0.1238,
+      "step": 4173
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9512308665181886,
+      "learning_rate": 7.119878276962818e-06,
+      "loss": 0.1188,
+      "step": 4174
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8742164132560155,
+      "learning_rate": 7.11854351116394e-06,
+      "loss": 0.1175,
+      "step": 4175
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9961944791213952,
+      "learning_rate": 7.11720856132846e-06,
+      "loss": 0.1126,
+      "step": 4176
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0353395389191196,
+      "learning_rate": 7.115873427572342e-06,
+      "loss": 0.1391,
+      "step": 4177
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0767031568616745,
+      "learning_rate": 7.114538110011573e-06,
+      "loss": 0.1444,
+      "step": 4178
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8728855260683073,
+      "learning_rate": 7.1132026087621485e-06,
+      "loss": 0.1254,
+      "step": 4179
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0147393998570635,
+      "learning_rate": 7.111866923940083e-06,
+      "loss": 0.1159,
+      "step": 4180
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.04122326472132,
+      "learning_rate": 7.110531055661406e-06,
+      "loss": 0.1302,
+      "step": 4181
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9710831441334116,
+      "learning_rate": 7.109195004042164e-06,
+      "loss": 0.1104,
+      "step": 4182
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8442436652944234,
+      "learning_rate": 7.10785876919842e-06,
+      "loss": 0.1059,
+      "step": 4183
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8174687215989198,
+      "learning_rate": 7.106522351246252e-06,
+      "loss": 0.1226,
+      "step": 4184
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7421700693815974,
+      "learning_rate": 7.105185750301751e-06,
+      "loss": 0.1223,
+      "step": 4185
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9449522372366252,
+      "learning_rate": 7.10384896648103e-06,
+      "loss": 0.1147,
+      "step": 4186
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.072921319594952,
+      "learning_rate": 7.102511999900213e-06,
+      "loss": 0.12,
+      "step": 4187
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7985882639390682,
+      "learning_rate": 7.101174850675442e-06,
+      "loss": 0.1165,
+      "step": 4188
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8685823692413241,
+      "learning_rate": 7.099837518922873e-06,
+      "loss": 0.1167,
+      "step": 4189
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.74136200040315,
+      "learning_rate": 7.098500004758682e-06,
+      "loss": 0.1135,
+      "step": 4190
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.961049470256503,
+      "learning_rate": 7.097162308299055e-06,
+      "loss": 0.1186,
+      "step": 4191
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.078256771796782,
+      "learning_rate": 7.095824429660199e-06,
+      "loss": 0.1431,
+      "step": 4192
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.774420858630228,
+      "learning_rate": 7.094486368958334e-06,
+      "loss": 0.1062,
+      "step": 4193
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9955226694247632,
+      "learning_rate": 7.093148126309697e-06,
+      "loss": 0.1259,
+      "step": 4194
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.205935054140034,
+      "learning_rate": 7.091809701830539e-06,
+      "loss": 0.1105,
+      "step": 4195
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.3782782663129436,
+      "learning_rate": 7.090471095637129e-06,
+      "loss": 0.1471,
+      "step": 4196
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.173317062903518,
+      "learning_rate": 7.0891323078457505e-06,
+      "loss": 0.1346,
+      "step": 4197
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.295335070599442,
+      "learning_rate": 7.087793338572705e-06,
+      "loss": 0.1523,
+      "step": 4198
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0479439613412205,
+      "learning_rate": 7.086454187934306e-06,
+      "loss": 0.1195,
+      "step": 4199
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.1312433479321022,
+      "learning_rate": 7.085114856046884e-06,
+      "loss": 0.1355,
+      "step": 4200
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9229956267624408,
+      "learning_rate": 7.083775343026789e-06,
+      "loss": 0.1167,
+      "step": 4201
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.1150266225406,
+      "learning_rate": 7.082435648990381e-06,
+      "loss": 0.1139,
+      "step": 4202
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8607211296616,
+      "learning_rate": 7.08109577405404e-06,
+      "loss": 0.128,
+      "step": 4203
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9579360566153368,
+      "learning_rate": 7.079755718334158e-06,
+      "loss": 0.1396,
+      "step": 4204
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.5958254655223028,
+      "learning_rate": 7.0784154819471484e-06,
+      "loss": 0.1068,
+      "step": 4205
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.370232076423813,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 0.1564,
+      "step": 4206
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0578095908609813,
+      "learning_rate": 7.075734467637454e-06,
+      "loss": 0.1346,
+      "step": 4207
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7834407060485793,
+      "learning_rate": 7.074393689947671e-06,
+      "loss": 0.1047,
+      "step": 4208
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.00065412560574,
+      "learning_rate": 7.073052732056553e-06,
+      "loss": 0.1205,
+      "step": 4209
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9288737725165355,
+      "learning_rate": 7.07171159408059e-06,
+      "loss": 0.1154,
+      "step": 4210
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0338613156759084,
+      "learning_rate": 7.070370276136287e-06,
+      "loss": 0.1514,
+      "step": 4211
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.046183984541071,
+      "learning_rate": 7.06902877834016e-06,
+      "loss": 0.1223,
+      "step": 4212
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6956898338198543,
+      "learning_rate": 7.0676871008087465e-06,
+      "loss": 0.1038,
+      "step": 4213
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.817554657649637,
+      "learning_rate": 7.066345243658598e-06,
+      "loss": 0.1205,
+      "step": 4214
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.050161417853629,
+      "learning_rate": 7.065003207006278e-06,
+      "loss": 0.1206,
+      "step": 4215
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8997314485130785,
+      "learning_rate": 7.06366099096837e-06,
+      "loss": 0.0958,
+      "step": 4216
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2671000738009854,
+      "learning_rate": 7.062318595661475e-06,
+      "loss": 0.1653,
+      "step": 4217
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.061360639217304,
+      "learning_rate": 7.0609760212021994e-06,
+      "loss": 0.1151,
+      "step": 4218
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.102723365197078,
+      "learning_rate": 7.059633267707176e-06,
+      "loss": 0.114,
+      "step": 4219
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.109063012503294,
+      "learning_rate": 7.058290335293048e-06,
+      "loss": 0.128,
+      "step": 4220
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.117995857791581,
+      "learning_rate": 7.056947224076475e-06,
+      "loss": 0.1337,
+      "step": 4221
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.158029791174112,
+      "learning_rate": 7.055603934174132e-06,
+      "loss": 0.1304,
+      "step": 4222
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0278188098729975,
+      "learning_rate": 7.054260465702712e-06,
+      "loss": 0.13,
+      "step": 4223
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.389617138982808,
+      "learning_rate": 7.052916818778918e-06,
+      "loss": 0.1422,
+      "step": 4224
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0315604299132906,
+      "learning_rate": 7.051572993519474e-06,
+      "loss": 0.0992,
+      "step": 4225
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.212832764798801,
+      "learning_rate": 7.050228990041117e-06,
+      "loss": 0.1428,
+      "step": 4226
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8986893222541308,
+      "learning_rate": 7.048884808460599e-06,
+      "loss": 0.1294,
+      "step": 4227
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9259685203070438,
+      "learning_rate": 7.047540448894687e-06,
+      "loss": 0.1018,
+      "step": 4228
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7217273337801775,
+      "learning_rate": 7.04619591146017e-06,
+      "loss": 0.1121,
+      "step": 4229
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8423791308770312,
+      "learning_rate": 7.044851196273841e-06,
+      "loss": 0.1184,
+      "step": 4230
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0711187783761025,
+      "learning_rate": 7.0435063034525164e-06,
+      "loss": 0.123,
+      "step": 4231
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.95835572032229,
+      "learning_rate": 7.042161233113029e-06,
+      "loss": 0.1407,
+      "step": 4232
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7439597108143134,
+      "learning_rate": 7.040815985372221e-06,
+      "loss": 0.1195,
+      "step": 4233
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7258780266423253,
+      "learning_rate": 7.039470560346955e-06,
+      "loss": 0.1067,
+      "step": 4234
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.344276126992682,
+      "learning_rate": 7.038124958154108e-06,
+      "loss": 0.1525,
+      "step": 4235
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9654926615433947,
+      "learning_rate": 7.036779178910569e-06,
+      "loss": 0.1396,
+      "step": 4236
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.943138094752504,
+      "learning_rate": 7.035433222733246e-06,
+      "loss": 0.1147,
+      "step": 4237
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9334169758925888,
+      "learning_rate": 7.0340870897390635e-06,
+      "loss": 0.1085,
+      "step": 4238
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.6837298765295836,
+      "learning_rate": 7.032740780044957e-06,
+      "loss": 0.1026,
+      "step": 4239
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9445648272349414,
+      "learning_rate": 7.031394293767879e-06,
+      "loss": 0.1286,
+      "step": 4240
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8476097235385962,
+      "learning_rate": 7.030047631024801e-06,
+      "loss": 0.116,
+      "step": 4241
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2111881567549516,
+      "learning_rate": 7.028700791932703e-06,
+      "loss": 0.1287,
+      "step": 4242
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1320852057264488,
+      "learning_rate": 7.027353776608587e-06,
+      "loss": 0.1183,
+      "step": 4243
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9700567037683123,
+      "learning_rate": 7.026006585169467e-06,
+      "loss": 0.1149,
+      "step": 4244
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.3428504867811513,
+      "learning_rate": 7.024659217732372e-06,
+      "loss": 0.1333,
+      "step": 4245
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.080400968766066,
+      "learning_rate": 7.023311674414346e-06,
+      "loss": 0.1295,
+      "step": 4246
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2448668670048293,
+      "learning_rate": 7.0219639553324525e-06,
+      "loss": 0.1343,
+      "step": 4247
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0593018432261805,
+      "learning_rate": 7.020616060603765e-06,
+      "loss": 0.124,
+      "step": 4248
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8586380750129603,
+      "learning_rate": 7.019267990345372e-06,
+      "loss": 0.1195,
+      "step": 4249
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.047150637397853,
+      "learning_rate": 7.017919744674384e-06,
+      "loss": 0.1012,
+      "step": 4250
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1457479593067754,
+      "learning_rate": 7.016571323707919e-06,
+      "loss": 0.1299,
+      "step": 4251
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1857211462970683,
+      "learning_rate": 7.0152227275631144e-06,
+      "loss": 0.1335,
+      "step": 4252
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.932837708680595,
+      "learning_rate": 7.013873956357123e-06,
+      "loss": 0.1211,
+      "step": 4253
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.5781721297960525,
+      "learning_rate": 7.0125250102071115e-06,
+      "loss": 0.0911,
+      "step": 4254
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9521296349302901,
+      "learning_rate": 7.011175889230261e-06,
+      "loss": 0.1222,
+      "step": 4255
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.076519599463524,
+      "learning_rate": 7.009826593543769e-06,
+      "loss": 0.0995,
+      "step": 4256
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8167851730626865,
+      "learning_rate": 7.008477123264849e-06,
+      "loss": 0.1253,
+      "step": 4257
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8448176798392104,
+      "learning_rate": 7.007127478510727e-06,
+      "loss": 0.1167,
+      "step": 4258
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2332308101911242,
+      "learning_rate": 7.005777659398647e-06,
+      "loss": 0.1486,
+      "step": 4259
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0190051945147904,
+      "learning_rate": 7.004427666045867e-06,
+      "loss": 0.1317,
+      "step": 4260
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8764438738661249,
+      "learning_rate": 7.00307749856966e-06,
+      "loss": 0.1208,
+      "step": 4261
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0725838482772905,
+      "learning_rate": 7.001727157087316e-06,
+      "loss": 0.1225,
+      "step": 4262
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0932489701436228,
+      "learning_rate": 7.0003766417161335e-06,
+      "loss": 0.1161,
+      "step": 4263
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.00441893964883,
+      "learning_rate": 6.999025952573435e-06,
+      "loss": 0.14,
+      "step": 4264
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.846084268613071,
+      "learning_rate": 6.997675089776554e-06,
+      "loss": 0.1287,
+      "step": 4265
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.85172270423492,
+      "learning_rate": 6.9963240534428374e-06,
+      "loss": 0.1022,
+      "step": 4266
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.3668144447478334,
+      "learning_rate": 6.994972843689651e-06,
+      "loss": 0.1612,
+      "step": 4267
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.6711942449879786,
+      "learning_rate": 6.993621460634371e-06,
+      "loss": 0.1033,
+      "step": 4268
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.712624127512383,
+      "learning_rate": 6.992269904394392e-06,
+      "loss": 0.0943,
+      "step": 4269
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.800327483910594,
+      "learning_rate": 6.990918175087124e-06,
+      "loss": 0.1015,
+      "step": 4270
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0961204019571076,
+      "learning_rate": 6.989566272829989e-06,
+      "loss": 0.1448,
+      "step": 4271
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.054936268423516,
+      "learning_rate": 6.98821419774043e-06,
+      "loss": 0.1341,
+      "step": 4272
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8976777208083724,
+      "learning_rate": 6.986861949935897e-06,
+      "loss": 0.1069,
+      "step": 4273
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0912721242077508,
+      "learning_rate": 6.985509529533859e-06,
+      "loss": 0.132,
+      "step": 4274
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9678223030561028,
+      "learning_rate": 6.984156936651802e-06,
+      "loss": 0.119,
+      "step": 4275
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7853394273208971,
+      "learning_rate": 6.982804171407225e-06,
+      "loss": 0.1188,
+      "step": 4276
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7908139763191069,
+      "learning_rate": 6.981451233917639e-06,
+      "loss": 0.1039,
+      "step": 4277
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.084509822331322,
+      "learning_rate": 6.980098124300576e-06,
+      "loss": 0.1206,
+      "step": 4278
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.052934168735636,
+      "learning_rate": 6.978744842673578e-06,
+      "loss": 0.1291,
+      "step": 4279
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.3931468954730373,
+      "learning_rate": 6.977391389154204e-06,
+      "loss": 0.1395,
+      "step": 4280
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9541633086080412,
+      "learning_rate": 6.9760377638600295e-06,
+      "loss": 0.1124,
+      "step": 4281
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.868302049787096,
+      "learning_rate": 6.974683966908642e-06,
+      "loss": 0.1187,
+      "step": 4282
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9710634722054383,
+      "learning_rate": 6.973329998417643e-06,
+      "loss": 0.1093,
+      "step": 4283
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1819134685202495,
+      "learning_rate": 6.971975858504653e-06,
+      "loss": 0.1345,
+      "step": 4284
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9557615338724021,
+      "learning_rate": 6.970621547287306e-06,
+      "loss": 0.1051,
+      "step": 4285
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0865632291687874,
+      "learning_rate": 6.969267064883247e-06,
+      "loss": 0.1213,
+      "step": 4286
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0096697189694095,
+      "learning_rate": 6.967912411410143e-06,
+      "loss": 0.1279,
+      "step": 4287
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.897429747021453,
+      "learning_rate": 6.966557586985671e-06,
+      "loss": 0.1304,
+      "step": 4288
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.6844514227148697,
+      "learning_rate": 6.965202591727521e-06,
+      "loss": 0.0887,
+      "step": 4289
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.938360386958679,
+      "learning_rate": 6.9638474257534025e-06,
+      "loss": 0.1354,
+      "step": 4290
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.047576512757608,
+      "learning_rate": 6.96249208918104e-06,
+      "loss": 0.1319,
+      "step": 4291
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8197269478309723,
+      "learning_rate": 6.961136582128167e-06,
+      "loss": 0.0974,
+      "step": 4292
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9062555607494636,
+      "learning_rate": 6.959780904712538e-06,
+      "loss": 0.1255,
+      "step": 4293
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0897857656234136,
+      "learning_rate": 6.95842505705192e-06,
+      "loss": 0.1296,
+      "step": 4294
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0633233737730983,
+      "learning_rate": 6.957069039264093e-06,
+      "loss": 0.1415,
+      "step": 4295
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.231369399217659,
+      "learning_rate": 6.9557128514668535e-06,
+      "loss": 0.142,
+      "step": 4296
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9044511720153494,
+      "learning_rate": 6.954356493778016e-06,
+      "loss": 0.1223,
+      "step": 4297
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9876611952437309,
+      "learning_rate": 6.952999966315402e-06,
+      "loss": 0.139,
+      "step": 4298
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.6674830669223852,
+      "learning_rate": 6.951643269196855e-06,
+      "loss": 0.1068,
+      "step": 4299
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9194246382596751,
+      "learning_rate": 6.950286402540231e-06,
+      "loss": 0.1266,
+      "step": 4300
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.879631995162897,
+      "learning_rate": 6.948929366463397e-06,
+      "loss": 0.1114,
+      "step": 4301
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7327504748182798,
+      "learning_rate": 6.94757216108424e-06,
+      "loss": 0.0962,
+      "step": 4302
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9077091626100586,
+      "learning_rate": 6.9462147865206616e-06,
+      "loss": 0.125,
+      "step": 4303
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.7212628925580487,
+      "learning_rate": 6.944857242890573e-06,
+      "loss": 0.1327,
+      "step": 4304
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0529262741899923,
+      "learning_rate": 6.943499530311903e-06,
+      "loss": 0.115,
+      "step": 4305
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.5875285032245656,
+      "learning_rate": 6.942141648902599e-06,
+      "loss": 0.0977,
+      "step": 4306
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.02722199605142,
+      "learning_rate": 6.940783598780613e-06,
+      "loss": 0.126,
+      "step": 4307
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0379055128484453,
+      "learning_rate": 6.939425380063924e-06,
+      "loss": 0.1193,
+      "step": 4308
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0838819666108805,
+      "learning_rate": 6.938066992870519e-06,
+      "loss": 0.1246,
+      "step": 4309
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1960063506491663,
+      "learning_rate": 6.936708437318397e-06,
+      "loss": 0.1393,
+      "step": 4310
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7414271553138678,
+      "learning_rate": 6.935349713525577e-06,
+      "loss": 0.1023,
+      "step": 4311
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9884633759299402,
+      "learning_rate": 6.93399082161009e-06,
+      "loss": 0.1168,
+      "step": 4312
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9144685669895434,
+      "learning_rate": 6.932631761689982e-06,
+      "loss": 0.1052,
+      "step": 4313
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8654138829745965,
+      "learning_rate": 6.931272533883313e-06,
+      "loss": 0.11,
+      "step": 4314
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.049301376570788,
+      "learning_rate": 6.929913138308162e-06,
+      "loss": 0.1103,
+      "step": 4315
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.061390570549711,
+      "learning_rate": 6.928553575082615e-06,
+      "loss": 0.1312,
+      "step": 4316
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9287779484180387,
+      "learning_rate": 6.927193844324777e-06,
+      "loss": 0.1078,
+      "step": 4317
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.231122509242185,
+      "learning_rate": 6.925833946152769e-06,
+      "loss": 0.144,
+      "step": 4318
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0769626714806115,
+      "learning_rate": 6.924473880684721e-06,
+      "loss": 0.11,
+      "step": 4319
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.479447013210604,
+      "learning_rate": 6.923113648038784e-06,
+      "loss": 0.1532,
+      "step": 4320
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0848371895590168,
+      "learning_rate": 6.921753248333122e-06,
+      "loss": 0.1185,
+      "step": 4321
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.4081735322733953,
+      "learning_rate": 6.920392681685908e-06,
+      "loss": 0.1547,
+      "step": 4322
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3946156723427046,
+      "learning_rate": 6.919031948215335e-06,
+      "loss": 0.1475,
+      "step": 4323
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8642727007928703,
+      "learning_rate": 6.917671048039611e-06,
+      "loss": 0.1212,
+      "step": 4324
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0712212641651067,
+      "learning_rate": 6.916309981276954e-06,
+      "loss": 0.1294,
+      "step": 4325
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.204484927308151,
+      "learning_rate": 6.9149487480456e-06,
+      "loss": 0.1312,
+      "step": 4326
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7955964672117872,
+      "learning_rate": 6.913587348463802e-06,
+      "loss": 0.1106,
+      "step": 4327
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8421985067183215,
+      "learning_rate": 6.912225782649818e-06,
+      "loss": 0.1175,
+      "step": 4328
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6671289885611584,
+      "learning_rate": 6.910864050721928e-06,
+      "loss": 0.1006,
+      "step": 4329
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.891298841138614,
+      "learning_rate": 6.909502152798428e-06,
+      "loss": 0.1287,
+      "step": 4330
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9311647863765549,
+      "learning_rate": 6.908140088997623e-06,
+      "loss": 0.1229,
+      "step": 4331
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6658519157331932,
+      "learning_rate": 6.906777859437835e-06,
+      "loss": 0.0829,
+      "step": 4332
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8796674805603164,
+      "learning_rate": 6.9054154642374e-06,
+      "loss": 0.1084,
+      "step": 4333
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7905634001115056,
+      "learning_rate": 6.904052903514668e-06,
+      "loss": 0.1046,
+      "step": 4334
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.051185888688871,
+      "learning_rate": 6.902690177388003e-06,
+      "loss": 0.1261,
+      "step": 4335
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9951267917742708,
+      "learning_rate": 6.901327285975787e-06,
+      "loss": 0.1299,
+      "step": 4336
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.931504460038208,
+      "learning_rate": 6.899964229396412e-06,
+      "loss": 0.1252,
+      "step": 4337
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.114802184147079,
+      "learning_rate": 6.898601007768285e-06,
+      "loss": 0.1236,
+      "step": 4338
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.174600757371958,
+      "learning_rate": 6.897237621209831e-06,
+      "loss": 0.121,
+      "step": 4339
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1075543090933153,
+      "learning_rate": 6.8958740698394835e-06,
+      "loss": 0.1432,
+      "step": 4340
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9116202887235025,
+      "learning_rate": 6.894510353775694e-06,
+      "loss": 0.1295,
+      "step": 4341
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.970827893263525,
+      "learning_rate": 6.89314647313693e-06,
+      "loss": 0.1131,
+      "step": 4342
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1604009117312084,
+      "learning_rate": 6.891782428041668e-06,
+      "loss": 0.1326,
+      "step": 4343
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9018141443868395,
+      "learning_rate": 6.890418218608403e-06,
+      "loss": 0.1079,
+      "step": 4344
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.077017323668445,
+      "learning_rate": 6.889053844955644e-06,
+      "loss": 0.1355,
+      "step": 4345
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8279842796933703,
+      "learning_rate": 6.887689307201911e-06,
+      "loss": 0.1157,
+      "step": 4346
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9804722832426336,
+      "learning_rate": 6.886324605465744e-06,
+      "loss": 0.1426,
+      "step": 4347
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6854976998558286,
+      "learning_rate": 6.884959739865691e-06,
+      "loss": 0.1068,
+      "step": 4348
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.271612163597498,
+      "learning_rate": 6.883594710520317e-06,
+      "loss": 0.159,
+      "step": 4349
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9124582695880326,
+      "learning_rate": 6.8822295175482024e-06,
+      "loss": 0.1106,
+      "step": 4350
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8436492715727826,
+      "learning_rate": 6.880864161067942e-06,
+      "loss": 0.1147,
+      "step": 4351
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8838711494505864,
+      "learning_rate": 6.879498641198141e-06,
+      "loss": 0.1101,
+      "step": 4352
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8680932854255603,
+      "learning_rate": 6.878132958057422e-06,
+      "loss": 0.127,
+      "step": 4353
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8755132497585303,
+      "learning_rate": 6.876767111764422e-06,
+      "loss": 0.1215,
+      "step": 4354
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.151839554497143,
+      "learning_rate": 6.87540110243779e-06,
+      "loss": 0.1451,
+      "step": 4355
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7696004789834525,
+      "learning_rate": 6.874034930196191e-06,
+      "loss": 0.0976,
+      "step": 4356
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9009103584189615,
+      "learning_rate": 6.872668595158304e-06,
+      "loss": 0.1304,
+      "step": 4357
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7888060217500739,
+      "learning_rate": 6.87130209744282e-06,
+      "loss": 0.116,
+      "step": 4358
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.172053258311761,
+      "learning_rate": 6.869935437168449e-06,
+      "loss": 0.1382,
+      "step": 4359
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8267618551655687,
+      "learning_rate": 6.86856861445391e-06,
+      "loss": 0.1185,
+      "step": 4360
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.108704095534012,
+      "learning_rate": 6.867201629417937e-06,
+      "loss": 0.1475,
+      "step": 4361
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2745701063767436,
+      "learning_rate": 6.865834482179279e-06,
+      "loss": 0.1637,
+      "step": 4362
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0638497492061103,
+      "learning_rate": 6.864467172856703e-06,
+      "loss": 0.1155,
+      "step": 4363
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.100832139470805,
+      "learning_rate": 6.863099701568982e-06,
+      "loss": 0.1225,
+      "step": 4364
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.046648632952665,
+      "learning_rate": 6.8617320684349105e-06,
+      "loss": 0.1383,
+      "step": 4365
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.181008204570768,
+      "learning_rate": 6.860364273573292e-06,
+      "loss": 0.1141,
+      "step": 4366
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9691299787988519,
+      "learning_rate": 6.8589963171029475e-06,
+      "loss": 0.1042,
+      "step": 4367
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7986259939742626,
+      "learning_rate": 6.85762819914271e-06,
+      "loss": 0.1121,
+      "step": 4368
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.14673264435981,
+      "learning_rate": 6.856259919811427e-06,
+      "loss": 0.1255,
+      "step": 4369
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0077026238615208,
+      "learning_rate": 6.854891479227959e-06,
+      "loss": 0.1286,
+      "step": 4370
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0601838545423297,
+      "learning_rate": 6.853522877511184e-06,
+      "loss": 0.1332,
+      "step": 4371
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2314633703606894,
+      "learning_rate": 6.85215411477999e-06,
+      "loss": 0.1392,
+      "step": 4372
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1076916919598023,
+      "learning_rate": 6.85078519115328e-06,
+      "loss": 0.1181,
+      "step": 4373
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.143598607788687,
+      "learning_rate": 6.849416106749973e-06,
+      "loss": 0.1307,
+      "step": 4374
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9671264914126156,
+      "learning_rate": 6.8480468616889994e-06,
+      "loss": 0.1325,
+      "step": 4375
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8964870655971007,
+      "learning_rate": 6.846677456089305e-06,
+      "loss": 0.1318,
+      "step": 4376
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9348104393709211,
+      "learning_rate": 6.845307890069851e-06,
+      "loss": 0.128,
+      "step": 4377
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.285546609484436,
+      "learning_rate": 6.843938163749608e-06,
+      "loss": 0.1333,
+      "step": 4378
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6462569039836328,
+      "learning_rate": 6.842568277247564e-06,
+      "loss": 0.1097,
+      "step": 4379
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.022182533112036,
+      "learning_rate": 6.841198230682723e-06,
+      "loss": 0.1363,
+      "step": 4380
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.074507769633965,
+      "learning_rate": 6.839828024174096e-06,
+      "loss": 0.135,
+      "step": 4381
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1442307452808556,
+      "learning_rate": 6.838457657840715e-06,
+      "loss": 0.1235,
+      "step": 4382
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9095702914900408,
+      "learning_rate": 6.837087131801622e-06,
+      "loss": 0.1121,
+      "step": 4383
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.34595380191112,
+      "learning_rate": 6.835716446175872e-06,
+      "loss": 0.1263,
+      "step": 4384
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1312092270184104,
+      "learning_rate": 6.834345601082538e-06,
+      "loss": 0.1251,
+      "step": 4385
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8315791869446716,
+      "learning_rate": 6.832974596640704e-06,
+      "loss": 0.1067,
+      "step": 4386
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.132148903617483,
+      "learning_rate": 6.831603432969468e-06,
+      "loss": 0.1418,
+      "step": 4387
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2407310624758154,
+      "learning_rate": 6.830232110187942e-06,
+      "loss": 0.1514,
+      "step": 4388
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.046107561092263,
+      "learning_rate": 6.8288606284152535e-06,
+      "loss": 0.1273,
+      "step": 4389
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.016467522782691,
+      "learning_rate": 6.827488987770539e-06,
+      "loss": 0.109,
+      "step": 4390
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0202312089343706,
+      "learning_rate": 6.826117188372956e-06,
+      "loss": 0.1171,
+      "step": 4391
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.083756879788876,
+      "learning_rate": 6.824745230341669e-06,
+      "loss": 0.1362,
+      "step": 4392
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.013940985614311,
+      "learning_rate": 6.82337311379586e-06,
+      "loss": 0.1114,
+      "step": 4393
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.100516784447122,
+      "learning_rate": 6.822000838854724e-06,
+      "loss": 0.1347,
+      "step": 4394
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8578601927901723,
+      "learning_rate": 6.82062840563747e-06,
+      "loss": 0.0915,
+      "step": 4395
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8780681222509128,
+      "learning_rate": 6.8192558142633215e-06,
+      "loss": 0.1069,
+      "step": 4396
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0061680788538667,
+      "learning_rate": 6.817883064851511e-06,
+      "loss": 0.1201,
+      "step": 4397
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9248814938532648,
+      "learning_rate": 6.816510157521295e-06,
+      "loss": 0.1147,
+      "step": 4398
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.539928457189343,
+      "learning_rate": 6.815137092391929e-06,
+      "loss": 0.1121,
+      "step": 4399
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0343832360752128,
+      "learning_rate": 6.813763869582694e-06,
+      "loss": 0.1396,
+      "step": 4400
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8321513543446084,
+      "learning_rate": 6.812390489212885e-06,
+      "loss": 0.101,
+      "step": 4401
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9640048305537825,
+      "learning_rate": 6.811016951401801e-06,
+      "loss": 0.1016,
+      "step": 4402
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9505712477330828,
+      "learning_rate": 6.809643256268762e-06,
+      "loss": 0.1119,
+      "step": 4403
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.193291071221819,
+      "learning_rate": 6.8082694039331006e-06,
+      "loss": 0.1083,
+      "step": 4404
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.315161476127364,
+      "learning_rate": 6.806895394514163e-06,
+      "loss": 0.1393,
+      "step": 4405
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6867695977023893,
+      "learning_rate": 6.8055212281313086e-06,
+      "loss": 0.1106,
+      "step": 4406
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.433538020006179,
+      "learning_rate": 6.80414690490391e-06,
+      "loss": 0.1654,
+      "step": 4407
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1369109692827317,
+      "learning_rate": 6.802772424951353e-06,
+      "loss": 0.1238,
+      "step": 4408
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.7254889286027804,
+      "learning_rate": 6.801397788393038e-06,
+      "loss": 0.1445,
+      "step": 4409
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0018589656865635,
+      "learning_rate": 6.800022995348381e-06,
+      "loss": 0.1409,
+      "step": 4410
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2430094240657827,
+      "learning_rate": 6.798648045936807e-06,
+      "loss": 0.1226,
+      "step": 4411
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6854282652407255,
+      "learning_rate": 6.797272940277757e-06,
+      "loss": 0.0996,
+      "step": 4412
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9272242398541077,
+      "learning_rate": 6.795897678490689e-06,
+      "loss": 0.1112,
+      "step": 4413
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9220261086832267,
+      "learning_rate": 6.7945222606950665e-06,
+      "loss": 0.1326,
+      "step": 4414
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0466058212255747,
+      "learning_rate": 6.7931466870103735e-06,
+      "loss": 0.1157,
+      "step": 4415
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8142705250918703,
+      "learning_rate": 6.791770957556106e-06,
+      "loss": 0.0968,
+      "step": 4416
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9916848763314907,
+      "learning_rate": 6.790395072451772e-06,
+      "loss": 0.1441,
+      "step": 4417
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.048626836698824,
+      "learning_rate": 6.789019031816893e-06,
+      "loss": 0.127,
+      "step": 4418
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8412008101757644,
+      "learning_rate": 6.787642835771006e-06,
+      "loss": 0.118,
+      "step": 4419
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9521555072990087,
+      "learning_rate": 6.78626648443366e-06,
+      "loss": 0.1051,
+      "step": 4420
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.285612216585969,
+      "learning_rate": 6.7848899779244175e-06,
+      "loss": 0.1526,
+      "step": 4421
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9731235622128136,
+      "learning_rate": 6.783513316362855e-06,
+      "loss": 0.1345,
+      "step": 4422
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9474456606035975,
+      "learning_rate": 6.782136499868562e-06,
+      "loss": 0.1262,
+      "step": 4423
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.248825170734204,
+      "learning_rate": 6.7807595285611425e-06,
+      "loss": 0.1347,
+      "step": 4424
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8434529464239122,
+      "learning_rate": 6.7793824025602125e-06,
+      "loss": 0.1054,
+      "step": 4425
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8063599054158566,
+      "learning_rate": 6.778005121985403e-06,
+      "loss": 0.1093,
+      "step": 4426
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.01343133286733,
+      "learning_rate": 6.776627686956354e-06,
+      "loss": 0.1213,
+      "step": 4427
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9929040654187302,
+      "learning_rate": 6.775250097592728e-06,
+      "loss": 0.1255,
+      "step": 4428
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.031601448213597,
+      "learning_rate": 6.773872354014193e-06,
+      "loss": 0.1236,
+      "step": 4429
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8493151409399733,
+      "learning_rate": 6.77249445634043e-06,
+      "loss": 0.1272,
+      "step": 4430
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9904998016818611,
+      "learning_rate": 6.77111640469114e-06,
+      "loss": 0.108,
+      "step": 4431
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2517350383159638,
+      "learning_rate": 6.769738199186031e-06,
+      "loss": 0.1503,
+      "step": 4432
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7866866698849637,
+      "learning_rate": 6.768359839944829e-06,
+      "loss": 0.0985,
+      "step": 4433
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9664737664433554,
+      "learning_rate": 6.766981327087271e-06,
+      "loss": 0.1319,
+      "step": 4434
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2746306981543842,
+      "learning_rate": 6.765602660733105e-06,
+      "loss": 0.1207,
+      "step": 4435
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1991484990747083,
+      "learning_rate": 6.764223841002096e-06,
+      "loss": 0.1282,
+      "step": 4436
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0863456919717156,
+      "learning_rate": 6.762844868014025e-06,
+      "loss": 0.1262,
+      "step": 4437
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1938037174378966,
+      "learning_rate": 6.761465741888678e-06,
+      "loss": 0.1008,
+      "step": 4438
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.350310303829384,
+      "learning_rate": 6.760086462745858e-06,
+      "loss": 0.1247,
+      "step": 4439
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7927045497688106,
+      "learning_rate": 6.758707030705387e-06,
+      "loss": 0.115,
+      "step": 4440
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.122607765706097,
+      "learning_rate": 6.757327445887092e-06,
+      "loss": 0.123,
+      "step": 4441
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9416306853427787,
+      "learning_rate": 6.7559477084108184e-06,
+      "loss": 0.1189,
+      "step": 4442
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.3198364360589645,
+      "learning_rate": 6.754567818396423e-06,
+      "loss": 0.1346,
+      "step": 4443
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.263072512603556,
+      "learning_rate": 6.753187775963773e-06,
+      "loss": 0.1186,
+      "step": 4444
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.831689006643091,
+      "learning_rate": 6.751807581232754e-06,
+      "loss": 0.1031,
+      "step": 4445
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.3303477428587,
+      "learning_rate": 6.750427234323266e-06,
+      "loss": 0.1554,
+      "step": 4446
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.4342204433425465,
+      "learning_rate": 6.749046735355213e-06,
+      "loss": 0.1384,
+      "step": 4447
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.968998918772651,
+      "learning_rate": 6.7476660844485234e-06,
+      "loss": 0.1156,
+      "step": 4448
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9961313538282164,
+      "learning_rate": 6.746285281723129e-06,
+      "loss": 0.1051,
+      "step": 4449
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9522877411661375,
+      "learning_rate": 6.744904327298982e-06,
+      "loss": 0.1145,
+      "step": 4450
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8731526581015785,
+      "learning_rate": 6.743523221296044e-06,
+      "loss": 0.1198,
+      "step": 4451
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.121210828299572,
+      "learning_rate": 6.742141963834294e-06,
+      "loss": 0.1237,
+      "step": 4452
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.270173295214354,
+      "learning_rate": 6.740760555033715e-06,
+      "loss": 0.1334,
+      "step": 4453
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.730138755628579,
+      "learning_rate": 6.739378995014314e-06,
+      "loss": 0.1126,
+      "step": 4454
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0201315626688863,
+      "learning_rate": 6.737997283896104e-06,
+      "loss": 0.1307,
+      "step": 4455
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8924361386075295,
+      "learning_rate": 6.7366154217991145e-06,
+      "loss": 0.1302,
+      "step": 4456
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.999772751425837,
+      "learning_rate": 6.735233408843387e-06,
+      "loss": 0.1156,
+      "step": 4457
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7814555281989872,
+      "learning_rate": 6.7338512451489745e-06,
+      "loss": 0.1153,
+      "step": 4458
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.776911393434251,
+      "learning_rate": 6.732468930835947e-06,
+      "loss": 0.101,
+      "step": 4459
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2107846725920846,
+      "learning_rate": 6.731086466024386e-06,
+      "loss": 0.1445,
+      "step": 4460
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9312449991394116,
+      "learning_rate": 6.729703850834381e-06,
+      "loss": 0.1288,
+      "step": 4461
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3165352765514315,
+      "learning_rate": 6.728321085386043e-06,
+      "loss": 0.1365,
+      "step": 4462
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6251709154956455,
+      "learning_rate": 6.726938169799492e-06,
+      "loss": 0.1023,
+      "step": 4463
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1837299822614273,
+      "learning_rate": 6.725555104194858e-06,
+      "loss": 0.1313,
+      "step": 4464
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8792243099198833,
+      "learning_rate": 6.724171888692288e-06,
+      "loss": 0.1146,
+      "step": 4465
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0122090309674374,
+      "learning_rate": 6.722788523411945e-06,
+      "loss": 0.1173,
+      "step": 4466
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0961079441183745,
+      "learning_rate": 6.7214050084739955e-06,
+      "loss": 0.1271,
+      "step": 4467
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.350865763792726,
+      "learning_rate": 6.720021343998627e-06,
+      "loss": 0.1434,
+      "step": 4468
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8461547466411057,
+      "learning_rate": 6.71863753010604e-06,
+      "loss": 0.1261,
+      "step": 4469
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8143747731356448,
+      "learning_rate": 6.717253566916442e-06,
+      "loss": 0.0841,
+      "step": 4470
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.335596706648105,
+      "learning_rate": 6.715869454550057e-06,
+      "loss": 0.1407,
+      "step": 4471
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0746353073848813,
+      "learning_rate": 6.714485193127126e-06,
+      "loss": 0.1265,
+      "step": 4472
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1667535795163753,
+      "learning_rate": 6.713100782767894e-06,
+      "loss": 0.1102,
+      "step": 4473
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.962765805030947,
+      "learning_rate": 6.711716223592628e-06,
+      "loss": 0.1264,
+      "step": 4474
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8180298215104513,
+      "learning_rate": 6.710331515721602e-06,
+      "loss": 0.1018,
+      "step": 4475
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.96033300282247,
+      "learning_rate": 6.708946659275104e-06,
+      "loss": 0.1144,
+      "step": 4476
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9867760162376893,
+      "learning_rate": 6.707561654373436e-06,
+      "loss": 0.1283,
+      "step": 4477
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.143567532233534,
+      "learning_rate": 6.706176501136914e-06,
+      "loss": 0.1278,
+      "step": 4478
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0813115829365687,
+      "learning_rate": 6.704791199685865e-06,
+      "loss": 0.1047,
+      "step": 4479
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.975664765552879,
+      "learning_rate": 6.703405750140627e-06,
+      "loss": 0.1352,
+      "step": 4480
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1127830997797807,
+      "learning_rate": 6.702020152621557e-06,
+      "loss": 0.1306,
+      "step": 4481
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0027344537696314,
+      "learning_rate": 6.700634407249017e-06,
+      "loss": 0.1216,
+      "step": 4482
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.974900933035852,
+      "learning_rate": 6.699248514143388e-06,
+      "loss": 0.1139,
+      "step": 4483
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9723876288659472,
+      "learning_rate": 6.697862473425063e-06,
+      "loss": 0.1247,
+      "step": 4484
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9187521069188633,
+      "learning_rate": 6.696476285214444e-06,
+      "loss": 0.105,
+      "step": 4485
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0557338544531527,
+      "learning_rate": 6.695089949631949e-06,
+      "loss": 0.1279,
+      "step": 4486
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.102788926394792,
+      "learning_rate": 6.69370346679801e-06,
+      "loss": 0.1359,
+      "step": 4487
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7734037695512,
+      "learning_rate": 6.692316836833066e-06,
+      "loss": 0.098,
+      "step": 4488
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0504134621274384,
+      "learning_rate": 6.6909300598575764e-06,
+      "loss": 0.1096,
+      "step": 4489
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.3353046327590627,
+      "learning_rate": 6.689543135992009e-06,
+      "loss": 0.1435,
+      "step": 4490
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.065793523917047,
+      "learning_rate": 6.688156065356845e-06,
+      "loss": 0.1428,
+      "step": 4491
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.037792243595762,
+      "learning_rate": 6.686768848072576e-06,
+      "loss": 0.1283,
+      "step": 4492
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.868527858804207,
+      "learning_rate": 6.685381484259712e-06,
+      "loss": 0.1081,
+      "step": 4493
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1683738640578665,
+      "learning_rate": 6.683993974038771e-06,
+      "loss": 0.1137,
+      "step": 4494
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8816017107639569,
+      "learning_rate": 6.682606317530284e-06,
+      "loss": 0.1046,
+      "step": 4495
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.86408222955228,
+      "learning_rate": 6.681218514854799e-06,
+      "loss": 0.1262,
+      "step": 4496
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9696894935007574,
+      "learning_rate": 6.67983056613287e-06,
+      "loss": 0.1261,
+      "step": 4497
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9056129615210844,
+      "learning_rate": 6.678442471485069e-06,
+      "loss": 0.1296,
+      "step": 4498
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.229792516457032,
+      "learning_rate": 6.677054231031981e-06,
+      "loss": 0.1499,
+      "step": 4499
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.3889220975976624,
+      "learning_rate": 6.675665844894197e-06,
+      "loss": 0.1388,
+      "step": 4500
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1118565572385384,
+      "learning_rate": 6.674277313192329e-06,
+      "loss": 0.1547,
+      "step": 4501
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9767194486206585,
+      "learning_rate": 6.672888636046997e-06,
+      "loss": 0.128,
+      "step": 4502
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1792831450720285,
+      "learning_rate": 6.671499813578835e-06,
+      "loss": 0.1241,
+      "step": 4503
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8366236456361555,
+      "learning_rate": 6.670110845908486e-06,
+      "loss": 0.1187,
+      "step": 4504
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.090273861052521,
+      "learning_rate": 6.668721733156613e-06,
+      "loss": 0.1343,
+      "step": 4505
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.4073599047775707,
+      "learning_rate": 6.667332475443885e-06,
+      "loss": 0.1295,
+      "step": 4506
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8723641935867583,
+      "learning_rate": 6.665943072890987e-06,
+      "loss": 0.1191,
+      "step": 4507
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.003749361870532,
+      "learning_rate": 6.664553525618616e-06,
+      "loss": 0.1225,
+      "step": 4508
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8798065665897452,
+      "learning_rate": 6.663163833747479e-06,
+      "loss": 0.1135,
+      "step": 4509
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6103720882798456,
+      "learning_rate": 6.6617739973982985e-06,
+      "loss": 0.1009,
+      "step": 4510
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8036595410315,
+      "learning_rate": 6.660384016691811e-06,
+      "loss": 0.1154,
+      "step": 4511
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.624544364109499,
+      "learning_rate": 6.65899389174876e-06,
+      "loss": 0.0888,
+      "step": 4512
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0055171201382787,
+      "learning_rate": 6.657603622689908e-06,
+      "loss": 0.1265,
+      "step": 4513
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0213285277094744,
+      "learning_rate": 6.656213209636024e-06,
+      "loss": 0.1322,
+      "step": 4514
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8428122938614997,
+      "learning_rate": 6.654822652707893e-06,
+      "loss": 0.1211,
+      "step": 4515
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9360818066329497,
+      "learning_rate": 6.6534319520263135e-06,
+      "loss": 0.1045,
+      "step": 4516
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1159799028108504,
+      "learning_rate": 6.652041107712094e-06,
+      "loss": 0.1353,
+      "step": 4517
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1977305158576805,
+      "learning_rate": 6.6506501198860555e-06,
+      "loss": 0.1369,
+      "step": 4518
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.901432541264185,
+      "learning_rate": 6.649258988669031e-06,
+      "loss": 0.0861,
+      "step": 4519
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8742972212180524,
+      "learning_rate": 6.647867714181872e-06,
+      "loss": 0.1152,
+      "step": 4520
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0367128854754273,
+      "learning_rate": 6.646476296545434e-06,
+      "loss": 0.1428,
+      "step": 4521
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1563396993556196,
+      "learning_rate": 6.645084735880589e-06,
+      "loss": 0.13,
+      "step": 4522
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.915582650070517,
+      "learning_rate": 6.6436930323082215e-06,
+      "loss": 0.119,
+      "step": 4523
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1115991832060654,
+      "learning_rate": 6.642301185949227e-06,
+      "loss": 0.1391,
+      "step": 4524
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.036518481559962,
+      "learning_rate": 6.640909196924516e-06,
+      "loss": 0.1245,
+      "step": 4525
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0602933131633576,
+      "learning_rate": 6.6395170653550085e-06,
+      "loss": 0.1276,
+      "step": 4526
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0035515946775195,
+      "learning_rate": 6.63812479136164e-06,
+      "loss": 0.1112,
+      "step": 4527
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8421903646175244,
+      "learning_rate": 6.636732375065353e-06,
+      "loss": 0.1179,
+      "step": 4528
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9315176364317235,
+      "learning_rate": 6.635339816587109e-06,
+      "loss": 0.129,
+      "step": 4529
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.996951127197386,
+      "learning_rate": 6.633947116047877e-06,
+      "loss": 0.118,
+      "step": 4530
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9967870165783594,
+      "learning_rate": 6.632554273568641e-06,
+      "loss": 0.0982,
+      "step": 4531
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8242548129565206,
+      "learning_rate": 6.631161289270398e-06,
+      "loss": 0.1074,
+      "step": 4532
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0970688283744847,
+      "learning_rate": 6.629768163274152e-06,
+      "loss": 0.1374,
+      "step": 4533
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.02810614464841,
+      "learning_rate": 6.628374895700924e-06,
+      "loss": 0.1275,
+      "step": 4534
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.869279469859292,
+      "learning_rate": 6.626981486671748e-06,
+      "loss": 0.1106,
+      "step": 4535
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7735494316284937,
+      "learning_rate": 6.6255879363076695e-06,
+      "loss": 0.1239,
+      "step": 4536
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2952517902463923,
+      "learning_rate": 6.62419424472974e-06,
+      "loss": 0.1486,
+      "step": 4537
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7188167631773674,
+      "learning_rate": 6.622800412059036e-06,
+      "loss": 0.1226,
+      "step": 4538
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7662766698023447,
+      "learning_rate": 6.621406438416633e-06,
+      "loss": 0.1268,
+      "step": 4539
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0916233145139174,
+      "learning_rate": 6.620012323923628e-06,
+      "loss": 0.1091,
+      "step": 4540
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7434188277919636,
+      "learning_rate": 6.618618068701126e-06,
+      "loss": 0.1212,
+      "step": 4541
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7071286717899237,
+      "learning_rate": 6.617223672870244e-06,
+      "loss": 0.0924,
+      "step": 4542
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7465804771266527,
+      "learning_rate": 6.615829136552112e-06,
+      "loss": 0.0933,
+      "step": 4543
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9147516866932275,
+      "learning_rate": 6.614434459867875e-06,
+      "loss": 0.1012,
+      "step": 4544
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7580181575377933,
+      "learning_rate": 6.613039642938687e-06,
+      "loss": 0.1102,
+      "step": 4545
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.923802986248132,
+      "learning_rate": 6.611644685885713e-06,
+      "loss": 0.124,
+      "step": 4546
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9493312657273785,
+      "learning_rate": 6.610249588830135e-06,
+      "loss": 0.1191,
+      "step": 4547
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.234923398815059,
+      "learning_rate": 6.60885435189314e-06,
+      "loss": 0.1354,
+      "step": 4548
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.032789365143602,
+      "learning_rate": 6.607458975195937e-06,
+      "loss": 0.1594,
+      "step": 4549
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9970849235932349,
+      "learning_rate": 6.606063458859737e-06,
+      "loss": 0.127,
+      "step": 4550
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8447412347452457,
+      "learning_rate": 6.60466780300577e-06,
+      "loss": 0.1176,
+      "step": 4551
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7550955490696036,
+      "learning_rate": 6.6032720077552744e-06,
+      "loss": 0.1224,
+      "step": 4552
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.3561969956422666,
+      "learning_rate": 6.601876073229504e-06,
+      "loss": 0.145,
+      "step": 4553
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6786349819561515,
+      "learning_rate": 6.600479999549721e-06,
+      "loss": 0.096,
+      "step": 4554
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0985536530743962,
+      "learning_rate": 6.599083786837202e-06,
+      "loss": 0.1428,
+      "step": 4555
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.79791758134227,
+      "learning_rate": 6.5976874352132336e-06,
+      "loss": 0.1205,
+      "step": 4556
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1033145094264425,
+      "learning_rate": 6.59629094479912e-06,
+      "loss": 0.1389,
+      "step": 4557
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9152687153157175,
+      "learning_rate": 6.59489431571617e-06,
+      "loss": 0.1227,
+      "step": 4558
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1568646318938414,
+      "learning_rate": 6.593497548085709e-06,
+      "loss": 0.1279,
+      "step": 4559
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9995263318737113,
+      "learning_rate": 6.592100642029073e-06,
+      "loss": 0.1282,
+      "step": 4560
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0037478086532774,
+      "learning_rate": 6.5907035976676116e-06,
+      "loss": 0.1138,
+      "step": 4561
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0818954728418078,
+      "learning_rate": 6.589306415122684e-06,
+      "loss": 0.1229,
+      "step": 4562
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.2580158878139285,
+      "learning_rate": 6.587909094515663e-06,
+      "loss": 0.1625,
+      "step": 4563
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9823576851391977,
+      "learning_rate": 6.586511635967934e-06,
+      "loss": 0.1402,
+      "step": 4564
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7923703203378352,
+      "learning_rate": 6.585114039600891e-06,
+      "loss": 0.114,
+      "step": 4565
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6243933220789477,
+      "learning_rate": 6.5837163055359435e-06,
+      "loss": 0.1055,
+      "step": 4566
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8643331194757933,
+      "learning_rate": 6.582318433894513e-06,
+      "loss": 0.1135,
+      "step": 4567
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7890743283091954,
+      "learning_rate": 6.580920424798031e-06,
+      "loss": 0.1023,
+      "step": 4568
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9794758132092567,
+      "learning_rate": 6.57952227836794e-06,
+      "loss": 0.1137,
+      "step": 4569
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.5970968659901188,
+      "learning_rate": 6.578123994725699e-06,
+      "loss": 0.0918,
+      "step": 4570
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.027923918355746,
+      "learning_rate": 6.576725573992775e-06,
+      "loss": 0.1291,
+      "step": 4571
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.2924311653420846,
+      "learning_rate": 6.575327016290647e-06,
+      "loss": 0.1517,
+      "step": 4572
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8996138273242262,
+      "learning_rate": 6.573928321740808e-06,
+      "loss": 0.127,
+      "step": 4573
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6567208691976831,
+      "learning_rate": 6.57252949046476e-06,
+      "loss": 0.0858,
+      "step": 4574
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.144612260631249,
+      "learning_rate": 6.571130522584022e-06,
+      "loss": 0.1178,
+      "step": 4575
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.208753456934384,
+      "learning_rate": 6.569731418220119e-06,
+      "loss": 0.1369,
+      "step": 4576
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8510551458589826,
+      "learning_rate": 6.56833217749459e-06,
+      "loss": 0.1025,
+      "step": 4577
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.587685679646334,
+      "learning_rate": 6.566932800528987e-06,
+      "loss": 0.1519,
+      "step": 4578
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0540714070637924,
+      "learning_rate": 6.565533287444874e-06,
+      "loss": 0.1373,
+      "step": 4579
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0071598776575894,
+      "learning_rate": 6.564133638363823e-06,
+      "loss": 0.1216,
+      "step": 4580
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7131240886252699,
+      "learning_rate": 6.5627338534074234e-06,
+      "loss": 0.0983,
+      "step": 4581
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.844502087994929,
+      "learning_rate": 6.561333932697275e-06,
+      "loss": 0.1172,
+      "step": 4582
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.925692826679753,
+      "learning_rate": 6.559933876354983e-06,
+      "loss": 0.1126,
+      "step": 4583
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.578651996034083,
+      "learning_rate": 6.558533684502174e-06,
+      "loss": 0.1335,
+      "step": 4584
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6361737641457619,
+      "learning_rate": 6.557133357260481e-06,
+      "loss": 0.101,
+      "step": 4585
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8477842780521918,
+      "learning_rate": 6.555732894751548e-06,
+      "loss": 0.1214,
+      "step": 4586
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9844077766006776,
+      "learning_rate": 6.554332297097032e-06,
+      "loss": 0.1228,
+      "step": 4587
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.059357208925206,
+      "learning_rate": 6.552931564418605e-06,
+      "loss": 0.1271,
+      "step": 4588
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.2640283374933885,
+      "learning_rate": 6.5515306968379445e-06,
+      "loss": 0.1365,
+      "step": 4589
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0067816907276357,
+      "learning_rate": 6.550129694476744e-06,
+      "loss": 0.1265,
+      "step": 4590
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7833592758340449,
+      "learning_rate": 6.54872855745671e-06,
+      "loss": 0.1057,
+      "step": 4591
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1661855172576696,
+      "learning_rate": 6.547327285899556e-06,
+      "loss": 0.1299,
+      "step": 4592
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0258065636598945,
+      "learning_rate": 6.54592587992701e-06,
+      "loss": 0.1201,
+      "step": 4593
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6926321408557277,
+      "learning_rate": 6.544524339660813e-06,
+      "loss": 0.0954,
+      "step": 4594
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.801013477938161,
+      "learning_rate": 6.543122665222713e-06,
+      "loss": 0.1124,
+      "step": 4595
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9662375106303902,
+      "learning_rate": 6.541720856734475e-06,
+      "loss": 0.1244,
+      "step": 4596
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8997344256893935,
+      "learning_rate": 6.5403189143178725e-06,
+      "loss": 0.1233,
+      "step": 4597
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8352948642934723,
+      "learning_rate": 6.538916838094691e-06,
+      "loss": 0.1206,
+      "step": 4598
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8571787063102723,
+      "learning_rate": 6.537514628186727e-06,
+      "loss": 0.1208,
+      "step": 4599
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.4023176474275263,
+      "learning_rate": 6.536112284715795e-06,
+      "loss": 0.154,
+      "step": 4600
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.82491652623926,
+      "learning_rate": 6.534709807803707e-06,
+      "loss": 0.1096,
+      "step": 4601
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.5756244569887157,
+      "learning_rate": 6.533307197572302e-06,
+      "loss": 0.0917,
+      "step": 4602
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0933927333588933,
+      "learning_rate": 6.5319044541434225e-06,
+      "loss": 0.149,
+      "step": 4603
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.049884612563024,
+      "learning_rate": 6.530501577638923e-06,
+      "loss": 0.1348,
+      "step": 4604
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.797460742808321,
+      "learning_rate": 6.529098568180672e-06,
+      "loss": 0.111,
+      "step": 4605
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8830847019792387,
+      "learning_rate": 6.527695425890547e-06,
+      "loss": 0.1214,
+      "step": 4606
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8481337259277173,
+      "learning_rate": 6.526292150890437e-06,
+      "loss": 0.1107,
+      "step": 4607
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8250724207809175,
+      "learning_rate": 6.5248887433022446e-06,
+      "loss": 0.1014,
+      "step": 4608
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8573307488975646,
+      "learning_rate": 6.523485203247886e-06,
+      "loss": 0.1087,
+      "step": 4609
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.012869002747278,
+      "learning_rate": 6.5220815308492805e-06,
+      "loss": 0.1353,
+      "step": 4610
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.227262553418994,
+      "learning_rate": 6.520677726228366e-06,
+      "loss": 0.1319,
+      "step": 4611
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9613893355761616,
+      "learning_rate": 6.519273789507094e-06,
+      "loss": 0.1158,
+      "step": 4612
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0339892630473075,
+      "learning_rate": 6.517869720807419e-06,
+      "loss": 0.1298,
+      "step": 4613
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.1302049448954494,
+      "learning_rate": 6.5164655202513135e-06,
+      "loss": 0.127,
+      "step": 4614
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0518336949199614,
+      "learning_rate": 6.51506118796076e-06,
+      "loss": 0.1316,
+      "step": 4615
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7715262406117718,
+      "learning_rate": 6.513656724057751e-06,
+      "loss": 0.1121,
+      "step": 4616
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.6882796016180939,
+      "learning_rate": 6.512252128664292e-06,
+      "loss": 0.0977,
+      "step": 4617
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.997244188201705,
+      "learning_rate": 6.510847401902398e-06,
+      "loss": 0.1366,
+      "step": 4618
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.297821682832337,
+      "learning_rate": 6.509442543894099e-06,
+      "loss": 0.1607,
+      "step": 4619
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7312851821001018,
+      "learning_rate": 6.5080375547614325e-06,
+      "loss": 0.1033,
+      "step": 4620
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7679070426232457,
+      "learning_rate": 6.50663243462645e-06,
+      "loss": 0.1182,
+      "step": 4621
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.137841963950832,
+      "learning_rate": 6.505227183611214e-06,
+      "loss": 0.1176,
+      "step": 4622
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.4801080126343276,
+      "learning_rate": 6.503821801837795e-06,
+      "loss": 0.086,
+      "step": 4623
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7965054491185959,
+      "learning_rate": 6.502416289428282e-06,
+      "loss": 0.1178,
+      "step": 4624
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9400787329162716,
+      "learning_rate": 6.501010646504766e-06,
+      "loss": 0.1324,
+      "step": 4625
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.6871720333265288,
+      "learning_rate": 6.499604873189358e-06,
+      "loss": 0.1034,
+      "step": 4626
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.6252937350861658,
+      "learning_rate": 6.498198969604177e-06,
+      "loss": 0.0894,
+      "step": 4627
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.1259664135452265,
+      "learning_rate": 6.49679293587135e-06,
+      "loss": 0.1284,
+      "step": 4628
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8299341192140002,
+      "learning_rate": 6.495386772113019e-06,
+      "loss": 0.103,
+      "step": 4629
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0800849613870493,
+      "learning_rate": 6.49398047845134e-06,
+      "loss": 0.1331,
+      "step": 4630
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7596640862474255,
+      "learning_rate": 6.492574055008474e-06,
+      "loss": 0.102,
+      "step": 4631
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.4210149601902162,
+      "learning_rate": 6.491167501906596e-06,
+      "loss": 0.1468,
+      "step": 4632
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0448032667643963,
+      "learning_rate": 6.489760819267893e-06,
+      "loss": 0.1122,
+      "step": 4633
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.1356338883247203,
+      "learning_rate": 6.488354007214562e-06,
+      "loss": 0.1385,
+      "step": 4634
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8896036858816418,
+      "learning_rate": 6.486947065868814e-06,
+      "loss": 0.1107,
+      "step": 4635
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.3322172941578887,
+      "learning_rate": 6.4855399953528675e-06,
+      "loss": 0.1362,
+      "step": 4636
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0256378509758317,
+      "learning_rate": 6.4841327957889535e-06,
+      "loss": 0.1298,
+      "step": 4637
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7836263212057055,
+      "learning_rate": 6.482725467299316e-06,
+      "loss": 0.1086,
+      "step": 4638
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8548198808943297,
+      "learning_rate": 6.481318010006208e-06,
+      "loss": 0.1247,
+      "step": 4639
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9205156491856794,
+      "learning_rate": 6.479910424031893e-06,
+      "loss": 0.1131,
+      "step": 4640
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.116143695498577,
+      "learning_rate": 6.478502709498649e-06,
+      "loss": 0.1285,
+      "step": 4641
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.784206771270871,
+      "learning_rate": 6.477094866528764e-06,
+      "loss": 0.1145,
+      "step": 4642
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.085787707914258,
+      "learning_rate": 6.475686895244534e-06,
+      "loss": 0.1342,
+      "step": 4643
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.009961567625247,
+      "learning_rate": 6.474278795768272e-06,
+      "loss": 0.1424,
+      "step": 4644
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9924155504243144,
+      "learning_rate": 6.472870568222295e-06,
+      "loss": 0.1214,
+      "step": 4645
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8415606436676872,
+      "learning_rate": 6.471462212728936e-06,
+      "loss": 0.1104,
+      "step": 4646
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7942278663976636,
+      "learning_rate": 6.470053729410541e-06,
+      "loss": 0.1049,
+      "step": 4647
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.827009715074889,
+      "learning_rate": 6.4686451183894604e-06,
+      "loss": 0.1176,
+      "step": 4648
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0335119483914674,
+      "learning_rate": 6.467236379788061e-06,
+      "loss": 0.1292,
+      "step": 4649
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9735863372655535,
+      "learning_rate": 6.4658275137287196e-06,
+      "loss": 0.1361,
+      "step": 4650
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.2233048808284557,
+      "learning_rate": 6.464418520333821e-06,
+      "loss": 0.1296,
+      "step": 4651
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7564361418175638,
+      "learning_rate": 6.463009399725767e-06,
+      "loss": 0.1313,
+      "step": 4652
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7767552095739203,
+      "learning_rate": 6.461600152026966e-06,
+      "loss": 0.1128,
+      "step": 4653
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7891359617897904,
+      "learning_rate": 6.460190777359836e-06,
+      "loss": 0.1091,
+      "step": 4654
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.938453573793179,
+      "learning_rate": 6.458781275846811e-06,
+      "loss": 0.1109,
+      "step": 4655
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.565924447426406,
+      "learning_rate": 6.457371647610334e-06,
+      "loss": 0.0818,
+      "step": 4656
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8599144577576698,
+      "learning_rate": 6.455961892772857e-06,
+      "loss": 0.1115,
+      "step": 4657
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0841588304314236,
+      "learning_rate": 6.454552011456845e-06,
+      "loss": 0.1304,
+      "step": 4658
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.138594330201214,
+      "learning_rate": 6.453142003784774e-06,
+      "loss": 0.1145,
+      "step": 4659
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.1280666802232138,
+      "learning_rate": 6.4517318698791294e-06,
+      "loss": 0.1192,
+      "step": 4660
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9899606842328335,
+      "learning_rate": 6.45032160986241e-06,
+      "loss": 0.1521,
+      "step": 4661
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0137320504038945,
+      "learning_rate": 6.448911223857124e-06,
+      "loss": 0.1295,
+      "step": 4662
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8831779477340354,
+      "learning_rate": 6.44750071198579e-06,
+      "loss": 0.1256,
+      "step": 4663
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9674529933839957,
+      "learning_rate": 6.446090074370939e-06,
+      "loss": 0.1103,
+      "step": 4664
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8373291262761091,
+      "learning_rate": 6.444679311135112e-06,
+      "loss": 0.1072,
+      "step": 4665
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.147294083901844,
+      "learning_rate": 6.4432684224008615e-06,
+      "loss": 0.1283,
+      "step": 4666
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.1431550886958326,
+      "learning_rate": 6.441857408290751e-06,
+      "loss": 0.1445,
+      "step": 4667
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.019178412078036,
+      "learning_rate": 6.440446268927352e-06,
+      "loss": 0.1403,
+      "step": 4668
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7512739075778292,
+      "learning_rate": 6.4390350044332514e-06,
+      "loss": 0.1172,
+      "step": 4669
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8060631994893221,
+      "learning_rate": 6.437623614931045e-06,
+      "loss": 0.0916,
+      "step": 4670
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9848273389112192,
+      "learning_rate": 6.43621210054334e-06,
+      "loss": 0.1239,
+      "step": 4671
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0222309183780234,
+      "learning_rate": 6.434800461392752e-06,
+      "loss": 0.1437,
+      "step": 4672
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9642637158232656,
+      "learning_rate": 6.4333886976019085e-06,
+      "loss": 0.1115,
+      "step": 4673
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8312092685760508,
+      "learning_rate": 6.431976809293452e-06,
+      "loss": 0.1178,
+      "step": 4674
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2139286211226783,
+      "learning_rate": 6.430564796590028e-06,
+      "loss": 0.1209,
+      "step": 4675
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9691512732112373,
+      "learning_rate": 6.429152659614302e-06,
+      "loss": 0.1089,
+      "step": 4676
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0093226940421354,
+      "learning_rate": 6.427740398488943e-06,
+      "loss": 0.1198,
+      "step": 4677
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8782675022294457,
+      "learning_rate": 6.4263280133366326e-06,
+      "loss": 0.1016,
+      "step": 4678
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.025372464707761,
+      "learning_rate": 6.424915504280065e-06,
+      "loss": 0.1221,
+      "step": 4679
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8942523592540887,
+      "learning_rate": 6.423502871441943e-06,
+      "loss": 0.1047,
+      "step": 4680
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8305540486447613,
+      "learning_rate": 6.422090114944982e-06,
+      "loss": 0.0999,
+      "step": 4681
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9824460351947288,
+      "learning_rate": 6.420677234911908e-06,
+      "loss": 0.1134,
+      "step": 4682
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.980943268880249,
+      "learning_rate": 6.4192642314654565e-06,
+      "loss": 0.1017,
+      "step": 4683
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.3973803672232,
+      "learning_rate": 6.417851104728372e-06,
+      "loss": 0.1454,
+      "step": 4684
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.269742852905679,
+      "learning_rate": 6.416437854823414e-06,
+      "loss": 0.1454,
+      "step": 4685
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8811875135857112,
+      "learning_rate": 6.415024481873352e-06,
+      "loss": 0.1041,
+      "step": 4686
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.678128253002755,
+      "learning_rate": 6.413610986000963e-06,
+      "loss": 0.0906,
+      "step": 4687
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0292386810676106,
+      "learning_rate": 6.412197367329036e-06,
+      "loss": 0.1285,
+      "step": 4688
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8265625715311626,
+      "learning_rate": 6.4107836259803745e-06,
+      "loss": 0.1115,
+      "step": 4689
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7279577169374565,
+      "learning_rate": 6.409369762077784e-06,
+      "loss": 0.1151,
+      "step": 4690
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8487289179792048,
+      "learning_rate": 6.40795577574409e-06,
+      "loss": 0.1145,
+      "step": 4691
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.058754207847507,
+      "learning_rate": 6.406541667102126e-06,
+      "loss": 0.1125,
+      "step": 4692
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.5655007882651693,
+      "learning_rate": 6.40512743627473e-06,
+      "loss": 0.0945,
+      "step": 4693
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.291245959863372,
+      "learning_rate": 6.403713083384758e-06,
+      "loss": 0.1495,
+      "step": 4694
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9898321885396917,
+      "learning_rate": 6.402298608555076e-06,
+      "loss": 0.1242,
+      "step": 4695
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.450414734862128,
+      "learning_rate": 6.4008840119085535e-06,
+      "loss": 0.1332,
+      "step": 4696
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8306260902591966,
+      "learning_rate": 6.399469293568079e-06,
+      "loss": 0.1152,
+      "step": 4697
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.845292472734369,
+      "learning_rate": 6.398054453656549e-06,
+      "loss": 0.1239,
+      "step": 4698
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9228471501464572,
+      "learning_rate": 6.396639492296868e-06,
+      "loss": 0.1157,
+      "step": 4699
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8702767947417749,
+      "learning_rate": 6.3952244096119535e-06,
+      "loss": 0.1188,
+      "step": 4700
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8729537584239604,
+      "learning_rate": 6.393809205724734e-06,
+      "loss": 0.1147,
+      "step": 4701
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8927800020065917,
+      "learning_rate": 6.392393880758144e-06,
+      "loss": 0.134,
+      "step": 4702
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.18543663183883,
+      "learning_rate": 6.390978434835135e-06,
+      "loss": 0.1329,
+      "step": 4703
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.109917707510275,
+      "learning_rate": 6.389562868078666e-06,
+      "loss": 0.1407,
+      "step": 4704
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2004734083488935,
+      "learning_rate": 6.388147180611705e-06,
+      "loss": 0.1439,
+      "step": 4705
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9186248989142887,
+      "learning_rate": 6.386731372557231e-06,
+      "loss": 0.124,
+      "step": 4706
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.292624624670308,
+      "learning_rate": 6.385315444038238e-06,
+      "loss": 0.1102,
+      "step": 4707
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7955428450545798,
+      "learning_rate": 6.383899395177724e-06,
+      "loss": 0.0923,
+      "step": 4708
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0387533499635215,
+      "learning_rate": 6.3824832260987e-06,
+      "loss": 0.1379,
+      "step": 4709
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0665160561807894,
+      "learning_rate": 6.381066936924189e-06,
+      "loss": 0.129,
+      "step": 4710
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7487152595087074,
+      "learning_rate": 6.379650527777224e-06,
+      "loss": 0.1035,
+      "step": 4711
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7387690567688239,
+      "learning_rate": 6.378233998780846e-06,
+      "loss": 0.117,
+      "step": 4712
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9632437273407148,
+      "learning_rate": 6.376817350058109e-06,
+      "loss": 0.1147,
+      "step": 4713
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2039138139857766,
+      "learning_rate": 6.375400581732076e-06,
+      "loss": 0.1139,
+      "step": 4714
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7624752146613907,
+      "learning_rate": 6.373983693925819e-06,
+      "loss": 0.103,
+      "step": 4715
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2372372192109204,
+      "learning_rate": 6.372566686762427e-06,
+      "loss": 0.1421,
+      "step": 4716
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.672289964724045,
+      "learning_rate": 6.37114956036499e-06,
+      "loss": 0.0828,
+      "step": 4717
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9898122915555132,
+      "learning_rate": 6.369732314856614e-06,
+      "loss": 0.1225,
+      "step": 4718
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1464142311126366,
+      "learning_rate": 6.368314950360416e-06,
+      "loss": 0.1197,
+      "step": 4719
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8841631657173143,
+      "learning_rate": 6.366897466999519e-06,
+      "loss": 0.1041,
+      "step": 4720
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0668124989360934,
+      "learning_rate": 6.3654798648970605e-06,
+      "loss": 0.1429,
+      "step": 4721
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.917378582219527,
+      "learning_rate": 6.364062144176188e-06,
+      "loss": 0.1105,
+      "step": 4722
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.296647153848635,
+      "learning_rate": 6.362644304960055e-06,
+      "loss": 0.1374,
+      "step": 4723
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9799489605382683,
+      "learning_rate": 6.36122634737183e-06,
+      "loss": 0.1369,
+      "step": 4724
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.888744617172784,
+      "learning_rate": 6.359808271534691e-06,
+      "loss": 0.1251,
+      "step": 4725
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.802733871180748,
+      "learning_rate": 6.358390077571823e-06,
+      "loss": 0.1019,
+      "step": 4726
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7722024760365744,
+      "learning_rate": 6.356971765606427e-06,
+      "loss": 0.1051,
+      "step": 4727
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.94503672819843,
+      "learning_rate": 6.355553335761708e-06,
+      "loss": 0.1128,
+      "step": 4728
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0860697685187395,
+      "learning_rate": 6.354134788160885e-06,
+      "loss": 0.1347,
+      "step": 4729
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1449484116051085,
+      "learning_rate": 6.352716122927187e-06,
+      "loss": 0.1222,
+      "step": 4730
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7781092502776699,
+      "learning_rate": 6.351297340183852e-06,
+      "loss": 0.1225,
+      "step": 4731
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0266709716155114,
+      "learning_rate": 6.349878440054129e-06,
+      "loss": 0.1295,
+      "step": 4732
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7432872652801654,
+      "learning_rate": 6.348459422661276e-06,
+      "loss": 0.1074,
+      "step": 4733
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1246569000664084,
+      "learning_rate": 6.3470402881285635e-06,
+      "loss": 0.1077,
+      "step": 4734
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9933213179922722,
+      "learning_rate": 6.34562103657927e-06,
+      "loss": 0.1159,
+      "step": 4735
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.79867290164734,
+      "learning_rate": 6.344201668136687e-06,
+      "loss": 0.1109,
+      "step": 4736
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7053131110618118,
+      "learning_rate": 6.342782182924112e-06,
+      "loss": 0.0828,
+      "step": 4737
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.994551232335655,
+      "learning_rate": 6.341362581064856e-06,
+      "loss": 0.1387,
+      "step": 4738
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9067712028932886,
+      "learning_rate": 6.3399428626822375e-06,
+      "loss": 0.1248,
+      "step": 4739
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.018007750684826,
+      "learning_rate": 6.338523027899589e-06,
+      "loss": 0.1453,
+      "step": 4740
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0104198781409166,
+      "learning_rate": 6.337103076840248e-06,
+      "loss": 0.1077,
+      "step": 4741
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7877783748911336,
+      "learning_rate": 6.3356830096275666e-06,
+      "loss": 0.0894,
+      "step": 4742
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.001070498243399,
+      "learning_rate": 6.334262826384905e-06,
+      "loss": 0.1181,
+      "step": 4743
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8804212981654032,
+      "learning_rate": 6.332842527235632e-06,
+      "loss": 0.1142,
+      "step": 4744
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8804941409848202,
+      "learning_rate": 6.331422112303132e-06,
+      "loss": 0.1202,
+      "step": 4745
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9911130392424214,
+      "learning_rate": 6.3300015817107895e-06,
+      "loss": 0.1278,
+      "step": 4746
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.038769948738525,
+      "learning_rate": 6.3285809355820106e-06,
+      "loss": 0.1301,
+      "step": 4747
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9623091732661166,
+      "learning_rate": 6.327160174040205e-06,
+      "loss": 0.1217,
+      "step": 4748
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.4064429967832086,
+      "learning_rate": 6.32573929720879e-06,
+      "loss": 0.1238,
+      "step": 4749
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0140086885287403,
+      "learning_rate": 6.324318305211201e-06,
+      "loss": 0.1318,
+      "step": 4750
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9705698499649666,
+      "learning_rate": 6.3228971981708765e-06,
+      "loss": 0.1336,
+      "step": 4751
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9915688180462514,
+      "learning_rate": 6.321475976211267e-06,
+      "loss": 0.1177,
+      "step": 4752
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0354044399723716,
+      "learning_rate": 6.320054639455832e-06,
+      "loss": 0.1178,
+      "step": 4753
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0613693979582624,
+      "learning_rate": 6.318633188028045e-06,
+      "loss": 0.1377,
+      "step": 4754
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.411880674874074,
+      "learning_rate": 6.317211622051384e-06,
+      "loss": 0.1322,
+      "step": 4755
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0086784414083096,
+      "learning_rate": 6.315789941649341e-06,
+      "loss": 0.126,
+      "step": 4756
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9471495685372828,
+      "learning_rate": 6.314368146945418e-06,
+      "loss": 0.11,
+      "step": 4757
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9444258779939045,
+      "learning_rate": 6.312946238063121e-06,
+      "loss": 0.114,
+      "step": 4758
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.119899727989978,
+      "learning_rate": 6.311524215125975e-06,
+      "loss": 0.1419,
+      "step": 4759
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.7902925628211697,
+      "learning_rate": 6.310102078257508e-06,
+      "loss": 0.1081,
+      "step": 4760
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9553573925504344,
+      "learning_rate": 6.30867982758126e-06,
+      "loss": 0.1292,
+      "step": 4761
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1461333123633977,
+      "learning_rate": 6.307257463220782e-06,
+      "loss": 0.1405,
+      "step": 4762
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8098940965239325,
+      "learning_rate": 6.3058349852996345e-06,
+      "loss": 0.1253,
+      "step": 4763
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.5227095387454714,
+      "learning_rate": 6.304412393941386e-06,
+      "loss": 0.1,
+      "step": 4764
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8452370328167844,
+      "learning_rate": 6.3029896892696155e-06,
+      "loss": 0.1132,
+      "step": 4765
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.659252963081831,
+      "learning_rate": 6.301566871407915e-06,
+      "loss": 0.1148,
+      "step": 4766
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9225785943092828,
+      "learning_rate": 6.300143940479881e-06,
+      "loss": 0.0986,
+      "step": 4767
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9244229679954448,
+      "learning_rate": 6.298720896609125e-06,
+      "loss": 0.1189,
+      "step": 4768
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9990178513490748,
+      "learning_rate": 6.297297739919266e-06,
+      "loss": 0.1085,
+      "step": 4769
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.85466110080012,
+      "learning_rate": 6.295874470533929e-06,
+      "loss": 0.1252,
+      "step": 4770
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8281269984137958,
+      "learning_rate": 6.294451088576757e-06,
+      "loss": 0.112,
+      "step": 4771
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8657391080439982,
+      "learning_rate": 6.293027594171397e-06,
+      "loss": 0.1117,
+      "step": 4772
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.054295315157647,
+      "learning_rate": 6.291603987441506e-06,
+      "loss": 0.1119,
+      "step": 4773
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.038378911260525,
+      "learning_rate": 6.290180268510753e-06,
+      "loss": 0.1393,
+      "step": 4774
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0465552647400203,
+      "learning_rate": 6.288756437502816e-06,
+      "loss": 0.1199,
+      "step": 4775
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9839970112236058,
+      "learning_rate": 6.28733249454138e-06,
+      "loss": 0.1302,
+      "step": 4776
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9735301545719488,
+      "learning_rate": 6.2859084397501434e-06,
+      "loss": 0.1121,
+      "step": 4777
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.933232469921969,
+      "learning_rate": 6.2844842732528145e-06,
+      "loss": 0.0981,
+      "step": 4778
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9711930077916098,
+      "learning_rate": 6.283059995173109e-06,
+      "loss": 0.128,
+      "step": 4779
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.157327713639551,
+      "learning_rate": 6.281635605634751e-06,
+      "loss": 0.1427,
+      "step": 4780
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0426310662453577,
+      "learning_rate": 6.280211104761479e-06,
+      "loss": 0.1234,
+      "step": 4781
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9756964040824303,
+      "learning_rate": 6.278786492677037e-06,
+      "loss": 0.1317,
+      "step": 4782
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.001724170659894,
+      "learning_rate": 6.2773617695051806e-06,
+      "loss": 0.1135,
+      "step": 4783
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9537172100205218,
+      "learning_rate": 6.275936935369675e-06,
+      "loss": 0.1278,
+      "step": 4784
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7665908496400504,
+      "learning_rate": 6.274511990394294e-06,
+      "loss": 0.1046,
+      "step": 4785
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9431417544789436,
+      "learning_rate": 6.273086934702823e-06,
+      "loss": 0.1194,
+      "step": 4786
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3004261283855563,
+      "learning_rate": 6.271661768419055e-06,
+      "loss": 0.1539,
+      "step": 4787
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9380146680491177,
+      "learning_rate": 6.270236491666792e-06,
+      "loss": 0.113,
+      "step": 4788
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8238071960149504,
+      "learning_rate": 6.268811104569849e-06,
+      "loss": 0.1024,
+      "step": 4789
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.981209537223023,
+      "learning_rate": 6.267385607252048e-06,
+      "loss": 0.1099,
+      "step": 4790
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9878156928560995,
+      "learning_rate": 6.265959999837219e-06,
+      "loss": 0.1154,
+      "step": 4791
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0035518830800165,
+      "learning_rate": 6.2645342824492065e-06,
+      "loss": 0.1375,
+      "step": 4792
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8018096634099765,
+      "learning_rate": 6.263108455211862e-06,
+      "loss": 0.1117,
+      "step": 4793
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7187858370181939,
+      "learning_rate": 6.261682518249043e-06,
+      "loss": 0.116,
+      "step": 4794
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8433209777807242,
+      "learning_rate": 6.260256471684622e-06,
+      "loss": 0.1284,
+      "step": 4795
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9554113777640811,
+      "learning_rate": 6.258830315642479e-06,
+      "loss": 0.1068,
+      "step": 4796
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9579278787057743,
+      "learning_rate": 6.257404050246503e-06,
+      "loss": 0.123,
+      "step": 4797
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1057885769440947,
+      "learning_rate": 6.255977675620592e-06,
+      "loss": 0.1166,
+      "step": 4798
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0729024591941334,
+      "learning_rate": 6.254551191888656e-06,
+      "loss": 0.119,
+      "step": 4799
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0832789371539056,
+      "learning_rate": 6.25312459917461e-06,
+      "loss": 0.1202,
+      "step": 4800
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8964979656182601,
+      "learning_rate": 6.251697897602384e-06,
+      "loss": 0.1182,
+      "step": 4801
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8041336466073548,
+      "learning_rate": 6.2502710872959134e-06,
+      "loss": 0.092,
+      "step": 4802
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9036793097133802,
+      "learning_rate": 6.248844168379144e-06,
+      "loss": 0.1271,
+      "step": 4803
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9410457536194146,
+      "learning_rate": 6.247417140976033e-06,
+      "loss": 0.1282,
+      "step": 4804
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9498557309340399,
+      "learning_rate": 6.2459900052105445e-06,
+      "loss": 0.1184,
+      "step": 4805
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6732274802251281,
+      "learning_rate": 6.2445627612066526e-06,
+      "loss": 0.0999,
+      "step": 4806
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8747749569817358,
+      "learning_rate": 6.243135409088341e-06,
+      "loss": 0.1139,
+      "step": 4807
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1677434683595416,
+      "learning_rate": 6.241707948979604e-06,
+      "loss": 0.1349,
+      "step": 4808
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8921749709926956,
+      "learning_rate": 6.240280381004444e-06,
+      "loss": 0.122,
+      "step": 4809
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8943805397377445,
+      "learning_rate": 6.23885270528687e-06,
+      "loss": 0.1197,
+      "step": 4810
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8493354871440684,
+      "learning_rate": 6.237424921950909e-06,
+      "loss": 0.1144,
+      "step": 4811
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8953022008850646,
+      "learning_rate": 6.235997031120585e-06,
+      "loss": 0.1343,
+      "step": 4812
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.552142248735999,
+      "learning_rate": 6.234569032919944e-06,
+      "loss": 0.1161,
+      "step": 4813
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.790407437838741,
+      "learning_rate": 6.233140927473033e-06,
+      "loss": 0.1222,
+      "step": 4814
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9373260027044,
+      "learning_rate": 6.231712714903909e-06,
+      "loss": 0.1214,
+      "step": 4815
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0553997847416605,
+      "learning_rate": 6.230284395336643e-06,
+      "loss": 0.116,
+      "step": 4816
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.075450076080619,
+      "learning_rate": 6.22885596889531e-06,
+      "loss": 0.114,
+      "step": 4817
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.062329088847458,
+      "learning_rate": 6.227427435703997e-06,
+      "loss": 0.141,
+      "step": 4818
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7435307307738757,
+      "learning_rate": 6.2259987958868005e-06,
+      "loss": 0.1136,
+      "step": 4819
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8099285043582265,
+      "learning_rate": 6.224570049567825e-06,
+      "loss": 0.1131,
+      "step": 4820
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.840382332357282,
+      "learning_rate": 6.223141196871185e-06,
+      "loss": 0.1132,
+      "step": 4821
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9753181904438506,
+      "learning_rate": 6.221712237921005e-06,
+      "loss": 0.1194,
+      "step": 4822
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0296908058577587,
+      "learning_rate": 6.220283172841416e-06,
+      "loss": 0.1519,
+      "step": 4823
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.738773031226602,
+      "learning_rate": 6.21885400175656e-06,
+      "loss": 0.1072,
+      "step": 4824
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9847561177958029,
+      "learning_rate": 6.217424724790592e-06,
+      "loss": 0.1317,
+      "step": 4825
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1882594840092846,
+      "learning_rate": 6.215995342067666e-06,
+      "loss": 0.0978,
+      "step": 4826
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7091302195872704,
+      "learning_rate": 6.214565853711956e-06,
+      "loss": 0.1073,
+      "step": 4827
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.758939439252944,
+      "learning_rate": 6.213136259847642e-06,
+      "loss": 0.1196,
+      "step": 4828
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.894513268226193,
+      "learning_rate": 6.211706560598909e-06,
+      "loss": 0.1165,
+      "step": 4829
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7927417877245728,
+      "learning_rate": 6.2102767560899545e-06,
+      "loss": 0.1233,
+      "step": 4830
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.031101088555856,
+      "learning_rate": 6.208846846444987e-06,
+      "loss": 0.1084,
+      "step": 4831
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9153611707796494,
+      "learning_rate": 6.207416831788219e-06,
+      "loss": 0.1313,
+      "step": 4832
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8517018418779887,
+      "learning_rate": 6.205986712243876e-06,
+      "loss": 0.1248,
+      "step": 4833
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.951504780850715,
+      "learning_rate": 6.204556487936193e-06,
+      "loss": 0.1205,
+      "step": 4834
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.116722513445758,
+      "learning_rate": 6.203126158989411e-06,
+      "loss": 0.1615,
+      "step": 4835
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9866247511889275,
+      "learning_rate": 6.201695725527781e-06,
+      "loss": 0.118,
+      "step": 4836
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.21921291709542,
+      "learning_rate": 6.200265187675568e-06,
+      "loss": 0.1273,
+      "step": 4837
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8261340532859198,
+      "learning_rate": 6.198834545557038e-06,
+      "loss": 0.1003,
+      "step": 4838
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8693238963561363,
+      "learning_rate": 6.197403799296471e-06,
+      "loss": 0.109,
+      "step": 4839
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0368100655309034,
+      "learning_rate": 6.195972949018157e-06,
+      "loss": 0.1318,
+      "step": 4840
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7440186239494408,
+      "learning_rate": 6.194541994846388e-06,
+      "loss": 0.0977,
+      "step": 4841
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7798737169981045,
+      "learning_rate": 6.193110936905476e-06,
+      "loss": 0.1191,
+      "step": 4842
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.098212759074431,
+      "learning_rate": 6.191679775319734e-06,
+      "loss": 0.1341,
+      "step": 4843
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8978721030925707,
+      "learning_rate": 6.190248510213486e-06,
+      "loss": 0.1049,
+      "step": 4844
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8161119314906053,
+      "learning_rate": 6.188817141711063e-06,
+      "loss": 0.1189,
+      "step": 4845
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.30653690693518,
+      "learning_rate": 6.1873856699368115e-06,
+      "loss": 0.1209,
+      "step": 4846
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8492552840920051,
+      "learning_rate": 6.185954095015079e-06,
+      "loss": 0.1105,
+      "step": 4847
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.017640080619557,
+      "learning_rate": 6.184522417070227e-06,
+      "loss": 0.1182,
+      "step": 4848
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9889419835980953,
+      "learning_rate": 6.183090636226625e-06,
+      "loss": 0.1276,
+      "step": 4849
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.105367690109629,
+      "learning_rate": 6.181658752608649e-06,
+      "loss": 0.1302,
+      "step": 4850
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8261736367652186,
+      "learning_rate": 6.180226766340688e-06,
+      "loss": 0.1231,
+      "step": 4851
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.5989238417814968,
+      "learning_rate": 6.178794677547138e-06,
+      "loss": 0.1101,
+      "step": 4852
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0655123239058333,
+      "learning_rate": 6.1773624863524e-06,
+      "loss": 0.1307,
+      "step": 4853
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7420354190303058,
+      "learning_rate": 6.175930192880891e-06,
+      "loss": 0.1234,
+      "step": 4854
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.983555966873605,
+      "learning_rate": 6.174497797257034e-06,
+      "loss": 0.1115,
+      "step": 4855
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7439263386363626,
+      "learning_rate": 6.173065299605257e-06,
+      "loss": 0.1035,
+      "step": 4856
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7460730836687877,
+      "learning_rate": 6.171632700050003e-06,
+      "loss": 0.1191,
+      "step": 4857
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.025897958102639,
+      "learning_rate": 6.1701999987157225e-06,
+      "loss": 0.1169,
+      "step": 4858
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8515608098733858,
+      "learning_rate": 6.168767195726868e-06,
+      "loss": 0.1092,
+      "step": 4859
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7902842933010585,
+      "learning_rate": 6.16733429120791e-06,
+      "loss": 0.1132,
+      "step": 4860
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0402755488426525,
+      "learning_rate": 6.165901285283326e-06,
+      "loss": 0.1312,
+      "step": 4861
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1572285788412575,
+      "learning_rate": 6.164468178077595e-06,
+      "loss": 0.1467,
+      "step": 4862
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9344974762607847,
+      "learning_rate": 6.163034969715214e-06,
+      "loss": 0.1189,
+      "step": 4863
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8207147522886569,
+      "learning_rate": 6.161601660320684e-06,
+      "loss": 0.1121,
+      "step": 4864
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.893640736028462,
+      "learning_rate": 6.160168250018516e-06,
+      "loss": 0.0965,
+      "step": 4865
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9955242086047664,
+      "learning_rate": 6.158734738933228e-06,
+      "loss": 0.1195,
+      "step": 4866
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1306176885166015,
+      "learning_rate": 6.1573011271893515e-06,
+      "loss": 0.1488,
+      "step": 4867
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7892104762253915,
+      "learning_rate": 6.15586741491142e-06,
+      "loss": 0.1192,
+      "step": 4868
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8917766805389953,
+      "learning_rate": 6.154433602223979e-06,
+      "loss": 0.1053,
+      "step": 4869
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.3611622194245716,
+      "learning_rate": 6.152999689251588e-06,
+      "loss": 0.1381,
+      "step": 4870
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9887611929851323,
+      "learning_rate": 6.151565676118805e-06,
+      "loss": 0.1291,
+      "step": 4871
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.050931157823187,
+      "learning_rate": 6.150131562950204e-06,
+      "loss": 0.1221,
+      "step": 4872
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9188897349699203,
+      "learning_rate": 6.148697349870364e-06,
+      "loss": 0.1193,
+      "step": 4873
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9004784309760325,
+      "learning_rate": 6.147263037003877e-06,
+      "loss": 0.1163,
+      "step": 4874
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.93423289184628,
+      "learning_rate": 6.145828624475337e-06,
+      "loss": 0.1231,
+      "step": 4875
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7730809587308083,
+      "learning_rate": 6.144394112409356e-06,
+      "loss": 0.1027,
+      "step": 4876
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7369376687779499,
+      "learning_rate": 6.142959500930543e-06,
+      "loss": 0.0981,
+      "step": 4877
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7293111687054623,
+      "learning_rate": 6.1415247901635256e-06,
+      "loss": 0.1066,
+      "step": 4878
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9816602644444095,
+      "learning_rate": 6.140089980232937e-06,
+      "loss": 0.1314,
+      "step": 4879
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9002630686190995,
+      "learning_rate": 6.138655071263415e-06,
+      "loss": 0.1305,
+      "step": 4880
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0219745755450025,
+      "learning_rate": 6.137220063379612e-06,
+      "loss": 0.1052,
+      "step": 4881
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.019427212779509,
+      "learning_rate": 6.135784956706186e-06,
+      "loss": 0.1143,
+      "step": 4882
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9769469324163873,
+      "learning_rate": 6.134349751367802e-06,
+      "loss": 0.1159,
+      "step": 4883
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.913974403284735,
+      "learning_rate": 6.132914447489137e-06,
+      "loss": 0.1263,
+      "step": 4884
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.288616969391857,
+      "learning_rate": 6.131479045194875e-06,
+      "loss": 0.1702,
+      "step": 4885
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9323919256215265,
+      "learning_rate": 6.130043544609707e-06,
+      "loss": 0.1388,
+      "step": 4886
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.809332127661882,
+      "learning_rate": 6.128607945858336e-06,
+      "loss": 0.1194,
+      "step": 4887
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.342999704043993,
+      "learning_rate": 6.127172249065471e-06,
+      "loss": 0.1403,
+      "step": 4888
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9773235167416923,
+      "learning_rate": 6.125736454355831e-06,
+      "loss": 0.1338,
+      "step": 4889
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7618945982640297,
+      "learning_rate": 6.124300561854139e-06,
+      "loss": 0.1021,
+      "step": 4890
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2018598817268393,
+      "learning_rate": 6.122864571685135e-06,
+      "loss": 0.1218,
+      "step": 4891
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8836135751638283,
+      "learning_rate": 6.121428483973559e-06,
+      "loss": 0.1295,
+      "step": 4892
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8837658093597585,
+      "learning_rate": 6.119992298844165e-06,
+      "loss": 0.106,
+      "step": 4893
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8272297312477663,
+      "learning_rate": 6.118556016421713e-06,
+      "loss": 0.1193,
+      "step": 4894
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9135369806054934,
+      "learning_rate": 6.117119636830971e-06,
+      "loss": 0.1196,
+      "step": 4895
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3119131342734995,
+      "learning_rate": 6.115683160196718e-06,
+      "loss": 0.1203,
+      "step": 4896
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8277687611017608,
+      "learning_rate": 6.114246586643739e-06,
+      "loss": 0.0985,
+      "step": 4897
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1388876005771267,
+      "learning_rate": 6.112809916296829e-06,
+      "loss": 0.1439,
+      "step": 4898
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.589459885541316,
+      "learning_rate": 6.11137314928079e-06,
+      "loss": 0.0947,
+      "step": 4899
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7601580968530142,
+      "learning_rate": 6.109936285720433e-06,
+      "loss": 0.1023,
+      "step": 4900
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.922006007107989,
+      "learning_rate": 6.108499325740577e-06,
+      "loss": 0.0975,
+      "step": 4901
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7838705097233027,
+      "learning_rate": 6.107062269466052e-06,
+      "loss": 0.1202,
+      "step": 4902
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.947670472120174,
+      "learning_rate": 6.105625117021692e-06,
+      "loss": 0.121,
+      "step": 4903
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9439048271487138,
+      "learning_rate": 6.104187868532341e-06,
+      "loss": 0.147,
+      "step": 4904
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.746878444271801,
+      "learning_rate": 6.102750524122856e-06,
+      "loss": 0.1122,
+      "step": 4905
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9094150992437706,
+      "learning_rate": 6.1013130839180936e-06,
+      "loss": 0.1154,
+      "step": 4906
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.018840239067516,
+      "learning_rate": 6.099875548042925e-06,
+      "loss": 0.1289,
+      "step": 4907
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.145174511066835,
+      "learning_rate": 6.098437916622231e-06,
+      "loss": 0.1138,
+      "step": 4908
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1275968938083847,
+      "learning_rate": 6.097000189780893e-06,
+      "loss": 0.1281,
+      "step": 4909
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.111484999647587,
+      "learning_rate": 6.095562367643807e-06,
+      "loss": 0.1204,
+      "step": 4910
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9052044936938572,
+      "learning_rate": 6.0941244503358776e-06,
+      "loss": 0.1056,
+      "step": 4911
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9951598828586277,
+      "learning_rate": 6.0926864379820135e-06,
+      "loss": 0.1433,
+      "step": 4912
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.0080645922255607,
+      "learning_rate": 6.091248330707136e-06,
+      "loss": 0.1265,
+      "step": 4913
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.851477172836219,
+      "learning_rate": 6.089810128636173e-06,
+      "loss": 0.1266,
+      "step": 4914
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9372677265156553,
+      "learning_rate": 6.088371831894057e-06,
+      "loss": 0.1171,
+      "step": 4915
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7032265818076067,
+      "learning_rate": 6.086933440605733e-06,
+      "loss": 0.1052,
+      "step": 4916
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.105536542212519,
+      "learning_rate": 6.085494954896156e-06,
+      "loss": 0.1508,
+      "step": 4917
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.000622536557249,
+      "learning_rate": 6.0840563748902836e-06,
+      "loss": 0.1152,
+      "step": 4918
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.275176361874426,
+      "learning_rate": 6.082617700713083e-06,
+      "loss": 0.1155,
+      "step": 4919
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.599289273385687,
+      "learning_rate": 6.0811789324895365e-06,
+      "loss": 0.1061,
+      "step": 4920
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.215750448800732,
+      "learning_rate": 6.079740070344625e-06,
+      "loss": 0.1605,
+      "step": 4921
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2791099409385382,
+      "learning_rate": 6.078301114403341e-06,
+      "loss": 0.1437,
+      "step": 4922
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.762802683676954,
+      "learning_rate": 6.07686206479069e-06,
+      "loss": 0.1039,
+      "step": 4923
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.555928325814913,
+      "learning_rate": 6.075422921631675e-06,
+      "loss": 0.1071,
+      "step": 4924
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.077040906360435,
+      "learning_rate": 6.073983685051319e-06,
+      "loss": 0.1315,
+      "step": 4925
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8635405108850642,
+      "learning_rate": 6.0725443551746454e-06,
+      "loss": 0.101,
+      "step": 4926
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0995796453109206,
+      "learning_rate": 6.071104932126687e-06,
+      "loss": 0.1197,
+      "step": 4927
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.009879275551528,
+      "learning_rate": 6.0696654160324875e-06,
+      "loss": 0.1195,
+      "step": 4928
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.2620708792957065,
+      "learning_rate": 6.068225807017096e-06,
+      "loss": 0.1412,
+      "step": 4929
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7952296679916147,
+      "learning_rate": 6.06678610520557e-06,
+      "loss": 0.1072,
+      "step": 4930
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9705423440939835,
+      "learning_rate": 6.065346310722976e-06,
+      "loss": 0.116,
+      "step": 4931
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6999090883129435,
+      "learning_rate": 6.063906423694389e-06,
+      "loss": 0.1066,
+      "step": 4932
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8796616706033624,
+      "learning_rate": 6.062466444244889e-06,
+      "loss": 0.1243,
+      "step": 4933
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0559884626043514,
+      "learning_rate": 6.061026372499568e-06,
+      "loss": 0.1116,
+      "step": 4934
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8845821330534942,
+      "learning_rate": 6.059586208583523e-06,
+      "loss": 0.1235,
+      "step": 4935
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7344141989550121,
+      "learning_rate": 6.058145952621861e-06,
+      "loss": 0.1058,
+      "step": 4936
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8872765670986487,
+      "learning_rate": 6.056705604739696e-06,
+      "loss": 0.122,
+      "step": 4937
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8218256328225666,
+      "learning_rate": 6.055265165062149e-06,
+      "loss": 0.1048,
+      "step": 4938
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8714701326514385,
+      "learning_rate": 6.053824633714352e-06,
+      "loss": 0.1154,
+      "step": 4939
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8014985656111493,
+      "learning_rate": 6.0523840108214425e-06,
+      "loss": 0.1073,
+      "step": 4940
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.855702892815786,
+      "learning_rate": 6.0509432965085665e-06,
+      "loss": 0.1098,
+      "step": 4941
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.017549518132829,
+      "learning_rate": 6.049502490900877e-06,
+      "loss": 0.1158,
+      "step": 4942
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7800169097330127,
+      "learning_rate": 6.048061594123536e-06,
+      "loss": 0.0962,
+      "step": 4943
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0495922912093616,
+      "learning_rate": 6.046620606301716e-06,
+      "loss": 0.1313,
+      "step": 4944
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.5866250909184794,
+      "learning_rate": 6.045179527560592e-06,
+      "loss": 0.0959,
+      "step": 4945
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.020444762927052,
+      "learning_rate": 6.04373835802535e-06,
+      "loss": 0.1187,
+      "step": 4946
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0854377433145013,
+      "learning_rate": 6.042297097821184e-06,
+      "loss": 0.1223,
+      "step": 4947
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.882900900099412,
+      "learning_rate": 6.040855747073294e-06,
+      "loss": 0.1149,
+      "step": 4948
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6224227318394695,
+      "learning_rate": 6.039414305906892e-06,
+      "loss": 0.099,
+      "step": 4949
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9503745876084193,
+      "learning_rate": 6.037972774447194e-06,
+      "loss": 0.1221,
+      "step": 4950
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6921267398935587,
+      "learning_rate": 6.036531152819425e-06,
+      "loss": 0.1125,
+      "step": 4951
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0196253522305327,
+      "learning_rate": 6.035089441148816e-06,
+      "loss": 0.1113,
+      "step": 4952
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9355851865094544,
+      "learning_rate": 6.03364763956061e-06,
+      "loss": 0.1123,
+      "step": 4953
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9510864477490875,
+      "learning_rate": 6.032205748180054e-06,
+      "loss": 0.1187,
+      "step": 4954
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.73266849218343,
+      "learning_rate": 6.030763767132406e-06,
+      "loss": 0.0941,
+      "step": 4955
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8752550765646894,
+      "learning_rate": 6.0293216965429294e-06,
+      "loss": 0.1155,
+      "step": 4956
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9121570703098103,
+      "learning_rate": 6.027879536536893e-06,
+      "loss": 0.1186,
+      "step": 4957
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3957029646050807,
+      "learning_rate": 6.026437287239581e-06,
+      "loss": 0.1403,
+      "step": 4958
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7789354591976543,
+      "learning_rate": 6.024994948776277e-06,
+      "loss": 0.1085,
+      "step": 4959
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7389394559466675,
+      "learning_rate": 6.023552521272278e-06,
+      "loss": 0.113,
+      "step": 4960
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6298333586523461,
+      "learning_rate": 6.0221100048528866e-06,
+      "loss": 0.0952,
+      "step": 4961
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7582323954970474,
+      "learning_rate": 6.020667399643414e-06,
+      "loss": 0.1094,
+      "step": 4962
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.070687729934323,
+      "learning_rate": 6.019224705769176e-06,
+      "loss": 0.115,
+      "step": 4963
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9327608736164168,
+      "learning_rate": 6.017781923355501e-06,
+      "loss": 0.1198,
+      "step": 4964
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9347310828801432,
+      "learning_rate": 6.016339052527723e-06,
+      "loss": 0.1198,
+      "step": 4965
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9289071650990133,
+      "learning_rate": 6.014896093411181e-06,
+      "loss": 0.1145,
+      "step": 4966
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8527529792686324,
+      "learning_rate": 6.013453046131224e-06,
+      "loss": 0.1157,
+      "step": 4967
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8542076734063087,
+      "learning_rate": 6.0120099108132126e-06,
+      "loss": 0.1129,
+      "step": 4968
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.902671784114292,
+      "learning_rate": 6.010566687582507e-06,
+      "loss": 0.1201,
+      "step": 4969
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.6504671171049503,
+      "learning_rate": 6.0091233765644796e-06,
+      "loss": 0.1037,
+      "step": 4970
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.6160452605524753,
+      "learning_rate": 6.0076799778845105e-06,
+      "loss": 0.1063,
+      "step": 4971
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.002626443814734,
+      "learning_rate": 6.0062364916679885e-06,
+      "loss": 0.1222,
+      "step": 4972
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8776185281144557,
+      "learning_rate": 6.0047929180403065e-06,
+      "loss": 0.1243,
+      "step": 4973
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.6494634514612223,
+      "learning_rate": 6.003349257126867e-06,
+      "loss": 0.0928,
+      "step": 4974
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7440905885785756,
+      "learning_rate": 6.00190550905308e-06,
+      "loss": 0.1024,
+      "step": 4975
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1378398123744247,
+      "learning_rate": 6.000461673944364e-06,
+      "loss": 0.15,
+      "step": 4976
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0204688178815013,
+      "learning_rate": 5.9990177519261435e-06,
+      "loss": 0.12,
+      "step": 4977
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1824224578324993,
+      "learning_rate": 5.997573743123852e-06,
+      "loss": 0.1366,
+      "step": 4978
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.5762183897756177,
+      "learning_rate": 5.996129647662928e-06,
+      "loss": 0.0829,
+      "step": 4979
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0254442812204454,
+      "learning_rate": 5.994685465668819e-06,
+      "loss": 0.1214,
+      "step": 4980
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.080769381856775,
+      "learning_rate": 5.993241197266982e-06,
+      "loss": 0.1304,
+      "step": 4981
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0758919167755088,
+      "learning_rate": 5.99179684258288e-06,
+      "loss": 0.1384,
+      "step": 4982
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7571459561927143,
+      "learning_rate": 5.990352401741981e-06,
+      "loss": 0.1139,
+      "step": 4983
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9482665749881678,
+      "learning_rate": 5.988907874869764e-06,
+      "loss": 0.133,
+      "step": 4984
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9885995283699256,
+      "learning_rate": 5.987463262091715e-06,
+      "loss": 0.1233,
+      "step": 4985
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7636513211660254,
+      "learning_rate": 5.986018563533325e-06,
+      "loss": 0.0794,
+      "step": 4986
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0538612953934816,
+      "learning_rate": 5.984573779320093e-06,
+      "loss": 0.1286,
+      "step": 4987
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9727186223260489,
+      "learning_rate": 5.983128909577532e-06,
+      "loss": 0.1469,
+      "step": 4988
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7452652132484658,
+      "learning_rate": 5.98168395443115e-06,
+      "loss": 0.0996,
+      "step": 4989
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.885851806757247,
+      "learning_rate": 5.980238914006473e-06,
+      "loss": 0.1199,
+      "step": 4990
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.612526849537247,
+      "learning_rate": 5.9787937884290325e-06,
+      "loss": 0.1026,
+      "step": 4991
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1189572322875163,
+      "learning_rate": 5.977348577824362e-06,
+      "loss": 0.1099,
+      "step": 4992
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8267328505100635,
+      "learning_rate": 5.975903282318009e-06,
+      "loss": 0.1188,
+      "step": 4993
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8256461121631937,
+      "learning_rate": 5.974457902035524e-06,
+      "loss": 0.1035,
+      "step": 4994
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8914515794387323,
+      "learning_rate": 5.973012437102466e-06,
+      "loss": 0.1199,
+      "step": 4995
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1919251568604503,
+      "learning_rate": 5.971566887644401e-06,
+      "loss": 0.1197,
+      "step": 4996
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0191614139301737,
+      "learning_rate": 5.970121253786907e-06,
+      "loss": 0.1022,
+      "step": 4997
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0038054881034335,
+      "learning_rate": 5.96867553565556e-06,
+      "loss": 0.1246,
+      "step": 4998
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8980156628446445,
+      "learning_rate": 5.967229733375952e-06,
+      "loss": 0.123,
+      "step": 4999
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9227323869861515,
+      "learning_rate": 5.965783847073679e-06,
+      "loss": 0.106,
+      "step": 5000
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8243048521610479,
+      "learning_rate": 5.964337876874343e-06,
+      "loss": 0.1196,
+      "step": 5001
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6680419197272611,
+      "learning_rate": 5.962891822903555e-06,
+      "loss": 0.1019,
+      "step": 5002
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8450948812067482,
+      "learning_rate": 5.961445685286933e-06,
+      "loss": 0.1146,
+      "step": 5003
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.108562972689307,
+      "learning_rate": 5.959999464150101e-06,
+      "loss": 0.1617,
+      "step": 5004
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9499365554759507,
+      "learning_rate": 5.958553159618693e-06,
+      "loss": 0.1079,
+      "step": 5005
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8859355103291315,
+      "learning_rate": 5.957106771818348e-06,
+      "loss": 0.1222,
+      "step": 5006
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8315372473982805,
+      "learning_rate": 5.955660300874712e-06,
+      "loss": 0.1146,
+      "step": 5007
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3307066437226154,
+      "learning_rate": 5.9542137469134405e-06,
+      "loss": 0.1355,
+      "step": 5008
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0522661107552698,
+      "learning_rate": 5.9527671100601956e-06,
+      "loss": 0.1228,
+      "step": 5009
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8579057079758383,
+      "learning_rate": 5.951320390440642e-06,
+      "loss": 0.1396,
+      "step": 5010
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8855469724348655,
+      "learning_rate": 5.949873588180458e-06,
+      "loss": 0.1171,
+      "step": 5011
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.144756090238146,
+      "learning_rate": 5.948426703405327e-06,
+      "loss": 0.1224,
+      "step": 5012
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1342181296058187,
+      "learning_rate": 5.946979736240938e-06,
+      "loss": 0.1581,
+      "step": 5013
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7121265361196136,
+      "learning_rate": 5.945532686812987e-06,
+      "loss": 0.117,
+      "step": 5014
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.810679433667134,
+      "learning_rate": 5.944085555247181e-06,
+      "loss": 0.1174,
+      "step": 5015
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6941252956226815,
+      "learning_rate": 5.94263834166923e-06,
+      "loss": 0.1008,
+      "step": 5016
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.074915230490796,
+      "learning_rate": 5.941191046204851e-06,
+      "loss": 0.1227,
+      "step": 5017
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8541399783188595,
+      "learning_rate": 5.939743668979774e-06,
+      "loss": 0.133,
+      "step": 5018
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.188606767937708,
+      "learning_rate": 5.938296210119727e-06,
+      "loss": 0.141,
+      "step": 5019
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7683583488225754,
+      "learning_rate": 5.9368486697504525e-06,
+      "loss": 0.098,
+      "step": 5020
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.950058543180246,
+      "learning_rate": 5.935401047997697e-06,
+      "loss": 0.1345,
+      "step": 5021
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.931108125559404,
+      "learning_rate": 5.933953344987215e-06,
+      "loss": 0.1246,
+      "step": 5022
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0354671410815186,
+      "learning_rate": 5.932505560844766e-06,
+      "loss": 0.104,
+      "step": 5023
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.022780983806588,
+      "learning_rate": 5.93105769569612e-06,
+      "loss": 0.136,
+      "step": 5024
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9322876665778022,
+      "learning_rate": 5.929609749667052e-06,
+      "loss": 0.1095,
+      "step": 5025
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1609916934143016,
+      "learning_rate": 5.928161722883341e-06,
+      "loss": 0.1399,
+      "step": 5026
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.851747300715781,
+      "learning_rate": 5.926713615470781e-06,
+      "loss": 0.119,
+      "step": 5027
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.988226219250091,
+      "learning_rate": 5.925265427555166e-06,
+      "loss": 0.1432,
+      "step": 5028
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9521867454549575,
+      "learning_rate": 5.923817159262297e-06,
+      "loss": 0.1206,
+      "step": 5029
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8559110066631774,
+      "learning_rate": 5.922368810717989e-06,
+      "loss": 0.1126,
+      "step": 5030
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7608750355664113,
+      "learning_rate": 5.9209203820480555e-06,
+      "loss": 0.1328,
+      "step": 5031
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7977843451529179,
+      "learning_rate": 5.919471873378322e-06,
+      "loss": 0.115,
+      "step": 5032
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.059582824376823,
+      "learning_rate": 5.91802328483462e-06,
+      "loss": 0.1407,
+      "step": 5033
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7589401785545038,
+      "learning_rate": 5.916574616542785e-06,
+      "loss": 0.0937,
+      "step": 5034
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8464722292905205,
+      "learning_rate": 5.915125868628664e-06,
+      "loss": 0.093,
+      "step": 5035
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.91809140446907,
+      "learning_rate": 5.913677041218111e-06,
+      "loss": 0.1277,
+      "step": 5036
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.147611748890377,
+      "learning_rate": 5.912228134436979e-06,
+      "loss": 0.1223,
+      "step": 5037
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7223954603002656,
+      "learning_rate": 5.910779148411139e-06,
+      "loss": 0.1007,
+      "step": 5038
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.6435248779725946,
+      "learning_rate": 5.9093300832664625e-06,
+      "loss": 0.1072,
+      "step": 5039
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8316810571933446,
+      "learning_rate": 5.907880939128826e-06,
+      "loss": 0.1239,
+      "step": 5040
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0288153789314536,
+      "learning_rate": 5.9064317161241185e-06,
+      "loss": 0.1241,
+      "step": 5041
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8976503491707963,
+      "learning_rate": 5.904982414378233e-06,
+      "loss": 0.1357,
+      "step": 5042
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.177573045228236,
+      "learning_rate": 5.903533034017068e-06,
+      "loss": 0.1226,
+      "step": 5043
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7863708962816582,
+      "learning_rate": 5.902083575166532e-06,
+      "loss": 0.095,
+      "step": 5044
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9578563020406434,
+      "learning_rate": 5.900634037952537e-06,
+      "loss": 0.1159,
+      "step": 5045
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.759787605817178,
+      "learning_rate": 5.899184422501005e-06,
+      "loss": 0.1091,
+      "step": 5046
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9860180552598778,
+      "learning_rate": 5.897734728937863e-06,
+      "loss": 0.1189,
+      "step": 5047
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.822050993233891,
+      "learning_rate": 5.896284957389042e-06,
+      "loss": 0.0991,
+      "step": 5048
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.13954128838974,
+      "learning_rate": 5.8948351079804875e-06,
+      "loss": 0.1257,
+      "step": 5049
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8856741099998384,
+      "learning_rate": 5.893385180838144e-06,
+      "loss": 0.1154,
+      "step": 5050
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.655622211696462,
+      "learning_rate": 5.891935176087967e-06,
+      "loss": 0.102,
+      "step": 5051
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8962847087911745,
+      "learning_rate": 5.890485093855916e-06,
+      "loss": 0.1319,
+      "step": 5052
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1459880049209663,
+      "learning_rate": 5.889034934267962e-06,
+      "loss": 0.1334,
+      "step": 5053
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9619211554407519,
+      "learning_rate": 5.887584697450075e-06,
+      "loss": 0.1281,
+      "step": 5054
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.855940385075262,
+      "learning_rate": 5.88613438352824e-06,
+      "loss": 0.1217,
+      "step": 5055
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.886327747691445,
+      "learning_rate": 5.8846839926284435e-06,
+      "loss": 0.1213,
+      "step": 5056
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8263491048157114,
+      "learning_rate": 5.883233524876681e-06,
+      "loss": 0.0998,
+      "step": 5057
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.4613830994080448,
+      "learning_rate": 5.88178298039895e-06,
+      "loss": 0.0851,
+      "step": 5058
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1843487471030403,
+      "learning_rate": 5.880332359321264e-06,
+      "loss": 0.1329,
+      "step": 5059
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.02411749728561,
+      "learning_rate": 5.878881661769633e-06,
+      "loss": 0.1152,
+      "step": 5060
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.305011313037653,
+      "learning_rate": 5.877430887870081e-06,
+      "loss": 0.1385,
+      "step": 5061
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.972171209393961,
+      "learning_rate": 5.875980037748635e-06,
+      "loss": 0.1591,
+      "step": 5062
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.992975974697504,
+      "learning_rate": 5.87452911153133e-06,
+      "loss": 0.1253,
+      "step": 5063
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7498121276083862,
+      "learning_rate": 5.873078109344204e-06,
+      "loss": 0.1072,
+      "step": 5064
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.64478378218017,
+      "learning_rate": 5.871627031313311e-06,
+      "loss": 0.1078,
+      "step": 5065
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0786553141524453,
+      "learning_rate": 5.870175877564699e-06,
+      "loss": 0.1197,
+      "step": 5066
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.536810066594668,
+      "learning_rate": 5.8687246482244306e-06,
+      "loss": 0.08,
+      "step": 5067
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.007317162977653,
+      "learning_rate": 5.867273343418577e-06,
+      "loss": 0.1375,
+      "step": 5068
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.823498260559666,
+      "learning_rate": 5.865821963273206e-06,
+      "loss": 0.1072,
+      "step": 5069
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9448057679008224,
+      "learning_rate": 5.864370507914403e-06,
+      "loss": 0.1272,
+      "step": 5070
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7932060406569077,
+      "learning_rate": 5.8629189774682524e-06,
+      "loss": 0.0989,
+      "step": 5071
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.173016208328216,
+      "learning_rate": 5.8614673720608495e-06,
+      "loss": 0.1314,
+      "step": 5072
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8715303874180422,
+      "learning_rate": 5.860015691818292e-06,
+      "loss": 0.1143,
+      "step": 5073
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.911421442866843,
+      "learning_rate": 5.858563936866691e-06,
+      "loss": 0.1387,
+      "step": 5074
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.6210919890453315,
+      "learning_rate": 5.857112107332155e-06,
+      "loss": 0.1096,
+      "step": 5075
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2499598607131857,
+      "learning_rate": 5.855660203340804e-06,
+      "loss": 0.1141,
+      "step": 5076
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.101609077176601,
+      "learning_rate": 5.854208225018767e-06,
+      "loss": 0.1293,
+      "step": 5077
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8359219084747167,
+      "learning_rate": 5.8527561724921735e-06,
+      "loss": 0.1056,
+      "step": 5078
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7277293827184643,
+      "learning_rate": 5.851304045887164e-06,
+      "loss": 0.0963,
+      "step": 5079
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9781716158758402,
+      "learning_rate": 5.849851845329884e-06,
+      "loss": 0.1213,
+      "step": 5080
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.710632924115769,
+      "learning_rate": 5.8483995709464845e-06,
+      "loss": 0.0971,
+      "step": 5081
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9650019287849914,
+      "learning_rate": 5.846947222863123e-06,
+      "loss": 0.1401,
+      "step": 5082
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7432319442192512,
+      "learning_rate": 5.845494801205967e-06,
+      "loss": 0.0989,
+      "step": 5083
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.11468034896895,
+      "learning_rate": 5.844042306101184e-06,
+      "loss": 0.1436,
+      "step": 5084
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.058623735536554,
+      "learning_rate": 5.842589737674954e-06,
+      "loss": 0.1328,
+      "step": 5085
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8170427656664632,
+      "learning_rate": 5.841137096053459e-06,
+      "loss": 0.1073,
+      "step": 5086
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.6444283767521197,
+      "learning_rate": 5.839684381362891e-06,
+      "loss": 0.1059,
+      "step": 5087
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0231515383693486,
+      "learning_rate": 5.8382315937294444e-06,
+      "loss": 0.1351,
+      "step": 5088
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7531764334516389,
+      "learning_rate": 5.836778733279322e-06,
+      "loss": 0.1005,
+      "step": 5089
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8506491627549855,
+      "learning_rate": 5.835325800138736e-06,
+      "loss": 0.1313,
+      "step": 5090
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8006825521940102,
+      "learning_rate": 5.833872794433897e-06,
+      "loss": 0.1082,
+      "step": 5091
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8237788117541966,
+      "learning_rate": 5.832419716291031e-06,
+      "loss": 0.1033,
+      "step": 5092
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1470922232331353,
+      "learning_rate": 5.830966565836365e-06,
+      "loss": 0.1338,
+      "step": 5093
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1695064850909884,
+      "learning_rate": 5.829513343196132e-06,
+      "loss": 0.1301,
+      "step": 5094
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1051031273457044,
+      "learning_rate": 5.828060048496573e-06,
+      "loss": 0.1345,
+      "step": 5095
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1356332004517324,
+      "learning_rate": 5.826606681863934e-06,
+      "loss": 0.1337,
+      "step": 5096
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.6984421975735762,
+      "learning_rate": 5.825153243424471e-06,
+      "loss": 0.1099,
+      "step": 5097
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.050796456952679,
+      "learning_rate": 5.823699733304441e-06,
+      "loss": 0.1381,
+      "step": 5098
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.764124458219414,
+      "learning_rate": 5.822246151630109e-06,
+      "loss": 0.1155,
+      "step": 5099
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.3479149520313665,
+      "learning_rate": 5.820792498527749e-06,
+      "loss": 0.1685,
+      "step": 5100
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7759005018329987,
+      "learning_rate": 5.819338774123638e-06,
+      "loss": 0.1239,
+      "step": 5101
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.883729435899512,
+      "learning_rate": 5.81788497854406e-06,
+      "loss": 0.1249,
+      "step": 5102
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.841413161515065,
+      "learning_rate": 5.816431111915304e-06,
+      "loss": 0.1189,
+      "step": 5103
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8217901430142631,
+      "learning_rate": 5.8149771743636675e-06,
+      "loss": 0.1285,
+      "step": 5104
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8284336105426517,
+      "learning_rate": 5.813523166015455e-06,
+      "loss": 0.1178,
+      "step": 5105
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0543267234027347,
+      "learning_rate": 5.812069086996972e-06,
+      "loss": 0.1294,
+      "step": 5106
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8859240779560933,
+      "learning_rate": 5.810614937434537e-06,
+      "loss": 0.1175,
+      "step": 5107
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9975831709056648,
+      "learning_rate": 5.8091607174544695e-06,
+      "loss": 0.1269,
+      "step": 5108
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9169591813566134,
+      "learning_rate": 5.807706427183096e-06,
+      "loss": 0.1166,
+      "step": 5109
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.858687849558745,
+      "learning_rate": 5.806252066746751e-06,
+      "loss": 0.1245,
+      "step": 5110
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.996083827760046,
+      "learning_rate": 5.804797636271772e-06,
+      "loss": 0.1335,
+      "step": 5111
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7409289111680077,
+      "learning_rate": 5.803343135884507e-06,
+      "loss": 0.1098,
+      "step": 5112
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0347170804943717,
+      "learning_rate": 5.801888565711308e-06,
+      "loss": 0.1353,
+      "step": 5113
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7591855239442908,
+      "learning_rate": 5.8004339258785296e-06,
+      "loss": 0.1007,
+      "step": 5114
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8522143084981304,
+      "learning_rate": 5.798979216512536e-06,
+      "loss": 0.1125,
+      "step": 5115
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8438428593620593,
+      "learning_rate": 5.797524437739699e-06,
+      "loss": 0.1083,
+      "step": 5116
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9663445844545666,
+      "learning_rate": 5.796069589686393e-06,
+      "loss": 0.1201,
+      "step": 5117
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7869958705209372,
+      "learning_rate": 5.794614672479e-06,
+      "loss": 0.1215,
+      "step": 5118
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.1040768537238215,
+      "learning_rate": 5.793159686243908e-06,
+      "loss": 0.1156,
+      "step": 5119
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9073501656739287,
+      "learning_rate": 5.791704631107511e-06,
+      "loss": 0.1279,
+      "step": 5120
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6889405649149396,
+      "learning_rate": 5.790249507196207e-06,
+      "loss": 0.104,
+      "step": 5121
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0225453173415753,
+      "learning_rate": 5.7887943146364045e-06,
+      "loss": 0.1335,
+      "step": 5122
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8392982426845952,
+      "learning_rate": 5.787339053554512e-06,
+      "loss": 0.0912,
+      "step": 5123
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.304125827390601,
+      "learning_rate": 5.78588372407695e-06,
+      "loss": 0.1409,
+      "step": 5124
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.128773474822439,
+      "learning_rate": 5.784428326330143e-06,
+      "loss": 0.1308,
+      "step": 5125
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9968719846992482,
+      "learning_rate": 5.782972860440517e-06,
+      "loss": 0.1358,
+      "step": 5126
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8717743962610498,
+      "learning_rate": 5.781517326534509e-06,
+      "loss": 0.1249,
+      "step": 5127
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8655682885447384,
+      "learning_rate": 5.780061724738559e-06,
+      "loss": 0.1248,
+      "step": 5128
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7267417061042534,
+      "learning_rate": 5.778606055179117e-06,
+      "loss": 0.0936,
+      "step": 5129
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7412119215793365,
+      "learning_rate": 5.777150317982636e-06,
+      "loss": 0.1023,
+      "step": 5130
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0767667159637377,
+      "learning_rate": 5.7756945132755715e-06,
+      "loss": 0.1144,
+      "step": 5131
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.124782594984012,
+      "learning_rate": 5.774238641184391e-06,
+      "loss": 0.149,
+      "step": 5132
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.2121331224697167,
+      "learning_rate": 5.7727827018355665e-06,
+      "loss": 0.13,
+      "step": 5133
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6512234751839003,
+      "learning_rate": 5.771326695355573e-06,
+      "loss": 0.0849,
+      "step": 5134
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8300428370996042,
+      "learning_rate": 5.76987062187089e-06,
+      "loss": 0.1235,
+      "step": 5135
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7915115070904726,
+      "learning_rate": 5.768414481508011e-06,
+      "loss": 0.1095,
+      "step": 5136
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8695414499671794,
+      "learning_rate": 5.766958274393428e-06,
+      "loss": 0.1046,
+      "step": 5137
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.038623042397329,
+      "learning_rate": 5.765502000653639e-06,
+      "loss": 0.1391,
+      "step": 5138
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7753507956216645,
+      "learning_rate": 5.764045660415153e-06,
+      "loss": 0.1336,
+      "step": 5139
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6353312881382196,
+      "learning_rate": 5.762589253804478e-06,
+      "loss": 0.095,
+      "step": 5140
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8982971180945565,
+      "learning_rate": 5.761132780948132e-06,
+      "loss": 0.1267,
+      "step": 5141
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6510989612078975,
+      "learning_rate": 5.75967624197264e-06,
+      "loss": 0.0979,
+      "step": 5142
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9611921514073136,
+      "learning_rate": 5.758219637004529e-06,
+      "loss": 0.1342,
+      "step": 5143
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7749491898904008,
+      "learning_rate": 5.756762966170334e-06,
+      "loss": 0.1204,
+      "step": 5144
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.08518080656657,
+      "learning_rate": 5.755306229596594e-06,
+      "loss": 0.1094,
+      "step": 5145
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8935107554233241,
+      "learning_rate": 5.753849427409857e-06,
+      "loss": 0.1295,
+      "step": 5146
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9037261979553235,
+      "learning_rate": 5.752392559736671e-06,
+      "loss": 0.1196,
+      "step": 5147
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0466734713293575,
+      "learning_rate": 5.750935626703598e-06,
+      "loss": 0.1136,
+      "step": 5148
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7773326058936851,
+      "learning_rate": 5.749478628437196e-06,
+      "loss": 0.102,
+      "step": 5149
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.972591120865003,
+      "learning_rate": 5.748021565064037e-06,
+      "loss": 0.1307,
+      "step": 5150
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8704000697522698,
+      "learning_rate": 5.746564436710694e-06,
+      "loss": 0.1203,
+      "step": 5151
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.016668253341138,
+      "learning_rate": 5.745107243503747e-06,
+      "loss": 0.1317,
+      "step": 5152
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8615736932006264,
+      "learning_rate": 5.74364998556978e-06,
+      "loss": 0.1192,
+      "step": 5153
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.821764223084276,
+      "learning_rate": 5.742192663035388e-06,
+      "loss": 0.1217,
+      "step": 5154
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.078251197216359,
+      "learning_rate": 5.740735276027164e-06,
+      "loss": 0.1228,
+      "step": 5155
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0492589209062237,
+      "learning_rate": 5.739277824671711e-06,
+      "loss": 0.1211,
+      "step": 5156
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9216932825568611,
+      "learning_rate": 5.737820309095639e-06,
+      "loss": 0.1148,
+      "step": 5157
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.931295093359262,
+      "learning_rate": 5.736362729425558e-06,
+      "loss": 0.1199,
+      "step": 5158
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.923224287788985,
+      "learning_rate": 5.734905085788091e-06,
+      "loss": 0.1226,
+      "step": 5159
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9387714382958043,
+      "learning_rate": 5.733447378309861e-06,
+      "loss": 0.1254,
+      "step": 5160
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7823131147660431,
+      "learning_rate": 5.731989607117497e-06,
+      "loss": 0.1001,
+      "step": 5161
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9993065746969982,
+      "learning_rate": 5.730531772337634e-06,
+      "loss": 0.1239,
+      "step": 5162
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9201976216481262,
+      "learning_rate": 5.729073874096917e-06,
+      "loss": 0.1192,
+      "step": 5163
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.6441507211020685,
+      "learning_rate": 5.72761591252199e-06,
+      "loss": 0.1298,
+      "step": 5164
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.139222525703095,
+      "learning_rate": 5.726157887739505e-06,
+      "loss": 0.147,
+      "step": 5165
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9035990479721026,
+      "learning_rate": 5.724699799876124e-06,
+      "loss": 0.1252,
+      "step": 5166
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8101621125422658,
+      "learning_rate": 5.723241649058503e-06,
+      "loss": 0.1125,
+      "step": 5167
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.171727132080312,
+      "learning_rate": 5.721783435413315e-06,
+      "loss": 0.1176,
+      "step": 5168
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8652802140485323,
+      "learning_rate": 5.7203251590672345e-06,
+      "loss": 0.1229,
+      "step": 5169
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0042443396582086,
+      "learning_rate": 5.71886682014694e-06,
+      "loss": 0.1226,
+      "step": 5170
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7492718707986108,
+      "learning_rate": 5.7174084187791165e-06,
+      "loss": 0.1144,
+      "step": 5171
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0745748578950893,
+      "learning_rate": 5.715949955090456e-06,
+      "loss": 0.1201,
+      "step": 5172
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8303897475427093,
+      "learning_rate": 5.714491429207651e-06,
+      "loss": 0.1123,
+      "step": 5173
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9188090421231063,
+      "learning_rate": 5.713032841257407e-06,
+      "loss": 0.1176,
+      "step": 5174
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.661572595767909,
+      "learning_rate": 5.711574191366427e-06,
+      "loss": 0.1141,
+      "step": 5175
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7389120063191656,
+      "learning_rate": 5.710115479661425e-06,
+      "loss": 0.1078,
+      "step": 5176
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7369512408618193,
+      "learning_rate": 5.708656706269117e-06,
+      "loss": 0.1037,
+      "step": 5177
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9359516778878942,
+      "learning_rate": 5.707197871316228e-06,
+      "loss": 0.1277,
+      "step": 5178
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.335203195175909,
+      "learning_rate": 5.705738974929484e-06,
+      "loss": 0.1544,
+      "step": 5179
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6346588492619258,
+      "learning_rate": 5.70428001723562e-06,
+      "loss": 0.0881,
+      "step": 5180
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.695901369889061,
+      "learning_rate": 5.702820998361374e-06,
+      "loss": 0.1018,
+      "step": 5181
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.5519986490440538,
+      "learning_rate": 5.701361918433489e-06,
+      "loss": 0.0934,
+      "step": 5182
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1031870955854868,
+      "learning_rate": 5.699902777578716e-06,
+      "loss": 0.1203,
+      "step": 5183
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1861747085122927,
+      "learning_rate": 5.69844357592381e-06,
+      "loss": 0.1411,
+      "step": 5184
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0106468313162167,
+      "learning_rate": 5.696984313595529e-06,
+      "loss": 0.1267,
+      "step": 5185
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8232743567063099,
+      "learning_rate": 5.69552499072064e-06,
+      "loss": 0.0933,
+      "step": 5186
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.5936963866086897,
+      "learning_rate": 5.694065607425914e-06,
+      "loss": 0.0904,
+      "step": 5187
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1360558712143707,
+      "learning_rate": 5.692606163838125e-06,
+      "loss": 0.1279,
+      "step": 5188
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.195234164364305,
+      "learning_rate": 5.6911466600840535e-06,
+      "loss": 0.105,
+      "step": 5189
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9116659246370282,
+      "learning_rate": 5.689687096290488e-06,
+      "loss": 0.1258,
+      "step": 5190
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6544165835655402,
+      "learning_rate": 5.688227472584218e-06,
+      "loss": 0.0969,
+      "step": 5191
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9026982124008578,
+      "learning_rate": 5.686767789092041e-06,
+      "loss": 0.1086,
+      "step": 5192
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0786105027061814,
+      "learning_rate": 5.68530804594076e-06,
+      "loss": 0.1317,
+      "step": 5193
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.2290110759518957,
+      "learning_rate": 5.683848243257181e-06,
+      "loss": 0.1141,
+      "step": 5194
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.748043193744473,
+      "learning_rate": 5.682388381168115e-06,
+      "loss": 0.1214,
+      "step": 5195
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9772083775253568,
+      "learning_rate": 5.68092845980038e-06,
+      "loss": 0.1161,
+      "step": 5196
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9116313747404785,
+      "learning_rate": 5.679468479280798e-06,
+      "loss": 0.1161,
+      "step": 5197
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9262304952269778,
+      "learning_rate": 5.678008439736198e-06,
+      "loss": 0.1099,
+      "step": 5198
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.029175851506085,
+      "learning_rate": 5.6765483412934144e-06,
+      "loss": 0.1344,
+      "step": 5199
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7106856206631176,
+      "learning_rate": 5.67508818407928e-06,
+      "loss": 0.1052,
+      "step": 5200
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.79970087680083,
+      "learning_rate": 5.673627968220642e-06,
+      "loss": 0.1058,
+      "step": 5201
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7264443631193565,
+      "learning_rate": 5.672167693844348e-06,
+      "loss": 0.105,
+      "step": 5202
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.5255982165042767,
+      "learning_rate": 5.670707361077249e-06,
+      "loss": 0.0941,
+      "step": 5203
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8350066059754944,
+      "learning_rate": 5.669246970046206e-06,
+      "loss": 0.115,
+      "step": 5204
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8149584151390863,
+      "learning_rate": 5.667786520878079e-06,
+      "loss": 0.1177,
+      "step": 5205
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.123866810707329,
+      "learning_rate": 5.666326013699739e-06,
+      "loss": 0.1476,
+      "step": 5206
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8146274675873144,
+      "learning_rate": 5.664865448638059e-06,
+      "loss": 0.1067,
+      "step": 5207
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8350339308023162,
+      "learning_rate": 5.663404825819916e-06,
+      "loss": 0.1035,
+      "step": 5208
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8973912403158681,
+      "learning_rate": 5.661944145372193e-06,
+      "loss": 0.1168,
+      "step": 5209
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0352549219325153,
+      "learning_rate": 5.660483407421783e-06,
+      "loss": 0.1231,
+      "step": 5210
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7064762501848323,
+      "learning_rate": 5.659022612095575e-06,
+      "loss": 0.1123,
+      "step": 5211
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7056421520573755,
+      "learning_rate": 5.657561759520467e-06,
+      "loss": 0.1092,
+      "step": 5212
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.061808326111,
+      "learning_rate": 5.656100849823366e-06,
+      "loss": 0.1299,
+      "step": 5213
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.820253565660934,
+      "learning_rate": 5.6546398831311774e-06,
+      "loss": 0.1085,
+      "step": 5214
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.066764402401858,
+      "learning_rate": 5.6531788595708155e-06,
+      "loss": 0.112,
+      "step": 5215
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.221039780122519,
+      "learning_rate": 5.6517177792692005e-06,
+      "loss": 0.1232,
+      "step": 5216
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8134641770559579,
+      "learning_rate": 5.650256642353251e-06,
+      "loss": 0.1184,
+      "step": 5217
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9443455206721616,
+      "learning_rate": 5.648795448949898e-06,
+      "loss": 0.1362,
+      "step": 5218
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.114599231343073,
+      "learning_rate": 5.6473341991860755e-06,
+      "loss": 0.1461,
+      "step": 5219
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8069603848787623,
+      "learning_rate": 5.645872893188718e-06,
+      "loss": 0.111,
+      "step": 5220
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.774438468258415,
+      "learning_rate": 5.644411531084771e-06,
+      "loss": 0.1212,
+      "step": 5221
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8343346473995494,
+      "learning_rate": 5.642950113001183e-06,
+      "loss": 0.1286,
+      "step": 5222
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.850253466989269,
+      "learning_rate": 5.641488639064904e-06,
+      "loss": 0.1255,
+      "step": 5223
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.853881721111044,
+      "learning_rate": 5.640027109402892e-06,
+      "loss": 0.1144,
+      "step": 5224
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.750781899846873,
+      "learning_rate": 5.638565524142111e-06,
+      "loss": 0.0837,
+      "step": 5225
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7737824940329674,
+      "learning_rate": 5.637103883409525e-06,
+      "loss": 0.0917,
+      "step": 5226
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.166112960748287,
+      "learning_rate": 5.635642187332108e-06,
+      "loss": 0.1438,
+      "step": 5227
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.994350547631619,
+      "learning_rate": 5.634180436036836e-06,
+      "loss": 0.1232,
+      "step": 5228
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8041519762508742,
+      "learning_rate": 5.63271862965069e-06,
+      "loss": 0.0873,
+      "step": 5229
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7876646858536376,
+      "learning_rate": 5.6312567683006565e-06,
+      "loss": 0.1044,
+      "step": 5230
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3974761630972368,
+      "learning_rate": 5.629794852113729e-06,
+      "loss": 0.1209,
+      "step": 5231
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1172445704772316,
+      "learning_rate": 5.628332881216899e-06,
+      "loss": 0.14,
+      "step": 5232
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.010250280278537,
+      "learning_rate": 5.6268708557371695e-06,
+      "loss": 0.1308,
+      "step": 5233
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.929899743094044,
+      "learning_rate": 5.625408775801546e-06,
+      "loss": 0.1851,
+      "step": 5234
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.700682830707001,
+      "learning_rate": 5.623946641537038e-06,
+      "loss": 0.1181,
+      "step": 5235
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.726260156262408,
+      "learning_rate": 5.622484453070659e-06,
+      "loss": 0.1188,
+      "step": 5236
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7428440268994987,
+      "learning_rate": 5.621022210529431e-06,
+      "loss": 0.107,
+      "step": 5237
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9425193704864747,
+      "learning_rate": 5.619559914040376e-06,
+      "loss": 0.1377,
+      "step": 5238
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.132174108983779,
+      "learning_rate": 5.618097563730522e-06,
+      "loss": 0.1358,
+      "step": 5239
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.026909821081086,
+      "learning_rate": 5.616635159726907e-06,
+      "loss": 0.1312,
+      "step": 5240
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7817257757433718,
+      "learning_rate": 5.615172702156564e-06,
+      "loss": 0.1109,
+      "step": 5241
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9056466491720712,
+      "learning_rate": 5.613710191146539e-06,
+      "loss": 0.1362,
+      "step": 5242
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8344045572626833,
+      "learning_rate": 5.612247626823878e-06,
+      "loss": 0.1206,
+      "step": 5243
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8614398371496157,
+      "learning_rate": 5.610785009315633e-06,
+      "loss": 0.1296,
+      "step": 5244
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9759860983485649,
+      "learning_rate": 5.609322338748861e-06,
+      "loss": 0.1252,
+      "step": 5245
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9552709724238888,
+      "learning_rate": 5.607859615250626e-06,
+      "loss": 0.1267,
+      "step": 5246
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6787095350622503,
+      "learning_rate": 5.606396838947988e-06,
+      "loss": 0.1035,
+      "step": 5247
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7447287687220425,
+      "learning_rate": 5.604934009968023e-06,
+      "loss": 0.1035,
+      "step": 5248
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.638886627808892,
+      "learning_rate": 5.603471128437804e-06,
+      "loss": 0.1046,
+      "step": 5249
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8234028366458395,
+      "learning_rate": 5.60200819448441e-06,
+      "loss": 0.0954,
+      "step": 5250
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1610716059279964,
+      "learning_rate": 5.600545208234927e-06,
+      "loss": 0.1385,
+      "step": 5251
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9966944765756987,
+      "learning_rate": 5.599082169816441e-06,
+      "loss": 0.12,
+      "step": 5252
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.310644494404322,
+      "learning_rate": 5.597619079356047e-06,
+      "loss": 0.1393,
+      "step": 5253
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7647891542698966,
+      "learning_rate": 5.596155936980844e-06,
+      "loss": 0.1011,
+      "step": 5254
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9427808823891612,
+      "learning_rate": 5.594692742817932e-06,
+      "loss": 0.1182,
+      "step": 5255
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8641421740813948,
+      "learning_rate": 5.593229496994419e-06,
+      "loss": 0.1102,
+      "step": 5256
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9689183204187046,
+      "learning_rate": 5.5917661996374155e-06,
+      "loss": 0.0946,
+      "step": 5257
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7226767739737063,
+      "learning_rate": 5.5903028508740385e-06,
+      "loss": 0.1103,
+      "step": 5258
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2352050278918365,
+      "learning_rate": 5.588839450831407e-06,
+      "loss": 0.1302,
+      "step": 5259
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9616214256530697,
+      "learning_rate": 5.587375999636645e-06,
+      "loss": 0.1342,
+      "step": 5260
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.997698391433047,
+      "learning_rate": 5.585912497416885e-06,
+      "loss": 0.1106,
+      "step": 5261
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2240891896275716,
+      "learning_rate": 5.5844489442992575e-06,
+      "loss": 0.1424,
+      "step": 5262
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6917312742149904,
+      "learning_rate": 5.582985340410901e-06,
+      "loss": 0.0982,
+      "step": 5263
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.208501316386937,
+      "learning_rate": 5.581521685878959e-06,
+      "loss": 0.1304,
+      "step": 5264
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6825963992276745,
+      "learning_rate": 5.5800579808305766e-06,
+      "loss": 0.1036,
+      "step": 5265
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6003004788966047,
+      "learning_rate": 5.578594225392906e-06,
+      "loss": 0.1045,
+      "step": 5266
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.925600542779073,
+      "learning_rate": 5.577130419693104e-06,
+      "loss": 0.125,
+      "step": 5267
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0109714005885895,
+      "learning_rate": 5.575666563858329e-06,
+      "loss": 0.1156,
+      "step": 5268
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8871176870291024,
+      "learning_rate": 5.574202658015744e-06,
+      "loss": 0.1179,
+      "step": 5269
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.916120523130814,
+      "learning_rate": 5.57273870229252e-06,
+      "loss": 0.0988,
+      "step": 5270
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.018212891320888,
+      "learning_rate": 5.571274696815828e-06,
+      "loss": 0.1385,
+      "step": 5271
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7706841540620224,
+      "learning_rate": 5.569810641712847e-06,
+      "loss": 0.1225,
+      "step": 5272
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0651736163841328,
+      "learning_rate": 5.568346537110759e-06,
+      "loss": 0.135,
+      "step": 5273
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6153609832755442,
+      "learning_rate": 5.566882383136748e-06,
+      "loss": 0.1077,
+      "step": 5274
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6231648979903837,
+      "learning_rate": 5.565418179918004e-06,
+      "loss": 0.105,
+      "step": 5275
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.286318659616542,
+      "learning_rate": 5.563953927581724e-06,
+      "loss": 0.1395,
+      "step": 5276
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8072308724091177,
+      "learning_rate": 5.562489626255104e-06,
+      "loss": 0.1174,
+      "step": 5277
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0271338722833763,
+      "learning_rate": 5.561025276065348e-06,
+      "loss": 0.1391,
+      "step": 5278
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9143869577044594,
+      "learning_rate": 5.559560877139665e-06,
+      "loss": 0.1189,
+      "step": 5279
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.732693333924942,
+      "learning_rate": 5.558096429605263e-06,
+      "loss": 0.113,
+      "step": 5280
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8761674973781715,
+      "learning_rate": 5.5566319335893604e-06,
+      "loss": 0.1131,
+      "step": 5281
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6265334383371182,
+      "learning_rate": 5.555167389219176e-06,
+      "loss": 0.1137,
+      "step": 5282
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7019982464878458,
+      "learning_rate": 5.553702796621933e-06,
+      "loss": 0.0989,
+      "step": 5283
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.90156828208498,
+      "learning_rate": 5.552238155924861e-06,
+      "loss": 0.1188,
+      "step": 5284
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9826773284013783,
+      "learning_rate": 5.550773467255195e-06,
+      "loss": 0.1182,
+      "step": 5285
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.076557665138287,
+      "learning_rate": 5.549308730740166e-06,
+      "loss": 0.1295,
+      "step": 5286
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1533821835296334,
+      "learning_rate": 5.5478439465070174e-06,
+      "loss": 0.153,
+      "step": 5287
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7381585741314176,
+      "learning_rate": 5.546379114682996e-06,
+      "loss": 0.1066,
+      "step": 5288
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6635927313462626,
+      "learning_rate": 5.544914235395347e-06,
+      "loss": 0.1073,
+      "step": 5289
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.781751405359924,
+      "learning_rate": 5.543449308771328e-06,
+      "loss": 0.1206,
+      "step": 5290
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.971139103329158,
+      "learning_rate": 5.541984334938193e-06,
+      "loss": 0.1346,
+      "step": 5291
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6186839785683387,
+      "learning_rate": 5.540519314023204e-06,
+      "loss": 0.0929,
+      "step": 5292
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1385548062551716,
+      "learning_rate": 5.5390542461536275e-06,
+      "loss": 0.1564,
+      "step": 5293
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7201549108815215,
+      "learning_rate": 5.5375891314567335e-06,
+      "loss": 0.1245,
+      "step": 5294
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.822533325792133,
+      "learning_rate": 5.536123970059793e-06,
+      "loss": 0.1236,
+      "step": 5295
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9278138382341734,
+      "learning_rate": 5.534658762090087e-06,
+      "loss": 0.1208,
+      "step": 5296
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.5521411642870067,
+      "learning_rate": 5.533193507674895e-06,
+      "loss": 0.0936,
+      "step": 5297
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8004949538420973,
+      "learning_rate": 5.531728206941502e-06,
+      "loss": 0.0979,
+      "step": 5298
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.5087233539099076,
+      "learning_rate": 5.5302628600172005e-06,
+      "loss": 0.0988,
+      "step": 5299
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6392810666795463,
+      "learning_rate": 5.5287974670292825e-06,
+      "loss": 0.0997,
+      "step": 5300
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.246531209545475,
+      "learning_rate": 5.527332028105046e-06,
+      "loss": 0.1441,
+      "step": 5301
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.5906231938863176,
+      "learning_rate": 5.525866543371794e-06,
+      "loss": 0.1063,
+      "step": 5302
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.641074152783996,
+      "learning_rate": 5.5244010129568294e-06,
+      "loss": 0.1252,
+      "step": 5303
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.096881893280961,
+      "learning_rate": 5.522935436987465e-06,
+      "loss": 0.1363,
+      "step": 5304
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7027707552441753,
+      "learning_rate": 5.521469815591014e-06,
+      "loss": 0.1063,
+      "step": 5305
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9685022450361709,
+      "learning_rate": 5.520004148894793e-06,
+      "loss": 0.1286,
+      "step": 5306
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7145771958239175,
+      "learning_rate": 5.518538437026123e-06,
+      "loss": 0.1123,
+      "step": 5307
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8889040155802865,
+      "learning_rate": 5.517072680112332e-06,
+      "loss": 0.1133,
+      "step": 5308
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8396074353811016,
+      "learning_rate": 5.515606878280747e-06,
+      "loss": 0.1337,
+      "step": 5309
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8626200442685266,
+      "learning_rate": 5.514141031658703e-06,
+      "loss": 0.1292,
+      "step": 5310
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.749966327344818,
+      "learning_rate": 5.512675140373537e-06,
+      "loss": 0.1161,
+      "step": 5311
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9994239729956802,
+      "learning_rate": 5.511209204552588e-06,
+      "loss": 0.139,
+      "step": 5312
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.81852097240687,
+      "learning_rate": 5.509743224323203e-06,
+      "loss": 0.1128,
+      "step": 5313
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6591640013897353,
+      "learning_rate": 5.508277199812732e-06,
+      "loss": 0.1029,
+      "step": 5314
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8647796237086316,
+      "learning_rate": 5.506811131148524e-06,
+      "loss": 0.119,
+      "step": 5315
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.924104771325096,
+      "learning_rate": 5.5053450184579374e-06,
+      "loss": 0.0997,
+      "step": 5316
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.821136979857864,
+      "learning_rate": 5.5038788618683335e-06,
+      "loss": 0.1199,
+      "step": 5317
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6883172787387781,
+      "learning_rate": 5.502412661507076e-06,
+      "loss": 0.1038,
+      "step": 5318
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7017783469327037,
+      "learning_rate": 5.500946417501532e-06,
+      "loss": 0.1075,
+      "step": 5319
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9339275791805233,
+      "learning_rate": 5.499480129979073e-06,
+      "loss": 0.1171,
+      "step": 5320
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9950668374318195,
+      "learning_rate": 5.498013799067077e-06,
+      "loss": 0.1181,
+      "step": 5321
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0864399431821004,
+      "learning_rate": 5.49654742489292e-06,
+      "loss": 0.1301,
+      "step": 5322
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0548946930359393,
+      "learning_rate": 5.495081007583986e-06,
+      "loss": 0.1422,
+      "step": 5323
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.746566338432751,
+      "learning_rate": 5.493614547267664e-06,
+      "loss": 0.1039,
+      "step": 5324
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6622084833057937,
+      "learning_rate": 5.492148044071342e-06,
+      "loss": 0.0979,
+      "step": 5325
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6182923331828287,
+      "learning_rate": 5.490681498122415e-06,
+      "loss": 0.1179,
+      "step": 5326
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2117874882886666,
+      "learning_rate": 5.4892149095482815e-06,
+      "loss": 0.1319,
+      "step": 5327
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9514931195335927,
+      "learning_rate": 5.487748278476342e-06,
+      "loss": 0.1075,
+      "step": 5328
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0211051629681203,
+      "learning_rate": 5.486281605034004e-06,
+      "loss": 0.1118,
+      "step": 5329
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.749586553830844,
+      "learning_rate": 5.484814889348673e-06,
+      "loss": 0.117,
+      "step": 5330
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7387106373355585,
+      "learning_rate": 5.483348131547765e-06,
+      "loss": 0.1102,
+      "step": 5331
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7869853654007763,
+      "learning_rate": 5.481881331758696e-06,
+      "loss": 0.119,
+      "step": 5332
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.713687749713591,
+      "learning_rate": 5.480414490108884e-06,
+      "loss": 0.1046,
+      "step": 5333
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0489839934274596,
+      "learning_rate": 5.478947606725754e-06,
+      "loss": 0.1246,
+      "step": 5334
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1280105722323204,
+      "learning_rate": 5.477480681736734e-06,
+      "loss": 0.1373,
+      "step": 5335
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.715127355121317,
+      "learning_rate": 5.476013715269254e-06,
+      "loss": 0.1147,
+      "step": 5336
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8742772229921738,
+      "learning_rate": 5.474546707450748e-06,
+      "loss": 0.1275,
+      "step": 5337
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7982285423370132,
+      "learning_rate": 5.473079658408655e-06,
+      "loss": 0.1172,
+      "step": 5338
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.900943795778086,
+      "learning_rate": 5.471612568270415e-06,
+      "loss": 0.1042,
+      "step": 5339
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7097739134935468,
+      "learning_rate": 5.4701454371634756e-06,
+      "loss": 0.0953,
+      "step": 5340
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6460378318747704,
+      "learning_rate": 5.468678265215286e-06,
+      "loss": 0.0935,
+      "step": 5341
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8542532618124885,
+      "learning_rate": 5.467211052553295e-06,
+      "loss": 0.1197,
+      "step": 5342
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6895578666012736,
+      "learning_rate": 5.465743799304961e-06,
+      "loss": 0.1084,
+      "step": 5343
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6244813519278594,
+      "learning_rate": 5.464276505597743e-06,
+      "loss": 0.0952,
+      "step": 5344
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7419565845525498,
+      "learning_rate": 5.462809171559104e-06,
+      "loss": 0.097,
+      "step": 5345
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.203499387502144,
+      "learning_rate": 5.46134179731651e-06,
+      "loss": 0.1267,
+      "step": 5346
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.5798623980431044,
+      "learning_rate": 5.4598743829974334e-06,
+      "loss": 0.0944,
+      "step": 5347
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.907882318608598,
+      "learning_rate": 5.458406928729343e-06,
+      "loss": 0.1287,
+      "step": 5348
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8825765376062369,
+      "learning_rate": 5.456939434639719e-06,
+      "loss": 0.1122,
+      "step": 5349
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9507807590122432,
+      "learning_rate": 5.455471900856041e-06,
+      "loss": 0.1251,
+      "step": 5350
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.038331018661256,
+      "learning_rate": 5.454004327505792e-06,
+      "loss": 0.1266,
+      "step": 5351
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8049492614449725,
+      "learning_rate": 5.45253671471646e-06,
+      "loss": 0.1193,
+      "step": 5352
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.5526121393090817,
+      "learning_rate": 5.451069062615536e-06,
+      "loss": 0.0997,
+      "step": 5353
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9678547710699303,
+      "learning_rate": 5.4496013713305126e-06,
+      "loss": 0.1216,
+      "step": 5354
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.141050783221842,
+      "learning_rate": 5.4481336409888886e-06,
+      "loss": 0.1249,
+      "step": 5355
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7455599854222645,
+      "learning_rate": 5.446665871718166e-06,
+      "loss": 0.117,
+      "step": 5356
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7707706774333558,
+      "learning_rate": 5.445198063645844e-06,
+      "loss": 0.1124,
+      "step": 5357
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.613915861689703,
+      "learning_rate": 5.443730216899437e-06,
+      "loss": 0.0971,
+      "step": 5358
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9841003393937986,
+      "learning_rate": 5.442262331606451e-06,
+      "loss": 0.1303,
+      "step": 5359
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9577249804294283,
+      "learning_rate": 5.440794407894403e-06,
+      "loss": 0.1195,
+      "step": 5360
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9162297061922315,
+      "learning_rate": 5.439326445890808e-06,
+      "loss": 0.1437,
+      "step": 5361
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.50385774696173,
+      "learning_rate": 5.437858445723191e-06,
+      "loss": 0.09,
+      "step": 5362
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.811790801267793,
+      "learning_rate": 5.436390407519072e-06,
+      "loss": 0.1179,
+      "step": 5363
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.139819314232434,
+      "learning_rate": 5.43492233140598e-06,
+      "loss": 0.1269,
+      "step": 5364
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6323557311836059,
+      "learning_rate": 5.4334542175114495e-06,
+      "loss": 0.0953,
+      "step": 5365
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8822564660148549,
+      "learning_rate": 5.431986065963008e-06,
+      "loss": 0.1268,
+      "step": 5366
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8287951562105682,
+      "learning_rate": 5.430517876888199e-06,
+      "loss": 0.1111,
+      "step": 5367
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9578642108356705,
+      "learning_rate": 5.4290496504145595e-06,
+      "loss": 0.1379,
+      "step": 5368
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8105815968441237,
+      "learning_rate": 5.427581386669635e-06,
+      "loss": 0.1196,
+      "step": 5369
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.7774140981727296,
+      "learning_rate": 5.426113085780971e-06,
+      "loss": 0.1085,
+      "step": 5370
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9187038356090786,
+      "learning_rate": 5.424644747876121e-06,
+      "loss": 0.117,
+      "step": 5371
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.0099795512932945,
+      "learning_rate": 5.423176373082636e-06,
+      "loss": 0.093,
+      "step": 5372
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8668897617992943,
+      "learning_rate": 5.421707961528073e-06,
+      "loss": 0.1123,
+      "step": 5373
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8400590378911752,
+      "learning_rate": 5.4202395133399955e-06,
+      "loss": 0.1244,
+      "step": 5374
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9791363922410767,
+      "learning_rate": 5.418771028645962e-06,
+      "loss": 0.117,
+      "step": 5375
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.7833718931959492,
+      "learning_rate": 5.41730250757354e-06,
+      "loss": 0.1096,
+      "step": 5376
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8667183246935424,
+      "learning_rate": 5.415833950250302e-06,
+      "loss": 0.1197,
+      "step": 5377
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.941056875846269,
+      "learning_rate": 5.414365356803817e-06,
+      "loss": 0.1238,
+      "step": 5378
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.6231093106127832,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.0989,
+      "step": 5379
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.622461326398856,
+      "learning_rate": 5.411428062051418e-06,
+      "loss": 0.099,
+      "step": 5380
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.6973197504775575,
+      "learning_rate": 5.409959361000665e-06,
+      "loss": 0.1103,
+      "step": 5381
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.036626134080738,
+      "learning_rate": 5.408490624336987e-06,
+      "loss": 0.124,
+      "step": 5382
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.0390093464123584,
+      "learning_rate": 5.407021852187976e-06,
+      "loss": 0.132,
+      "step": 5383
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8688049493497199,
+      "learning_rate": 5.40555304468122e-06,
+      "loss": 0.1087,
+      "step": 5384
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.7657774746978248,
+      "learning_rate": 5.404084201944315e-06,
+      "loss": 0.1167,
+      "step": 5385
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.667686551566891,
+      "learning_rate": 5.402615324104858e-06,
+      "loss": 0.0977,
+      "step": 5386
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8424574401429858,
+      "learning_rate": 5.40114641129045e-06,
+      "loss": 0.1233,
+      "step": 5387
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9061416063010608,
+      "learning_rate": 5.399677463628695e-06,
+      "loss": 0.1194,
+      "step": 5388
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.6374474404333954,
+      "learning_rate": 5.398208481247198e-06,
+      "loss": 0.1071,
+      "step": 5389
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.0099240870564454,
+      "learning_rate": 5.396739464273569e-06,
+      "loss": 0.1207,
+      "step": 5390
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.938987331654749,
+      "learning_rate": 5.395270412835423e-06,
+      "loss": 0.1181,
+      "step": 5391
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9258258082537223,
+      "learning_rate": 5.393801327060372e-06,
+      "loss": 0.1225,
+      "step": 5392
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9710305906378367,
+      "learning_rate": 5.392332207076036e-06,
+      "loss": 0.1317,
+      "step": 5393
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.1308237195391513,
+      "learning_rate": 5.390863053010038e-06,
+      "loss": 0.1345,
+      "step": 5394
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8339198457387127,
+      "learning_rate": 5.389393864990001e-06,
+      "loss": 0.1229,
+      "step": 5395
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.0524854673575383,
+      "learning_rate": 5.387924643143553e-06,
+      "loss": 0.1245,
+      "step": 5396
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9154452065107503,
+      "learning_rate": 5.386455387598325e-06,
+      "loss": 0.1235,
+      "step": 5397
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.0372973714155407,
+      "learning_rate": 5.384986098481948e-06,
+      "loss": 0.12,
+      "step": 5398
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.870346753126263,
+      "learning_rate": 5.383516775922061e-06,
+      "loss": 0.1203,
+      "step": 5399
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8854025881316814,
+      "learning_rate": 5.382047420046302e-06,
+      "loss": 0.115,
+      "step": 5400
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.768386777215909,
+      "learning_rate": 5.380578030982313e-06,
+      "loss": 0.1232,
+      "step": 5401
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8956770024484721,
+      "learning_rate": 5.379108608857739e-06,
+      "loss": 0.1187,
+      "step": 5402
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.230955789474564,
+      "learning_rate": 5.377639153800229e-06,
+      "loss": 0.1317,
+      "step": 5403
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8160872160067627,
+      "learning_rate": 5.3761696659374315e-06,
+      "loss": 0.1016,
+      "step": 5404
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8644833216940295,
+      "learning_rate": 5.3747001453970005e-06,
+      "loss": 0.1218,
+      "step": 5405
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8640313511085593,
+      "learning_rate": 5.373230592306595e-06,
+      "loss": 0.1299,
+      "step": 5406
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.7137013584991465,
+      "learning_rate": 5.371761006793871e-06,
+      "loss": 0.1005,
+      "step": 5407
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9408380939281658,
+      "learning_rate": 5.370291388986491e-06,
+      "loss": 0.1246,
+      "step": 5408
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9167641967436153,
+      "learning_rate": 5.368821739012122e-06,
+      "loss": 0.1259,
+      "step": 5409
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.6288732396568644,
+      "learning_rate": 5.367352056998429e-06,
+      "loss": 0.0982,
+      "step": 5410
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9667016016364929,
+      "learning_rate": 5.3658823430730834e-06,
+      "loss": 0.1351,
+      "step": 5411
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8630565890108357,
+      "learning_rate": 5.36441259736376e-06,
+      "loss": 0.1145,
+      "step": 5412
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.051144487439951,
+      "learning_rate": 5.362942819998131e-06,
+      "loss": 0.1268,
+      "step": 5413
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9013794972576763,
+      "learning_rate": 5.361473011103879e-06,
+      "loss": 0.1059,
+      "step": 5414
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9188026307920527,
+      "learning_rate": 5.360003170808684e-06,
+      "loss": 0.1276,
+      "step": 5415
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9047557059065274,
+      "learning_rate": 5.358533299240228e-06,
+      "loss": 0.1279,
+      "step": 5416
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8222789137510702,
+      "learning_rate": 5.357063396526201e-06,
+      "loss": 0.1086,
+      "step": 5417
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9020706774030296,
+      "learning_rate": 5.355593462794292e-06,
+      "loss": 0.1197,
+      "step": 5418
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.075733079654561,
+      "learning_rate": 5.354123498172191e-06,
+      "loss": 0.1251,
+      "step": 5419
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8623482595294856,
+      "learning_rate": 5.352653502787595e-06,
+      "loss": 0.1105,
+      "step": 5420
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8098337804556652,
+      "learning_rate": 5.351183476768202e-06,
+      "loss": 0.1003,
+      "step": 5421
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8494545752735825,
+      "learning_rate": 5.34971342024171e-06,
+      "loss": 0.1258,
+      "step": 5422
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8290360296814934,
+      "learning_rate": 5.348243333335823e-06,
+      "loss": 0.1254,
+      "step": 5423
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.043789887935686,
+      "learning_rate": 5.346773216178248e-06,
+      "loss": 0.1078,
+      "step": 5424
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.0542332837201034,
+      "learning_rate": 5.345303068896692e-06,
+      "loss": 0.1328,
+      "step": 5425
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8579128298271494,
+      "learning_rate": 5.3438328916188655e-06,
+      "loss": 0.113,
+      "step": 5426
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.0727797046882337,
+      "learning_rate": 5.342362684472483e-06,
+      "loss": 0.1279,
+      "step": 5427
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9803406155549197,
+      "learning_rate": 5.3408924475852585e-06,
+      "loss": 0.1353,
+      "step": 5428
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.5916085315665354,
+      "learning_rate": 5.3394221810849125e-06,
+      "loss": 0.0966,
+      "step": 5429
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.7014633358515945,
+      "learning_rate": 5.337951885099167e-06,
+      "loss": 0.1052,
+      "step": 5430
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8184008827614397,
+      "learning_rate": 5.336481559755742e-06,
+      "loss": 0.1032,
+      "step": 5431
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.6992542137518225,
+      "learning_rate": 5.335011205182366e-06,
+      "loss": 0.1022,
+      "step": 5432
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.002014863379674,
+      "learning_rate": 5.33354082150677e-06,
+      "loss": 0.1258,
+      "step": 5433
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8476666819000511,
+      "learning_rate": 5.332070408856681e-06,
+      "loss": 0.1279,
+      "step": 5434
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.106112551655337,
+      "learning_rate": 5.330599967359836e-06,
+      "loss": 0.1267,
+      "step": 5435
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9703838168497059,
+      "learning_rate": 5.329129497143971e-06,
+      "loss": 0.1281,
+      "step": 5436
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.016946601179282,
+      "learning_rate": 5.327658998336825e-06,
+      "loss": 0.1389,
+      "step": 5437
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.42039575408107,
+      "learning_rate": 5.326188471066136e-06,
+      "loss": 0.1169,
+      "step": 5438
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.7268366868352651,
+      "learning_rate": 5.3247179154596525e-06,
+      "loss": 0.1164,
+      "step": 5439
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.7369725057881076,
+      "learning_rate": 5.323247331645118e-06,
+      "loss": 0.0978,
+      "step": 5440
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.6167532800095725,
+      "learning_rate": 5.321776719750283e-06,
+      "loss": 0.1004,
+      "step": 5441
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.0935810634785885,
+      "learning_rate": 5.3203060799028976e-06,
+      "loss": 0.1256,
+      "step": 5442
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8103419327591084,
+      "learning_rate": 5.318835412230714e-06,
+      "loss": 0.1025,
+      "step": 5443
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.957669970356974,
+      "learning_rate": 5.3173647168614906e-06,
+      "loss": 0.1043,
+      "step": 5444
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.185145221144955,
+      "learning_rate": 5.3158939939229855e-06,
+      "loss": 0.1231,
+      "step": 5445
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.5721845787137647,
+      "learning_rate": 5.314423243542959e-06,
+      "loss": 0.0971,
+      "step": 5446
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.0181474568045314,
+      "learning_rate": 5.312952465849173e-06,
+      "loss": 0.1081,
+      "step": 5447
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8344109746025858,
+      "learning_rate": 5.311481660969395e-06,
+      "loss": 0.1107,
+      "step": 5448
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.0499702113211424,
+      "learning_rate": 5.310010829031392e-06,
+      "loss": 0.1245,
+      "step": 5449
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.0650918523973703,
+      "learning_rate": 5.3085399701629344e-06,
+      "loss": 0.1311,
+      "step": 5450
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.111119967535554,
+      "learning_rate": 5.307069084491797e-06,
+      "loss": 0.1441,
+      "step": 5451
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.603316496380195,
+      "learning_rate": 5.305598172145751e-06,
+      "loss": 0.0938,
+      "step": 5452
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8980223416976045,
+      "learning_rate": 5.304127233252574e-06,
+      "loss": 0.1207,
+      "step": 5453
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.7889506035417992,
+      "learning_rate": 5.30265626794005e-06,
+      "loss": 0.0929,
+      "step": 5454
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8918028478029527,
+      "learning_rate": 5.301185276335956e-06,
+      "loss": 0.1127,
+      "step": 5455
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.2012681602529223,
+      "learning_rate": 5.299714258568077e-06,
+      "loss": 0.1196,
+      "step": 5456
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8833081500806705,
+      "learning_rate": 5.298243214764203e-06,
+      "loss": 0.1195,
+      "step": 5457
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8309389790151482,
+      "learning_rate": 5.296772145052118e-06,
+      "loss": 0.1077,
+      "step": 5458
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8552252204011952,
+      "learning_rate": 5.295301049559616e-06,
+      "loss": 0.1059,
+      "step": 5459
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.2029322268580778,
+      "learning_rate": 5.29382992841449e-06,
+      "loss": 0.1223,
+      "step": 5460
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.886531701622489,
+      "learning_rate": 5.292358781744533e-06,
+      "loss": 0.1114,
+      "step": 5461
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.7201363791895972,
+      "learning_rate": 5.290887609677545e-06,
+      "loss": 0.1125,
+      "step": 5462
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.9765320171125889,
+      "learning_rate": 5.289416412341326e-06,
+      "loss": 0.121,
+      "step": 5463
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.9735494735066739,
+      "learning_rate": 5.287945189863676e-06,
+      "loss": 0.1185,
+      "step": 5464
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.7265716090888994,
+      "learning_rate": 5.2864739423723996e-06,
+      "loss": 0.1089,
+      "step": 5465
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.9063092883715924,
+      "learning_rate": 5.285002669995306e-06,
+      "loss": 0.1226,
+      "step": 5466
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.6350154751736292,
+      "learning_rate": 5.283531372860201e-06,
+      "loss": 0.1087,
+      "step": 5467
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8433539037068998,
+      "learning_rate": 5.282060051094895e-06,
+      "loss": 0.1226,
+      "step": 5468
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.6937732976531295,
+      "learning_rate": 5.2805887048272035e-06,
+      "loss": 0.1138,
+      "step": 5469
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8574343730451959,
+      "learning_rate": 5.279117334184939e-06,
+      "loss": 0.124,
+      "step": 5470
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.88357716621281,
+      "learning_rate": 5.2776459392959186e-06,
+      "loss": 0.1077,
+      "step": 5471
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.7460263466076074,
+      "learning_rate": 5.2761745202879636e-06,
+      "loss": 0.1146,
+      "step": 5472
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.13290242786307,
+      "learning_rate": 5.274703077288893e-06,
+      "loss": 0.115,
+      "step": 5473
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.077001707197766,
+      "learning_rate": 5.27323161042653e-06,
+      "loss": 0.1343,
+      "step": 5474
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8596094435233093,
+      "learning_rate": 5.271760119828703e-06,
+      "loss": 0.1035,
+      "step": 5475
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.6517817795141545,
+      "learning_rate": 5.270288605623237e-06,
+      "loss": 0.1019,
+      "step": 5476
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.7913906566920585,
+      "learning_rate": 5.268817067937962e-06,
+      "loss": 0.1068,
+      "step": 5477
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.1806506240737864,
+      "learning_rate": 5.26734550690071e-06,
+      "loss": 0.1496,
+      "step": 5478
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.7354540816232031,
+      "learning_rate": 5.265873922639315e-06,
+      "loss": 0.1152,
+      "step": 5479
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.8956449065392988,
+      "learning_rate": 5.264402315281613e-06,
+      "loss": 0.1385,
+      "step": 5480
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.6082443965958646,
+      "learning_rate": 5.262930684955439e-06,
+      "loss": 0.0929,
+      "step": 5481
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9332125092649675,
+      "learning_rate": 5.261459031788634e-06,
+      "loss": 0.1298,
+      "step": 5482
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9033050135458704,
+      "learning_rate": 5.259987355909042e-06,
+      "loss": 0.1143,
+      "step": 5483
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.5923888783129834,
+      "learning_rate": 5.258515657444503e-06,
+      "loss": 0.1064,
+      "step": 5484
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.005318272274225,
+      "learning_rate": 5.257043936522864e-06,
+      "loss": 0.1352,
+      "step": 5485
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.182297560930908,
+      "learning_rate": 5.255572193271974e-06,
+      "loss": 0.1677,
+      "step": 5486
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.668251637478478,
+      "learning_rate": 5.254100427819681e-06,
+      "loss": 0.1114,
+      "step": 5487
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.8564126608541183,
+      "learning_rate": 5.252628640293834e-06,
+      "loss": 0.1114,
+      "step": 5488
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.7932579286837058,
+      "learning_rate": 5.251156830822293e-06,
+      "loss": 0.1048,
+      "step": 5489
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9381573882463772,
+      "learning_rate": 5.249684999532906e-06,
+      "loss": 0.1361,
+      "step": 5490
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.833008376701025,
+      "learning_rate": 5.248213146553533e-06,
+      "loss": 0.1176,
+      "step": 5491
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.085573112663426,
+      "learning_rate": 5.2467412720120345e-06,
+      "loss": 0.1482,
+      "step": 5492
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9076362078359177,
+      "learning_rate": 5.245269376036269e-06,
+      "loss": 0.1274,
+      "step": 5493
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.8303109961597201,
+      "learning_rate": 5.2437974587540994e-06,
+      "loss": 0.1137,
+      "step": 5494
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.6465210405881163,
+      "learning_rate": 5.242325520293393e-06,
+      "loss": 0.1069,
+      "step": 5495
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9385391538450683,
+      "learning_rate": 5.240853560782013e-06,
+      "loss": 0.1228,
+      "step": 5496
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.831495159273908,
+      "learning_rate": 5.23938158034783e-06,
+      "loss": 0.1108,
+      "step": 5497
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.8415022217007568,
+      "learning_rate": 5.237909579118713e-06,
+      "loss": 0.1189,
+      "step": 5498
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9894904444522798,
+      "learning_rate": 5.236437557222533e-06,
+      "loss": 0.1205,
+      "step": 5499
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.017582276453585,
+      "learning_rate": 5.234965514787164e-06,
+      "loss": 0.1208,
+      "step": 5500
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.5813129923391347,
+      "learning_rate": 5.233493451940483e-06,
+      "loss": 0.1044,
+      "step": 5501
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9495441774044682,
+      "learning_rate": 5.2320213688103645e-06,
+      "loss": 0.12,
+      "step": 5502
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.680220736710512,
+      "learning_rate": 5.230549265524689e-06,
+      "loss": 0.0992,
+      "step": 5503
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.1525253997865135,
+      "learning_rate": 5.22907714221134e-06,
+      "loss": 0.1161,
+      "step": 5504
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.8645637232857368,
+      "learning_rate": 5.227604998998195e-06,
+      "loss": 0.1031,
+      "step": 5505
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.925527683425804,
+      "learning_rate": 5.226132836013142e-06,
+      "loss": 0.1254,
+      "step": 5506
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.7177162600787759,
+      "learning_rate": 5.224660653384064e-06,
+      "loss": 0.1086,
+      "step": 5507
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.084156999693669,
+      "learning_rate": 5.2231884512388505e-06,
+      "loss": 0.1308,
+      "step": 5508
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9647063306905448,
+      "learning_rate": 5.22171622970539e-06,
+      "loss": 0.1243,
+      "step": 5509
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.782484955165553,
+      "learning_rate": 5.2202439889115755e-06,
+      "loss": 0.1112,
+      "step": 5510
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.439899298606176,
+      "learning_rate": 5.218771728985296e-06,
+      "loss": 0.1257,
+      "step": 5511
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9452047451192829,
+      "learning_rate": 5.2172994500544485e-06,
+      "loss": 0.1273,
+      "step": 5512
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.5380729424952688,
+      "learning_rate": 5.215827152246928e-06,
+      "loss": 0.0768,
+      "step": 5513
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.911422107254962,
+      "learning_rate": 5.2143548356906336e-06,
+      "loss": 0.1196,
+      "step": 5514
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.0369349170237143,
+      "learning_rate": 5.212882500513462e-06,
+      "loss": 0.1254,
+      "step": 5515
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8467128802293964,
+      "learning_rate": 5.211410146843316e-06,
+      "loss": 0.1147,
+      "step": 5516
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.0227455725619263,
+      "learning_rate": 5.209937774808098e-06,
+      "loss": 0.1192,
+      "step": 5517
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8280557628629182,
+      "learning_rate": 5.208465384535711e-06,
+      "loss": 0.1172,
+      "step": 5518
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.943499247667336,
+      "learning_rate": 5.206992976154063e-06,
+      "loss": 0.1209,
+      "step": 5519
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.7525501474821539,
+      "learning_rate": 5.205520549791058e-06,
+      "loss": 0.1215,
+      "step": 5520
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.7005970932037453,
+      "learning_rate": 5.204048105574606e-06,
+      "loss": 0.1144,
+      "step": 5521
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.6496039410322916,
+      "learning_rate": 5.202575643632619e-06,
+      "loss": 0.1031,
+      "step": 5522
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8243207406825588,
+      "learning_rate": 5.201103164093007e-06,
+      "loss": 0.1176,
+      "step": 5523
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.661748958656446,
+      "learning_rate": 5.199630667083682e-06,
+      "loss": 0.0996,
+      "step": 5524
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.6875505161867994,
+      "learning_rate": 5.198158152732564e-06,
+      "loss": 0.1202,
+      "step": 5525
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.07128548085433,
+      "learning_rate": 5.196685621167564e-06,
+      "loss": 0.1284,
+      "step": 5526
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.9612023158244825,
+      "learning_rate": 5.195213072516603e-06,
+      "loss": 0.1034,
+      "step": 5527
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.1650998447366296,
+      "learning_rate": 5.193740506907601e-06,
+      "loss": 0.145,
+      "step": 5528
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8585114812108394,
+      "learning_rate": 5.192267924468476e-06,
+      "loss": 0.1177,
+      "step": 5529
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.5740653915234362,
+      "learning_rate": 5.1907953253271514e-06,
+      "loss": 0.0898,
+      "step": 5530
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.949270206733329,
+      "learning_rate": 5.189322709611552e-06,
+      "loss": 0.1107,
+      "step": 5531
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.7201998366613591,
+      "learning_rate": 5.187850077449604e-06,
+      "loss": 0.1215,
+      "step": 5532
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8774995750030488,
+      "learning_rate": 5.186377428969232e-06,
+      "loss": 0.1236,
+      "step": 5533
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.9386935512614523,
+      "learning_rate": 5.184904764298364e-06,
+      "loss": 0.1205,
+      "step": 5534
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.9732850854733865,
+      "learning_rate": 5.183432083564931e-06,
+      "loss": 0.1245,
+      "step": 5535
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.766023149949048,
+      "learning_rate": 5.181959386896862e-06,
+      "loss": 0.121,
+      "step": 5536
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8855184058671322,
+      "learning_rate": 5.180486674422091e-06,
+      "loss": 0.1235,
+      "step": 5537
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.0236063177403056,
+      "learning_rate": 5.179013946268552e-06,
+      "loss": 0.1154,
+      "step": 5538
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.106226598638917,
+      "learning_rate": 5.177541202564177e-06,
+      "loss": 0.1177,
+      "step": 5539
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.9068869324121809,
+      "learning_rate": 5.176068443436907e-06,
+      "loss": 0.1133,
+      "step": 5540
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8342978640108614,
+      "learning_rate": 5.174595669014675e-06,
+      "loss": 0.113,
+      "step": 5541
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.6821753810546645,
+      "learning_rate": 5.173122879425423e-06,
+      "loss": 0.1109,
+      "step": 5542
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.0889860484782243,
+      "learning_rate": 5.17165007479709e-06,
+      "loss": 0.1212,
+      "step": 5543
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8245016757926622,
+      "learning_rate": 5.170177255257618e-06,
+      "loss": 0.1125,
+      "step": 5544
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.027718152803371,
+      "learning_rate": 5.16870442093495e-06,
+      "loss": 0.1344,
+      "step": 5545
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.1516525393791275,
+      "learning_rate": 5.167231571957032e-06,
+      "loss": 0.1246,
+      "step": 5546
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8753629460150982,
+      "learning_rate": 5.165758708451807e-06,
+      "loss": 0.1282,
+      "step": 5547
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.9724831638143476,
+      "learning_rate": 5.164285830547221e-06,
+      "loss": 0.1308,
+      "step": 5548
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8003082765960625,
+      "learning_rate": 5.162812938371226e-06,
+      "loss": 0.1346,
+      "step": 5549
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.9449248516573328,
+      "learning_rate": 5.161340032051767e-06,
+      "loss": 0.1145,
+      "step": 5550
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.8266982634710451,
+      "learning_rate": 5.159867111716797e-06,
+      "loss": 0.1065,
+      "step": 5551
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.9620052554637786,
+      "learning_rate": 5.158394177494268e-06,
+      "loss": 0.1093,
+      "step": 5552
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.711501362078701,
+      "learning_rate": 5.156921229512131e-06,
+      "loss": 0.1113,
+      "step": 5553
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.828331898998082,
+      "learning_rate": 5.15544826789834e-06,
+      "loss": 0.1163,
+      "step": 5554
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.2031074109439657,
+      "learning_rate": 5.153975292780852e-06,
+      "loss": 0.1285,
+      "step": 5555
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.1428979507569488,
+      "learning_rate": 5.1525023042876245e-06,
+      "loss": 0.1521,
+      "step": 5556
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.7640243699101763,
+      "learning_rate": 5.151029302546612e-06,
+      "loss": 0.0998,
+      "step": 5557
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.7058676701755744,
+      "learning_rate": 5.149556287685775e-06,
+      "loss": 0.1114,
+      "step": 5558
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.749880608814022,
+      "learning_rate": 5.148083259833073e-06,
+      "loss": 0.1247,
+      "step": 5559
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.042925597188746,
+      "learning_rate": 5.146610219116467e-06,
+      "loss": 0.1281,
+      "step": 5560
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.7715506526778562,
+      "learning_rate": 5.145137165663921e-06,
+      "loss": 0.1166,
+      "step": 5561
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.8047194074529922,
+      "learning_rate": 5.143664099603394e-06,
+      "loss": 0.1001,
+      "step": 5562
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.990602952832416,
+      "learning_rate": 5.142191021062854e-06,
+      "loss": 0.1144,
+      "step": 5563
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.7409881419880036,
+      "learning_rate": 5.140717930170267e-06,
+      "loss": 0.1105,
+      "step": 5564
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.721682316889882,
+      "learning_rate": 5.139244827053595e-06,
+      "loss": 0.1088,
+      "step": 5565
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.8850251993259846,
+      "learning_rate": 5.137771711840811e-06,
+      "loss": 0.1366,
+      "step": 5566
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.8311969043490457,
+      "learning_rate": 5.13629858465988e-06,
+      "loss": 0.105,
+      "step": 5567
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.8250626920858655,
+      "learning_rate": 5.134825445638772e-06,
+      "loss": 0.1217,
+      "step": 5568
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.9165437287039446,
+      "learning_rate": 5.133352294905461e-06,
+      "loss": 0.1368,
+      "step": 5569
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.648731143092894,
+      "learning_rate": 5.131879132587915e-06,
+      "loss": 0.1062,
+      "step": 5570
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.305510525623916,
+      "learning_rate": 5.130405958814108e-06,
+      "loss": 0.1332,
+      "step": 5571
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.9158620855468196,
+      "learning_rate": 5.1289327737120145e-06,
+      "loss": 0.1111,
+      "step": 5572
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.6979154737301312,
+      "learning_rate": 5.1274595774096055e-06,
+      "loss": 0.1129,
+      "step": 5573
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.861734382316711,
+      "learning_rate": 5.125986370034862e-06,
+      "loss": 0.1244,
+      "step": 5574
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.7052257697986306,
+      "learning_rate": 5.124513151715759e-06,
+      "loss": 0.0949,
+      "step": 5575
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.9022504903614739,
+      "learning_rate": 5.1230399225802715e-06,
+      "loss": 0.1116,
+      "step": 5576
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.9294389115024777,
+      "learning_rate": 5.12156668275638e-06,
+      "loss": 0.1343,
+      "step": 5577
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.7962084987355424,
+      "learning_rate": 5.120093432372065e-06,
+      "loss": 0.114,
+      "step": 5578
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.9131206564039642,
+      "learning_rate": 5.1186201715553055e-06,
+      "loss": 0.129,
+      "step": 5579
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.9541140275212818,
+      "learning_rate": 5.117146900434082e-06,
+      "loss": 0.1254,
+      "step": 5580
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.041725234566848,
+      "learning_rate": 5.115673619136378e-06,
+      "loss": 0.1228,
+      "step": 5581
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.7287929040843057,
+      "learning_rate": 5.114200327790178e-06,
+      "loss": 0.1088,
+      "step": 5582
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.0976604980732034,
+      "learning_rate": 5.112727026523461e-06,
+      "loss": 0.1357,
+      "step": 5583
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.8011316653008067,
+      "learning_rate": 5.111253715464217e-06,
+      "loss": 0.1092,
+      "step": 5584
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.7089208745209623,
+      "learning_rate": 5.109780394740429e-06,
+      "loss": 0.1195,
+      "step": 5585
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.8850089193004487,
+      "learning_rate": 5.108307064480084e-06,
+      "loss": 0.1204,
+      "step": 5586
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7199709097192781,
+      "learning_rate": 5.10683372481117e-06,
+      "loss": 0.1048,
+      "step": 5587
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.684038680390065,
+      "learning_rate": 5.105360375861673e-06,
+      "loss": 0.1071,
+      "step": 5588
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.884816659817741,
+      "learning_rate": 5.103887017759585e-06,
+      "loss": 0.1186,
+      "step": 5589
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8049175103913289,
+      "learning_rate": 5.1024136506328935e-06,
+      "loss": 0.1114,
+      "step": 5590
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.6393052797805543,
+      "learning_rate": 5.10094027460959e-06,
+      "loss": 0.1139,
+      "step": 5591
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8008624285750674,
+      "learning_rate": 5.099466889817664e-06,
+      "loss": 0.1269,
+      "step": 5592
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.9565035504860313,
+      "learning_rate": 5.097993496385112e-06,
+      "loss": 0.1121,
+      "step": 5593
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.668896305008462,
+      "learning_rate": 5.0965200944399215e-06,
+      "loss": 0.1079,
+      "step": 5594
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7751093056103935,
+      "learning_rate": 5.09504668411009e-06,
+      "loss": 0.1165,
+      "step": 5595
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7781712490982982,
+      "learning_rate": 5.093573265523609e-06,
+      "loss": 0.1118,
+      "step": 5596
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7516702649679112,
+      "learning_rate": 5.0920998388084755e-06,
+      "loss": 0.1124,
+      "step": 5597
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.674652659612386,
+      "learning_rate": 5.090626404092682e-06,
+      "loss": 0.1153,
+      "step": 5598
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.6724900799610873,
+      "learning_rate": 5.0891529615042305e-06,
+      "loss": 0.1014,
+      "step": 5599
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.9206051067678724,
+      "learning_rate": 5.087679511171113e-06,
+      "loss": 0.116,
+      "step": 5600
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8036062937861914,
+      "learning_rate": 5.086206053221328e-06,
+      "loss": 0.1132,
+      "step": 5601
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.2424271425643707,
+      "learning_rate": 5.084732587782878e-06,
+      "loss": 0.1182,
+      "step": 5602
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8681286876223646,
+      "learning_rate": 5.083259114983757e-06,
+      "loss": 0.1114,
+      "step": 5603
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8417970612281476,
+      "learning_rate": 5.081785634951967e-06,
+      "loss": 0.0925,
+      "step": 5604
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8652003201407068,
+      "learning_rate": 5.0803121478155085e-06,
+      "loss": 0.122,
+      "step": 5605
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8224579755735701,
+      "learning_rate": 5.078838653702381e-06,
+      "loss": 0.1248,
+      "step": 5606
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8192520281293496,
+      "learning_rate": 5.077365152740587e-06,
+      "loss": 0.1112,
+      "step": 5607
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.925189457085331,
+      "learning_rate": 5.075891645058129e-06,
+      "loss": 0.1218,
+      "step": 5608
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.714661287225587,
+      "learning_rate": 5.0744181307830095e-06,
+      "loss": 0.0987,
+      "step": 5609
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.9243594989547417,
+      "learning_rate": 5.0729446100432326e-06,
+      "loss": 0.1049,
+      "step": 5610
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.092910354194046,
+      "learning_rate": 5.0714710829668004e-06,
+      "loss": 0.128,
+      "step": 5611
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8465912217009812,
+      "learning_rate": 5.069997549681718e-06,
+      "loss": 0.1322,
+      "step": 5612
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.5969199700198156,
+      "learning_rate": 5.068524010315989e-06,
+      "loss": 0.1087,
+      "step": 5613
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7168261258946884,
+      "learning_rate": 5.067050464997624e-06,
+      "loss": 0.0953,
+      "step": 5614
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.074251717778969,
+      "learning_rate": 5.065576913854623e-06,
+      "loss": 0.1432,
+      "step": 5615
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7528318966637877,
+      "learning_rate": 5.064103357014995e-06,
+      "loss": 0.1128,
+      "step": 5616
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.813077194357431,
+      "learning_rate": 5.062629794606748e-06,
+      "loss": 0.1208,
+      "step": 5617
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8105683627264062,
+      "learning_rate": 5.061156226757887e-06,
+      "loss": 0.1159,
+      "step": 5618
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.0692919953294555,
+      "learning_rate": 5.059682653596422e-06,
+      "loss": 0.1335,
+      "step": 5619
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7553729412505454,
+      "learning_rate": 5.058209075250361e-06,
+      "loss": 0.1147,
+      "step": 5620
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.047342904492541,
+      "learning_rate": 5.056735491847712e-06,
+      "loss": 0.1465,
+      "step": 5621
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.708023377980643,
+      "learning_rate": 5.055261903516485e-06,
+      "loss": 0.1004,
+      "step": 5622
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.7425084013139187,
+      "learning_rate": 5.053788310384691e-06,
+      "loss": 0.102,
+      "step": 5623
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.8104132235779342,
+      "learning_rate": 5.052314712580336e-06,
+      "loss": 0.1255,
+      "step": 5624
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6862834881513558,
+      "learning_rate": 5.050841110231435e-06,
+      "loss": 0.0885,
+      "step": 5625
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.590692119395041,
+      "learning_rate": 5.049367503465998e-06,
+      "loss": 0.0884,
+      "step": 5626
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.92226024012906,
+      "learning_rate": 5.047893892412035e-06,
+      "loss": 0.1331,
+      "step": 5627
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.835177765250274,
+      "learning_rate": 5.046420277197558e-06,
+      "loss": 0.1014,
+      "step": 5628
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.4299238542388544,
+      "learning_rate": 5.04494665795058e-06,
+      "loss": 0.13,
+      "step": 5629
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6232884903013816,
+      "learning_rate": 5.043473034799112e-06,
+      "loss": 0.1099,
+      "step": 5630
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.7168779115045096,
+      "learning_rate": 5.041999407871168e-06,
+      "loss": 0.1009,
+      "step": 5631
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.9260979250250478,
+      "learning_rate": 5.040525777294762e-06,
+      "loss": 0.1328,
+      "step": 5632
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6021866287067223,
+      "learning_rate": 5.039052143197904e-06,
+      "loss": 0.1021,
+      "step": 5633
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6973376066730286,
+      "learning_rate": 5.03757850570861e-06,
+      "loss": 0.1023,
+      "step": 5634
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.8539831428433575,
+      "learning_rate": 5.036104864954895e-06,
+      "loss": 0.1075,
+      "step": 5635
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.7903668627281948,
+      "learning_rate": 5.034631221064771e-06,
+      "loss": 0.1131,
+      "step": 5636
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.861225324590869,
+      "learning_rate": 5.033157574166254e-06,
+      "loss": 0.0891,
+      "step": 5637
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.8802712174841574,
+      "learning_rate": 5.031683924387359e-06,
+      "loss": 0.1227,
+      "step": 5638
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.9809688520309148,
+      "learning_rate": 5.0302102718561e-06,
+      "loss": 0.1139,
+      "step": 5639
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.8584979550167655,
+      "learning_rate": 5.0287366167004925e-06,
+      "loss": 0.1073,
+      "step": 5640
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.296553622552679,
+      "learning_rate": 5.027262959048554e-06,
+      "loss": 0.1104,
+      "step": 5641
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.291486010804984,
+      "learning_rate": 5.0257892990282965e-06,
+      "loss": 0.1532,
+      "step": 5642
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.0904098542175036,
+      "learning_rate": 5.024315636767738e-06,
+      "loss": 0.141,
+      "step": 5643
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.8363973076198659,
+      "learning_rate": 5.0228419723948976e-06,
+      "loss": 0.125,
+      "step": 5644
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.033014575605109,
+      "learning_rate": 5.021368306037786e-06,
+      "loss": 0.1297,
+      "step": 5645
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.0625003428347393,
+      "learning_rate": 5.019894637824423e-06,
+      "loss": 0.1065,
+      "step": 5646
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.9558092050965044,
+      "learning_rate": 5.0184209678828265e-06,
+      "loss": 0.1515,
+      "step": 5647
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6733726194144343,
+      "learning_rate": 5.016947296341009e-06,
+      "loss": 0.0969,
+      "step": 5648
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.01385127510367,
+      "learning_rate": 5.015473623326992e-06,
+      "loss": 0.1108,
+      "step": 5649
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.068321031230667,
+      "learning_rate": 5.01399994896879e-06,
+      "loss": 0.1473,
+      "step": 5650
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.7548321598878058,
+      "learning_rate": 5.01252627339442e-06,
+      "loss": 0.1107,
+      "step": 5651
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.8857505464503226,
+      "learning_rate": 5.0110525967319014e-06,
+      "loss": 0.1213,
+      "step": 5652
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.730373025987402,
+      "learning_rate": 5.00957891910925e-06,
+      "loss": 0.1174,
+      "step": 5653
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6724570413217763,
+      "learning_rate": 5.008105240654484e-06,
+      "loss": 0.0997,
+      "step": 5654
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6378072926054223,
+      "learning_rate": 5.006631561495619e-06,
+      "loss": 0.0948,
+      "step": 5655
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.7758720818611609,
+      "learning_rate": 5.005157881760676e-06,
+      "loss": 0.0964,
+      "step": 5656
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.7962637614290708,
+      "learning_rate": 5.003684201577671e-06,
+      "loss": 0.1186,
+      "step": 5657
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.2655690768766648,
+      "learning_rate": 5.00221052107462e-06,
+      "loss": 0.1393,
+      "step": 5658
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.7777998570584177,
+      "learning_rate": 5.0007368403795445e-06,
+      "loss": 0.1143,
+      "step": 5659
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.7558897600856012,
+      "learning_rate": 4.999263159620457e-06,
+      "loss": 0.1045,
+      "step": 5660
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.031210830016424,
+      "learning_rate": 4.997789478925381e-06,
+      "loss": 0.1235,
+      "step": 5661
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.934091233045082,
+      "learning_rate": 4.996315798422331e-06,
+      "loss": 0.1295,
+      "step": 5662
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.7280359998980814,
+      "learning_rate": 4.9948421182393255e-06,
+      "loss": 0.1068,
+      "step": 5663
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8785839450477815,
+      "learning_rate": 4.993368438504381e-06,
+      "loss": 0.1148,
+      "step": 5664
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.9720584283676688,
+      "learning_rate": 4.991894759345519e-06,
+      "loss": 0.1236,
+      "step": 5665
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.0121207616333905,
+      "learning_rate": 4.990421080890751e-06,
+      "loss": 0.1233,
+      "step": 5666
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.1782927040738307,
+      "learning_rate": 4.9889474032681e-06,
+      "loss": 0.138,
+      "step": 5667
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8129014974143547,
+      "learning_rate": 4.987473726605581e-06,
+      "loss": 0.1096,
+      "step": 5668
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.803896452146066,
+      "learning_rate": 4.986000051031212e-06,
+      "loss": 0.118,
+      "step": 5669
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.067597732413137,
+      "learning_rate": 4.98452637667301e-06,
+      "loss": 0.152,
+      "step": 5670
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.9111122051385818,
+      "learning_rate": 4.983052703658993e-06,
+      "loss": 0.1319,
+      "step": 5671
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.9427129436199546,
+      "learning_rate": 4.981579032117175e-06,
+      "loss": 0.1187,
+      "step": 5672
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.658142341628866,
+      "learning_rate": 4.980105362175579e-06,
+      "loss": 0.1097,
+      "step": 5673
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.6511248088283692,
+      "learning_rate": 4.978631693962216e-06,
+      "loss": 0.0994,
+      "step": 5674
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.6356809147118867,
+      "learning_rate": 4.977158027605105e-06,
+      "loss": 0.0946,
+      "step": 5675
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.9975584833690718,
+      "learning_rate": 4.975684363232263e-06,
+      "loss": 0.1248,
+      "step": 5676
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8369968635811522,
+      "learning_rate": 4.974210700971706e-06,
+      "loss": 0.1246,
+      "step": 5677
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8593771112850501,
+      "learning_rate": 4.972737040951448e-06,
+      "loss": 0.1309,
+      "step": 5678
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.747340510580516,
+      "learning_rate": 4.971263383299509e-06,
+      "loss": 0.0978,
+      "step": 5679
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.6215517704741043,
+      "learning_rate": 4.969789728143902e-06,
+      "loss": 0.0942,
+      "step": 5680
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.75672850115425,
+      "learning_rate": 4.968316075612643e-06,
+      "loss": 0.102,
+      "step": 5681
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.9669161735713907,
+      "learning_rate": 4.966842425833748e-06,
+      "loss": 0.1299,
+      "step": 5682
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.901679111324784,
+      "learning_rate": 4.965368778935231e-06,
+      "loss": 0.1469,
+      "step": 5683
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.7768482313776948,
+      "learning_rate": 4.963895135045106e-06,
+      "loss": 0.1036,
+      "step": 5684
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.6277421141176784,
+      "learning_rate": 4.9624214942913916e-06,
+      "loss": 0.1008,
+      "step": 5685
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8337856441661124,
+      "learning_rate": 4.960947856802097e-06,
+      "loss": 0.1192,
+      "step": 5686
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.780159456757161,
+      "learning_rate": 4.959474222705241e-06,
+      "loss": 0.1044,
+      "step": 5687
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.6007562662347046,
+      "learning_rate": 4.958000592128834e-06,
+      "loss": 0.0974,
+      "step": 5688
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.6694203916932255,
+      "learning_rate": 4.956526965200891e-06,
+      "loss": 0.1044,
+      "step": 5689
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.653217059591154,
+      "learning_rate": 4.9550533420494216e-06,
+      "loss": 0.0854,
+      "step": 5690
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8819241939426945,
+      "learning_rate": 4.953579722802444e-06,
+      "loss": 0.1137,
+      "step": 5691
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8651468200409407,
+      "learning_rate": 4.952106107587967e-06,
+      "loss": 0.1148,
+      "step": 5692
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.7526053626772513,
+      "learning_rate": 4.950632496534004e-06,
+      "loss": 0.0919,
+      "step": 5693
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8310157198213042,
+      "learning_rate": 4.949158889768566e-06,
+      "loss": 0.106,
+      "step": 5694
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.1219930115365644,
+      "learning_rate": 4.9476852874196665e-06,
+      "loss": 0.1125,
+      "step": 5695
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.837587959049806,
+      "learning_rate": 4.9462116896153115e-06,
+      "loss": 0.0934,
+      "step": 5696
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9998489226606917,
+      "learning_rate": 4.9447380964835165e-06,
+      "loss": 0.1238,
+      "step": 5697
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.3671861825266167,
+      "learning_rate": 4.94326450815229e-06,
+      "loss": 0.1389,
+      "step": 5698
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.1497990502985274,
+      "learning_rate": 4.9417909247496415e-06,
+      "loss": 0.1142,
+      "step": 5699
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.033372234976445,
+      "learning_rate": 4.94031734640358e-06,
+      "loss": 0.1321,
+      "step": 5700
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.7093406378908598,
+      "learning_rate": 4.938843773242115e-06,
+      "loss": 0.101,
+      "step": 5701
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9523284450374463,
+      "learning_rate": 4.9373702053932534e-06,
+      "loss": 0.1178,
+      "step": 5702
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.0611238166998405,
+      "learning_rate": 4.935896642985006e-06,
+      "loss": 0.1358,
+      "step": 5703
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.8091374609670188,
+      "learning_rate": 4.934423086145379e-06,
+      "loss": 0.1021,
+      "step": 5704
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.8315061251008664,
+      "learning_rate": 4.932949535002379e-06,
+      "loss": 0.1307,
+      "step": 5705
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.5752788518998808,
+      "learning_rate": 4.9314759896840115e-06,
+      "loss": 0.0932,
+      "step": 5706
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.8289836160182324,
+      "learning_rate": 4.930002450318282e-06,
+      "loss": 0.124,
+      "step": 5707
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9310853166360196,
+      "learning_rate": 4.928528917033201e-06,
+      "loss": 0.1062,
+      "step": 5708
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.818647996449721,
+      "learning_rate": 4.927055389956768e-06,
+      "loss": 0.1291,
+      "step": 5709
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9569161687472154,
+      "learning_rate": 4.925581869216991e-06,
+      "loss": 0.1173,
+      "step": 5710
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.667077688058881,
+      "learning_rate": 4.9241083549418714e-06,
+      "loss": 0.0874,
+      "step": 5711
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.762875435620747,
+      "learning_rate": 4.922634847259415e-06,
+      "loss": 0.1176,
+      "step": 5712
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.903497213373195,
+      "learning_rate": 4.92116134629762e-06,
+      "loss": 0.1141,
+      "step": 5713
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.8133803376171806,
+      "learning_rate": 4.919687852184493e-06,
+      "loss": 0.1169,
+      "step": 5714
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.8570255570873222,
+      "learning_rate": 4.918214365048034e-06,
+      "loss": 0.1258,
+      "step": 5715
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9009246373204074,
+      "learning_rate": 4.916740885016244e-06,
+      "loss": 0.1163,
+      "step": 5716
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.6720574676633757,
+      "learning_rate": 4.9152674122171235e-06,
+      "loss": 0.0971,
+      "step": 5717
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.8093878511718644,
+      "learning_rate": 4.9137939467786724e-06,
+      "loss": 0.1169,
+      "step": 5718
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.7209826144512443,
+      "learning_rate": 4.912320488828887e-06,
+      "loss": 0.1037,
+      "step": 5719
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.609029004903627,
+      "learning_rate": 4.910847038495771e-06,
+      "loss": 0.1028,
+      "step": 5720
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9421893024123071,
+      "learning_rate": 4.909373595907317e-06,
+      "loss": 0.1122,
+      "step": 5721
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.7453372901433868,
+      "learning_rate": 4.907900161191527e-06,
+      "loss": 0.1199,
+      "step": 5722
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.7006496328797378,
+      "learning_rate": 4.9064267344763924e-06,
+      "loss": 0.1082,
+      "step": 5723
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.7374672099068977,
+      "learning_rate": 4.904953315889912e-06,
+      "loss": 0.1127,
+      "step": 5724
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.828307663872893,
+      "learning_rate": 4.9034799055600785e-06,
+      "loss": 0.117,
+      "step": 5725
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.7282872858421505,
+      "learning_rate": 4.9020065036148885e-06,
+      "loss": 0.1232,
+      "step": 5726
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.8154254284960079,
+      "learning_rate": 4.900533110182335e-06,
+      "loss": 0.1162,
+      "step": 5727
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.7500849027564973,
+      "learning_rate": 4.899059725390412e-06,
+      "loss": 0.1127,
+      "step": 5728
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9075675736820878,
+      "learning_rate": 4.897586349367107e-06,
+      "loss": 0.0996,
+      "step": 5729
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9859964943920791,
+      "learning_rate": 4.896112982240417e-06,
+      "loss": 0.133,
+      "step": 5730
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.684623806492088,
+      "learning_rate": 4.894639624138327e-06,
+      "loss": 0.112,
+      "step": 5731
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.7481052846918421,
+      "learning_rate": 4.893166275188831e-06,
+      "loss": 0.0996,
+      "step": 5732
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8416961533704146,
+      "learning_rate": 4.891692935519917e-06,
+      "loss": 0.12,
+      "step": 5733
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7881593786151584,
+      "learning_rate": 4.8902196052595725e-06,
+      "loss": 0.1069,
+      "step": 5734
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.9102205765429747,
+      "learning_rate": 4.888746284535784e-06,
+      "loss": 0.1134,
+      "step": 5735
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6806145859210895,
+      "learning_rate": 4.88727297347654e-06,
+      "loss": 0.1062,
+      "step": 5736
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.5584932563044123,
+      "learning_rate": 4.885799672209823e-06,
+      "loss": 0.0968,
+      "step": 5737
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7434458052038524,
+      "learning_rate": 4.8843263808636225e-06,
+      "loss": 0.122,
+      "step": 5738
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7821129088055114,
+      "learning_rate": 4.8828530995659185e-06,
+      "loss": 0.109,
+      "step": 5739
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.9620004057843237,
+      "learning_rate": 4.881379828444696e-06,
+      "loss": 0.1327,
+      "step": 5740
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7635306355476654,
+      "learning_rate": 4.8799065676279354e-06,
+      "loss": 0.1267,
+      "step": 5741
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7682174779191964,
+      "learning_rate": 4.878433317243621e-06,
+      "loss": 0.1241,
+      "step": 5742
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.713173004251655,
+      "learning_rate": 4.8769600774197285e-06,
+      "loss": 0.1133,
+      "step": 5743
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.753934743386582,
+      "learning_rate": 4.875486848284243e-06,
+      "loss": 0.0962,
+      "step": 5744
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.694013358194294,
+      "learning_rate": 4.874013629965138e-06,
+      "loss": 0.0966,
+      "step": 5745
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6081367182112725,
+      "learning_rate": 4.872540422590395e-06,
+      "loss": 0.1016,
+      "step": 5746
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6497513420626597,
+      "learning_rate": 4.871067226287988e-06,
+      "loss": 0.1063,
+      "step": 5747
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.9007595261958676,
+      "learning_rate": 4.869594041185895e-06,
+      "loss": 0.1314,
+      "step": 5748
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8466426882191314,
+      "learning_rate": 4.868120867412085e-06,
+      "loss": 0.1069,
+      "step": 5749
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8612378106663066,
+      "learning_rate": 4.866647705094541e-06,
+      "loss": 0.1216,
+      "step": 5750
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8405144033828842,
+      "learning_rate": 4.865174554361228e-06,
+      "loss": 0.1044,
+      "step": 5751
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6995991271554405,
+      "learning_rate": 4.863701415340122e-06,
+      "loss": 0.0995,
+      "step": 5752
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7552008637991825,
+      "learning_rate": 4.862228288159191e-06,
+      "loss": 0.1108,
+      "step": 5753
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6071088609528863,
+      "learning_rate": 4.8607551729464066e-06,
+      "loss": 0.1033,
+      "step": 5754
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8477585095269573,
+      "learning_rate": 4.859282069829735e-06,
+      "loss": 0.1084,
+      "step": 5755
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8963890238903187,
+      "learning_rate": 4.8578089789371476e-06,
+      "loss": 0.127,
+      "step": 5756
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6656099080347362,
+      "learning_rate": 4.856335900396607e-06,
+      "loss": 0.109,
+      "step": 5757
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7733985607823848,
+      "learning_rate": 4.854862834336082e-06,
+      "loss": 0.1092,
+      "step": 5758
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8727722672515896,
+      "learning_rate": 4.853389780883535e-06,
+      "loss": 0.1228,
+      "step": 5759
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8176649994104128,
+      "learning_rate": 4.85191674016693e-06,
+      "loss": 0.133,
+      "step": 5760
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.798999066859521,
+      "learning_rate": 4.850443712314226e-06,
+      "loss": 0.1235,
+      "step": 5761
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8658596334148467,
+      "learning_rate": 4.84897069745339e-06,
+      "loss": 0.1091,
+      "step": 5762
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.634225858589599,
+      "learning_rate": 4.847497695712378e-06,
+      "loss": 0.1003,
+      "step": 5763
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.5401341581020818,
+      "learning_rate": 4.846024707219149e-06,
+      "loss": 0.1069,
+      "step": 5764
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6183584850698047,
+      "learning_rate": 4.844551732101662e-06,
+      "loss": 0.0941,
+      "step": 5765
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6543020518878626,
+      "learning_rate": 4.8430787704878725e-06,
+      "loss": 0.1039,
+      "step": 5766
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.057836347452287,
+      "learning_rate": 4.841605822505734e-06,
+      "loss": 0.1348,
+      "step": 5767
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7612813463632422,
+      "learning_rate": 4.840132888283205e-06,
+      "loss": 0.1073,
+      "step": 5768
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7488480715678698,
+      "learning_rate": 4.838659967948234e-06,
+      "loss": 0.1007,
+      "step": 5769
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.949305034389504,
+      "learning_rate": 4.837187061628777e-06,
+      "loss": 0.1195,
+      "step": 5770
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8768754169423492,
+      "learning_rate": 4.835714169452781e-06,
+      "loss": 0.0951,
+      "step": 5771
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.6872155994730111,
+      "learning_rate": 4.8342412915481965e-06,
+      "loss": 0.1053,
+      "step": 5772
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.757100945986478,
+      "learning_rate": 4.832768428042969e-06,
+      "loss": 0.1136,
+      "step": 5773
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8266618111570159,
+      "learning_rate": 4.83129557906505e-06,
+      "loss": 0.1006,
+      "step": 5774
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8800910902856227,
+      "learning_rate": 4.829822744742383e-06,
+      "loss": 0.1024,
+      "step": 5775
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7789115076420592,
+      "learning_rate": 4.828349925202912e-06,
+      "loss": 0.1152,
+      "step": 5776
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.64202071672074,
+      "learning_rate": 4.826877120574579e-06,
+      "loss": 0.1136,
+      "step": 5777
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.0242727417068327,
+      "learning_rate": 4.825404330985328e-06,
+      "loss": 0.1367,
+      "step": 5778
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7802500752376826,
+      "learning_rate": 4.823931556563094e-06,
+      "loss": 0.1041,
+      "step": 5779
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.0323776150631527,
+      "learning_rate": 4.822458797435824e-06,
+      "loss": 0.1312,
+      "step": 5780
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8693817759850502,
+      "learning_rate": 4.8209860537314504e-06,
+      "loss": 0.1168,
+      "step": 5781
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.9952595945622615,
+      "learning_rate": 4.819513325577911e-06,
+      "loss": 0.119,
+      "step": 5782
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7791461906356476,
+      "learning_rate": 4.818040613103139e-06,
+      "loss": 0.1128,
+      "step": 5783
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.9572449691796707,
+      "learning_rate": 4.816567916435072e-06,
+      "loss": 0.1208,
+      "step": 5784
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.0010546884517972,
+      "learning_rate": 4.815095235701637e-06,
+      "loss": 0.1228,
+      "step": 5785
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.9309152984072147,
+      "learning_rate": 4.81362257103077e-06,
+      "loss": 0.1345,
+      "step": 5786
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8454420843006936,
+      "learning_rate": 4.8121499225503974e-06,
+      "loss": 0.0998,
+      "step": 5787
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.778736418162696,
+      "learning_rate": 4.810677290388449e-06,
+      "loss": 0.1199,
+      "step": 5788
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.6879138017801798,
+      "learning_rate": 4.80920467467285e-06,
+      "loss": 0.0982,
+      "step": 5789
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7918692659552269,
+      "learning_rate": 4.807732075531527e-06,
+      "loss": 0.1271,
+      "step": 5790
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7341349136829614,
+      "learning_rate": 4.8062594930924015e-06,
+      "loss": 0.1221,
+      "step": 5791
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.60675201915511,
+      "learning_rate": 4.804786927483399e-06,
+      "loss": 0.1066,
+      "step": 5792
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7868072677455564,
+      "learning_rate": 4.803314378832437e-06,
+      "loss": 0.1107,
+      "step": 5793
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.9284059809003662,
+      "learning_rate": 4.801841847267439e-06,
+      "loss": 0.1371,
+      "step": 5794
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7601085819950688,
+      "learning_rate": 4.800369332916319e-06,
+      "loss": 0.1083,
+      "step": 5795
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8269185248881448,
+      "learning_rate": 4.7988968359069965e-06,
+      "loss": 0.1242,
+      "step": 5796
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.6697723497919919,
+      "learning_rate": 4.797424356367383e-06,
+      "loss": 0.1108,
+      "step": 5797
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8184438258718252,
+      "learning_rate": 4.795951894425396e-06,
+      "loss": 0.1209,
+      "step": 5798
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8167479408872471,
+      "learning_rate": 4.794479450208944e-06,
+      "loss": 0.0974,
+      "step": 5799
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7001557135695522,
+      "learning_rate": 4.793007023845939e-06,
+      "loss": 0.094,
+      "step": 5800
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.9240326244901658,
+      "learning_rate": 4.79153461546429e-06,
+      "loss": 0.1194,
+      "step": 5801
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.900376832500559,
+      "learning_rate": 4.790062225191902e-06,
+      "loss": 0.1189,
+      "step": 5802
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8746660758595346,
+      "learning_rate": 4.788589853156685e-06,
+      "loss": 0.1318,
+      "step": 5803
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.5530386367504285,
+      "learning_rate": 4.787117499486539e-06,
+      "loss": 0.0888,
+      "step": 5804
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.9014417800591763,
+      "learning_rate": 4.785645164309368e-06,
+      "loss": 0.1265,
+      "step": 5805
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8352995595755826,
+      "learning_rate": 4.784172847753073e-06,
+      "loss": 0.1172,
+      "step": 5806
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.7906503847992112,
+      "learning_rate": 4.782700549945554e-06,
+      "loss": 0.1043,
+      "step": 5807
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.6622055077269564,
+      "learning_rate": 4.781228271014704e-06,
+      "loss": 0.0897,
+      "step": 5808
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.9378819097600386,
+      "learning_rate": 4.779756011088427e-06,
+      "loss": 0.123,
+      "step": 5809
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8986105258992039,
+      "learning_rate": 4.778283770294611e-06,
+      "loss": 0.1349,
+      "step": 5810
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.9373484793514644,
+      "learning_rate": 4.776811548761151e-06,
+      "loss": 0.12,
+      "step": 5811
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8488450062922406,
+      "learning_rate": 4.775339346615937e-06,
+      "loss": 0.1196,
+      "step": 5812
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.9135308776486715,
+      "learning_rate": 4.773867163986861e-06,
+      "loss": 0.1189,
+      "step": 5813
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.653273620121883,
+      "learning_rate": 4.772395001001805e-06,
+      "loss": 0.1092,
+      "step": 5814
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8324525133036889,
+      "learning_rate": 4.770922857788662e-06,
+      "loss": 0.106,
+      "step": 5815
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8954599507751637,
+      "learning_rate": 4.769450734475311e-06,
+      "loss": 0.1129,
+      "step": 5816
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.885755000372393,
+      "learning_rate": 4.767978631189637e-06,
+      "loss": 0.1081,
+      "step": 5817
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.272605380914167,
+      "learning_rate": 4.766506548059519e-06,
+      "loss": 0.1305,
+      "step": 5818
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.687725613498644,
+      "learning_rate": 4.765034485212838e-06,
+      "loss": 0.0974,
+      "step": 5819
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.1352277805682323,
+      "learning_rate": 4.763562442777468e-06,
+      "loss": 0.1372,
+      "step": 5820
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.9097599706014994,
+      "learning_rate": 4.762090420881289e-06,
+      "loss": 0.1128,
+      "step": 5821
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.5309999626312183,
+      "learning_rate": 4.760618419652171e-06,
+      "loss": 0.096,
+      "step": 5822
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.746586246681869,
+      "learning_rate": 4.759146439217988e-06,
+      "loss": 0.0986,
+      "step": 5823
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.5348836153514618,
+      "learning_rate": 4.757674479706608e-06,
+      "loss": 0.1065,
+      "step": 5824
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8055086328004646,
+      "learning_rate": 4.756202541245901e-06,
+      "loss": 0.1004,
+      "step": 5825
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.7819694254964096,
+      "learning_rate": 4.7547306239637314e-06,
+      "loss": 0.1217,
+      "step": 5826
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.213888238903891,
+      "learning_rate": 4.753258727987967e-06,
+      "loss": 0.142,
+      "step": 5827
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.9114142111373016,
+      "learning_rate": 4.751786853446467e-06,
+      "loss": 0.1298,
+      "step": 5828
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.1320990677164704,
+      "learning_rate": 4.750315000467096e-06,
+      "loss": 0.1249,
+      "step": 5829
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8533968633744407,
+      "learning_rate": 4.74884316917771e-06,
+      "loss": 0.1153,
+      "step": 5830
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.88040365841761,
+      "learning_rate": 4.747371359706167e-06,
+      "loss": 0.1237,
+      "step": 5831
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.6644013164049285,
+      "learning_rate": 4.74589957218032e-06,
+      "loss": 0.0834,
+      "step": 5832
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.763009551105408,
+      "learning_rate": 4.7444278067280275e-06,
+      "loss": 0.1009,
+      "step": 5833
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8901838968526794,
+      "learning_rate": 4.742956063477136e-06,
+      "loss": 0.1086,
+      "step": 5834
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8874826789291914,
+      "learning_rate": 4.741484342555498e-06,
+      "loss": 0.1203,
+      "step": 5835
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.2014894923974317,
+      "learning_rate": 4.7400126440909595e-06,
+      "loss": 0.1168,
+      "step": 5836
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.0727115741286912,
+      "learning_rate": 4.738540968211367e-06,
+      "loss": 0.1119,
+      "step": 5837
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.7702872967757892,
+      "learning_rate": 4.737069315044562e-06,
+      "loss": 0.1127,
+      "step": 5838
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.974011891411579,
+      "learning_rate": 4.735597684718389e-06,
+      "loss": 0.1352,
+      "step": 5839
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.7997201859376395,
+      "learning_rate": 4.734126077360685e-06,
+      "loss": 0.1156,
+      "step": 5840
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.6977115954645179,
+      "learning_rate": 4.7326544930992905e-06,
+      "loss": 0.1208,
+      "step": 5841
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.6120726414039799,
+      "learning_rate": 4.7311829320620384e-06,
+      "loss": 0.0975,
+      "step": 5842
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9216844667439537,
+      "learning_rate": 4.729711394376765e-06,
+      "loss": 0.1281,
+      "step": 5843
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.5732382219691132,
+      "learning_rate": 4.728239880171298e-06,
+      "loss": 0.0946,
+      "step": 5844
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.5587617194953125,
+      "learning_rate": 4.726768389573471e-06,
+      "loss": 0.1082,
+      "step": 5845
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.807657434962819,
+      "learning_rate": 4.725296922711109e-06,
+      "loss": 0.101,
+      "step": 5846
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7152001206403256,
+      "learning_rate": 4.723825479712039e-06,
+      "loss": 0.1264,
+      "step": 5847
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9446942358779278,
+      "learning_rate": 4.722354060704083e-06,
+      "loss": 0.1295,
+      "step": 5848
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7303020307657297,
+      "learning_rate": 4.720882665815064e-06,
+      "loss": 0.1137,
+      "step": 5849
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7861039588568814,
+      "learning_rate": 4.719411295172797e-06,
+      "loss": 0.0962,
+      "step": 5850
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.764737303748098,
+      "learning_rate": 4.717939948905106e-06,
+      "loss": 0.1076,
+      "step": 5851
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9657570521919652,
+      "learning_rate": 4.7164686271398005e-06,
+      "loss": 0.1249,
+      "step": 5852
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.8603412528297958,
+      "learning_rate": 4.714997330004696e-06,
+      "loss": 0.126,
+      "step": 5853
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9178296228055158,
+      "learning_rate": 4.713526057627601e-06,
+      "loss": 0.1032,
+      "step": 5854
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7848120414476905,
+      "learning_rate": 4.712054810136327e-06,
+      "loss": 0.1088,
+      "step": 5855
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.227612031919105,
+      "learning_rate": 4.710583587658675e-06,
+      "loss": 0.1602,
+      "step": 5856
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9106229480430856,
+      "learning_rate": 4.709112390322456e-06,
+      "loss": 0.1187,
+      "step": 5857
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7438963947927832,
+      "learning_rate": 4.707641218255468e-06,
+      "loss": 0.1088,
+      "step": 5858
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9839585538762703,
+      "learning_rate": 4.706170071585513e-06,
+      "loss": 0.1248,
+      "step": 5859
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.6212696133177271,
+      "learning_rate": 4.704698950440386e-06,
+      "loss": 0.0981,
+      "step": 5860
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.8134499348071371,
+      "learning_rate": 4.703227854947884e-06,
+      "loss": 0.1167,
+      "step": 5861
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7978405292640642,
+      "learning_rate": 4.701756785235798e-06,
+      "loss": 0.0957,
+      "step": 5862
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.6951695522737673,
+      "learning_rate": 4.700285741431924e-06,
+      "loss": 0.1036,
+      "step": 5863
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.8436728595925593,
+      "learning_rate": 4.698814723664046e-06,
+      "loss": 0.1159,
+      "step": 5864
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.5854219462219072,
+      "learning_rate": 4.697343732059953e-06,
+      "loss": 0.0985,
+      "step": 5865
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.8573442487280667,
+      "learning_rate": 4.695872766747427e-06,
+      "loss": 0.0983,
+      "step": 5866
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7640963539350536,
+      "learning_rate": 4.694401827854252e-06,
+      "loss": 0.1161,
+      "step": 5867
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.8738824373646115,
+      "learning_rate": 4.6929309155082045e-06,
+      "loss": 0.1245,
+      "step": 5868
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.8557139389586945,
+      "learning_rate": 4.691460029837066e-06,
+      "loss": 0.1277,
+      "step": 5869
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.5719878205675166,
+      "learning_rate": 4.689989170968609e-06,
+      "loss": 0.1064,
+      "step": 5870
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9914432579846897,
+      "learning_rate": 4.688518339030607e-06,
+      "loss": 0.1344,
+      "step": 5871
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.5228425687207567,
+      "learning_rate": 4.687047534150829e-06,
+      "loss": 0.0802,
+      "step": 5872
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.8003761930528475,
+      "learning_rate": 4.685576756457044e-06,
+      "loss": 0.1148,
+      "step": 5873
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.650483536474264,
+      "learning_rate": 4.684106006077015e-06,
+      "loss": 0.1021,
+      "step": 5874
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.6761870246654584,
+      "learning_rate": 4.682635283138511e-06,
+      "loss": 0.1089,
+      "step": 5875
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9133238556564902,
+      "learning_rate": 4.681164587769287e-06,
+      "loss": 0.1065,
+      "step": 5876
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.6507591831526423,
+      "learning_rate": 4.679693920097105e-06,
+      "loss": 0.1014,
+      "step": 5877
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.8866229060608464,
+      "learning_rate": 4.678223280249718e-06,
+      "loss": 0.1058,
+      "step": 5878
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9760817917867326,
+      "learning_rate": 4.676752668354884e-06,
+      "loss": 0.1156,
+      "step": 5879
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7839330291960105,
+      "learning_rate": 4.675282084540348e-06,
+      "loss": 0.1127,
+      "step": 5880
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.4124860362722247,
+      "learning_rate": 4.673811528933865e-06,
+      "loss": 0.0838,
+      "step": 5881
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.9095096221208152,
+      "learning_rate": 4.672341001663178e-06,
+      "loss": 0.1168,
+      "step": 5882
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.6278396460492184,
+      "learning_rate": 4.670870502856031e-06,
+      "loss": 0.1091,
+      "step": 5883
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.6741461912256008,
+      "learning_rate": 4.669400032640165e-06,
+      "loss": 0.1135,
+      "step": 5884
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.482529975352485,
+      "learning_rate": 4.6679295911433215e-06,
+      "loss": 0.0866,
+      "step": 5885
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7964494809704494,
+      "learning_rate": 4.666459178493232e-06,
+      "loss": 0.1097,
+      "step": 5886
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.6443770655683032,
+      "learning_rate": 4.664988794817637e-06,
+      "loss": 0.1117,
+      "step": 5887
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.989703416048947,
+      "learning_rate": 4.66351844024426e-06,
+      "loss": 0.1316,
+      "step": 5888
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8195507552922463,
+      "learning_rate": 4.662048114900837e-06,
+      "loss": 0.1068,
+      "step": 5889
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8281156687574072,
+      "learning_rate": 4.66057781891509e-06,
+      "loss": 0.1125,
+      "step": 5890
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8648349936389876,
+      "learning_rate": 4.659107552414744e-06,
+      "loss": 0.1343,
+      "step": 5891
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8575779782882569,
+      "learning_rate": 4.657637315527519e-06,
+      "loss": 0.1106,
+      "step": 5892
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8685792018228158,
+      "learning_rate": 4.656167108381135e-06,
+      "loss": 0.1264,
+      "step": 5893
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.9297831148882634,
+      "learning_rate": 4.65469693110331e-06,
+      "loss": 0.121,
+      "step": 5894
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.921258402958159,
+      "learning_rate": 4.653226783821753e-06,
+      "loss": 0.1128,
+      "step": 5895
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.5968018096518126,
+      "learning_rate": 4.651756666664178e-06,
+      "loss": 0.1043,
+      "step": 5896
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7831648443247416,
+      "learning_rate": 4.650286579758291e-06,
+      "loss": 0.112,
+      "step": 5897
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8452065215787334,
+      "learning_rate": 4.6488165232318e-06,
+      "loss": 0.1046,
+      "step": 5898
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8055015068544278,
+      "learning_rate": 4.647346497212406e-06,
+      "loss": 0.1177,
+      "step": 5899
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8288518671024598,
+      "learning_rate": 4.6458765018278104e-06,
+      "loss": 0.1025,
+      "step": 5900
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7897386813041352,
+      "learning_rate": 4.64440653720571e-06,
+      "loss": 0.1027,
+      "step": 5901
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7526472122341537,
+      "learning_rate": 4.6429366034738005e-06,
+      "loss": 0.1102,
+      "step": 5902
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7037760862198994,
+      "learning_rate": 4.641466700759772e-06,
+      "loss": 0.1128,
+      "step": 5903
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.6130010370747099,
+      "learning_rate": 4.6399968291913175e-06,
+      "loss": 0.0972,
+      "step": 5904
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.5708719910355782,
+      "learning_rate": 4.638526988896122e-06,
+      "loss": 0.0998,
+      "step": 5905
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8686336439053501,
+      "learning_rate": 4.6370571800018695e-06,
+      "loss": 0.124,
+      "step": 5906
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.733026578232837,
+      "learning_rate": 4.635587402636241e-06,
+      "loss": 0.0974,
+      "step": 5907
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.9655421975424643,
+      "learning_rate": 4.634117656926917e-06,
+      "loss": 0.1256,
+      "step": 5908
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.081120244565284,
+      "learning_rate": 4.6326479430015715e-06,
+      "loss": 0.1312,
+      "step": 5909
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.0682548297454897,
+      "learning_rate": 4.631178260987879e-06,
+      "loss": 0.126,
+      "step": 5910
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7631987164388545,
+      "learning_rate": 4.629708611013509e-06,
+      "loss": 0.1147,
+      "step": 5911
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7951499068402208,
+      "learning_rate": 4.628238993206131e-06,
+      "loss": 0.1108,
+      "step": 5912
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.737180170388322,
+      "learning_rate": 4.6267694076934066e-06,
+      "loss": 0.0889,
+      "step": 5913
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.952864510253775,
+      "learning_rate": 4.625299854603e-06,
+      "loss": 0.1317,
+      "step": 5914
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.0077823719376795,
+      "learning_rate": 4.623830334062569e-06,
+      "loss": 0.1296,
+      "step": 5915
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8460052896720212,
+      "learning_rate": 4.622360846199772e-06,
+      "loss": 0.0962,
+      "step": 5916
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.7313392476325535,
+      "learning_rate": 4.620891391142262e-06,
+      "loss": 0.1101,
+      "step": 5917
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8173263523567296,
+      "learning_rate": 4.619421969017688e-06,
+      "loss": 0.1241,
+      "step": 5918
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.5820178223866355,
+      "learning_rate": 4.617952579953699e-06,
+      "loss": 0.0992,
+      "step": 5919
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8015485617719473,
+      "learning_rate": 4.6164832240779405e-06,
+      "loss": 0.1206,
+      "step": 5920
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.857087754955123,
+      "learning_rate": 4.615013901518052e-06,
+      "loss": 0.126,
+      "step": 5921
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.1065851785426304,
+      "learning_rate": 4.613544612401677e-06,
+      "loss": 0.1333,
+      "step": 5922
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.6800644092348458,
+      "learning_rate": 4.612075356856447e-06,
+      "loss": 0.1098,
+      "step": 5923
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.6796151338359744,
+      "learning_rate": 4.61060613501e-06,
+      "loss": 0.0964,
+      "step": 5924
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8169100907761206,
+      "learning_rate": 4.6091369469899634e-06,
+      "loss": 0.1202,
+      "step": 5925
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.5137748084633729,
+      "learning_rate": 4.6076677929239656e-06,
+      "loss": 0.0873,
+      "step": 5926
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.764877837925718,
+      "learning_rate": 4.606198672939628e-06,
+      "loss": 0.1029,
+      "step": 5927
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.7179899907351193,
+      "learning_rate": 4.6047295871645785e-06,
+      "loss": 0.1062,
+      "step": 5928
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.789792340906551,
+      "learning_rate": 4.603260535726432e-06,
+      "loss": 0.1103,
+      "step": 5929
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.6541793464139767,
+      "learning_rate": 4.6017915187528036e-06,
+      "loss": 0.0951,
+      "step": 5930
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.725131736538696,
+      "learning_rate": 4.6003225363713065e-06,
+      "loss": 0.0984,
+      "step": 5931
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.101253603030808,
+      "learning_rate": 4.598853588709552e-06,
+      "loss": 0.1488,
+      "step": 5932
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.9099775590148313,
+      "learning_rate": 4.597384675895142e-06,
+      "loss": 0.1193,
+      "step": 5933
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.842068850354602,
+      "learning_rate": 4.595915798055686e-06,
+      "loss": 0.124,
+      "step": 5934
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.714918250138469,
+      "learning_rate": 4.594446955318781e-06,
+      "loss": 0.0961,
+      "step": 5935
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.9167393976566174,
+      "learning_rate": 4.592978147812026e-06,
+      "loss": 0.1113,
+      "step": 5936
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.613239502967477,
+      "learning_rate": 4.591509375663014e-06,
+      "loss": 0.1009,
+      "step": 5937
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.0809948036082773,
+      "learning_rate": 4.590040638999338e-06,
+      "loss": 0.134,
+      "step": 5938
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.6574741298949107,
+      "learning_rate": 4.588571937948583e-06,
+      "loss": 0.1062,
+      "step": 5939
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.662263012786599,
+      "learning_rate": 4.587103272638339e-06,
+      "loss": 0.1094,
+      "step": 5940
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.0448275718192277,
+      "learning_rate": 4.585634643196185e-06,
+      "loss": 0.0957,
+      "step": 5941
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8607738921100478,
+      "learning_rate": 4.584166049749701e-06,
+      "loss": 0.1167,
+      "step": 5942
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8788623559379083,
+      "learning_rate": 4.582697492426461e-06,
+      "loss": 0.1229,
+      "step": 5943
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.7507636599053813,
+      "learning_rate": 4.581228971354042e-06,
+      "loss": 0.117,
+      "step": 5944
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.597577253295051,
+      "learning_rate": 4.579760486660006e-06,
+      "loss": 0.0977,
+      "step": 5945
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8037949381340144,
+      "learning_rate": 4.578292038471928e-06,
+      "loss": 0.0961,
+      "step": 5946
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.730924550344127,
+      "learning_rate": 4.576823626917365e-06,
+      "loss": 0.1108,
+      "step": 5947
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.750923119483021,
+      "learning_rate": 4.575355252123881e-06,
+      "loss": 0.1078,
+      "step": 5948
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.6578763482900079,
+      "learning_rate": 4.573886914219031e-06,
+      "loss": 0.1073,
+      "step": 5949
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.7506037793493139,
+      "learning_rate": 4.572418613330368e-06,
+      "loss": 0.1227,
+      "step": 5950
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8480968311463317,
+      "learning_rate": 4.570950349585442e-06,
+      "loss": 0.1203,
+      "step": 5951
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.724518155793137,
+      "learning_rate": 4.569482123111804e-06,
+      "loss": 0.1009,
+      "step": 5952
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.0633310051106135,
+      "learning_rate": 4.568013934036993e-06,
+      "loss": 0.1077,
+      "step": 5953
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.9928162663049163,
+      "learning_rate": 4.566545782488554e-06,
+      "loss": 0.1129,
+      "step": 5954
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.8053118334412859,
+      "learning_rate": 4.56507766859402e-06,
+      "loss": 0.1105,
+      "step": 5955
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.4233780665855478,
+      "learning_rate": 4.563609592480931e-06,
+      "loss": 0.0878,
+      "step": 5956
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.8812415416352042,
+      "learning_rate": 4.562141554276811e-06,
+      "loss": 0.1193,
+      "step": 5957
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6792852742555906,
+      "learning_rate": 4.5606735541091925e-06,
+      "loss": 0.1015,
+      "step": 5958
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.072203027140658,
+      "learning_rate": 4.559205592105599e-06,
+      "loss": 0.1069,
+      "step": 5959
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.7459184580478888,
+      "learning_rate": 4.557737668393551e-06,
+      "loss": 0.1041,
+      "step": 5960
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.5168771234380187,
+      "learning_rate": 4.556269783100565e-06,
+      "loss": 0.0768,
+      "step": 5961
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.991111831668939,
+      "learning_rate": 4.554801936354157e-06,
+      "loss": 0.1244,
+      "step": 5962
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.9807953753972587,
+      "learning_rate": 4.553334128281836e-06,
+      "loss": 0.1413,
+      "step": 5963
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.9377278658324935,
+      "learning_rate": 4.551866359011114e-06,
+      "loss": 0.1249,
+      "step": 5964
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.2586336573239487,
+      "learning_rate": 4.550398628669489e-06,
+      "loss": 0.1288,
+      "step": 5965
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6957542631992188,
+      "learning_rate": 4.548930937384466e-06,
+      "loss": 0.0984,
+      "step": 5966
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.1670118422770734,
+      "learning_rate": 4.547463285283542e-06,
+      "loss": 0.1399,
+      "step": 5967
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.5530313380967515,
+      "learning_rate": 4.54599567249421e-06,
+      "loss": 0.0844,
+      "step": 5968
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6775201891284022,
+      "learning_rate": 4.544528099143961e-06,
+      "loss": 0.1158,
+      "step": 5969
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.9541166998928459,
+      "learning_rate": 4.543060565360284e-06,
+      "loss": 0.1293,
+      "step": 5970
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.8419156432081785,
+      "learning_rate": 4.541593071270658e-06,
+      "loss": 0.1118,
+      "step": 5971
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.5742995758436404,
+      "learning_rate": 4.54012561700257e-06,
+      "loss": 0.1118,
+      "step": 5972
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.930682830432296,
+      "learning_rate": 4.53865820268349e-06,
+      "loss": 0.1184,
+      "step": 5973
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.8406021395369825,
+      "learning_rate": 4.537190828440898e-06,
+      "loss": 0.1266,
+      "step": 5974
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.5844384959033837,
+      "learning_rate": 4.535723494402258e-06,
+      "loss": 0.1122,
+      "step": 5975
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6337274901187584,
+      "learning_rate": 4.534256200695042e-06,
+      "loss": 0.1107,
+      "step": 5976
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.5682099887127616,
+      "learning_rate": 4.532788947446706e-06,
+      "loss": 0.0921,
+      "step": 5977
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6174022104201136,
+      "learning_rate": 4.531321734784717e-06,
+      "loss": 0.1043,
+      "step": 5978
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6407503054264556,
+      "learning_rate": 4.529854562836525e-06,
+      "loss": 0.095,
+      "step": 5979
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.4033207707062956,
+      "learning_rate": 4.528387431729587e-06,
+      "loss": 0.0898,
+      "step": 5980
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.1389391281939623,
+      "learning_rate": 4.5269203415913465e-06,
+      "loss": 0.1601,
+      "step": 5981
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6035528289081962,
+      "learning_rate": 4.525453292549255e-06,
+      "loss": 0.1102,
+      "step": 5982
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.7262380214177933,
+      "learning_rate": 4.523986284730747e-06,
+      "loss": 0.1108,
+      "step": 5983
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6105912783085636,
+      "learning_rate": 4.5225193182632675e-06,
+      "loss": 0.0968,
+      "step": 5984
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.710759912978609,
+      "learning_rate": 4.5210523932742475e-06,
+      "loss": 0.1038,
+      "step": 5985
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.5727984318899904,
+      "learning_rate": 4.5195855098911165e-06,
+      "loss": 0.0878,
+      "step": 5986
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.901003196193916,
+      "learning_rate": 4.518118668241306e-06,
+      "loss": 0.1112,
+      "step": 5987
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.8374057805227946,
+      "learning_rate": 4.516651868452236e-06,
+      "loss": 0.1118,
+      "step": 5988
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.8069375972387882,
+      "learning_rate": 4.515185110651328e-06,
+      "loss": 0.1082,
+      "step": 5989
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.7369838707424228,
+      "learning_rate": 4.513718394965998e-06,
+      "loss": 0.1022,
+      "step": 5990
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8609155947739908,
+      "learning_rate": 4.512251721523659e-06,
+      "loss": 0.109,
+      "step": 5991
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.6686200305283587,
+      "learning_rate": 4.510785090451719e-06,
+      "loss": 0.0918,
+      "step": 5992
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.2044515457404033,
+      "learning_rate": 4.509318501877586e-06,
+      "loss": 0.1423,
+      "step": 5993
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.7281595478403926,
+      "learning_rate": 4.507851955928659e-06,
+      "loss": 0.1036,
+      "step": 5994
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.9442202317120507,
+      "learning_rate": 4.506385452732338e-06,
+      "loss": 0.1218,
+      "step": 5995
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.1510209923823966,
+      "learning_rate": 4.5049189924160144e-06,
+      "loss": 0.1091,
+      "step": 5996
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.7095541956394982,
+      "learning_rate": 4.5034525751070825e-06,
+      "loss": 0.1061,
+      "step": 5997
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.890543216708684,
+      "learning_rate": 4.501986200932924e-06,
+      "loss": 0.1172,
+      "step": 5998
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8049146671292007,
+      "learning_rate": 4.500519870020928e-06,
+      "loss": 0.1045,
+      "step": 5999
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.9778979271571064,
+      "learning_rate": 4.499053582498469e-06,
+      "loss": 0.1336,
+      "step": 6000
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8396988897962223,
+      "learning_rate": 4.497587338492926e-06,
+      "loss": 0.1063,
+      "step": 6001
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.5298610093213953,
+      "learning_rate": 4.496121138131667e-06,
+      "loss": 0.0876,
+      "step": 6002
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.6735007896285592,
+      "learning_rate": 4.494654981542064e-06,
+      "loss": 0.1017,
+      "step": 6003
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.5797400770801575,
+      "learning_rate": 4.493188868851477e-06,
+      "loss": 0.0955,
+      "step": 6004
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.9734114454759775,
+      "learning_rate": 4.491722800187271e-06,
+      "loss": 0.1278,
+      "step": 6005
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.057493024741134,
+      "learning_rate": 4.4902567756767976e-06,
+      "loss": 0.1146,
+      "step": 6006
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8276332986971147,
+      "learning_rate": 4.488790795447414e-06,
+      "loss": 0.1047,
+      "step": 6007
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.879961989908558,
+      "learning_rate": 4.487324859626465e-06,
+      "loss": 0.1119,
+      "step": 6008
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.963071487620247,
+      "learning_rate": 4.485858968341299e-06,
+      "loss": 0.1252,
+      "step": 6009
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.7058874521568435,
+      "learning_rate": 4.484393121719253e-06,
+      "loss": 0.1192,
+      "step": 6010
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8168198587149336,
+      "learning_rate": 4.482927319887669e-06,
+      "loss": 0.1037,
+      "step": 6011
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.6436685829908715,
+      "learning_rate": 4.481461562973877e-06,
+      "loss": 0.0992,
+      "step": 6012
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.6811039330883193,
+      "learning_rate": 4.479995851105209e-06,
+      "loss": 0.1003,
+      "step": 6013
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.6460242189581848,
+      "learning_rate": 4.478530184408987e-06,
+      "loss": 0.1019,
+      "step": 6014
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.9694314273719495,
+      "learning_rate": 4.477064563012536e-06,
+      "loss": 0.1088,
+      "step": 6015
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.9429413119297423,
+      "learning_rate": 4.4755989870431705e-06,
+      "loss": 0.1264,
+      "step": 6016
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.588587431121469,
+      "learning_rate": 4.474133456628208e-06,
+      "loss": 0.1072,
+      "step": 6017
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.772559048295154,
+      "learning_rate": 4.472667971894955e-06,
+      "loss": 0.1039,
+      "step": 6018
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.4110562646523765,
+      "learning_rate": 4.471202532970719e-06,
+      "loss": 0.085,
+      "step": 6019
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.932895919774141,
+      "learning_rate": 4.469737139982801e-06,
+      "loss": 0.1321,
+      "step": 6020
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8008064895673273,
+      "learning_rate": 4.4682717930585e-06,
+      "loss": 0.1043,
+      "step": 6021
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8801228819357656,
+      "learning_rate": 4.466806492325106e-06,
+      "loss": 0.1191,
+      "step": 6022
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.732623795018266,
+      "learning_rate": 4.465341237909915e-06,
+      "loss": 0.1039,
+      "step": 6023
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8151171943216724,
+      "learning_rate": 4.463876029940207e-06,
+      "loss": 0.1131,
+      "step": 6024
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.02751617743123,
+      "learning_rate": 4.462410868543268e-06,
+      "loss": 0.1294,
+      "step": 6025
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.8267941223455404,
+      "learning_rate": 4.460945753846373e-06,
+      "loss": 0.1185,
+      "step": 6026
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.8182058831336576,
+      "learning_rate": 4.459480685976798e-06,
+      "loss": 0.1148,
+      "step": 6027
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.9317619077462784,
+      "learning_rate": 4.458015665061807e-06,
+      "loss": 0.1168,
+      "step": 6028
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.014717354376884,
+      "learning_rate": 4.456550691228673e-06,
+      "loss": 0.1217,
+      "step": 6029
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.5455633799754318,
+      "learning_rate": 4.455085764604653e-06,
+      "loss": 0.0993,
+      "step": 6030
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7226404025610575,
+      "learning_rate": 4.453620885317006e-06,
+      "loss": 0.1102,
+      "step": 6031
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.6968506397729772,
+      "learning_rate": 4.452156053492983e-06,
+      "loss": 0.1122,
+      "step": 6032
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.775571699121647,
+      "learning_rate": 4.450691269259837e-06,
+      "loss": 0.1066,
+      "step": 6033
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.0232766543832352,
+      "learning_rate": 4.449226532744807e-06,
+      "loss": 0.1276,
+      "step": 6034
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.58766934271213,
+      "learning_rate": 4.4477618440751395e-06,
+      "loss": 0.0965,
+      "step": 6035
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.8608029797363759,
+      "learning_rate": 4.4462972033780675e-06,
+      "loss": 0.1195,
+      "step": 6036
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.68656101615884,
+      "learning_rate": 4.444832610780827e-06,
+      "loss": 0.1114,
+      "step": 6037
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.661519549708528,
+      "learning_rate": 4.443368066410641e-06,
+      "loss": 0.1054,
+      "step": 6038
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7118041490560474,
+      "learning_rate": 4.441903570394739e-06,
+      "loss": 0.1214,
+      "step": 6039
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.0135784324447377,
+      "learning_rate": 4.4404391228603366e-06,
+      "loss": 0.125,
+      "step": 6040
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.8327260817054687,
+      "learning_rate": 4.438974723934654e-06,
+      "loss": 0.11,
+      "step": 6041
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.6715937331501853,
+      "learning_rate": 4.437510373744897e-06,
+      "loss": 0.0945,
+      "step": 6042
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7992523445708128,
+      "learning_rate": 4.436046072418278e-06,
+      "loss": 0.1165,
+      "step": 6043
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7225612603246383,
+      "learning_rate": 4.4345818200819974e-06,
+      "loss": 0.103,
+      "step": 6044
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.8926515529643086,
+      "learning_rate": 4.433117616863255e-06,
+      "loss": 0.1222,
+      "step": 6045
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.5403042936333862,
+      "learning_rate": 4.4316534628892425e-06,
+      "loss": 0.0907,
+      "step": 6046
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.9478355059591983,
+      "learning_rate": 4.430189358287155e-06,
+      "loss": 0.1237,
+      "step": 6047
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.8144992775176658,
+      "learning_rate": 4.4287253031841725e-06,
+      "loss": 0.1159,
+      "step": 6048
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.6139718530573681,
+      "learning_rate": 4.427261297707482e-06,
+      "loss": 0.1061,
+      "step": 6049
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.840463539097827,
+      "learning_rate": 4.425797341984258e-06,
+      "loss": 0.1077,
+      "step": 6050
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.5731634796924834,
+      "learning_rate": 4.424333436141675e-06,
+      "loss": 0.0893,
+      "step": 6051
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.4412368540337546,
+      "learning_rate": 4.422869580306897e-06,
+      "loss": 0.0939,
+      "step": 6052
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.6443479368626843,
+      "learning_rate": 4.421405774607096e-06,
+      "loss": 0.0992,
+      "step": 6053
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7529625038341259,
+      "learning_rate": 4.419942019169424e-06,
+      "loss": 0.1086,
+      "step": 6054
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7095652375383714,
+      "learning_rate": 4.418478314121043e-06,
+      "loss": 0.1051,
+      "step": 6055
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.8300389945089752,
+      "learning_rate": 4.4170146595891006e-06,
+      "loss": 0.1068,
+      "step": 6056
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7265692656538434,
+      "learning_rate": 4.415551055700745e-06,
+      "loss": 0.1175,
+      "step": 6057
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.9590154169254117,
+      "learning_rate": 4.414087502583116e-06,
+      "loss": 0.1112,
+      "step": 6058
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7273431418739582,
+      "learning_rate": 4.4126240003633565e-06,
+      "loss": 0.1119,
+      "step": 6059
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.0566043145868207,
+      "learning_rate": 4.411160549168595e-06,
+      "loss": 0.1356,
+      "step": 6060
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.041543829836337,
+      "learning_rate": 4.409697149125964e-06,
+      "loss": 0.1454,
+      "step": 6061
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.5205109912013173,
+      "learning_rate": 4.408233800362586e-06,
+      "loss": 0.0899,
+      "step": 6062
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.6649255006118326,
+      "learning_rate": 4.406770503005584e-06,
+      "loss": 0.1065,
+      "step": 6063
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.0694680538071646,
+      "learning_rate": 4.405307257182069e-06,
+      "loss": 0.1441,
+      "step": 6064
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.774343052389184,
+      "learning_rate": 4.403844063019159e-06,
+      "loss": 0.1306,
+      "step": 6065
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.5592185909555658,
+      "learning_rate": 4.402380920643954e-06,
+      "loss": 0.1083,
+      "step": 6066
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.6371961749187038,
+      "learning_rate": 4.400917830183561e-06,
+      "loss": 0.1034,
+      "step": 6067
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.755101902647616,
+      "learning_rate": 4.399454791765076e-06,
+      "loss": 0.0922,
+      "step": 6068
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.868333138884964,
+      "learning_rate": 4.397991805515592e-06,
+      "loss": 0.1285,
+      "step": 6069
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7689899627588024,
+      "learning_rate": 4.3965288715621965e-06,
+      "loss": 0.1063,
+      "step": 6070
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7490369364710514,
+      "learning_rate": 4.395065990031979e-06,
+      "loss": 0.1057,
+      "step": 6071
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7585700914928077,
+      "learning_rate": 4.3936031610520126e-06,
+      "loss": 0.1136,
+      "step": 6072
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.8462882337026858,
+      "learning_rate": 4.3921403847493775e-06,
+      "loss": 0.095,
+      "step": 6073
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.1086245895975595,
+      "learning_rate": 4.39067766125114e-06,
+      "loss": 0.1357,
+      "step": 6074
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.4771881857171978,
+      "learning_rate": 4.389214990684369e-06,
+      "loss": 0.0854,
+      "step": 6075
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.780137085344616,
+      "learning_rate": 4.387752373176123e-06,
+      "loss": 0.1048,
+      "step": 6076
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.6358329176303328,
+      "learning_rate": 4.386289808853462e-06,
+      "loss": 0.1087,
+      "step": 6077
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.8109212818991554,
+      "learning_rate": 4.384827297843437e-06,
+      "loss": 0.1139,
+      "step": 6078
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.6343558317055689,
+      "learning_rate": 4.383364840273094e-06,
+      "loss": 0.0988,
+      "step": 6079
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7288117879231606,
+      "learning_rate": 4.381902436269479e-06,
+      "loss": 0.1084,
+      "step": 6080
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.688601538329786,
+      "learning_rate": 4.380440085959625e-06,
+      "loss": 0.1109,
+      "step": 6081
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.6878755192456445,
+      "learning_rate": 4.3789777894705706e-06,
+      "loss": 0.1106,
+      "step": 6082
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.8208732034313206,
+      "learning_rate": 4.377515546929341e-06,
+      "loss": 0.1225,
+      "step": 6083
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.8024846244095072,
+      "learning_rate": 4.3760533584629636e-06,
+      "loss": 0.1154,
+      "step": 6084
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.629381080009447,
+      "learning_rate": 4.374591224198455e-06,
+      "loss": 0.0998,
+      "step": 6085
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7362893865488154,
+      "learning_rate": 4.373129144262832e-06,
+      "loss": 0.1036,
+      "step": 6086
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.9259703915171458,
+      "learning_rate": 4.371667118783101e-06,
+      "loss": 0.1179,
+      "step": 6087
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.8622867567014016,
+      "learning_rate": 4.370205147886273e-06,
+      "loss": 0.1158,
+      "step": 6088
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.8340059736349363,
+      "learning_rate": 4.3687432316993434e-06,
+      "loss": 0.1111,
+      "step": 6089
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7043827099226356,
+      "learning_rate": 4.367281370349311e-06,
+      "loss": 0.0984,
+      "step": 6090
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.81928292081651,
+      "learning_rate": 4.365819563963166e-06,
+      "loss": 0.1059,
+      "step": 6091
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7309597713194826,
+      "learning_rate": 4.364357812667894e-06,
+      "loss": 0.0958,
+      "step": 6092
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.9684908703575503,
+      "learning_rate": 4.362896116590475e-06,
+      "loss": 0.1428,
+      "step": 6093
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.5332570645171262,
+      "learning_rate": 4.361434475857891e-06,
+      "loss": 0.0997,
+      "step": 6094
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7070096177973109,
+      "learning_rate": 4.3599728905971086e-06,
+      "loss": 0.1163,
+      "step": 6095
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.8725119737451372,
+      "learning_rate": 4.358511360935097e-06,
+      "loss": 0.1083,
+      "step": 6096
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7176913733880348,
+      "learning_rate": 4.357049886998818e-06,
+      "loss": 0.1146,
+      "step": 6097
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.9151234736466296,
+      "learning_rate": 4.35558846891523e-06,
+      "loss": 0.1331,
+      "step": 6098
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6407195234162124,
+      "learning_rate": 4.354127106811282e-06,
+      "loss": 0.1041,
+      "step": 6099
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6552260683068514,
+      "learning_rate": 4.352665800813926e-06,
+      "loss": 0.1038,
+      "step": 6100
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6879649788991324,
+      "learning_rate": 4.351204551050102e-06,
+      "loss": 0.0989,
+      "step": 6101
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6656070110711363,
+      "learning_rate": 4.349743357646751e-06,
+      "loss": 0.1008,
+      "step": 6102
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.7620849376688874,
+      "learning_rate": 4.348282220730802e-06,
+      "loss": 0.1197,
+      "step": 6103
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.1230782299117936,
+      "learning_rate": 4.346821140429186e-06,
+      "loss": 0.1236,
+      "step": 6104
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.9394631374902518,
+      "learning_rate": 4.3453601168688225e-06,
+      "loss": 0.1362,
+      "step": 6105
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.7893323098127385,
+      "learning_rate": 4.343899150176635e-06,
+      "loss": 0.1137,
+      "step": 6106
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.5678627477445248,
+      "learning_rate": 4.342438240479533e-06,
+      "loss": 0.0974,
+      "step": 6107
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.5471340414745498,
+      "learning_rate": 4.340977387904427e-06,
+      "loss": 0.0877,
+      "step": 6108
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6239086620667968,
+      "learning_rate": 4.339516592578218e-06,
+      "loss": 0.0973,
+      "step": 6109
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.617461885183093,
+      "learning_rate": 4.3380558546278075e-06,
+      "loss": 0.0876,
+      "step": 6110
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.5980019663907559,
+      "learning_rate": 4.336595174180085e-06,
+      "loss": 0.1074,
+      "step": 6111
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.666743744786788,
+      "learning_rate": 4.335134551361942e-06,
+      "loss": 0.0981,
+      "step": 6112
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.848565637864274,
+      "learning_rate": 4.333673986300262e-06,
+      "loss": 0.1312,
+      "step": 6113
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.090829653417758,
+      "learning_rate": 4.332213479121922e-06,
+      "loss": 0.1214,
+      "step": 6114
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.0550833350570934,
+      "learning_rate": 4.330753029953796e-06,
+      "loss": 0.1177,
+      "step": 6115
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.93060655947151,
+      "learning_rate": 4.329292638922753e-06,
+      "loss": 0.1084,
+      "step": 6116
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.94170093657727,
+      "learning_rate": 4.327832306155652e-06,
+      "loss": 0.1227,
+      "step": 6117
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.7468851736640723,
+      "learning_rate": 4.326372031779359e-06,
+      "loss": 0.1037,
+      "step": 6118
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6982515206555788,
+      "learning_rate": 4.32491181592072e-06,
+      "loss": 0.1048,
+      "step": 6119
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.5755799510466173,
+      "learning_rate": 4.323451658706587e-06,
+      "loss": 0.0882,
+      "step": 6120
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.703143848448169,
+      "learning_rate": 4.321991560263802e-06,
+      "loss": 0.0975,
+      "step": 6121
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.9970759020855016,
+      "learning_rate": 4.320531520719203e-06,
+      "loss": 0.1234,
+      "step": 6122
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6293693062088537,
+      "learning_rate": 4.319071540199621e-06,
+      "loss": 0.1132,
+      "step": 6123
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.9609684899922433,
+      "learning_rate": 4.317611618831888e-06,
+      "loss": 0.1107,
+      "step": 6124
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6835006063149112,
+      "learning_rate": 4.316151756742821e-06,
+      "loss": 0.11,
+      "step": 6125
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.7601749229382702,
+      "learning_rate": 4.314691954059242e-06,
+      "loss": 0.1145,
+      "step": 6126
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6172299395662388,
+      "learning_rate": 4.313232210907959e-06,
+      "loss": 0.0882,
+      "step": 6127
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.5784701854701733,
+      "learning_rate": 4.311772527415784e-06,
+      "loss": 0.1148,
+      "step": 6128
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6543611196364094,
+      "learning_rate": 4.310312903709513e-06,
+      "loss": 0.103,
+      "step": 6129
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.8485781630595715,
+      "learning_rate": 4.308853339915949e-06,
+      "loss": 0.1096,
+      "step": 6130
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.5805406977936411,
+      "learning_rate": 4.307393836161877e-06,
+      "loss": 0.0969,
+      "step": 6131
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.953262426161697,
+      "learning_rate": 4.305934392574088e-06,
+      "loss": 0.1355,
+      "step": 6132
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6764775999032013,
+      "learning_rate": 4.304475009279361e-06,
+      "loss": 0.0969,
+      "step": 6133
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.5674610518184402,
+      "learning_rate": 4.303015686404473e-06,
+      "loss": 0.0987,
+      "step": 6134
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.7838356928953378,
+      "learning_rate": 4.301556424076191e-06,
+      "loss": 0.106,
+      "step": 6135
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.8373877789563595,
+      "learning_rate": 4.300097222421287e-06,
+      "loss": 0.1194,
+      "step": 6136
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.6884284285996094,
+      "learning_rate": 4.298638081566513e-06,
+      "loss": 0.0857,
+      "step": 6137
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.621829373862572,
+      "learning_rate": 4.297179001638629e-06,
+      "loss": 0.0988,
+      "step": 6138
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.922966772177148,
+      "learning_rate": 4.295719982764382e-06,
+      "loss": 0.1172,
+      "step": 6139
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.0303617599944297,
+      "learning_rate": 4.294261025070519e-06,
+      "loss": 0.1256,
+      "step": 6140
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.8962979237111295,
+      "learning_rate": 4.292802128683773e-06,
+      "loss": 0.1301,
+      "step": 6141
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.8550087231272105,
+      "learning_rate": 4.291343293730885e-06,
+      "loss": 0.1094,
+      "step": 6142
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.62859130809329,
+      "learning_rate": 4.289884520338577e-06,
+      "loss": 0.1074,
+      "step": 6143
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.70412936651213,
+      "learning_rate": 4.2884258086335755e-06,
+      "loss": 0.0997,
+      "step": 6144
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.7740179603308133,
+      "learning_rate": 4.286967158742596e-06,
+      "loss": 0.1012,
+      "step": 6145
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.6360805729378918,
+      "learning_rate": 4.285508570792351e-06,
+      "loss": 0.112,
+      "step": 6146
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.6797344581401086,
+      "learning_rate": 4.2840500449095455e-06,
+      "loss": 0.1103,
+      "step": 6147
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.7092724354614606,
+      "learning_rate": 4.282591581220886e-06,
+      "loss": 0.1031,
+      "step": 6148
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.7744060713396335,
+      "learning_rate": 4.281133179853061e-06,
+      "loss": 0.1156,
+      "step": 6149
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.9371245417592706,
+      "learning_rate": 4.279674840932767e-06,
+      "loss": 0.1298,
+      "step": 6150
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.5977810422862884,
+      "learning_rate": 4.278216564586687e-06,
+      "loss": 0.1094,
+      "step": 6151
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.74495715300954,
+      "learning_rate": 4.2767583509415e-06,
+      "loss": 0.1144,
+      "step": 6152
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.8100158669145083,
+      "learning_rate": 4.275300200123879e-06,
+      "loss": 0.1095,
+      "step": 6153
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.7526332567600043,
+      "learning_rate": 4.2738421122604964e-06,
+      "loss": 0.1137,
+      "step": 6154
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.7975184162380935,
+      "learning_rate": 4.272384087478011e-06,
+      "loss": 0.1126,
+      "step": 6155
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.7337847735619645,
+      "learning_rate": 4.270926125903085e-06,
+      "loss": 0.106,
+      "step": 6156
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.7744175740297166,
+      "learning_rate": 4.2694682276623675e-06,
+      "loss": 0.1096,
+      "step": 6157
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.6513612086909404,
+      "learning_rate": 4.268010392882506e-06,
+      "loss": 0.1114,
+      "step": 6158
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.8202119447634264,
+      "learning_rate": 4.266552621690141e-06,
+      "loss": 0.118,
+      "step": 6159
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.969437549994304,
+      "learning_rate": 4.2650949142119116e-06,
+      "loss": 0.1116,
+      "step": 6160
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.0954889027229138,
+      "learning_rate": 4.2636372705744425e-06,
+      "loss": 0.1187,
+      "step": 6161
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.5991152732132552,
+      "learning_rate": 4.262179690904363e-06,
+      "loss": 0.1055,
+      "step": 6162
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.6651971914598611,
+      "learning_rate": 4.26072217532829e-06,
+      "loss": 0.1009,
+      "step": 6163
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.8137942352817806,
+      "learning_rate": 4.259264723972839e-06,
+      "loss": 0.0978,
+      "step": 6164
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.901322185348973,
+      "learning_rate": 4.2578073369646135e-06,
+      "loss": 0.119,
+      "step": 6165
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.5771608730415214,
+      "learning_rate": 4.256350014430221e-06,
+      "loss": 0.0991,
+      "step": 6166
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.725720906861579,
+      "learning_rate": 4.254892756496255e-06,
+      "loss": 0.1064,
+      "step": 6167
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.8779194416022726,
+      "learning_rate": 4.2534355632893085e-06,
+      "loss": 0.1079,
+      "step": 6168
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.7127076584945626,
+      "learning_rate": 4.251978434935964e-06,
+      "loss": 0.1056,
+      "step": 6169
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.7260917632296027,
+      "learning_rate": 4.250521371562803e-06,
+      "loss": 0.1038,
+      "step": 6170
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.6001759860508347,
+      "learning_rate": 4.249064373296403e-06,
+      "loss": 0.0946,
+      "step": 6171
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.872591333169689,
+      "learning_rate": 4.247607440263329e-06,
+      "loss": 0.1096,
+      "step": 6172
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.7913107046741024,
+      "learning_rate": 4.246150572590145e-06,
+      "loss": 0.1143,
+      "step": 6173
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.6940806817026999,
+      "learning_rate": 4.2446937704034065e-06,
+      "loss": 0.0934,
+      "step": 6174
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.6213743474547668,
+      "learning_rate": 4.243237033829668e-06,
+      "loss": 0.1053,
+      "step": 6175
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9097946400190038,
+      "learning_rate": 4.241780362995471e-06,
+      "loss": 0.1106,
+      "step": 6176
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.8433203070746311,
+      "learning_rate": 4.240323758027361e-06,
+      "loss": 0.1119,
+      "step": 6177
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9083970681942128,
+      "learning_rate": 4.238867219051868e-06,
+      "loss": 0.1081,
+      "step": 6178
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.896381519393454,
+      "learning_rate": 4.237410746195524e-06,
+      "loss": 0.1149,
+      "step": 6179
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.5956697135336217,
+      "learning_rate": 4.235954339584849e-06,
+      "loss": 0.1079,
+      "step": 6180
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.7288693670369817,
+      "learning_rate": 4.234497999346363e-06,
+      "loss": 0.1125,
+      "step": 6181
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.74024813877436,
+      "learning_rate": 4.233041725606573e-06,
+      "loss": 0.1063,
+      "step": 6182
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.6642118478211958,
+      "learning_rate": 4.231585518491989e-06,
+      "loss": 0.1166,
+      "step": 6183
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.7322645731058426,
+      "learning_rate": 4.23012937812911e-06,
+      "loss": 0.1052,
+      "step": 6184
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.8102222726026993,
+      "learning_rate": 4.22867330464443e-06,
+      "loss": 0.1106,
+      "step": 6185
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9057024966265923,
+      "learning_rate": 4.227217298164434e-06,
+      "loss": 0.1047,
+      "step": 6186
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.0122740418217004,
+      "learning_rate": 4.22576135881561e-06,
+      "loss": 0.1277,
+      "step": 6187
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.0036301671695225,
+      "learning_rate": 4.2243054867244285e-06,
+      "loss": 0.1255,
+      "step": 6188
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9717285805002271,
+      "learning_rate": 4.222849682017366e-06,
+      "loss": 0.1212,
+      "step": 6189
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.5790668876520453,
+      "learning_rate": 4.221393944820883e-06,
+      "loss": 0.1045,
+      "step": 6190
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.7476060625551413,
+      "learning_rate": 4.219938275261442e-06,
+      "loss": 0.1086,
+      "step": 6191
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.5785725736088405,
+      "learning_rate": 4.2184826734654925e-06,
+      "loss": 0.1033,
+      "step": 6192
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.8288703972164801,
+      "learning_rate": 4.2170271395594855e-06,
+      "loss": 0.117,
+      "step": 6193
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.8825803403152777,
+      "learning_rate": 4.215571673669857e-06,
+      "loss": 0.1199,
+      "step": 6194
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.6492432007552422,
+      "learning_rate": 4.214116275923051e-06,
+      "loss": 0.0913,
+      "step": 6195
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.7909989098790684,
+      "learning_rate": 4.2126609464454876e-06,
+      "loss": 0.1209,
+      "step": 6196
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.8332592624498136,
+      "learning_rate": 4.211205685363597e-06,
+      "loss": 0.1033,
+      "step": 6197
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9176775788282023,
+      "learning_rate": 4.209750492803794e-06,
+      "loss": 0.1384,
+      "step": 6198
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9712404700660504,
+      "learning_rate": 4.208295368892491e-06,
+      "loss": 0.1189,
+      "step": 6199
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.6637269645537947,
+      "learning_rate": 4.206840313756092e-06,
+      "loss": 0.0971,
+      "step": 6200
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.5477873849277628,
+      "learning_rate": 4.205385327521002e-06,
+      "loss": 0.1024,
+      "step": 6201
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.8596908728724282,
+      "learning_rate": 4.203930410313608e-06,
+      "loss": 0.0951,
+      "step": 6202
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.4778054525938429,
+      "learning_rate": 4.202475562260302e-06,
+      "loss": 0.0866,
+      "step": 6203
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.0126047816033177,
+      "learning_rate": 4.201020783487465e-06,
+      "loss": 0.1171,
+      "step": 6204
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.8389749463207241,
+      "learning_rate": 4.199566074121473e-06,
+      "loss": 0.1115,
+      "step": 6205
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.5828426566207194,
+      "learning_rate": 4.198111434288693e-06,
+      "loss": 0.1021,
+      "step": 6206
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.0356217362206266,
+      "learning_rate": 4.196656864115494e-06,
+      "loss": 0.1134,
+      "step": 6207
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.727423384079484,
+      "learning_rate": 4.195202363728227e-06,
+      "loss": 0.1132,
+      "step": 6208
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.887721590345589,
+      "learning_rate": 4.19374793325325e-06,
+      "loss": 0.1031,
+      "step": 6209
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.8682436645469198,
+      "learning_rate": 4.1922935728169045e-06,
+      "loss": 0.1304,
+      "step": 6210
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.6056136810925175,
+      "learning_rate": 4.190839282545532e-06,
+      "loss": 0.0898,
+      "step": 6211
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.834031052172413,
+      "learning_rate": 4.1893850625654626e-06,
+      "loss": 0.1218,
+      "step": 6212
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.7464525978972036,
+      "learning_rate": 4.187930913003029e-06,
+      "loss": 0.1162,
+      "step": 6213
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.8107501587966144,
+      "learning_rate": 4.186476833984546e-06,
+      "loss": 0.1112,
+      "step": 6214
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.6075800943630099,
+      "learning_rate": 4.185022825636334e-06,
+      "loss": 0.1191,
+      "step": 6215
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.545418033846747,
+      "learning_rate": 4.183568888084698e-06,
+      "loss": 0.0998,
+      "step": 6216
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.5586167232423198,
+      "learning_rate": 4.182115021455944e-06,
+      "loss": 0.1006,
+      "step": 6217
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.9304884502825523,
+      "learning_rate": 4.180661225876363e-06,
+      "loss": 0.1391,
+      "step": 6218
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.849257718600328,
+      "learning_rate": 4.179207501472254e-06,
+      "loss": 0.1306,
+      "step": 6219
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.6273398337223621,
+      "learning_rate": 4.177753848369892e-06,
+      "loss": 0.1085,
+      "step": 6220
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.6257542497069029,
+      "learning_rate": 4.1763002666955615e-06,
+      "loss": 0.11,
+      "step": 6221
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.0146443830332688,
+      "learning_rate": 4.174846756575531e-06,
+      "loss": 0.1186,
+      "step": 6222
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.399684554794835,
+      "learning_rate": 4.1733933181360685e-06,
+      "loss": 0.0891,
+      "step": 6223
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.8405455736401044,
+      "learning_rate": 4.1719399515034285e-06,
+      "loss": 0.1354,
+      "step": 6224
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.6416844235507686,
+      "learning_rate": 4.1704866568038715e-06,
+      "loss": 0.1185,
+      "step": 6225
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.609336411819547,
+      "learning_rate": 4.169033434163637e-06,
+      "loss": 0.1125,
+      "step": 6226
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.19935327321842,
+      "learning_rate": 4.167580283708971e-06,
+      "loss": 0.1125,
+      "step": 6227
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.7915923785389125,
+      "learning_rate": 4.166127205566104e-06,
+      "loss": 0.1146,
+      "step": 6228
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.5245011255218217,
+      "learning_rate": 4.1646741998612676e-06,
+      "loss": 0.0893,
+      "step": 6229
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.7861856473251987,
+      "learning_rate": 4.1632212667206786e-06,
+      "loss": 0.1136,
+      "step": 6230
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.9102802804857326,
+      "learning_rate": 4.161768406270559e-06,
+      "loss": 0.1039,
+      "step": 6231
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.744142733969087,
+      "learning_rate": 4.1603156186371106e-06,
+      "loss": 0.1081,
+      "step": 6232
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.9990226150833714,
+      "learning_rate": 4.158862903946543e-06,
+      "loss": 0.1283,
+      "step": 6233
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.9946853239127929,
+      "learning_rate": 4.1574102623250476e-06,
+      "loss": 0.1078,
+      "step": 6234
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.7170015494425657,
+      "learning_rate": 4.155957693898817e-06,
+      "loss": 0.1133,
+      "step": 6235
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.8090467061126,
+      "learning_rate": 4.154505198794034e-06,
+      "loss": 0.1224,
+      "step": 6236
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.830096191286136,
+      "learning_rate": 4.153052777136879e-06,
+      "loss": 0.1338,
+      "step": 6237
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.6866106988384477,
+      "learning_rate": 4.151600429053517e-06,
+      "loss": 0.1175,
+      "step": 6238
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.8175927489964359,
+      "learning_rate": 4.1501481546701185e-06,
+      "loss": 0.1116,
+      "step": 6239
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.4777329478896946,
+      "learning_rate": 4.148695954112838e-06,
+      "loss": 0.0923,
+      "step": 6240
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.5771567064194922,
+      "learning_rate": 4.147243827507829e-06,
+      "loss": 0.0854,
+      "step": 6241
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.9061231967446501,
+      "learning_rate": 4.1457917749812345e-06,
+      "loss": 0.1239,
+      "step": 6242
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.4517485857028007,
+      "learning_rate": 4.1443397966591985e-06,
+      "loss": 0.1015,
+      "step": 6243
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.314425253776543,
+      "learning_rate": 4.142887892667848e-06,
+      "loss": 0.1322,
+      "step": 6244
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.722177867766709,
+      "learning_rate": 4.141436063133312e-06,
+      "loss": 0.1049,
+      "step": 6245
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.6777799467020105,
+      "learning_rate": 4.1399843081817085e-06,
+      "loss": 0.0967,
+      "step": 6246
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7876454286767574,
+      "learning_rate": 4.138532627939153e-06,
+      "loss": 0.115,
+      "step": 6247
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7104228494378149,
+      "learning_rate": 4.137081022531748e-06,
+      "loss": 0.1169,
+      "step": 6248
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.8195999084262289,
+      "learning_rate": 4.1356294920856e-06,
+      "loss": 0.1316,
+      "step": 6249
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.012071523505236,
+      "learning_rate": 4.134178036726795e-06,
+      "loss": 0.1345,
+      "step": 6250
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.9317357464122598,
+      "learning_rate": 4.132726656581426e-06,
+      "loss": 0.1175,
+      "step": 6251
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.6234949767479492,
+      "learning_rate": 4.13127535177557e-06,
+      "loss": 0.0964,
+      "step": 6252
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.6201085088735945,
+      "learning_rate": 4.129824122435304e-06,
+      "loss": 0.1219,
+      "step": 6253
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.6107762805712103,
+      "learning_rate": 4.128372968686691e-06,
+      "loss": 0.0935,
+      "step": 6254
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.589530462542716,
+      "learning_rate": 4.126921890655797e-06,
+      "loss": 0.0922,
+      "step": 6255
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7950944160366196,
+      "learning_rate": 4.125470888468672e-06,
+      "loss": 0.107,
+      "step": 6256
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.9591376933550466,
+      "learning_rate": 4.124019962251366e-06,
+      "loss": 0.1043,
+      "step": 6257
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.9207632201314828,
+      "learning_rate": 4.12256911212992e-06,
+      "loss": 0.1087,
+      "step": 6258
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.5569263988899922,
+      "learning_rate": 4.121118338230369e-06,
+      "loss": 0.0943,
+      "step": 6259
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.1941362611572037,
+      "learning_rate": 4.119667640678737e-06,
+      "loss": 0.1174,
+      "step": 6260
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.674340734627718,
+      "learning_rate": 4.118217019601053e-06,
+      "loss": 0.0842,
+      "step": 6261
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.3077173771590105,
+      "learning_rate": 4.116766475123322e-06,
+      "loss": 0.0736,
+      "step": 6262
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7016590249287207,
+      "learning_rate": 4.115316007371557e-06,
+      "loss": 0.1004,
+      "step": 6263
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.9484416720576012,
+      "learning_rate": 4.113865616471761e-06,
+      "loss": 0.1383,
+      "step": 6264
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.4888373401562678,
+      "learning_rate": 4.112415302549925e-06,
+      "loss": 0.0862,
+      "step": 6265
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.6757488447946416,
+      "learning_rate": 4.11096506573204e-06,
+      "loss": 0.0991,
+      "step": 6266
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.797865324735706,
+      "learning_rate": 4.109514906144084e-06,
+      "loss": 0.1112,
+      "step": 6267
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.8179678288211754,
+      "learning_rate": 4.108064823912035e-06,
+      "loss": 0.116,
+      "step": 6268
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.6192344758158899,
+      "learning_rate": 4.106614819161857e-06,
+      "loss": 0.0995,
+      "step": 6269
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.9529892900646009,
+      "learning_rate": 4.105164892019514e-06,
+      "loss": 0.1287,
+      "step": 6270
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7831377880203494,
+      "learning_rate": 4.103715042610958e-06,
+      "loss": 0.113,
+      "step": 6271
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.71200879740593,
+      "learning_rate": 4.102265271062139e-06,
+      "loss": 0.1111,
+      "step": 6272
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7451634224136954,
+      "learning_rate": 4.100815577498995e-06,
+      "loss": 0.0938,
+      "step": 6273
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.658340972407541,
+      "learning_rate": 4.099365962047464e-06,
+      "loss": 0.1152,
+      "step": 6274
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.9082733883621454,
+      "learning_rate": 4.097916424833469e-06,
+      "loss": 0.1207,
+      "step": 6275
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.8794396958763797,
+      "learning_rate": 4.0964669659829335e-06,
+      "loss": 0.1144,
+      "step": 6276
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.8992218776703043,
+      "learning_rate": 4.095017585621767e-06,
+      "loss": 0.1137,
+      "step": 6277
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.4857048235580483,
+      "learning_rate": 4.093568283875882e-06,
+      "loss": 0.0982,
+      "step": 6278
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.822075966997169,
+      "learning_rate": 4.0921190608711745e-06,
+      "loss": 0.1305,
+      "step": 6279
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7028811322326387,
+      "learning_rate": 4.090669916733539e-06,
+      "loss": 0.1175,
+      "step": 6280
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.5323279042293845,
+      "learning_rate": 4.089220851588861e-06,
+      "loss": 0.1033,
+      "step": 6281
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.8749770243241926,
+      "learning_rate": 4.087771865563022e-06,
+      "loss": 0.1146,
+      "step": 6282
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.6218832551442097,
+      "learning_rate": 4.08632295878189e-06,
+      "loss": 0.1001,
+      "step": 6283
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8377943380263704,
+      "learning_rate": 4.084874131371337e-06,
+      "loss": 0.1201,
+      "step": 6284
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.693273814372446,
+      "learning_rate": 4.083425383457215e-06,
+      "loss": 0.1161,
+      "step": 6285
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.4607612580036378,
+      "learning_rate": 4.081976715165382e-06,
+      "loss": 0.0947,
+      "step": 6286
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.9474545859606078,
+      "learning_rate": 4.080528126621679e-06,
+      "loss": 0.1205,
+      "step": 6287
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8513150835051702,
+      "learning_rate": 4.079079617951946e-06,
+      "loss": 0.1188,
+      "step": 6288
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.0285810892990783,
+      "learning_rate": 4.077631189282011e-06,
+      "loss": 0.1349,
+      "step": 6289
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7125713816727879,
+      "learning_rate": 4.0761828407377035e-06,
+      "loss": 0.1132,
+      "step": 6290
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8330073900229522,
+      "learning_rate": 4.074734572444835e-06,
+      "loss": 0.1004,
+      "step": 6291
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.107192627083278,
+      "learning_rate": 4.0732863845292204e-06,
+      "loss": 0.1347,
+      "step": 6292
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.009899782269485,
+      "learning_rate": 4.071838277116659e-06,
+      "loss": 0.1341,
+      "step": 6293
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7518023673265388,
+      "learning_rate": 4.070390250332951e-06,
+      "loss": 0.1179,
+      "step": 6294
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7840522966070804,
+      "learning_rate": 4.06894230430388e-06,
+      "loss": 0.1139,
+      "step": 6295
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.85980833486709,
+      "learning_rate": 4.067494439155236e-06,
+      "loss": 0.1225,
+      "step": 6296
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.6839369729734928,
+      "learning_rate": 4.066046655012786e-06,
+      "loss": 0.1116,
+      "step": 6297
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8625577205152355,
+      "learning_rate": 4.0645989520023035e-06,
+      "loss": 0.1116,
+      "step": 6298
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8290424019671168,
+      "learning_rate": 4.0631513302495475e-06,
+      "loss": 0.1022,
+      "step": 6299
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.6136230365369777,
+      "learning_rate": 4.0617037898802744e-06,
+      "loss": 0.1068,
+      "step": 6300
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.6510545205686094,
+      "learning_rate": 4.060256331020226e-06,
+      "loss": 0.0912,
+      "step": 6301
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.522876984073007,
+      "learning_rate": 4.058808953795149e-06,
+      "loss": 0.0995,
+      "step": 6302
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8819209908448353,
+      "learning_rate": 4.0573616583307705e-06,
+      "loss": 0.1159,
+      "step": 6303
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.0839851262176987,
+      "learning_rate": 4.05591444475282e-06,
+      "loss": 0.1364,
+      "step": 6304
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.5986155566970244,
+      "learning_rate": 4.054467313187013e-06,
+      "loss": 0.1056,
+      "step": 6305
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.916139763030213,
+      "learning_rate": 4.053020263759064e-06,
+      "loss": 0.1223,
+      "step": 6306
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7410239582425142,
+      "learning_rate": 4.051573296594673e-06,
+      "loss": 0.1071,
+      "step": 6307
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7294536644988945,
+      "learning_rate": 4.050126411819544e-06,
+      "loss": 0.1157,
+      "step": 6308
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.714476912047543,
+      "learning_rate": 4.048679609559359e-06,
+      "loss": 0.0967,
+      "step": 6309
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.4838636893457047,
+      "learning_rate": 4.047232889939807e-06,
+      "loss": 0.0891,
+      "step": 6310
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.5880553438132763,
+      "learning_rate": 4.04578625308656e-06,
+      "loss": 0.0943,
+      "step": 6311
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.6927265223940133,
+      "learning_rate": 4.044339699125289e-06,
+      "loss": 0.1038,
+      "step": 6312
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7261844509031412,
+      "learning_rate": 4.0428932281816524e-06,
+      "loss": 0.0913,
+      "step": 6313
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7361019847026014,
+      "learning_rate": 4.041446840381309e-06,
+      "loss": 0.1098,
+      "step": 6314
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.012847330060325,
+      "learning_rate": 4.0400005358499e-06,
+      "loss": 0.1068,
+      "step": 6315
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.967308941807177,
+      "learning_rate": 4.0385543147130694e-06,
+      "loss": 0.1091,
+      "step": 6316
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8784430748475012,
+      "learning_rate": 4.037108177096447e-06,
+      "loss": 0.1253,
+      "step": 6317
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8152356342976743,
+      "learning_rate": 4.03566212312566e-06,
+      "loss": 0.1091,
+      "step": 6318
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6640510985766093,
+      "learning_rate": 4.034216152926322e-06,
+      "loss": 0.1039,
+      "step": 6319
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7465476414731584,
+      "learning_rate": 4.032770266624051e-06,
+      "loss": 0.1144,
+      "step": 6320
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6961642023465495,
+      "learning_rate": 4.031324464344441e-06,
+      "loss": 0.1057,
+      "step": 6321
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6043180054873025,
+      "learning_rate": 4.029878746213096e-06,
+      "loss": 0.104,
+      "step": 6322
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.8908366038279583,
+      "learning_rate": 4.0284331123556e-06,
+      "loss": 0.1204,
+      "step": 6323
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.9136045368627994,
+      "learning_rate": 4.026987562897537e-06,
+      "loss": 0.133,
+      "step": 6324
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.867368555275212,
+      "learning_rate": 4.025542097964478e-06,
+      "loss": 0.1194,
+      "step": 6325
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7613660714842654,
+      "learning_rate": 4.024096717681994e-06,
+      "loss": 0.0983,
+      "step": 6326
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6330437116195555,
+      "learning_rate": 4.022651422175639e-06,
+      "loss": 0.1103,
+      "step": 6327
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6727600710904147,
+      "learning_rate": 4.02120621157097e-06,
+      "loss": 0.0972,
+      "step": 6328
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7047219839487577,
+      "learning_rate": 4.0197610859935275e-06,
+      "loss": 0.1178,
+      "step": 6329
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6410471793023342,
+      "learning_rate": 4.018316045568853e-06,
+      "loss": 0.1025,
+      "step": 6330
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6415898324600682,
+      "learning_rate": 4.016871090422471e-06,
+      "loss": 0.1015,
+      "step": 6331
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6361449685598493,
+      "learning_rate": 4.015426220679909e-06,
+      "loss": 0.0948,
+      "step": 6332
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7076354447215008,
+      "learning_rate": 4.013981436466677e-06,
+      "loss": 0.1195,
+      "step": 6333
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.9945633506345075,
+      "learning_rate": 4.012536737908288e-06,
+      "loss": 0.1232,
+      "step": 6334
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.016750448680824,
+      "learning_rate": 4.011092125130238e-06,
+      "loss": 0.1337,
+      "step": 6335
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6595075714977763,
+      "learning_rate": 4.009647598258022e-06,
+      "loss": 0.1015,
+      "step": 6336
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.5939061565182813,
+      "learning_rate": 4.008203157417122e-06,
+      "loss": 0.1097,
+      "step": 6337
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.5599216239504967,
+      "learning_rate": 4.00675880273302e-06,
+      "loss": 0.0992,
+      "step": 6338
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.116739625463112,
+      "learning_rate": 4.005314534331181e-06,
+      "loss": 0.1201,
+      "step": 6339
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.5403185486047273,
+      "learning_rate": 4.003870352337075e-06,
+      "loss": 0.0921,
+      "step": 6340
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7762830646816974,
+      "learning_rate": 4.00242625687615e-06,
+      "loss": 0.1129,
+      "step": 6341
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.011306506880436,
+      "learning_rate": 4.000982248073858e-06,
+      "loss": 0.1199,
+      "step": 6342
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.578675225508738,
+      "learning_rate": 3.999538326055636e-06,
+      "loss": 0.0885,
+      "step": 6343
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.8467848082906557,
+      "learning_rate": 3.998094490946922e-06,
+      "loss": 0.1165,
+      "step": 6344
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7278962800385307,
+      "learning_rate": 3.996650742873135e-06,
+      "loss": 0.1072,
+      "step": 6345
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7610107380318833,
+      "learning_rate": 3.995207081959696e-06,
+      "loss": 0.1027,
+      "step": 6346
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7415802279283394,
+      "learning_rate": 3.993763508332014e-06,
+      "loss": 0.0932,
+      "step": 6347
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7691901781920214,
+      "learning_rate": 3.992320022115492e-06,
+      "loss": 0.1077,
+      "step": 6348
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.8045781793822298,
+      "learning_rate": 3.990876623435522e-06,
+      "loss": 0.1329,
+      "step": 6349
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.318751822292224,
+      "learning_rate": 3.989433312417497e-06,
+      "loss": 0.116,
+      "step": 6350
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.8800514274353077,
+      "learning_rate": 3.987990089186789e-06,
+      "loss": 0.1198,
+      "step": 6351
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.834117538687703,
+      "learning_rate": 3.9865469538687765e-06,
+      "loss": 0.1005,
+      "step": 6352
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.8260038407520658,
+      "learning_rate": 3.985103906588821e-06,
+      "loss": 0.105,
+      "step": 6353
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.4648568153538277,
+      "learning_rate": 3.983660947472279e-06,
+      "loss": 0.0859,
+      "step": 6354
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6718674516106917,
+      "learning_rate": 3.9822180766445e-06,
+      "loss": 0.1002,
+      "step": 6355
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.8148230452477176,
+      "learning_rate": 3.980775294230824e-06,
+      "loss": 0.125,
+      "step": 6356
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.038399096135445,
+      "learning_rate": 3.979332600356587e-06,
+      "loss": 0.1073,
+      "step": 6357
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.6525805189891618,
+      "learning_rate": 3.977889995147114e-06,
+      "loss": 0.1006,
+      "step": 6358
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.6020102534148162,
+      "learning_rate": 3.976447478727723e-06,
+      "loss": 0.0945,
+      "step": 6359
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.8308861435571353,
+      "learning_rate": 3.9750050512237224e-06,
+      "loss": 0.1098,
+      "step": 6360
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.49953364426434,
+      "learning_rate": 3.973562712760421e-06,
+      "loss": 0.1067,
+      "step": 6361
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.8007984238567853,
+      "learning_rate": 3.9721204634631075e-06,
+      "loss": 0.1019,
+      "step": 6362
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.7925130265180567,
+      "learning_rate": 3.970678303457073e-06,
+      "loss": 0.1267,
+      "step": 6363
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.6890879945945154,
+      "learning_rate": 3.969236232867594e-06,
+      "loss": 0.1071,
+      "step": 6364
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.7418063945039548,
+      "learning_rate": 3.9677942518199465e-06,
+      "loss": 0.103,
+      "step": 6365
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.699227774074572,
+      "learning_rate": 3.96635236043939e-06,
+      "loss": 0.0994,
+      "step": 6366
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.9933384691082228,
+      "learning_rate": 3.9649105588511854e-06,
+      "loss": 0.1265,
+      "step": 6367
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.759316701935023,
+      "learning_rate": 3.963468847180576e-06,
+      "loss": 0.1003,
+      "step": 6368
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.6357790990251504,
+      "learning_rate": 3.962027225552807e-06,
+      "loss": 0.0954,
+      "step": 6369
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.9473634041898158,
+      "learning_rate": 3.960585694093108e-06,
+      "loss": 0.1313,
+      "step": 6370
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.643362854663145,
+      "learning_rate": 3.9591442529267065e-06,
+      "loss": 0.0983,
+      "step": 6371
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.7562258538083544,
+      "learning_rate": 3.957702902178816e-06,
+      "loss": 0.1054,
+      "step": 6372
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.870502096037075,
+      "learning_rate": 3.956261641974653e-06,
+      "loss": 0.1146,
+      "step": 6373
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.8749019154037525,
+      "learning_rate": 3.954820472439409e-06,
+      "loss": 0.1133,
+      "step": 6374
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.879597819384659,
+      "learning_rate": 3.953379393698286e-06,
+      "loss": 0.1193,
+      "step": 6375
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.7697407602776931,
+      "learning_rate": 3.951938405876464e-06,
+      "loss": 0.1071,
+      "step": 6376
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.6537547650649638,
+      "learning_rate": 3.950497509099124e-06,
+      "loss": 0.1029,
+      "step": 6377
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.5299361716564939,
+      "learning_rate": 3.9490567034914335e-06,
+      "loss": 0.091,
+      "step": 6378
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.027042229692232,
+      "learning_rate": 3.947615989178558e-06,
+      "loss": 0.1217,
+      "step": 6379
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.6522728425931497,
+      "learning_rate": 3.946175366285647e-06,
+      "loss": 0.0941,
+      "step": 6380
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.7413956911920896,
+      "learning_rate": 3.9447348349378514e-06,
+      "loss": 0.0967,
+      "step": 6381
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.7770048899054487,
+      "learning_rate": 3.943294395260305e-06,
+      "loss": 0.1025,
+      "step": 6382
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.3966216314330149,
+      "learning_rate": 3.94185404737814e-06,
+      "loss": 0.0819,
+      "step": 6383
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.9205587659461516,
+      "learning_rate": 3.940413791416477e-06,
+      "loss": 0.131,
+      "step": 6384
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.6689883052053822,
+      "learning_rate": 3.938973627500434e-06,
+      "loss": 0.1076,
+      "step": 6385
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.660496860500344,
+      "learning_rate": 3.937533555755111e-06,
+      "loss": 0.1046,
+      "step": 6386
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.733693328448356,
+      "learning_rate": 3.936093576305613e-06,
+      "loss": 0.1135,
+      "step": 6387
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.9321748449062366,
+      "learning_rate": 3.9346536892770245e-06,
+      "loss": 0.1186,
+      "step": 6388
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.618835787426827,
+      "learning_rate": 3.933213894794432e-06,
+      "loss": 0.1037,
+      "step": 6389
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.581015278960825,
+      "learning_rate": 3.9317741929829036e-06,
+      "loss": 0.1024,
+      "step": 6390
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.7747615736064235,
+      "learning_rate": 3.930334583967514e-06,
+      "loss": 0.1076,
+      "step": 6391
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.4866400982489236,
+      "learning_rate": 3.928895067873313e-06,
+      "loss": 0.0903,
+      "step": 6392
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.8653124645872947,
+      "learning_rate": 3.927455644825356e-06,
+      "loss": 0.1255,
+      "step": 6393
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.9008728108769946,
+      "learning_rate": 3.926016314948682e-06,
+      "loss": 0.1218,
+      "step": 6394
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7312758460558038,
+      "learning_rate": 3.924577078368326e-06,
+      "loss": 0.1055,
+      "step": 6395
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7965784299992846,
+      "learning_rate": 3.923137935209311e-06,
+      "loss": 0.0999,
+      "step": 6396
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7056743541019883,
+      "learning_rate": 3.9216988855966595e-06,
+      "loss": 0.103,
+      "step": 6397
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.6843522577964418,
+      "learning_rate": 3.920259929655376e-06,
+      "loss": 0.1074,
+      "step": 6398
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.6599475380326825,
+      "learning_rate": 3.918821067510464e-06,
+      "loss": 0.1116,
+      "step": 6399
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.8124204355444313,
+      "learning_rate": 3.9173822992869166e-06,
+      "loss": 0.1197,
+      "step": 6400
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.672096977376338,
+      "learning_rate": 3.915943625109719e-06,
+      "loss": 0.1004,
+      "step": 6401
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.812939294905621,
+      "learning_rate": 3.914505045103845e-06,
+      "loss": 0.114,
+      "step": 6402
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.5263461052725322,
+      "learning_rate": 3.9130665593942695e-06,
+      "loss": 0.0801,
+      "step": 6403
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.9436320032795122,
+      "learning_rate": 3.911628168105946e-06,
+      "loss": 0.1096,
+      "step": 6404
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.045023265162794,
+      "learning_rate": 3.91018987136383e-06,
+      "loss": 0.1174,
+      "step": 6405
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7693608262918126,
+      "learning_rate": 3.908751669292865e-06,
+      "loss": 0.1186,
+      "step": 6406
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.8094180156407509,
+      "learning_rate": 3.907313562017988e-06,
+      "loss": 0.0993,
+      "step": 6407
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.9548927597012942,
+      "learning_rate": 3.905875549664123e-06,
+      "loss": 0.1085,
+      "step": 6408
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.3712759906977015,
+      "learning_rate": 3.9044376323561955e-06,
+      "loss": 0.1346,
+      "step": 6409
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.6796882524081158,
+      "learning_rate": 3.902999810219109e-06,
+      "loss": 0.1167,
+      "step": 6410
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.0736014706249133,
+      "learning_rate": 3.901562083377772e-06,
+      "loss": 0.14,
+      "step": 6411
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.6749096651001998,
+      "learning_rate": 3.900124451957076e-06,
+      "loss": 0.1022,
+      "step": 6412
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.8005797540982094,
+      "learning_rate": 3.898686916081909e-06,
+      "loss": 0.1088,
+      "step": 6413
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.6764234464253458,
+      "learning_rate": 3.8972494758771455e-06,
+      "loss": 0.108,
+      "step": 6414
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.577037439838749,
+      "learning_rate": 3.895812131467661e-06,
+      "loss": 0.103,
+      "step": 6415
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7168569442585115,
+      "learning_rate": 3.89437488297831e-06,
+      "loss": 0.0908,
+      "step": 6416
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.779965992727559,
+      "learning_rate": 3.892937730533951e-06,
+      "loss": 0.1103,
+      "step": 6417
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.5831992631281655,
+      "learning_rate": 3.891500674259425e-06,
+      "loss": 0.0938,
+      "step": 6418
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.6234704476342032,
+      "learning_rate": 3.89006371427957e-06,
+      "loss": 0.1003,
+      "step": 6419
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.319157269093631,
+      "learning_rate": 3.8886268507192116e-06,
+      "loss": 0.0728,
+      "step": 6420
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.5664773761108455,
+      "learning_rate": 3.887190083703174e-06,
+      "loss": 0.0936,
+      "step": 6421
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.795138637449241,
+      "learning_rate": 3.8857534133562625e-06,
+      "loss": 0.1329,
+      "step": 6422
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.0248952350527865,
+      "learning_rate": 3.884316839803284e-06,
+      "loss": 0.1221,
+      "step": 6423
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7496588609967472,
+      "learning_rate": 3.88288036316903e-06,
+      "loss": 0.1066,
+      "step": 6424
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.753005107062362,
+      "learning_rate": 3.8814439835782895e-06,
+      "loss": 0.112,
+      "step": 6425
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.8922359529202695,
+      "learning_rate": 3.8800077011558354e-06,
+      "loss": 0.1146,
+      "step": 6426
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7295976390726013,
+      "learning_rate": 3.8785715160264435e-06,
+      "loss": 0.0882,
+      "step": 6427
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7230143363248638,
+      "learning_rate": 3.877135428314867e-06,
+      "loss": 0.1074,
+      "step": 6428
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7207681416753564,
+      "learning_rate": 3.875699438145862e-06,
+      "loss": 0.1078,
+      "step": 6429
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6256700685162027,
+      "learning_rate": 3.874263545644172e-06,
+      "loss": 0.0961,
+      "step": 6430
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8171232578098409,
+      "learning_rate": 3.872827750934531e-06,
+      "loss": 0.1172,
+      "step": 6431
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.464700928514557,
+      "learning_rate": 3.871392054141665e-06,
+      "loss": 0.09,
+      "step": 6432
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8859538243111091,
+      "learning_rate": 3.869956455390295e-06,
+      "loss": 0.1136,
+      "step": 6433
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6665449169631579,
+      "learning_rate": 3.868520954805126e-06,
+      "loss": 0.1124,
+      "step": 6434
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.654570204763802,
+      "learning_rate": 3.867085552510865e-06,
+      "loss": 0.0909,
+      "step": 6435
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7954724488798945,
+      "learning_rate": 3.865650248632199e-06,
+      "loss": 0.1161,
+      "step": 6436
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6960803927604093,
+      "learning_rate": 3.864215043293817e-06,
+      "loss": 0.1084,
+      "step": 6437
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8189493126448733,
+      "learning_rate": 3.86277993662039e-06,
+      "loss": 0.119,
+      "step": 6438
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.925060860223729,
+      "learning_rate": 3.861344928736588e-06,
+      "loss": 0.1141,
+      "step": 6439
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6462982336027874,
+      "learning_rate": 3.859910019767065e-06,
+      "loss": 0.0889,
+      "step": 6440
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6880888071064972,
+      "learning_rate": 3.858475209836476e-06,
+      "loss": 0.0965,
+      "step": 6441
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7741557268217825,
+      "learning_rate": 3.8570404990694585e-06,
+      "loss": 0.1078,
+      "step": 6442
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6136132586681615,
+      "learning_rate": 3.855605887590648e-06,
+      "loss": 0.0867,
+      "step": 6443
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.9227031191709991,
+      "learning_rate": 3.854171375524664e-06,
+      "loss": 0.1253,
+      "step": 6444
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.608952264758711,
+      "learning_rate": 3.8527369629961264e-06,
+      "loss": 0.0998,
+      "step": 6445
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.9284906945854983,
+      "learning_rate": 3.851302650129637e-06,
+      "loss": 0.1144,
+      "step": 6446
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.720398041903394,
+      "learning_rate": 3.849868437049799e-06,
+      "loss": 0.0982,
+      "step": 6447
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8307191966404819,
+      "learning_rate": 3.8484343238811976e-06,
+      "loss": 0.1037,
+      "step": 6448
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7194028803150143,
+      "learning_rate": 3.847000310748412e-06,
+      "loss": 0.1026,
+      "step": 6449
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.5147489211052783,
+      "learning_rate": 3.845566397776022e-06,
+      "loss": 0.0872,
+      "step": 6450
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7126656857846014,
+      "learning_rate": 3.844132585088581e-06,
+      "loss": 0.106,
+      "step": 6451
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8425324651866652,
+      "learning_rate": 3.84269887281065e-06,
+      "loss": 0.1265,
+      "step": 6452
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.860898585348654,
+      "learning_rate": 3.8412652610667725e-06,
+      "loss": 0.1082,
+      "step": 6453
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.3912992664421173,
+      "learning_rate": 3.839831749981486e-06,
+      "loss": 0.1411,
+      "step": 6454
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7318135671864818,
+      "learning_rate": 3.838398339679316e-06,
+      "loss": 0.1221,
+      "step": 6455
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7509032358501055,
+      "learning_rate": 3.836965030284788e-06,
+      "loss": 0.1041,
+      "step": 6456
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.5399513053699183,
+      "learning_rate": 3.835531821922405e-06,
+      "loss": 0.0866,
+      "step": 6457
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.660993941393976,
+      "learning_rate": 3.834098714716676e-06,
+      "loss": 0.0935,
+      "step": 6458
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8758154456516754,
+      "learning_rate": 3.83266570879209e-06,
+      "loss": 0.1264,
+      "step": 6459
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.5212660116844818,
+      "learning_rate": 3.831232804273133e-06,
+      "loss": 0.099,
+      "step": 6460
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7630505757765147,
+      "learning_rate": 3.829800001284278e-06,
+      "loss": 0.1005,
+      "step": 6461
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.9531576469215322,
+      "learning_rate": 3.828367299949998e-06,
+      "loss": 0.0996,
+      "step": 6462
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7749718001740593,
+      "learning_rate": 3.826934700394743e-06,
+      "loss": 0.1134,
+      "step": 6463
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6729206379580095,
+      "learning_rate": 3.8255022027429675e-06,
+      "loss": 0.1092,
+      "step": 6464
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.5505312976891268,
+      "learning_rate": 3.8240698071191096e-06,
+      "loss": 0.0948,
+      "step": 6465
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.844643445515912,
+      "learning_rate": 3.822637513647601e-06,
+      "loss": 0.1244,
+      "step": 6466
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.490980276687843,
+      "learning_rate": 3.821205322452863e-06,
+      "loss": 0.0956,
+      "step": 6467
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.7459080697077198,
+      "learning_rate": 3.819773233659314e-06,
+      "loss": 0.0938,
+      "step": 6468
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.8999375104193943,
+      "learning_rate": 3.818341247391351e-06,
+      "loss": 0.1289,
+      "step": 6469
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5190056298261871,
+      "learning_rate": 3.816909363773377e-06,
+      "loss": 0.0855,
+      "step": 6470
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.7734254151293125,
+      "learning_rate": 3.815477582929773e-06,
+      "loss": 0.1021,
+      "step": 6471
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.4988269917553074,
+      "learning_rate": 3.814045904984922e-06,
+      "loss": 0.0935,
+      "step": 6472
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.6231220216540578,
+      "learning_rate": 3.812614330063189e-06,
+      "loss": 0.0937,
+      "step": 6473
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.6387112535914998,
+      "learning_rate": 3.811182858288938e-06,
+      "loss": 0.096,
+      "step": 6474
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.926236321673452,
+      "learning_rate": 3.809751489786515e-06,
+      "loss": 0.1167,
+      "step": 6475
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.7127438060327105,
+      "learning_rate": 3.8083202246802675e-06,
+      "loss": 0.1005,
+      "step": 6476
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.4265618357966356,
+      "learning_rate": 3.8068890630945244e-06,
+      "loss": 0.0672,
+      "step": 6477
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.925076064315851,
+      "learning_rate": 3.8054580051536127e-06,
+      "loss": 0.1018,
+      "step": 6478
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.789362869258968,
+      "learning_rate": 3.8040270509818446e-06,
+      "loss": 0.1226,
+      "step": 6479
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.6415253230207796,
+      "learning_rate": 3.802596200703531e-06,
+      "loss": 0.091,
+      "step": 6480
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.0605143728519746,
+      "learning_rate": 3.8011654544429626e-06,
+      "loss": 0.1268,
+      "step": 6481
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.7230097741804673,
+      "learning_rate": 3.799734812324434e-06,
+      "loss": 0.0952,
+      "step": 6482
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.6546040043062766,
+      "learning_rate": 3.798304274472219e-06,
+      "loss": 0.1003,
+      "step": 6483
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.7718198969168648,
+      "learning_rate": 3.796873841010591e-06,
+      "loss": 0.1297,
+      "step": 6484
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.7966167868562755,
+      "learning_rate": 3.795443512063808e-06,
+      "loss": 0.1041,
+      "step": 6485
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5890620885769258,
+      "learning_rate": 3.794013287756125e-06,
+      "loss": 0.0979,
+      "step": 6486
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.886271351498799,
+      "learning_rate": 3.792583168211782e-06,
+      "loss": 0.1072,
+      "step": 6487
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.616674247202884,
+      "learning_rate": 3.7911531535550145e-06,
+      "loss": 0.0931,
+      "step": 6488
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.9051382378988642,
+      "learning_rate": 3.7897232439100455e-06,
+      "loss": 0.1182,
+      "step": 6489
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5975629571359193,
+      "learning_rate": 3.788293439401093e-06,
+      "loss": 0.0872,
+      "step": 6490
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.8772544938166351,
+      "learning_rate": 3.7868637401523582e-06,
+      "loss": 0.1236,
+      "step": 6491
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.7772421325054895,
+      "learning_rate": 3.785434146288045e-06,
+      "loss": 0.098,
+      "step": 6492
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.6555157819435316,
+      "learning_rate": 3.7840046579323346e-06,
+      "loss": 0.0962,
+      "step": 6493
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5272594060672573,
+      "learning_rate": 3.7825752752094113e-06,
+      "loss": 0.0925,
+      "step": 6494
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5116185690533241,
+      "learning_rate": 3.7811459982434414e-06,
+      "loss": 0.0943,
+      "step": 6495
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.6697508126442953,
+      "learning_rate": 3.779716827158587e-06,
+      "loss": 0.0974,
+      "step": 6496
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.6515327339448527,
+      "learning_rate": 3.7782877620789966e-06,
+      "loss": 0.0985,
+      "step": 6497
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.7202019819050054,
+      "learning_rate": 3.776858803128818e-06,
+      "loss": 0.1089,
+      "step": 6498
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.036772158605855,
+      "learning_rate": 3.775429950432176e-06,
+      "loss": 0.1081,
+      "step": 6499
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.738125713457519,
+      "learning_rate": 3.7740012041132016e-06,
+      "loss": 0.1037,
+      "step": 6500
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.8471002559587975,
+      "learning_rate": 3.7725725642960047e-06,
+      "loss": 0.1102,
+      "step": 6501
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.6684390044862578,
+      "learning_rate": 3.7711440311046928e-06,
+      "loss": 0.1105,
+      "step": 6502
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.7633960570193847,
+      "learning_rate": 3.769715604663358e-06,
+      "loss": 0.104,
+      "step": 6503
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.7255808746356345,
+      "learning_rate": 3.7682872850960933e-06,
+      "loss": 0.1038,
+      "step": 6504
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.7367369773595072,
+      "learning_rate": 3.766859072526969e-06,
+      "loss": 0.0968,
+      "step": 6505
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.923610787255813,
+      "learning_rate": 3.7654309670800575e-06,
+      "loss": 0.1173,
+      "step": 6506
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.014371567413567,
+      "learning_rate": 3.7640029688794155e-06,
+      "loss": 0.1176,
+      "step": 6507
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.6845506914753037,
+      "learning_rate": 3.7625750780490942e-06,
+      "loss": 0.1261,
+      "step": 6508
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.4925341793173412,
+      "learning_rate": 3.761147294713131e-06,
+      "loss": 0.0762,
+      "step": 6509
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.771630727913245,
+      "learning_rate": 3.7597196189955597e-06,
+      "loss": 0.1035,
+      "step": 6510
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.8120225083224282,
+      "learning_rate": 3.7582920510203976e-06,
+      "loss": 0.1178,
+      "step": 6511
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5605328304971582,
+      "learning_rate": 3.7568645909116608e-06,
+      "loss": 0.1069,
+      "step": 6512
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.960781930147555,
+      "learning_rate": 3.755437238793349e-06,
+      "loss": 0.1126,
+      "step": 6513
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.548207826420507,
+      "learning_rate": 3.7540099947894576e-06,
+      "loss": 0.1023,
+      "step": 6514
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5802087678438415,
+      "learning_rate": 3.752582859023968e-06,
+      "loss": 0.0942,
+      "step": 6515
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.6094298276627206,
+      "learning_rate": 3.751155831620858e-06,
+      "loss": 0.1077,
+      "step": 6516
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5037657871299246,
+      "learning_rate": 3.7497289127040882e-06,
+      "loss": 0.091,
+      "step": 6517
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.8227546607597456,
+      "learning_rate": 3.748302102397618e-06,
+      "loss": 0.1205,
+      "step": 6518
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.9321856952072975,
+      "learning_rate": 3.7468754008253915e-06,
+      "loss": 0.1338,
+      "step": 6519
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5676908198961796,
+      "learning_rate": 3.7454488081113473e-06,
+      "loss": 0.0866,
+      "step": 6520
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5494013831523572,
+      "learning_rate": 3.7440223243794095e-06,
+      "loss": 0.1098,
+      "step": 6521
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5999213286153722,
+      "learning_rate": 3.7425959497534997e-06,
+      "loss": 0.1094,
+      "step": 6522
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.6183921382450126,
+      "learning_rate": 3.741169684357522e-06,
+      "loss": 0.106,
+      "step": 6523
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.6362738729307476,
+      "learning_rate": 3.7397435283153795e-06,
+      "loss": 0.1035,
+      "step": 6524
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.751160447794452,
+      "learning_rate": 3.7383174817509583e-06,
+      "loss": 0.0996,
+      "step": 6525
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.067889380664673,
+      "learning_rate": 3.7368915447881404e-06,
+      "loss": 0.1359,
+      "step": 6526
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.9671021991120594,
+      "learning_rate": 3.7354657175507947e-06,
+      "loss": 0.1298,
+      "step": 6527
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.9459488418555648,
+      "learning_rate": 3.7340400001627832e-06,
+      "loss": 0.1276,
+      "step": 6528
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.8097339360372815,
+      "learning_rate": 3.732614392747954e-06,
+      "loss": 0.1138,
+      "step": 6529
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.9262320793680647,
+      "learning_rate": 3.7311888954301534e-06,
+      "loss": 0.1298,
+      "step": 6530
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.602407186642757,
+      "learning_rate": 3.7297635083332097e-06,
+      "loss": 0.1134,
+      "step": 6531
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.4676414332743384,
+      "learning_rate": 3.728338231580948e-06,
+      "loss": 0.0989,
+      "step": 6532
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.851882378144772,
+      "learning_rate": 3.7269130652971787e-06,
+      "loss": 0.1036,
+      "step": 6533
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.556133849475282,
+      "learning_rate": 3.725488009605708e-06,
+      "loss": 0.0962,
+      "step": 6534
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.894602342366671,
+      "learning_rate": 3.7240630646303262e-06,
+      "loss": 0.1247,
+      "step": 6535
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.6930784951732443,
+      "learning_rate": 3.7226382304948215e-06,
+      "loss": 0.0863,
+      "step": 6536
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.7036900265105728,
+      "learning_rate": 3.721213507322965e-06,
+      "loss": 0.116,
+      "step": 6537
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.7029355240497255,
+      "learning_rate": 3.7197888952385236e-06,
+      "loss": 0.1019,
+      "step": 6538
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.6303880208432018,
+      "learning_rate": 3.7183643943652513e-06,
+      "loss": 0.0951,
+      "step": 6539
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7987873872603957,
+      "learning_rate": 3.7169400048268945e-06,
+      "loss": 0.1199,
+      "step": 6540
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7566828248635518,
+      "learning_rate": 3.7155157267471863e-06,
+      "loss": 0.1067,
+      "step": 6541
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.9579602180485876,
+      "learning_rate": 3.7140915602498574e-06,
+      "loss": 0.1112,
+      "step": 6542
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.733989837374824,
+      "learning_rate": 3.712667505458622e-06,
+      "loss": 0.1213,
+      "step": 6543
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.6021641639621227,
+      "learning_rate": 3.7112435624971855e-06,
+      "loss": 0.091,
+      "step": 6544
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.5399654835767467,
+      "learning_rate": 3.7098197314892493e-06,
+      "loss": 0.0853,
+      "step": 6545
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.5430774342072628,
+      "learning_rate": 3.7083960125584944e-06,
+      "loss": 0.0906,
+      "step": 6546
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.8571089976689217,
+      "learning_rate": 3.7069724058286045e-06,
+      "loss": 0.1092,
+      "step": 6547
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.498037968860935,
+      "learning_rate": 3.7055489114232433e-06,
+      "loss": 0.0996,
+      "step": 6548
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.6837886361606174,
+      "learning_rate": 3.7041255294660723e-06,
+      "loss": 0.1008,
+      "step": 6549
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.6798957014764648,
+      "learning_rate": 3.702702260080735e-06,
+      "loss": 0.0995,
+      "step": 6550
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.9628604863377934,
+      "learning_rate": 3.7012791033908766e-06,
+      "loss": 0.1125,
+      "step": 6551
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.4901768766104952,
+      "learning_rate": 3.6998560595201188e-06,
+      "loss": 0.0963,
+      "step": 6552
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.8730088759216363,
+      "learning_rate": 3.698433128592086e-06,
+      "loss": 0.1094,
+      "step": 6553
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7366970038812257,
+      "learning_rate": 3.6970103107303845e-06,
+      "loss": 0.1126,
+      "step": 6554
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.6866143118892862,
+      "learning_rate": 3.695587606058616e-06,
+      "loss": 0.1044,
+      "step": 6555
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.6264082540919602,
+      "learning_rate": 3.6941650147003655e-06,
+      "loss": 0.0993,
+      "step": 6556
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.832193638718385,
+      "learning_rate": 3.692742536779219e-06,
+      "loss": 0.1214,
+      "step": 6557
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7615475855151723,
+      "learning_rate": 3.6913201724187397e-06,
+      "loss": 0.0921,
+      "step": 6558
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.6281581195082457,
+      "learning_rate": 3.6898979217424934e-06,
+      "loss": 0.0949,
+      "step": 6559
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.5124977547077132,
+      "learning_rate": 3.688475784874026e-06,
+      "loss": 0.09,
+      "step": 6560
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.6343000926567257,
+      "learning_rate": 3.68705376193688e-06,
+      "loss": 0.097,
+      "step": 6561
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.643227847149309,
+      "learning_rate": 3.685631853054583e-06,
+      "loss": 0.0934,
+      "step": 6562
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.8801033198210357,
+      "learning_rate": 3.6842100583506607e-06,
+      "loss": 0.1252,
+      "step": 6563
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.8324821365752662,
+      "learning_rate": 3.682788377948617e-06,
+      "loss": 0.1105,
+      "step": 6564
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.756068432387505,
+      "learning_rate": 3.681366811971957e-06,
+      "loss": 0.1207,
+      "step": 6565
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.8340457712239377,
+      "learning_rate": 3.6799453605441695e-06,
+      "loss": 0.1221,
+      "step": 6566
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.190789473925438,
+      "learning_rate": 3.6785240237887355e-06,
+      "loss": 0.0958,
+      "step": 6567
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.8477917898717289,
+      "learning_rate": 3.6771028018291244e-06,
+      "loss": 0.1053,
+      "step": 6568
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.8543499557881225,
+      "learning_rate": 3.675681694788801e-06,
+      "loss": 0.1084,
+      "step": 6569
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7149131192807923,
+      "learning_rate": 3.6742607027912093e-06,
+      "loss": 0.1064,
+      "step": 6570
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7968793843237767,
+      "learning_rate": 3.6728398259597965e-06,
+      "loss": 0.1198,
+      "step": 6571
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.5996142133101694,
+      "learning_rate": 3.6714190644179894e-06,
+      "loss": 0.0902,
+      "step": 6572
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.8663270939451262,
+      "learning_rate": 3.6699984182892113e-06,
+      "loss": 0.1189,
+      "step": 6573
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.763011937746098,
+      "learning_rate": 3.668577887696869e-06,
+      "loss": 0.1045,
+      "step": 6574
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7683328939236824,
+      "learning_rate": 3.6671574727643694e-06,
+      "loss": 0.1124,
+      "step": 6575
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.7431740139869394,
+      "learning_rate": 3.665737173615096e-06,
+      "loss": 0.1235,
+      "step": 6576
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.9963003822918417,
+      "learning_rate": 3.664316990372434e-06,
+      "loss": 0.0999,
+      "step": 6577
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6112830497399264,
+      "learning_rate": 3.662896923159752e-06,
+      "loss": 0.085,
+      "step": 6578
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.4831875313805127,
+      "learning_rate": 3.6614769721004127e-06,
+      "loss": 0.087,
+      "step": 6579
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7635098423897353,
+      "learning_rate": 3.6600571373177616e-06,
+      "loss": 0.11,
+      "step": 6580
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.9707760650085377,
+      "learning_rate": 3.658637418935146e-06,
+      "loss": 0.1207,
+      "step": 6581
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.9438004794542514,
+      "learning_rate": 3.6572178170758874e-06,
+      "loss": 0.1049,
+      "step": 6582
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.5553485003266168,
+      "learning_rate": 3.655798331863314e-06,
+      "loss": 0.075,
+      "step": 6583
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.550515940894469,
+      "learning_rate": 3.65437896342073e-06,
+      "loss": 0.0997,
+      "step": 6584
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6332365949418675,
+      "learning_rate": 3.6529597118714377e-06,
+      "loss": 0.0984,
+      "step": 6585
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.729818947015424,
+      "learning_rate": 3.6515405773387257e-06,
+      "loss": 0.1157,
+      "step": 6586
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.9000652838205647,
+      "learning_rate": 3.650121559945874e-06,
+      "loss": 0.1368,
+      "step": 6587
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7485175813389107,
+      "learning_rate": 3.648702659816149e-06,
+      "loss": 0.1219,
+      "step": 6588
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.8980224493213316,
+      "learning_rate": 3.647283877072815e-06,
+      "loss": 0.1257,
+      "step": 6589
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.935666943375356,
+      "learning_rate": 3.6458652118391164e-06,
+      "loss": 0.1325,
+      "step": 6590
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6782058162610771,
+      "learning_rate": 3.644446664238294e-06,
+      "loss": 0.1089,
+      "step": 6591
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7103254299325488,
+      "learning_rate": 3.6430282343935754e-06,
+      "loss": 0.1062,
+      "step": 6592
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.69967685470295,
+      "learning_rate": 3.6416099224281787e-06,
+      "loss": 0.1137,
+      "step": 6593
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.5088365327281856,
+      "learning_rate": 3.64019172846531e-06,
+      "loss": 0.0913,
+      "step": 6594
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7313580734523966,
+      "learning_rate": 3.6387736526281714e-06,
+      "loss": 0.1135,
+      "step": 6595
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6469220829331832,
+      "learning_rate": 3.637355695039947e-06,
+      "loss": 0.1023,
+      "step": 6596
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6179128832182075,
+      "learning_rate": 3.6359378558238145e-06,
+      "loss": 0.107,
+      "step": 6597
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.9003522616355872,
+      "learning_rate": 3.634520135102941e-06,
+      "loss": 0.1149,
+      "step": 6598
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7485551759034508,
+      "learning_rate": 3.6331025330004834e-06,
+      "loss": 0.1171,
+      "step": 6599
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7077247462811909,
+      "learning_rate": 3.6316850496395863e-06,
+      "loss": 0.1165,
+      "step": 6600
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.8310493916699688,
+      "learning_rate": 3.630267685143388e-06,
+      "loss": 0.1227,
+      "step": 6601
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7876844188489902,
+      "learning_rate": 3.628850439635012e-06,
+      "loss": 0.1169,
+      "step": 6602
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.9918548636072828,
+      "learning_rate": 3.627433313237576e-06,
+      "loss": 0.1083,
+      "step": 6603
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6044377292332295,
+      "learning_rate": 3.6260163060741816e-06,
+      "loss": 0.1004,
+      "step": 6604
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.5365220455039013,
+      "learning_rate": 3.624599418267927e-06,
+      "loss": 0.106,
+      "step": 6605
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6148062073085345,
+      "learning_rate": 3.623182649941892e-06,
+      "loss": 0.1016,
+      "step": 6606
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.719216853740819,
+      "learning_rate": 3.621766001219156e-06,
+      "loss": 0.1117,
+      "step": 6607
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.4723967861150682,
+      "learning_rate": 3.620349472222777e-06,
+      "loss": 0.0869,
+      "step": 6608
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7147562278307946,
+      "learning_rate": 3.6189330630758124e-06,
+      "loss": 0.116,
+      "step": 6609
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.5904777964315342,
+      "learning_rate": 3.6175167739013018e-06,
+      "loss": 0.1086,
+      "step": 6610
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7210591185891906,
+      "learning_rate": 3.616100604822279e-06,
+      "loss": 0.1074,
+      "step": 6611
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6278747572304295,
+      "learning_rate": 3.6146845559617634e-06,
+      "loss": 0.1015,
+      "step": 6612
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.7660752873198828,
+      "learning_rate": 3.6132686274427695e-06,
+      "loss": 0.1112,
+      "step": 6613
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6511927942425266,
+      "learning_rate": 3.6118528193882974e-06,
+      "loss": 0.0967,
+      "step": 6614
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.4793212175717843,
+      "learning_rate": 3.610437131921336e-06,
+      "loss": 0.084,
+      "step": 6615
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6749803063431985,
+      "learning_rate": 3.6090215651648664e-06,
+      "loss": 0.0923,
+      "step": 6616
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6607795175037037,
+      "learning_rate": 3.6076061192418582e-06,
+      "loss": 0.0954,
+      "step": 6617
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6957107538803322,
+      "learning_rate": 3.6061907942752677e-06,
+      "loss": 0.0998,
+      "step": 6618
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6934360390074377,
+      "learning_rate": 3.6047755903880478e-06,
+      "loss": 0.1052,
+      "step": 6619
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.8880189780865968,
+      "learning_rate": 3.603360507703133e-06,
+      "loss": 0.1146,
+      "step": 6620
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.5854975026932092,
+      "learning_rate": 3.601945546343453e-06,
+      "loss": 0.1057,
+      "step": 6621
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.8044575713500113,
+      "learning_rate": 3.600530706431922e-06,
+      "loss": 0.1092,
+      "step": 6622
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.7912558113702912,
+      "learning_rate": 3.599115988091449e-06,
+      "loss": 0.1095,
+      "step": 6623
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6965103998930002,
+      "learning_rate": 3.5977013914449264e-06,
+      "loss": 0.1094,
+      "step": 6624
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.744635719859434,
+      "learning_rate": 3.596286916615244e-06,
+      "loss": 0.1047,
+      "step": 6625
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.5825141605672404,
+      "learning_rate": 3.5948725637252713e-06,
+      "loss": 0.0978,
+      "step": 6626
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.8310636951831805,
+      "learning_rate": 3.5934583328978766e-06,
+      "loss": 0.1228,
+      "step": 6627
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.5778545050024169,
+      "learning_rate": 3.5920442242559107e-06,
+      "loss": 0.0991,
+      "step": 6628
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6525629357444294,
+      "learning_rate": 3.590630237922218e-06,
+      "loss": 0.101,
+      "step": 6629
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.4814570503972715,
+      "learning_rate": 3.5892163740196272e-06,
+      "loss": 0.0959,
+      "step": 6630
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.8054704239368022,
+      "learning_rate": 3.587802632670965e-06,
+      "loss": 0.1148,
+      "step": 6631
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6428554507862272,
+      "learning_rate": 3.586389013999039e-06,
+      "loss": 0.0906,
+      "step": 6632
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.507040253721973,
+      "learning_rate": 3.584975518126648e-06,
+      "loss": 0.0893,
+      "step": 6633
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.65860531247142,
+      "learning_rate": 3.5835621451765866e-06,
+      "loss": 0.116,
+      "step": 6634
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.4872547970542909,
+      "learning_rate": 3.5821488952716286e-06,
+      "loss": 0.0844,
+      "step": 6635
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.8842555276067297,
+      "learning_rate": 3.5807357685345456e-06,
+      "loss": 0.1347,
+      "step": 6636
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.896502281290437,
+      "learning_rate": 3.5793227650880928e-06,
+      "loss": 0.1314,
+      "step": 6637
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6397942078319327,
+      "learning_rate": 3.577909885055019e-06,
+      "loss": 0.1094,
+      "step": 6638
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6323810148116837,
+      "learning_rate": 3.576497128558057e-06,
+      "loss": 0.0978,
+      "step": 6639
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.721832034126878,
+      "learning_rate": 3.575084495719937e-06,
+      "loss": 0.1122,
+      "step": 6640
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6700910304015377,
+      "learning_rate": 3.573671986663368e-06,
+      "loss": 0.103,
+      "step": 6641
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.524416676418242,
+      "learning_rate": 3.572259601511058e-06,
+      "loss": 0.0963,
+      "step": 6642
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.9774518984208158,
+      "learning_rate": 3.570847340385698e-06,
+      "loss": 0.1222,
+      "step": 6643
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.3925679723058886,
+      "learning_rate": 3.569435203409972e-06,
+      "loss": 0.0773,
+      "step": 6644
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.064218949827412,
+      "learning_rate": 3.5680231907065487e-06,
+      "loss": 0.1123,
+      "step": 6645
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6749895464489748,
+      "learning_rate": 3.566611302398093e-06,
+      "loss": 0.1004,
+      "step": 6646
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.8976003279843356,
+      "learning_rate": 3.565199538607249e-06,
+      "loss": 0.1296,
+      "step": 6647
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.9483420294138387,
+      "learning_rate": 3.5637878994566616e-06,
+      "loss": 0.1182,
+      "step": 6648
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8236215575821053,
+      "learning_rate": 3.562376385068955e-06,
+      "loss": 0.1196,
+      "step": 6649
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.9116193902494234,
+      "learning_rate": 3.560964995566749e-06,
+      "loss": 0.1234,
+      "step": 6650
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.942537623052391,
+      "learning_rate": 3.559553731072648e-06,
+      "loss": 0.1004,
+      "step": 6651
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.9468654041217217,
+      "learning_rate": 3.5581425917092515e-06,
+      "loss": 0.1177,
+      "step": 6652
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.5204536546123528,
+      "learning_rate": 3.5567315775991384e-06,
+      "loss": 0.0888,
+      "step": 6653
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.569417922314605,
+      "learning_rate": 3.555320688864889e-06,
+      "loss": 0.0987,
+      "step": 6654
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.449190496989596,
+      "learning_rate": 3.5539099256290616e-06,
+      "loss": 0.0876,
+      "step": 6655
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.610140858560358,
+      "learning_rate": 3.5524992880142118e-06,
+      "loss": 0.1037,
+      "step": 6656
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.497627545502157,
+      "learning_rate": 3.5510887761428764e-06,
+      "loss": 0.0956,
+      "step": 6657
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8890554553883838,
+      "learning_rate": 3.549678390137592e-06,
+      "loss": 0.1276,
+      "step": 6658
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.1464939030328303,
+      "learning_rate": 3.548268130120871e-06,
+      "loss": 0.1447,
+      "step": 6659
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8510713925067024,
+      "learning_rate": 3.5468579962152272e-06,
+      "loss": 0.1277,
+      "step": 6660
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.6886607335146475,
+      "learning_rate": 3.545447988543156e-06,
+      "loss": 0.1196,
+      "step": 6661
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.6459459642977876,
+      "learning_rate": 3.5440381072271447e-06,
+      "loss": 0.0989,
+      "step": 6662
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.6308313115987418,
+      "learning_rate": 3.5426283523896675e-06,
+      "loss": 0.1003,
+      "step": 6663
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.714758082652541,
+      "learning_rate": 3.5412187241531904e-06,
+      "loss": 0.1053,
+      "step": 6664
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8486763781642428,
+      "learning_rate": 3.5398092226401644e-06,
+      "loss": 0.1256,
+      "step": 6665
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.7317898142549648,
+      "learning_rate": 3.5383998479730357e-06,
+      "loss": 0.1101,
+      "step": 6666
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.6293508911559902,
+      "learning_rate": 3.5369906002742332e-06,
+      "loss": 0.1054,
+      "step": 6667
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.502620274140018,
+      "learning_rate": 3.535581479666179e-06,
+      "loss": 0.0952,
+      "step": 6668
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8571531643163228,
+      "learning_rate": 3.5341724862712817e-06,
+      "loss": 0.1194,
+      "step": 6669
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.529365441293258,
+      "learning_rate": 3.5327636202119404e-06,
+      "loss": 0.1022,
+      "step": 6670
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.5997098329975605,
+      "learning_rate": 3.531354881610539e-06,
+      "loss": 0.132,
+      "step": 6671
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.0634077512790037,
+      "learning_rate": 3.5299462705894598e-06,
+      "loss": 0.1277,
+      "step": 6672
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.712346021264327,
+      "learning_rate": 3.5285377872710634e-06,
+      "loss": 0.1109,
+      "step": 6673
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8082908928594958,
+      "learning_rate": 3.5271294317777065e-06,
+      "loss": 0.1143,
+      "step": 6674
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.7496781001347508,
+      "learning_rate": 3.5257212042317302e-06,
+      "loss": 0.1019,
+      "step": 6675
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8523397397966093,
+      "learning_rate": 3.524313104755468e-06,
+      "loss": 0.1016,
+      "step": 6676
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8880953859232175,
+      "learning_rate": 3.522905133471237e-06,
+      "loss": 0.1141,
+      "step": 6677
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.7093372756155623,
+      "learning_rate": 3.5214972905013522e-06,
+      "loss": 0.1093,
+      "step": 6678
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.7808568283203687,
+      "learning_rate": 3.5200895759681086e-06,
+      "loss": 0.1163,
+      "step": 6679
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.6069185492716866,
+      "learning_rate": 3.518681989993795e-06,
+      "loss": 0.0817,
+      "step": 6680
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.7441918794042037,
+      "learning_rate": 3.517274532700686e-06,
+      "loss": 0.1067,
+      "step": 6681
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.7518825389278112,
+      "learning_rate": 3.5158672042110485e-06,
+      "loss": 0.1147,
+      "step": 6682
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.6563856845623466,
+      "learning_rate": 3.5144600046471338e-06,
+      "loss": 0.0844,
+      "step": 6683
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.72670655235064,
+      "learning_rate": 3.513052934131188e-06,
+      "loss": 0.0914,
+      "step": 6684
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.8156312178897744,
+      "learning_rate": 3.5116459927854383e-06,
+      "loss": 0.1162,
+      "step": 6685
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.6438493316017246,
+      "learning_rate": 3.510239180732109e-06,
+      "loss": 0.1042,
+      "step": 6686
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.8100702720265167,
+      "learning_rate": 3.5088324980934063e-06,
+      "loss": 0.1193,
+      "step": 6687
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.1870433129210727,
+      "learning_rate": 3.507425944991529e-06,
+      "loss": 0.1472,
+      "step": 6688
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.723683519529843,
+      "learning_rate": 3.506019521548661e-06,
+      "loss": 0.1187,
+      "step": 6689
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.7641032390605524,
+      "learning_rate": 3.5046132278869817e-06,
+      "loss": 0.1153,
+      "step": 6690
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.7268412598914904,
+      "learning_rate": 3.503207064128652e-06,
+      "loss": 0.1052,
+      "step": 6691
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.819943604247568,
+      "learning_rate": 3.501801030395826e-06,
+      "loss": 0.1065,
+      "step": 6692
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.5897525342271133,
+      "learning_rate": 3.5003951268106434e-06,
+      "loss": 0.1057,
+      "step": 6693
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.6509375702566622,
+      "learning_rate": 3.498989353495236e-06,
+      "loss": 0.1103,
+      "step": 6694
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.517911308973224,
+      "learning_rate": 3.4975837105717203e-06,
+      "loss": 0.0984,
+      "step": 6695
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.4985240952344228,
+      "learning_rate": 3.496178198162207e-06,
+      "loss": 0.0952,
+      "step": 6696
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.7303422948099,
+      "learning_rate": 3.4947728163887886e-06,
+      "loss": 0.1083,
+      "step": 6697
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.5751225801727782,
+      "learning_rate": 3.493367565373552e-06,
+      "loss": 0.0896,
+      "step": 6698
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.713480744981879,
+      "learning_rate": 3.491962445238569e-06,
+      "loss": 0.108,
+      "step": 6699
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.8205060359609513,
+      "learning_rate": 3.490557456105904e-06,
+      "loss": 0.1178,
+      "step": 6700
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.8315003704488404,
+      "learning_rate": 3.4891525980976034e-06,
+      "loss": 0.0991,
+      "step": 6701
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.5252380469899487,
+      "learning_rate": 3.4877478713357103e-06,
+      "loss": 0.0875,
+      "step": 6702
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.567304591782693,
+      "learning_rate": 3.4863432759422512e-06,
+      "loss": 0.1016,
+      "step": 6703
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.6834994019410636,
+      "learning_rate": 3.4849388120392422e-06,
+      "loss": 0.1112,
+      "step": 6704
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.6498090709397835,
+      "learning_rate": 3.483534479748688e-06,
+      "loss": 0.1102,
+      "step": 6705
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.7217256942227208,
+      "learning_rate": 3.482130279192584e-06,
+      "loss": 0.1125,
+      "step": 6706
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.6065673306333461,
+      "learning_rate": 3.4807262104929075e-06,
+      "loss": 0.0936,
+      "step": 6707
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.6695670649788883,
+      "learning_rate": 3.479322273771635e-06,
+      "loss": 0.108,
+      "step": 6708
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.7062090325701962,
+      "learning_rate": 3.4779184691507216e-06,
+      "loss": 0.0953,
+      "step": 6709
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.7483330215077921,
+      "learning_rate": 3.4765147967521174e-06,
+      "loss": 0.1201,
+      "step": 6710
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.7602761782172296,
+      "learning_rate": 3.4751112566977563e-06,
+      "loss": 0.1014,
+      "step": 6711
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.8437897018666467,
+      "learning_rate": 3.4737078491095657e-06,
+      "loss": 0.1259,
+      "step": 6712
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.9879516369433305,
+      "learning_rate": 3.4723045741094545e-06,
+      "loss": 0.1167,
+      "step": 6713
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.761322537900197,
+      "learning_rate": 3.4709014318193298e-06,
+      "loss": 0.0986,
+      "step": 6714
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.7758254462963727,
+      "learning_rate": 3.4694984223610774e-06,
+      "loss": 0.105,
+      "step": 6715
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.9740879121199117,
+      "learning_rate": 3.468095545856579e-06,
+      "loss": 0.1246,
+      "step": 6716
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.7970209449196572,
+      "learning_rate": 3.4666928024276993e-06,
+      "loss": 0.0848,
+      "step": 6717
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.5823427681698838,
+      "learning_rate": 3.4652901921962945e-06,
+      "loss": 0.1068,
+      "step": 6718
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.649645613739333,
+      "learning_rate": 3.4638877152842075e-06,
+      "loss": 0.0922,
+      "step": 6719
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.7807371026905454,
+      "learning_rate": 3.462485371813274e-06,
+      "loss": 0.0992,
+      "step": 6720
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.601764657644291,
+      "learning_rate": 3.461083161905311e-06,
+      "loss": 0.095,
+      "step": 6721
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.7156968621413224,
+      "learning_rate": 3.4596810856821304e-06,
+      "loss": 0.104,
+      "step": 6722
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.7658327665849918,
+      "learning_rate": 3.4582791432655273e-06,
+      "loss": 0.1151,
+      "step": 6723
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5504247168858252,
+      "learning_rate": 3.45687733477729e-06,
+      "loss": 0.0883,
+      "step": 6724
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6160233594553108,
+      "learning_rate": 3.4554756603391893e-06,
+      "loss": 0.0916,
+      "step": 6725
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.8722890110879344,
+      "learning_rate": 3.4540741200729903e-06,
+      "loss": 0.1122,
+      "step": 6726
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.7485686337377087,
+      "learning_rate": 3.4526727141004457e-06,
+      "loss": 0.1199,
+      "step": 6727
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.880307211237019,
+      "learning_rate": 3.45127144254329e-06,
+      "loss": 0.1216,
+      "step": 6728
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.34255137706671,
+      "learning_rate": 3.4498703055232575e-06,
+      "loss": 0.0848,
+      "step": 6729
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.677612511180962,
+      "learning_rate": 3.4484693031620563e-06,
+      "loss": 0.1051,
+      "step": 6730
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.7683846101717453,
+      "learning_rate": 3.447068435581398e-06,
+      "loss": 0.0975,
+      "step": 6731
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6382726809643489,
+      "learning_rate": 3.4456677029029687e-06,
+      "loss": 0.0961,
+      "step": 6732
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5388093681534645,
+      "learning_rate": 3.4442671052484545e-06,
+      "loss": 0.0839,
+      "step": 6733
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.1401575872429013,
+      "learning_rate": 3.4428666427395195e-06,
+      "loss": 0.1392,
+      "step": 6734
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5660715220995813,
+      "learning_rate": 3.441466315497828e-06,
+      "loss": 0.1029,
+      "step": 6735
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.9656216287487835,
+      "learning_rate": 3.440066123645017e-06,
+      "loss": 0.1153,
+      "step": 6736
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6102068054946224,
+      "learning_rate": 3.4386660673027267e-06,
+      "loss": 0.0984,
+      "step": 6737
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.824043807478563,
+      "learning_rate": 3.437266146592576e-06,
+      "loss": 0.1075,
+      "step": 6738
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5960429365643805,
+      "learning_rate": 3.4358663616361775e-06,
+      "loss": 0.0987,
+      "step": 6739
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.7422951708239955,
+      "learning_rate": 3.434466712555128e-06,
+      "loss": 0.0992,
+      "step": 6740
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.527141800737443,
+      "learning_rate": 3.433067199471015e-06,
+      "loss": 0.0958,
+      "step": 6741
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.8840447194428962,
+      "learning_rate": 3.4316678225054106e-06,
+      "loss": 0.1184,
+      "step": 6742
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.7874758430607982,
+      "learning_rate": 3.430268581779883e-06,
+      "loss": 0.1135,
+      "step": 6743
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6386548939184866,
+      "learning_rate": 3.428869477415979e-06,
+      "loss": 0.1063,
+      "step": 6744
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6393504532572505,
+      "learning_rate": 3.427470509535241e-06,
+      "loss": 0.0927,
+      "step": 6745
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5403534182437106,
+      "learning_rate": 3.4260716782591934e-06,
+      "loss": 0.0864,
+      "step": 6746
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6084930617423232,
+      "learning_rate": 3.424672983709355e-06,
+      "loss": 0.1006,
+      "step": 6747
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.730454347575827,
+      "learning_rate": 3.423274426007226e-06,
+      "loss": 0.0989,
+      "step": 6748
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.788461589945982,
+      "learning_rate": 3.4218760052743018e-06,
+      "loss": 0.1162,
+      "step": 6749
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.676860688678788,
+      "learning_rate": 3.4204777216320607e-06,
+      "loss": 0.1047,
+      "step": 6750
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6787575061921869,
+      "learning_rate": 3.4190795752019713e-06,
+      "loss": 0.0912,
+      "step": 6751
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6940592382887492,
+      "learning_rate": 3.4176815661054884e-06,
+      "loss": 0.1123,
+      "step": 6752
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6587299778761608,
+      "learning_rate": 3.416283694464058e-06,
+      "loss": 0.1064,
+      "step": 6753
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5874451683595527,
+      "learning_rate": 3.41488596039911e-06,
+      "loss": 0.1006,
+      "step": 6754
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5552910395266073,
+      "learning_rate": 3.413488364032068e-06,
+      "loss": 0.0989,
+      "step": 6755
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6974333100605856,
+      "learning_rate": 3.4120909054843375e-06,
+      "loss": 0.1116,
+      "step": 6756
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6456892862782926,
+      "learning_rate": 3.410693584877317e-06,
+      "loss": 0.0986,
+      "step": 6757
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.7511445853784662,
+      "learning_rate": 3.4092964023323893e-06,
+      "loss": 0.1107,
+      "step": 6758
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.5106536993423596,
+      "learning_rate": 3.4078993579709286e-06,
+      "loss": 0.0871,
+      "step": 6759
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7754188949103504,
+      "learning_rate": 3.406502451914292e-06,
+      "loss": 0.1148,
+      "step": 6760
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.825775282221653,
+      "learning_rate": 3.4051056842838315e-06,
+      "loss": 0.1201,
+      "step": 6761
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7339885427687431,
+      "learning_rate": 3.403709055200881e-06,
+      "loss": 0.1053,
+      "step": 6762
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.6463951494161604,
+      "learning_rate": 3.4023125647867673e-06,
+      "loss": 0.102,
+      "step": 6763
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.3877564016938142,
+      "learning_rate": 3.4009162131628e-06,
+      "loss": 0.0856,
+      "step": 6764
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7386206191211375,
+      "learning_rate": 3.3995200004502814e-06,
+      "loss": 0.106,
+      "step": 6765
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.9656627117902095,
+      "learning_rate": 3.398123926770497e-06,
+      "loss": 0.1283,
+      "step": 6766
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.6283524778107452,
+      "learning_rate": 3.396727992244726e-06,
+      "loss": 0.1059,
+      "step": 6767
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.6808075951216175,
+      "learning_rate": 3.395332196994231e-06,
+      "loss": 0.1034,
+      "step": 6768
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7202104174475907,
+      "learning_rate": 3.393936541140264e-06,
+      "loss": 0.1025,
+      "step": 6769
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.674657219966489,
+      "learning_rate": 3.3925410248040645e-06,
+      "loss": 0.0938,
+      "step": 6770
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.9521902291646216,
+      "learning_rate": 3.3911456481068613e-06,
+      "loss": 0.1305,
+      "step": 6771
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7114943318200078,
+      "learning_rate": 3.3897504111698665e-06,
+      "loss": 0.1096,
+      "step": 6772
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.6001913104656171,
+      "learning_rate": 3.3883553141142884e-06,
+      "loss": 0.1043,
+      "step": 6773
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.5947898853341833,
+      "learning_rate": 3.386960357061315e-06,
+      "loss": 0.0863,
+      "step": 6774
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.8351455715428466,
+      "learning_rate": 3.3855655401321267e-06,
+      "loss": 0.1116,
+      "step": 6775
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.8376204098320752,
+      "learning_rate": 3.38417086344789e-06,
+      "loss": 0.1259,
+      "step": 6776
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.8467105437924536,
+      "learning_rate": 3.3827763271297598e-06,
+      "loss": 0.1138,
+      "step": 6777
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7156743946748028,
+      "learning_rate": 3.381381931298876e-06,
+      "loss": 0.1122,
+      "step": 6778
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.6827558133408522,
+      "learning_rate": 3.379987676076374e-06,
+      "loss": 0.1089,
+      "step": 6779
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.725544625964348,
+      "learning_rate": 3.378593561583368e-06,
+      "loss": 0.1059,
+      "step": 6780
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7748463688601241,
+      "learning_rate": 3.3771995879409663e-06,
+      "loss": 0.114,
+      "step": 6781
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.604328116183872,
+      "learning_rate": 3.3758057552702604e-06,
+      "loss": 0.0968,
+      "step": 6782
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.5000926943483102,
+      "learning_rate": 3.374412063692334e-06,
+      "loss": 0.0887,
+      "step": 6783
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.6050507388338577,
+      "learning_rate": 3.3730185133282522e-06,
+      "loss": 0.0843,
+      "step": 6784
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.481865550066402,
+      "learning_rate": 3.3716251042990772e-06,
+      "loss": 0.0858,
+      "step": 6785
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.4953701854721975,
+      "learning_rate": 3.3702318367258503e-06,
+      "loss": 0.0803,
+      "step": 6786
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.6274610969114172,
+      "learning_rate": 3.368838710729605e-06,
+      "loss": 0.1098,
+      "step": 6787
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7689381443687886,
+      "learning_rate": 3.36744572643136e-06,
+      "loss": 0.1238,
+      "step": 6788
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.658357837517485,
+      "learning_rate": 3.3660528839521245e-06,
+      "loss": 0.1002,
+      "step": 6789
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7680712754961638,
+      "learning_rate": 3.3646601834128924e-06,
+      "loss": 0.1287,
+      "step": 6790
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.6905441531854648,
+      "learning_rate": 3.3632676249346487e-06,
+      "loss": 0.096,
+      "step": 6791
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.496392918270836,
+      "learning_rate": 3.361875208638362e-06,
+      "loss": 0.0842,
+      "step": 6792
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.9876049722721103,
+      "learning_rate": 3.360482934644993e-06,
+      "loss": 0.1219,
+      "step": 6793
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.706428525564243,
+      "learning_rate": 3.3590908030754854e-06,
+      "loss": 0.1087,
+      "step": 6794
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.7055572553450906,
+      "learning_rate": 3.3576988140507747e-06,
+      "loss": 0.1154,
+      "step": 6795
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.7494640154157222,
+      "learning_rate": 3.3563069676917798e-06,
+      "loss": 0.1033,
+      "step": 6796
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.0180819105617163,
+      "learning_rate": 3.3549152641194127e-06,
+      "loss": 0.116,
+      "step": 6797
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.8299503947169884,
+      "learning_rate": 3.3535237034545677e-06,
+      "loss": 0.0998,
+      "step": 6798
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.507241111723557,
+      "learning_rate": 3.3521322858181294e-06,
+      "loss": 0.0833,
+      "step": 6799
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.8878963576144874,
+      "learning_rate": 3.350741011330969e-06,
+      "loss": 0.1244,
+      "step": 6800
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.6785979482293591,
+      "learning_rate": 3.3493498801139466e-06,
+      "loss": 0.1002,
+      "step": 6801
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.8639060323024437,
+      "learning_rate": 3.347958892287907e-06,
+      "loss": 0.1313,
+      "step": 6802
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.8046891470179227,
+      "learning_rate": 3.3465680479736878e-06,
+      "loss": 0.1116,
+      "step": 6803
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.8955289557663788,
+      "learning_rate": 3.345177347292108e-06,
+      "loss": 0.1079,
+      "step": 6804
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.8798308865742541,
+      "learning_rate": 3.3437867903639787e-06,
+      "loss": 0.124,
+      "step": 6805
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.6528085084593902,
+      "learning_rate": 3.3423963773100944e-06,
+      "loss": 0.1051,
+      "step": 6806
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.580552815884661,
+      "learning_rate": 3.3410061082512422e-06,
+      "loss": 0.0982,
+      "step": 6807
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.86509479680996,
+      "learning_rate": 3.3396159833081902e-06,
+      "loss": 0.1157,
+      "step": 6808
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.8735586597079534,
+      "learning_rate": 3.3382260026017027e-06,
+      "loss": 0.1102,
+      "step": 6809
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.6709029130553332,
+      "learning_rate": 3.3368361662525226e-06,
+      "loss": 0.11,
+      "step": 6810
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.7001275565236607,
+      "learning_rate": 3.3354464743813864e-06,
+      "loss": 0.121,
+      "step": 6811
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.6119000328540138,
+      "learning_rate": 3.3340569271090145e-06,
+      "loss": 0.1114,
+      "step": 6812
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.6104322547194587,
+      "learning_rate": 3.3326675245561167e-06,
+      "loss": 0.0895,
+      "step": 6813
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.719466972673618,
+      "learning_rate": 3.331278266843388e-06,
+      "loss": 0.1024,
+      "step": 6814
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.0418920781174905,
+      "learning_rate": 3.329889154091515e-06,
+      "loss": 0.1202,
+      "step": 6815
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.47751345716607,
+      "learning_rate": 3.3285001864211672e-06,
+      "loss": 0.0898,
+      "step": 6816
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.55658568438265,
+      "learning_rate": 3.327111363953005e-06,
+      "loss": 0.1079,
+      "step": 6817
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.7886776349829745,
+      "learning_rate": 3.325722686807672e-06,
+      "loss": 0.1151,
+      "step": 6818
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.9101037533608973,
+      "learning_rate": 3.324334155105803e-06,
+      "loss": 0.1069,
+      "step": 6819
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.9207785469052485,
+      "learning_rate": 3.322945768968021e-06,
+      "loss": 0.1133,
+      "step": 6820
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.6463705587351587,
+      "learning_rate": 3.321557528514931e-06,
+      "loss": 0.1083,
+      "step": 6821
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.1152200369863676,
+      "learning_rate": 3.3201694338671313e-06,
+      "loss": 0.105,
+      "step": 6822
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.8677461396172759,
+      "learning_rate": 3.3187814851452026e-06,
+      "loss": 0.1142,
+      "step": 6823
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.4222431743836528,
+      "learning_rate": 3.3173936824697174e-06,
+      "loss": 0.084,
+      "step": 6824
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.7380465707245294,
+      "learning_rate": 3.3160060259612298e-06,
+      "loss": 0.1171,
+      "step": 6825
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.862013647316006,
+      "learning_rate": 3.314618515740289e-06,
+      "loss": 0.1146,
+      "step": 6826
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.7160244788871146,
+      "learning_rate": 3.313231151927424e-06,
+      "loss": 0.1057,
+      "step": 6827
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.7265476436506584,
+      "learning_rate": 3.311843934643157e-06,
+      "loss": 0.106,
+      "step": 6828
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.600374850850436,
+      "learning_rate": 3.3104568640079915e-06,
+      "loss": 0.0989,
+      "step": 6829
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.7268363061930119,
+      "learning_rate": 3.3090699401424244e-06,
+      "loss": 0.1006,
+      "step": 6830
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.5495903443835775,
+      "learning_rate": 3.307683163166934e-06,
+      "loss": 0.1,
+      "step": 6831
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.814542023946377,
+      "learning_rate": 3.306296533201992e-06,
+      "loss": 0.1161,
+      "step": 6832
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6512762212964869,
+      "learning_rate": 3.3049100503680516e-06,
+      "loss": 0.1111,
+      "step": 6833
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6552681381415215,
+      "learning_rate": 3.3035237147855575e-06,
+      "loss": 0.1064,
+      "step": 6834
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.56218545296008,
+      "learning_rate": 3.3021375265749385e-06,
+      "loss": 0.0964,
+      "step": 6835
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8150909131616568,
+      "learning_rate": 3.300751485856613e-06,
+      "loss": 0.1127,
+      "step": 6836
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.7825674061133077,
+      "learning_rate": 3.299365592750984e-06,
+      "loss": 0.1173,
+      "step": 6837
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.4670223102223359,
+      "learning_rate": 3.2979798473784453e-06,
+      "loss": 0.0791,
+      "step": 6838
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6969598997128585,
+      "learning_rate": 3.2965942498593735e-06,
+      "loss": 0.1034,
+      "step": 6839
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6801289997806028,
+      "learning_rate": 3.295208800314137e-06,
+      "loss": 0.1118,
+      "step": 6840
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8444462937226822,
+      "learning_rate": 3.293823498863087e-06,
+      "loss": 0.1251,
+      "step": 6841
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.4389994226722203,
+      "learning_rate": 3.292438345626565e-06,
+      "loss": 0.0846,
+      "step": 6842
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6907247443520161,
+      "learning_rate": 3.2910533407248966e-06,
+      "loss": 0.0992,
+      "step": 6843
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.0136715774701583,
+      "learning_rate": 3.2896684842784e-06,
+      "loss": 0.1333,
+      "step": 6844
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.5303895930962945,
+      "learning_rate": 3.288283776407373e-06,
+      "loss": 0.0959,
+      "step": 6845
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8727724056795176,
+      "learning_rate": 3.2868992172321068e-06,
+      "loss": 0.1241,
+      "step": 6846
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.5475972304417096,
+      "learning_rate": 3.2855148068728753e-06,
+      "loss": 0.1123,
+      "step": 6847
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.9063395517175663,
+      "learning_rate": 3.284130545449944e-06,
+      "loss": 0.1038,
+      "step": 6848
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.735991127323755,
+      "learning_rate": 3.282746433083559e-06,
+      "loss": 0.1107,
+      "step": 6849
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.598154349493814,
+      "learning_rate": 3.2813624698939617e-06,
+      "loss": 0.1018,
+      "step": 6850
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8207927539107558,
+      "learning_rate": 3.279978656001373e-06,
+      "loss": 0.1133,
+      "step": 6851
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.5615855127088596,
+      "learning_rate": 3.278594991526006e-06,
+      "loss": 0.092,
+      "step": 6852
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.4163635610115943,
+      "learning_rate": 3.277211476588057e-06,
+      "loss": 0.0855,
+      "step": 6853
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6170113768243384,
+      "learning_rate": 3.2758281113077127e-06,
+      "loss": 0.0965,
+      "step": 6854
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8171888820318796,
+      "learning_rate": 3.2744448958051428e-06,
+      "loss": 0.1197,
+      "step": 6855
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.933727587063898,
+      "learning_rate": 3.2730618302005104e-06,
+      "loss": 0.1261,
+      "step": 6856
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.5126104068440398,
+      "learning_rate": 3.2716789146139573e-06,
+      "loss": 0.0843,
+      "step": 6857
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8983530562537037,
+      "learning_rate": 3.2702961491656197e-06,
+      "loss": 0.1221,
+      "step": 6858
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.9751572939693205,
+      "learning_rate": 3.2689135339756155e-06,
+      "loss": 0.1056,
+      "step": 6859
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.7397171724403664,
+      "learning_rate": 3.2675310691640538e-06,
+      "loss": 0.1169,
+      "step": 6860
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.550915514810408,
+      "learning_rate": 3.266148754851025e-06,
+      "loss": 0.0931,
+      "step": 6861
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6633387563643878,
+      "learning_rate": 3.2647665911566144e-06,
+      "loss": 0.0987,
+      "step": 6862
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.7180240015595727,
+      "learning_rate": 3.2633845782008867e-06,
+      "loss": 0.1077,
+      "step": 6863
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.5671731229007753,
+      "learning_rate": 3.2620027161038975e-06,
+      "loss": 0.1044,
+      "step": 6864
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6017729930546374,
+      "learning_rate": 3.2606210049856877e-06,
+      "loss": 0.1127,
+      "step": 6865
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.871195901785977,
+      "learning_rate": 3.2592394449662867e-06,
+      "loss": 0.1157,
+      "step": 6866
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6457069842527312,
+      "learning_rate": 3.2578580361657076e-06,
+      "loss": 0.1132,
+      "step": 6867
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.76728088189756,
+      "learning_rate": 3.2564767787039563e-06,
+      "loss": 0.1192,
+      "step": 6868
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7373662832248904,
+      "learning_rate": 3.2550956727010184e-06,
+      "loss": 0.1174,
+      "step": 6869
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7081058608431943,
+      "learning_rate": 3.2537147182768723e-06,
+      "loss": 0.1,
+      "step": 6870
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.9868797112812537,
+      "learning_rate": 3.2523339155514787e-06,
+      "loss": 0.1186,
+      "step": 6871
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.6118508883446636,
+      "learning_rate": 3.2509532646447883e-06,
+      "loss": 0.0981,
+      "step": 6872
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.8292972866024735,
+      "learning_rate": 3.2495727656767353e-06,
+      "loss": 0.1091,
+      "step": 6873
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7001894485427724,
+      "learning_rate": 3.2481924187672466e-06,
+      "loss": 0.0975,
+      "step": 6874
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.9354283960524175,
+      "learning_rate": 3.2468122240362287e-06,
+      "loss": 0.1206,
+      "step": 6875
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.5962509053364602,
+      "learning_rate": 3.2454321816035805e-06,
+      "loss": 0.0875,
+      "step": 6876
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7213833141442425,
+      "learning_rate": 3.2440522915891837e-06,
+      "loss": 0.1254,
+      "step": 6877
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7155166775442827,
+      "learning_rate": 3.24267255411291e-06,
+      "loss": 0.1073,
+      "step": 6878
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.862891458898532,
+      "learning_rate": 3.2412929692946137e-06,
+      "loss": 0.1141,
+      "step": 6879
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7096947054374239,
+      "learning_rate": 3.239913537254143e-06,
+      "loss": 0.1127,
+      "step": 6880
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.6691562203007395,
+      "learning_rate": 3.2385342581113242e-06,
+      "loss": 0.1079,
+      "step": 6881
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.4631399596596462,
+      "learning_rate": 3.2371551319859778e-06,
+      "loss": 0.1017,
+      "step": 6882
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.6755398241893824,
+      "learning_rate": 3.235776158997904e-06,
+      "loss": 0.1084,
+      "step": 6883
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.6435249147301128,
+      "learning_rate": 3.2343973392668976e-06,
+      "loss": 0.1002,
+      "step": 6884
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.4647376928237656,
+      "learning_rate": 3.233018672912731e-06,
+      "loss": 0.0967,
+      "step": 6885
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.872224585131727,
+      "learning_rate": 3.231640160055172e-06,
+      "loss": 0.1149,
+      "step": 6886
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.8777079857247378,
+      "learning_rate": 3.2302618008139696e-06,
+      "loss": 0.1196,
+      "step": 6887
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7160732762228272,
+      "learning_rate": 3.228883595308862e-06,
+      "loss": 0.116,
+      "step": 6888
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.5493231838412793,
+      "learning_rate": 3.2275055436595713e-06,
+      "loss": 0.0995,
+      "step": 6889
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.9442931432806336,
+      "learning_rate": 3.2261276459858105e-06,
+      "loss": 0.1031,
+      "step": 6890
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7361892327631556,
+      "learning_rate": 3.2247499024072727e-06,
+      "loss": 0.1184,
+      "step": 6891
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.9338517786211538,
+      "learning_rate": 3.223372313043647e-06,
+      "loss": 0.1174,
+      "step": 6892
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.5935189432005834,
+      "learning_rate": 3.221994878014599e-06,
+      "loss": 0.0794,
+      "step": 6893
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.9029338360992318,
+      "learning_rate": 3.2206175974397896e-06,
+      "loss": 0.1089,
+      "step": 6894
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7965669050756166,
+      "learning_rate": 3.219240471438859e-06,
+      "loss": 0.1104,
+      "step": 6895
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.816868620020627,
+      "learning_rate": 3.21786350013144e-06,
+      "loss": 0.1107,
+      "step": 6896
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.6820068858523178,
+      "learning_rate": 3.216486683637146e-06,
+      "loss": 0.0848,
+      "step": 6897
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.9882521225695604,
+      "learning_rate": 3.2151100220755842e-06,
+      "loss": 0.1233,
+      "step": 6898
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.4728983447933541,
+      "learning_rate": 3.213733515566342e-06,
+      "loss": 0.0807,
+      "step": 6899
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7044523297856788,
+      "learning_rate": 3.212357164228996e-06,
+      "loss": 0.0871,
+      "step": 6900
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.8718722138930204,
+      "learning_rate": 3.2109809681831084e-06,
+      "loss": 0.128,
+      "step": 6901
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.6206777934234293,
+      "learning_rate": 3.2096049275482306e-06,
+      "loss": 0.087,
+      "step": 6902
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.8382214285601413,
+      "learning_rate": 3.2082290424438945e-06,
+      "loss": 0.0984,
+      "step": 6903
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.5410179782453917,
+      "learning_rate": 3.2068533129896273e-06,
+      "loss": 0.099,
+      "step": 6904
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.828132191117788,
+      "learning_rate": 3.205477739304935e-06,
+      "loss": 0.1089,
+      "step": 6905
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.7859253247322253,
+      "learning_rate": 3.2041023215093135e-06,
+      "loss": 0.1168,
+      "step": 6906
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5848177211059267,
+      "learning_rate": 3.2027270597222437e-06,
+      "loss": 0.0877,
+      "step": 6907
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.8894989462765621,
+      "learning_rate": 3.2013519540631954e-06,
+      "loss": 0.101,
+      "step": 6908
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5189238124892686,
+      "learning_rate": 3.1999770046516198e-06,
+      "loss": 0.0908,
+      "step": 6909
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.875630597933029,
+      "learning_rate": 3.1986022116069625e-06,
+      "loss": 0.1065,
+      "step": 6910
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.4857733602600187,
+      "learning_rate": 3.1972275750486483e-06,
+      "loss": 0.0918,
+      "step": 6911
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6356902172707195,
+      "learning_rate": 3.1958530950960908e-06,
+      "loss": 0.0938,
+      "step": 6912
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6060028659775716,
+      "learning_rate": 3.194478771868693e-06,
+      "loss": 0.1025,
+      "step": 6913
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.68595767488497,
+      "learning_rate": 3.1931046054858366e-06,
+      "loss": 0.1056,
+      "step": 6914
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6444308021429594,
+      "learning_rate": 3.1917305960669e-06,
+      "loss": 0.1037,
+      "step": 6915
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6434753994172973,
+      "learning_rate": 3.1903567437312388e-06,
+      "loss": 0.1044,
+      "step": 6916
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.3472447225523396,
+      "learning_rate": 3.188983048598201e-06,
+      "loss": 0.0801,
+      "step": 6917
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.9790042246075576,
+      "learning_rate": 3.187609510787116e-06,
+      "loss": 0.131,
+      "step": 6918
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5585054547457806,
+      "learning_rate": 3.186236130417306e-06,
+      "loss": 0.0969,
+      "step": 6919
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.8272425438772226,
+      "learning_rate": 3.184862907608072e-06,
+      "loss": 0.115,
+      "step": 6920
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.8906317645910922,
+      "learning_rate": 3.1834898424787073e-06,
+      "loss": 0.1178,
+      "step": 6921
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.7885853006471713,
+      "learning_rate": 3.1821169351484884e-06,
+      "loss": 0.1155,
+      "step": 6922
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.716941106531984,
+      "learning_rate": 3.1807441857366798e-06,
+      "loss": 0.1152,
+      "step": 6923
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6039718984157232,
+      "learning_rate": 3.17937159436253e-06,
+      "loss": 0.1123,
+      "step": 6924
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.4786552558677983,
+      "learning_rate": 3.177999161145277e-06,
+      "loss": 0.0856,
+      "step": 6925
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.7379766483963266,
+      "learning_rate": 3.1766268862041406e-06,
+      "loss": 0.0993,
+      "step": 6926
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.546009770461926,
+      "learning_rate": 3.1752547696583323e-06,
+      "loss": 0.0911,
+      "step": 6927
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6973855445111465,
+      "learning_rate": 3.1738828116270447e-06,
+      "loss": 0.1064,
+      "step": 6928
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6708884730166054,
+      "learning_rate": 3.1725110122294615e-06,
+      "loss": 0.1006,
+      "step": 6929
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.7379159314712331,
+      "learning_rate": 3.1711393715847477e-06,
+      "loss": 0.1173,
+      "step": 6930
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.704621509150292,
+      "learning_rate": 3.1697678898120585e-06,
+      "loss": 0.1124,
+      "step": 6931
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5838012680969749,
+      "learning_rate": 3.1683965670305317e-06,
+      "loss": 0.098,
+      "step": 6932
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.7252879524218507,
+      "learning_rate": 3.167025403359297e-06,
+      "loss": 0.1219,
+      "step": 6933
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.681272206366005,
+      "learning_rate": 3.1656543989174625e-06,
+      "loss": 0.0988,
+      "step": 6934
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.748537925662701,
+      "learning_rate": 3.164283553824129e-06,
+      "loss": 0.0972,
+      "step": 6935
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.8193882482632355,
+      "learning_rate": 3.16291286819838e-06,
+      "loss": 0.1029,
+      "step": 6936
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.8323696417451965,
+      "learning_rate": 3.1615423421592873e-06,
+      "loss": 0.1029,
+      "step": 6937
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.9206713347219417,
+      "learning_rate": 3.160171975825904e-06,
+      "loss": 0.1165,
+      "step": 6938
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.7745707253327103,
+      "learning_rate": 3.158801769317279e-06,
+      "loss": 0.1055,
+      "step": 6939
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5522016622902524,
+      "learning_rate": 3.157431722752436e-06,
+      "loss": 0.0903,
+      "step": 6940
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.8047051109478471,
+      "learning_rate": 3.1560618362503937e-06,
+      "loss": 0.1091,
+      "step": 6941
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.7826666259666064,
+      "learning_rate": 3.1546921099301507e-06,
+      "loss": 0.1015,
+      "step": 6942
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.612127477196659,
+      "learning_rate": 3.1533225439106965e-06,
+      "loss": 0.0938,
+      "step": 6943
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.5780096930072713,
+      "learning_rate": 3.1519531383110014e-06,
+      "loss": 0.1064,
+      "step": 6944
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.501936803215791,
+      "learning_rate": 3.1505838932500287e-06,
+      "loss": 0.1029,
+      "step": 6945
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6618034266569448,
+      "learning_rate": 3.149214808846721e-06,
+      "loss": 0.1131,
+      "step": 6946
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.5867358281510369,
+      "learning_rate": 3.1478458852200122e-06,
+      "loss": 0.0907,
+      "step": 6947
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.5536016734892864,
+      "learning_rate": 3.1464771224888173e-06,
+      "loss": 0.1054,
+      "step": 6948
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6548856868193258,
+      "learning_rate": 3.1451085207720423e-06,
+      "loss": 0.1118,
+      "step": 6949
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.9005836277447175,
+      "learning_rate": 3.143740080188574e-06,
+      "loss": 0.1161,
+      "step": 6950
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6199172476610106,
+      "learning_rate": 3.1423718008572913e-06,
+      "loss": 0.1017,
+      "step": 6951
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6319113502795275,
+      "learning_rate": 3.1410036828970525e-06,
+      "loss": 0.108,
+      "step": 6952
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.7871738552739331,
+      "learning_rate": 3.1396357264267087e-06,
+      "loss": 0.1041,
+      "step": 6953
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.880436779032392,
+      "learning_rate": 3.1382679315650903e-06,
+      "loss": 0.118,
+      "step": 6954
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.7066230896831414,
+      "learning_rate": 3.136900298431019e-06,
+      "loss": 0.0996,
+      "step": 6955
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.610475184087908,
+      "learning_rate": 3.135532827143298e-06,
+      "loss": 0.0936,
+      "step": 6956
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.735675363617396,
+      "learning_rate": 3.134165517820722e-06,
+      "loss": 0.0978,
+      "step": 6957
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6280006872336938,
+      "learning_rate": 3.132798370582065e-06,
+      "loss": 0.1054,
+      "step": 6958
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6754636255346593,
+      "learning_rate": 3.131431385546093e-06,
+      "loss": 0.1134,
+      "step": 6959
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.5471663158482993,
+      "learning_rate": 3.130064562831553e-06,
+      "loss": 0.0968,
+      "step": 6960
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.7074249087913036,
+      "learning_rate": 3.1286979025571817e-06,
+      "loss": 0.1061,
+      "step": 6961
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.627628144743767,
+      "learning_rate": 3.1273314048416967e-06,
+      "loss": 0.1088,
+      "step": 6962
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6451333996415225,
+      "learning_rate": 3.1259650698038106e-06,
+      "loss": 0.1019,
+      "step": 6963
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6130077271533523,
+      "learning_rate": 3.1245988975622116e-06,
+      "loss": 0.1086,
+      "step": 6964
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.733701179048697,
+      "learning_rate": 3.12323288823558e-06,
+      "loss": 0.1098,
+      "step": 6965
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.596404975205049,
+      "learning_rate": 3.1218670419425794e-06,
+      "loss": 0.095,
+      "step": 6966
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.8598970436389741,
+      "learning_rate": 3.1205013588018616e-06,
+      "loss": 0.1049,
+      "step": 6967
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.4870449403527588,
+      "learning_rate": 3.119135838932059e-06,
+      "loss": 0.0713,
+      "step": 6968
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6233278555117967,
+      "learning_rate": 3.1177704824517984e-06,
+      "loss": 0.0971,
+      "step": 6969
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.3149177490753887,
+      "learning_rate": 3.1164052894796836e-06,
+      "loss": 0.0813,
+      "step": 6970
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.8082337005439453,
+      "learning_rate": 3.1150402601343116e-06,
+      "loss": 0.106,
+      "step": 6971
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6447516039693149,
+      "learning_rate": 3.113675394534258e-06,
+      "loss": 0.0952,
+      "step": 6972
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.8057585759112431,
+      "learning_rate": 3.1123106927980906e-06,
+      "loss": 0.1069,
+      "step": 6973
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.8155349133272556,
+      "learning_rate": 3.1109461550443574e-06,
+      "loss": 0.1072,
+      "step": 6974
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.8465096949739028,
+      "learning_rate": 3.1095817813915983e-06,
+      "loss": 0.1133,
+      "step": 6975
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.7988461303064716,
+      "learning_rate": 3.1082175719583336e-06,
+      "loss": 0.1033,
+      "step": 6976
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.4299173057939911,
+      "learning_rate": 3.106853526863073e-06,
+      "loss": 0.0855,
+      "step": 6977
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6435344335906938,
+      "learning_rate": 3.105489646224307e-06,
+      "loss": 0.1043,
+      "step": 6978
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6678930535227892,
+      "learning_rate": 3.1041259301605194e-06,
+      "loss": 0.0895,
+      "step": 6979
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.652446888045159,
+      "learning_rate": 3.1027623787901706e-06,
+      "loss": 0.1038,
+      "step": 6980
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.4263905734156785,
+      "learning_rate": 3.1013989922317154e-06,
+      "loss": 0.0937,
+      "step": 6981
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6349801745918973,
+      "learning_rate": 3.100035770603589e-06,
+      "loss": 0.1116,
+      "step": 6982
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.7871167154992145,
+      "learning_rate": 3.0986727140242145e-06,
+      "loss": 0.1129,
+      "step": 6983
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.8049718839766495,
+      "learning_rate": 3.097309822611998e-06,
+      "loss": 0.1121,
+      "step": 6984
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6548684562502816,
+      "learning_rate": 3.095947096485335e-06,
+      "loss": 0.1017,
+      "step": 6985
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.576807029105432,
+      "learning_rate": 3.0945845357626014e-06,
+      "loss": 0.0873,
+      "step": 6986
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6137110597195599,
+      "learning_rate": 3.093222140562167e-06,
+      "loss": 0.1059,
+      "step": 6987
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6082029287492177,
+      "learning_rate": 3.0918599110023784e-06,
+      "loss": 0.1051,
+      "step": 6988
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6739381436304963,
+      "learning_rate": 3.090497847201574e-06,
+      "loss": 0.0994,
+      "step": 6989
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.8879407364067824,
+      "learning_rate": 3.0891359492780734e-06,
+      "loss": 0.1151,
+      "step": 6990
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6382514402051553,
+      "learning_rate": 3.0877742173501857e-06,
+      "loss": 0.1095,
+      "step": 6991
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.361987093540842,
+      "learning_rate": 3.0864126515362003e-06,
+      "loss": 0.0796,
+      "step": 6992
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.7451611067461479,
+      "learning_rate": 3.0850512519544005e-06,
+      "loss": 0.1032,
+      "step": 6993
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.5852254759250786,
+      "learning_rate": 3.0836900187230475e-06,
+      "loss": 0.0919,
+      "step": 6994
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.5321003892666154,
+      "learning_rate": 3.0823289519603916e-06,
+      "loss": 0.0875,
+      "step": 6995
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6254591943080208,
+      "learning_rate": 3.0809680517846664e-06,
+      "loss": 0.0991,
+      "step": 6996
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.5992944150003598,
+      "learning_rate": 3.0796073183140953e-06,
+      "loss": 0.1016,
+      "step": 6997
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6096473091359744,
+      "learning_rate": 3.07824675166688e-06,
+      "loss": 0.0977,
+      "step": 6998
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6111611160954276,
+      "learning_rate": 3.076886351961217e-06,
+      "loss": 0.111,
+      "step": 6999
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.5514590117798215,
+      "learning_rate": 3.0755261193152797e-06,
+      "loss": 0.0786,
+      "step": 7000
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6714633802157248,
+      "learning_rate": 3.074166053847234e-06,
+      "loss": 0.1142,
+      "step": 7001
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.8388493808640693,
+      "learning_rate": 3.0728061556752246e-06,
+      "loss": 0.1145,
+      "step": 7002
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.3953102600993432,
+      "learning_rate": 3.071446424917388e-06,
+      "loss": 0.0785,
+      "step": 7003
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.9322247883564228,
+      "learning_rate": 3.070086861691839e-06,
+      "loss": 0.1084,
+      "step": 7004
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.4723004556035557,
+      "learning_rate": 3.0687274661166867e-06,
+      "loss": 0.0839,
+      "step": 7005
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.8213631652597726,
+      "learning_rate": 3.0673682383100194e-06,
+      "loss": 0.1074,
+      "step": 7006
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.5734232397531398,
+      "learning_rate": 3.0660091783899117e-06,
+      "loss": 0.0958,
+      "step": 7007
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.8841931408607158,
+      "learning_rate": 3.064650286474425e-06,
+      "loss": 0.1149,
+      "step": 7008
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.4887384915213429,
+      "learning_rate": 3.063291562681604e-06,
+      "loss": 0.0843,
+      "step": 7009
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.5701159055833478,
+      "learning_rate": 3.061933007129483e-06,
+      "loss": 0.0986,
+      "step": 7010
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.484992609880974,
+      "learning_rate": 3.0605746199360755e-06,
+      "loss": 0.092,
+      "step": 7011
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.7772399858693864,
+      "learning_rate": 3.059216401219387e-06,
+      "loss": 0.1096,
+      "step": 7012
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.7899516216797535,
+      "learning_rate": 3.0578583510974035e-06,
+      "loss": 0.1022,
+      "step": 7013
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.7679347349770604,
+      "learning_rate": 3.0565004696880984e-06,
+      "loss": 0.1153,
+      "step": 7014
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.8008798141824234,
+      "learning_rate": 3.055142757109428e-06,
+      "loss": 0.1065,
+      "step": 7015
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.941361596932643,
+      "learning_rate": 3.0537852134793393e-06,
+      "loss": 0.1127,
+      "step": 7016
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.6060316351705297,
+      "learning_rate": 3.0524278389157593e-06,
+      "loss": 0.1009,
+      "step": 7017
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.6427162540464562,
+      "learning_rate": 3.0510706335366034e-06,
+      "loss": 0.1166,
+      "step": 7018
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.6775323250167518,
+      "learning_rate": 3.04971359745977e-06,
+      "loss": 0.0978,
+      "step": 7019
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.4108792143245346,
+      "learning_rate": 3.048356730803146e-06,
+      "loss": 0.0829,
+      "step": 7020
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.437339266982179,
+      "learning_rate": 3.0470000336845977e-06,
+      "loss": 0.0895,
+      "step": 7021
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7229092485652902,
+      "learning_rate": 3.045643506221985e-06,
+      "loss": 0.1103,
+      "step": 7022
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.8188157080555727,
+      "learning_rate": 3.044287148533146e-06,
+      "loss": 0.1102,
+      "step": 7023
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.6249059374701658,
+      "learning_rate": 3.0429309607359088e-06,
+      "loss": 0.1038,
+      "step": 7024
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.295843262473041,
+      "learning_rate": 3.041574942948081e-06,
+      "loss": 0.0747,
+      "step": 7025
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.9370610366960002,
+      "learning_rate": 3.040219095287463e-06,
+      "loss": 0.1318,
+      "step": 7026
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.750306593491175,
+      "learning_rate": 3.0388634178718336e-06,
+      "loss": 0.0983,
+      "step": 7027
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.5656530025081041,
+      "learning_rate": 3.0375079108189613e-06,
+      "loss": 0.105,
+      "step": 7028
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.4161619701991173,
+      "learning_rate": 3.0361525742465975e-06,
+      "loss": 0.0781,
+      "step": 7029
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.5462961931064405,
+      "learning_rate": 3.034797408272481e-06,
+      "loss": 0.0941,
+      "step": 7030
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7048849935411121,
+      "learning_rate": 3.033442413014331e-06,
+      "loss": 0.0997,
+      "step": 7031
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7572741191390129,
+      "learning_rate": 3.032087588589858e-06,
+      "loss": 0.1108,
+      "step": 7032
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.5407038212087154,
+      "learning_rate": 3.0307329351167527e-06,
+      "loss": 0.0873,
+      "step": 7033
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.6744575782969193,
+      "learning_rate": 3.0293784527126956e-06,
+      "loss": 0.1089,
+      "step": 7034
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.6388947149107616,
+      "learning_rate": 3.0280241414953477e-06,
+      "loss": 0.1002,
+      "step": 7035
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.5309349531814886,
+      "learning_rate": 3.0266700015823585e-06,
+      "loss": 0.0954,
+      "step": 7036
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.3950952540378945,
+      "learning_rate": 3.02531603309136e-06,
+      "loss": 0.0791,
+      "step": 7037
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.566810636419363,
+      "learning_rate": 3.023962236139972e-06,
+      "loss": 0.0977,
+      "step": 7038
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7232781221321731,
+      "learning_rate": 3.022608610845795e-06,
+      "loss": 0.1103,
+      "step": 7039
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.6029285729937697,
+      "learning_rate": 3.0212551573264224e-06,
+      "loss": 0.0915,
+      "step": 7040
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.8620922428555597,
+      "learning_rate": 3.0199018756994245e-06,
+      "loss": 0.1229,
+      "step": 7041
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7063409879938742,
+      "learning_rate": 3.018548766082362e-06,
+      "loss": 0.1025,
+      "step": 7042
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7495428540027698,
+      "learning_rate": 3.017195828592777e-06,
+      "loss": 0.0949,
+      "step": 7043
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.8628941034322832,
+      "learning_rate": 3.0158430633481996e-06,
+      "loss": 0.0888,
+      "step": 7044
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.9278426676759393,
+      "learning_rate": 3.0144904704661413e-06,
+      "loss": 0.126,
+      "step": 7045
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7139311996866518,
+      "learning_rate": 3.013138050064105e-06,
+      "loss": 0.1021,
+      "step": 7046
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7922380336956099,
+      "learning_rate": 3.011785802259571e-06,
+      "loss": 0.0959,
+      "step": 7047
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.779174011549657,
+      "learning_rate": 3.0104337271700114e-06,
+      "loss": 0.0989,
+      "step": 7048
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.6547499421196536,
+      "learning_rate": 3.0090818249128773e-06,
+      "loss": 0.0992,
+      "step": 7049
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.579963859393145,
+      "learning_rate": 3.00773009560561e-06,
+      "loss": 0.0955,
+      "step": 7050
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.8738536023269088,
+      "learning_rate": 3.006378539365631e-06,
+      "loss": 0.1286,
+      "step": 7051
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.6629894998857333,
+      "learning_rate": 3.005027156310352e-06,
+      "loss": 0.1054,
+      "step": 7052
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.7133419134919652,
+      "learning_rate": 3.0036759465571634e-06,
+      "loss": 0.1188,
+      "step": 7053
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.7280722086986222,
+      "learning_rate": 3.0023249102234477e-06,
+      "loss": 0.1086,
+      "step": 7054
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.5332679865794903,
+      "learning_rate": 3.000974047426566e-06,
+      "loss": 0.0869,
+      "step": 7055
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.0626182737333623,
+      "learning_rate": 2.9996233582838686e-06,
+      "loss": 0.1231,
+      "step": 7056
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.74456585778073,
+      "learning_rate": 2.998272842912686e-06,
+      "loss": 0.0893,
+      "step": 7057
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.6054574997147781,
+      "learning_rate": 2.996922501430341e-06,
+      "loss": 0.1025,
+      "step": 7058
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.4603425254002318,
+      "learning_rate": 2.9955723339541336e-06,
+      "loss": 0.0912,
+      "step": 7059
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.8128231524086977,
+      "learning_rate": 2.994222340601355e-06,
+      "loss": 0.1205,
+      "step": 7060
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.4363029164280554,
+      "learning_rate": 2.992872521489275e-06,
+      "loss": 0.0884,
+      "step": 7061
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.6674638649933378,
+      "learning_rate": 2.991522876735154e-06,
+      "loss": 0.1093,
+      "step": 7062
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.788910511454472,
+      "learning_rate": 2.9901734064562328e-06,
+      "loss": 0.109,
+      "step": 7063
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.5631003909681371,
+      "learning_rate": 2.9888241107697413e-06,
+      "loss": 0.0992,
+      "step": 7064
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.7520997242192788,
+      "learning_rate": 2.98747498979289e-06,
+      "loss": 0.1156,
+      "step": 7065
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.5628607948013824,
+      "learning_rate": 2.9861260436428783e-06,
+      "loss": 0.1003,
+      "step": 7066
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.6715099728768117,
+      "learning_rate": 2.984777272436887e-06,
+      "loss": 0.1044,
+      "step": 7067
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.8790883466645505,
+      "learning_rate": 2.983428676292084e-06,
+      "loss": 0.1465,
+      "step": 7068
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.716192938107207,
+      "learning_rate": 2.982080255325618e-06,
+      "loss": 0.11,
+      "step": 7069
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.606108637550764,
+      "learning_rate": 2.98073200965463e-06,
+      "loss": 0.1026,
+      "step": 7070
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.8387548041110526,
+      "learning_rate": 2.9793839393962374e-06,
+      "loss": 0.1023,
+      "step": 7071
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.9010902473341666,
+      "learning_rate": 2.978036044667549e-06,
+      "loss": 0.1223,
+      "step": 7072
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.5654136815738038,
+      "learning_rate": 2.976688325585655e-06,
+      "loss": 0.1047,
+      "step": 7073
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.4847395671122539,
+      "learning_rate": 2.9753407822676307e-06,
+      "loss": 0.0997,
+      "step": 7074
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.3726704549669104,
+      "learning_rate": 2.973993414830534e-06,
+      "loss": 0.0786,
+      "step": 7075
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.6516326915108621,
+      "learning_rate": 2.9726462233914146e-06,
+      "loss": 0.1045,
+      "step": 7076
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.8592349258282432,
+      "learning_rate": 2.971299208067298e-06,
+      "loss": 0.1204,
+      "step": 7077
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.6906990805707725,
+      "learning_rate": 2.9699523689752017e-06,
+      "loss": 0.1036,
+      "step": 7078
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.744519183401334,
+      "learning_rate": 2.9686057062321226e-06,
+      "loss": 0.0861,
+      "step": 7079
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.5018210137823347,
+      "learning_rate": 2.9672592199550465e-06,
+      "loss": 0.081,
+      "step": 7080
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.4990104102537358,
+      "learning_rate": 2.965912910260938e-06,
+      "loss": 0.1013,
+      "step": 7081
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.8526982602162374,
+      "learning_rate": 2.9645667772667553e-06,
+      "loss": 0.109,
+      "step": 7082
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.7620767486072342,
+      "learning_rate": 2.9632208210894326e-06,
+      "loss": 0.1068,
+      "step": 7083
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.7072611840119034,
+      "learning_rate": 2.961875041845894e-06,
+      "loss": 0.1022,
+      "step": 7084
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.7262828298989001,
+      "learning_rate": 2.960529439653046e-06,
+      "loss": 0.1224,
+      "step": 7085
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.665912974284882,
+      "learning_rate": 2.959184014627781e-06,
+      "loss": 0.1084,
+      "step": 7086
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.5474788104025523,
+      "learning_rate": 2.957838766886972e-06,
+      "loss": 0.0949,
+      "step": 7087
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5689255610722481,
+      "learning_rate": 2.9564936965474844e-06,
+      "loss": 0.1013,
+      "step": 7088
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.9302090817034776,
+      "learning_rate": 2.955148803726161e-06,
+      "loss": 0.1243,
+      "step": 7089
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5052761051096597,
+      "learning_rate": 2.953804088539833e-06,
+      "loss": 0.0784,
+      "step": 7090
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6331099651091168,
+      "learning_rate": 2.9524595511053137e-06,
+      "loss": 0.1098,
+      "step": 7091
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.8628799541688463,
+      "learning_rate": 2.9511151915394043e-06,
+      "loss": 0.1067,
+      "step": 7092
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6886222147935586,
+      "learning_rate": 2.949771009958885e-06,
+      "loss": 0.1144,
+      "step": 7093
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.570771081163554,
+      "learning_rate": 2.948427006480528e-06,
+      "loss": 0.0888,
+      "step": 7094
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.9764596395837009,
+      "learning_rate": 2.9470831812210836e-06,
+      "loss": 0.11,
+      "step": 7095
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.979109919147417,
+      "learning_rate": 2.9457395342972904e-06,
+      "loss": 0.1233,
+      "step": 7096
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5829595729326278,
+      "learning_rate": 2.9443960658258696e-06,
+      "loss": 0.1024,
+      "step": 7097
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.7584027764203016,
+      "learning_rate": 2.943052775923526e-06,
+      "loss": 0.1058,
+      "step": 7098
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.7852003215974552,
+      "learning_rate": 2.9417096647069532e-06,
+      "loss": 0.1092,
+      "step": 7099
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5973069884742903,
+      "learning_rate": 2.9403667322928255e-06,
+      "loss": 0.0999,
+      "step": 7100
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.715281742152762,
+      "learning_rate": 2.9390239787978026e-06,
+      "loss": 0.1039,
+      "step": 7101
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6556594209082036,
+      "learning_rate": 2.937681404338527e-06,
+      "loss": 0.0947,
+      "step": 7102
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.7588433082327835,
+      "learning_rate": 2.93633900903163e-06,
+      "loss": 0.0865,
+      "step": 7103
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6301660289070452,
+      "learning_rate": 2.9349967929937218e-06,
+      "loss": 0.0983,
+      "step": 7104
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5966796797292506,
+      "learning_rate": 2.9336547563414036e-06,
+      "loss": 0.1049,
+      "step": 7105
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.738022142401436,
+      "learning_rate": 2.9323128991912543e-06,
+      "loss": 0.1095,
+      "step": 7106
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.0028512616026064,
+      "learning_rate": 2.9309712216598413e-06,
+      "loss": 0.1318,
+      "step": 7107
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6712006670282848,
+      "learning_rate": 2.929629723863715e-06,
+      "loss": 0.0933,
+      "step": 7108
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6122640228582024,
+      "learning_rate": 2.9282884059194112e-06,
+      "loss": 0.0932,
+      "step": 7109
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.7899905360257955,
+      "learning_rate": 2.926947267943447e-06,
+      "loss": 0.1103,
+      "step": 7110
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.4885080427170663,
+      "learning_rate": 2.9256063100523303e-06,
+      "loss": 0.0865,
+      "step": 7111
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.8351737586147492,
+      "learning_rate": 2.9242655323625458e-06,
+      "loss": 0.1189,
+      "step": 7112
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5013271185402737,
+      "learning_rate": 2.9229249349905686e-06,
+      "loss": 0.0944,
+      "step": 7113
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6504760832684162,
+      "learning_rate": 2.9215845180528537e-06,
+      "loss": 0.1041,
+      "step": 7114
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6096959939428497,
+      "learning_rate": 2.9202442816658433e-06,
+      "loss": 0.1038,
+      "step": 7115
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.0228252946783805,
+      "learning_rate": 2.918904225945961e-06,
+      "loss": 0.1232,
+      "step": 7116
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.4798887516881394,
+      "learning_rate": 2.9175643510096195e-06,
+      "loss": 0.0807,
+      "step": 7117
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6991095850444975,
+      "learning_rate": 2.916224656973211e-06,
+      "loss": 0.1014,
+      "step": 7118
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.8736220977574374,
+      "learning_rate": 2.9148851439531177e-06,
+      "loss": 0.1295,
+      "step": 7119
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.4613378710563603,
+      "learning_rate": 2.9135458120656958e-06,
+      "loss": 0.0948,
+      "step": 7120
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.632854429077691,
+      "learning_rate": 2.912206661427297e-06,
+      "loss": 0.1244,
+      "step": 7121
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6188542185006656,
+      "learning_rate": 2.910867692154249e-06,
+      "loss": 0.1047,
+      "step": 7122
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5997744629828488,
+      "learning_rate": 2.909528904362872e-06,
+      "loss": 0.1055,
+      "step": 7123
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.4820527828936578,
+      "learning_rate": 2.908190298169461e-06,
+      "loss": 0.0981,
+      "step": 7124
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.7883341284731695,
+      "learning_rate": 2.9068518736903063e-06,
+      "loss": 0.1155,
+      "step": 7125
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6594220733904497,
+      "learning_rate": 2.9055136310416664e-06,
+      "loss": 0.0894,
+      "step": 7126
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.846004590050599,
+      "learning_rate": 2.9041755703398023e-06,
+      "loss": 0.1177,
+      "step": 7127
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5881793862083315,
+      "learning_rate": 2.9028376917009448e-06,
+      "loss": 0.0853,
+      "step": 7128
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6827123383021485,
+      "learning_rate": 2.901499995241319e-06,
+      "loss": 0.099,
+      "step": 7129
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5060885124736059,
+      "learning_rate": 2.900162481077126e-06,
+      "loss": 0.0907,
+      "step": 7130
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.825513601583001,
+      "learning_rate": 2.8988251493245607e-06,
+      "loss": 0.1062,
+      "step": 7131
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.7653753315030944,
+      "learning_rate": 2.897488000099788e-06,
+      "loss": 0.0973,
+      "step": 7132
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.8267783962810458,
+      "learning_rate": 2.896151033518971e-06,
+      "loss": 0.1162,
+      "step": 7133
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5168075840055988,
+      "learning_rate": 2.8948142496982488e-06,
+      "loss": 0.0916,
+      "step": 7134
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6094190110775615,
+      "learning_rate": 2.8934776487537498e-06,
+      "loss": 0.093,
+      "step": 7135
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.0060222531550997,
+      "learning_rate": 2.8921412308015795e-06,
+      "loss": 0.1365,
+      "step": 7136
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.8704228851427382,
+      "learning_rate": 2.8908049959578375e-06,
+      "loss": 0.1078,
+      "step": 7137
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.580241100998725,
+      "learning_rate": 2.8894689443385947e-06,
+      "loss": 0.0934,
+      "step": 7138
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.651453207452964,
+      "learning_rate": 2.888133076059919e-06,
+      "loss": 0.0977,
+      "step": 7139
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6746311230271989,
+      "learning_rate": 2.8867973912378524e-06,
+      "loss": 0.1004,
+      "step": 7140
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5175627637414244,
+      "learning_rate": 2.885461889988428e-06,
+      "loss": 0.0806,
+      "step": 7141
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6337083980116929,
+      "learning_rate": 2.8841265724276566e-06,
+      "loss": 0.1178,
+      "step": 7142
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.556722073013614,
+      "learning_rate": 2.882791438671543e-06,
+      "loss": 0.0847,
+      "step": 7143
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5952924466319007,
+      "learning_rate": 2.8814564888360617e-06,
+      "loss": 0.0885,
+      "step": 7144
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5444463706038987,
+      "learning_rate": 2.8801217230371838e-06,
+      "loss": 0.1063,
+      "step": 7145
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.7075500760743183,
+      "learning_rate": 2.8787871413908563e-06,
+      "loss": 0.1012,
+      "step": 7146
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.523736793271295,
+      "learning_rate": 2.8774527440130173e-06,
+      "loss": 0.0831,
+      "step": 7147
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6414465474566335,
+      "learning_rate": 2.8761185310195803e-06,
+      "loss": 0.0977,
+      "step": 7148
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.619888272049271,
+      "learning_rate": 2.874784502526456e-06,
+      "loss": 0.092,
+      "step": 7149
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.4825273685704115,
+      "learning_rate": 2.87345065864952e-06,
+      "loss": 0.0823,
+      "step": 7150
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.8710760853350006,
+      "learning_rate": 2.8721169995046503e-06,
+      "loss": 0.1217,
+      "step": 7151
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.688078830066312,
+      "learning_rate": 2.8707835252076967e-06,
+      "loss": 0.0943,
+      "step": 7152
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5154091912650776,
+      "learning_rate": 2.8694502358745003e-06,
+      "loss": 0.079,
+      "step": 7153
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6424958786202457,
+      "learning_rate": 2.86811713162088e-06,
+      "loss": 0.1062,
+      "step": 7154
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.4125838589757584,
+      "learning_rate": 2.8667842125626473e-06,
+      "loss": 0.08,
+      "step": 7155
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.7350123309603294,
+      "learning_rate": 2.8654514788155846e-06,
+      "loss": 0.0955,
+      "step": 7156
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.65850665524578,
+      "learning_rate": 2.864118930495472e-06,
+      "loss": 0.0964,
+      "step": 7157
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6446333632298167,
+      "learning_rate": 2.862786567718062e-06,
+      "loss": 0.103,
+      "step": 7158
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.755229985312071,
+      "learning_rate": 2.861454390599101e-06,
+      "loss": 0.0988,
+      "step": 7159
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5730082503018132,
+      "learning_rate": 2.860122399254312e-06,
+      "loss": 0.0919,
+      "step": 7160
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5481810222978016,
+      "learning_rate": 2.858790593799405e-06,
+      "loss": 0.0892,
+      "step": 7161
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.6721221997241988,
+      "learning_rate": 2.85745897435007e-06,
+      "loss": 0.1024,
+      "step": 7162
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.582851270403692,
+      "learning_rate": 2.856127541021989e-06,
+      "loss": 0.0959,
+      "step": 7163
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.683197511584693,
+      "learning_rate": 2.8547962939308187e-06,
+      "loss": 0.0922,
+      "step": 7164
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.8433705924611326,
+      "learning_rate": 2.8534652331922073e-06,
+      "loss": 0.1227,
+      "step": 7165
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.090068189556892,
+      "learning_rate": 2.8521343589217816e-06,
+      "loss": 0.1083,
+      "step": 7166
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.9299027435630736,
+      "learning_rate": 2.8508036712351535e-06,
+      "loss": 0.1131,
+      "step": 7167
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.0628045657195453,
+      "learning_rate": 2.849473170247917e-06,
+      "loss": 0.1156,
+      "step": 7168
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.888024961377578,
+      "learning_rate": 2.8481428560756565e-06,
+      "loss": 0.117,
+      "step": 7169
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.971710023191058,
+      "learning_rate": 2.846812728833931e-06,
+      "loss": 0.129,
+      "step": 7170
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.797874835398455,
+      "learning_rate": 2.8454827886382918e-06,
+      "loss": 0.1106,
+      "step": 7171
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.4959055368596517,
+      "learning_rate": 2.844153035604269e-06,
+      "loss": 0.0928,
+      "step": 7172
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.560051113486972,
+      "learning_rate": 2.842823469847376e-06,
+      "loss": 0.1007,
+      "step": 7173
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.5667186296783124,
+      "learning_rate": 2.841494091483111e-06,
+      "loss": 0.0927,
+      "step": 7174
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.8206023033816052,
+      "learning_rate": 2.840164900626958e-06,
+      "loss": 0.1109,
+      "step": 7175
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.9050058938585528,
+      "learning_rate": 2.838835897394382e-06,
+      "loss": 0.1155,
+      "step": 7176
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.883249590174709,
+      "learning_rate": 2.8375070819008345e-06,
+      "loss": 0.1065,
+      "step": 7177
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.5069487818357925,
+      "learning_rate": 2.8361784542617476e-06,
+      "loss": 0.0978,
+      "step": 7178
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.9298931735142786,
+      "learning_rate": 2.8348500145925384e-06,
+      "loss": 0.1059,
+      "step": 7179
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.6508220445571002,
+      "learning_rate": 2.8335217630086053e-06,
+      "loss": 0.1064,
+      "step": 7180
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.716748732576026,
+      "learning_rate": 2.8321936996253368e-06,
+      "loss": 0.0998,
+      "step": 7181
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.6736281390205572,
+      "learning_rate": 2.8308658245580977e-06,
+      "loss": 0.1021,
+      "step": 7182
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.6536782061474171,
+      "learning_rate": 2.829538137922243e-06,
+      "loss": 0.0974,
+      "step": 7183
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.6557844422473402,
+      "learning_rate": 2.828210639833106e-06,
+      "loss": 0.0971,
+      "step": 7184
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.756145142490854,
+      "learning_rate": 2.826883330406006e-06,
+      "loss": 0.1074,
+      "step": 7185
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.6257806350083217,
+      "learning_rate": 2.8255562097562437e-06,
+      "loss": 0.1079,
+      "step": 7186
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.519961551973838,
+      "learning_rate": 2.8242292779991086e-06,
+      "loss": 0.0857,
+      "step": 7187
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.5920220831505345,
+      "learning_rate": 2.822902535249867e-06,
+      "loss": 0.1021,
+      "step": 7188
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.432716073255345,
+      "learning_rate": 2.8215759816237748e-06,
+      "loss": 0.0875,
+      "step": 7189
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.6203884156941701,
+      "learning_rate": 2.8202496172360715e-06,
+      "loss": 0.1061,
+      "step": 7190
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.8191959811143301,
+      "learning_rate": 2.8189234422019707e-06,
+      "loss": 0.1036,
+      "step": 7191
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.5326807414032029,
+      "learning_rate": 2.817597456636682e-06,
+      "loss": 0.0977,
+      "step": 7192
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.7563990423728753,
+      "learning_rate": 2.8162716606553885e-06,
+      "loss": 0.1068,
+      "step": 7193
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.6657378268546792,
+      "learning_rate": 2.8149460543732666e-06,
+      "loss": 0.0922,
+      "step": 7194
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.0817520235831894,
+      "learning_rate": 2.8136206379054658e-06,
+      "loss": 0.0836,
+      "step": 7195
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.7318692643651203,
+      "learning_rate": 2.812295411367131e-06,
+      "loss": 0.1101,
+      "step": 7196
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.490375277170295,
+      "learning_rate": 2.8109703748733746e-06,
+      "loss": 0.0998,
+      "step": 7197
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.4836005685472418,
+      "learning_rate": 2.8096455285393094e-06,
+      "loss": 0.0945,
+      "step": 7198
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.4480946662179084,
+      "learning_rate": 2.808320872480018e-06,
+      "loss": 0.0969,
+      "step": 7199
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.5673698210646434,
+      "learning_rate": 2.8069964068105786e-06,
+      "loss": 0.0918,
+      "step": 7200
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.7804432275168103,
+      "learning_rate": 2.8056721316460417e-06,
+      "loss": 0.1082,
+      "step": 7201
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.7593922536940547,
+      "learning_rate": 2.8043480471014524e-06,
+      "loss": 0.0974,
+      "step": 7202
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.9262990226737913,
+      "learning_rate": 2.8030241532918244e-06,
+      "loss": 0.1074,
+      "step": 7203
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.79953210518286,
+      "learning_rate": 2.801700450332171e-06,
+      "loss": 0.1072,
+      "step": 7204
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.4445661151242126,
+      "learning_rate": 2.8003769383374765e-06,
+      "loss": 0.0735,
+      "step": 7205
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.7606078172438218,
+      "learning_rate": 2.7990536174227175e-06,
+      "loss": 0.0959,
+      "step": 7206
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.713978977561146,
+      "learning_rate": 2.7977304877028464e-06,
+      "loss": 0.1072,
+      "step": 7207
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.689511396125973,
+      "learning_rate": 2.796407549292809e-06,
+      "loss": 0.101,
+      "step": 7208
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.113539642080664,
+      "learning_rate": 2.7950848023075194e-06,
+      "loss": 0.1263,
+      "step": 7209
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.587324320892791,
+      "learning_rate": 2.7937622468618906e-06,
+      "loss": 0.0868,
+      "step": 7210
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.861382463575405,
+      "learning_rate": 2.792439883070808e-06,
+      "loss": 0.0958,
+      "step": 7211
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.592869652685055,
+      "learning_rate": 2.7911177110491485e-06,
+      "loss": 0.0879,
+      "step": 7212
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.3316585634886697,
+      "learning_rate": 2.789795730911764e-06,
+      "loss": 0.0839,
+      "step": 7213
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.6097170946132777,
+      "learning_rate": 2.788473942773501e-06,
+      "loss": 0.0924,
+      "step": 7214
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.7245079775835004,
+      "learning_rate": 2.787152346749173e-06,
+      "loss": 0.1085,
+      "step": 7215
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.6176809503393725,
+      "learning_rate": 2.7858309429535934e-06,
+      "loss": 0.0971,
+      "step": 7216
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.6880686883653846,
+      "learning_rate": 2.7845097315015477e-06,
+      "loss": 0.112,
+      "step": 7217
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.5862474980487622,
+      "learning_rate": 2.7831887125078128e-06,
+      "loss": 0.1006,
+      "step": 7218
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.7483993647632274,
+      "learning_rate": 2.7818678860871395e-06,
+      "loss": 0.1192,
+      "step": 7219
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.556856555694627,
+      "learning_rate": 2.7805472523542755e-06,
+      "loss": 0.0862,
+      "step": 7220
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.5409671225193207,
+      "learning_rate": 2.7792268114239336e-06,
+      "loss": 0.089,
+      "step": 7221
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.4690074078029538,
+      "learning_rate": 2.7779065634108265e-06,
+      "loss": 0.0894,
+      "step": 7222
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.054361122286564,
+      "learning_rate": 2.776586508429639e-06,
+      "loss": 0.1127,
+      "step": 7223
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.7072596980160584,
+      "learning_rate": 2.7752666465950477e-06,
+      "loss": 0.1003,
+      "step": 7224
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.5843975643675292,
+      "learning_rate": 2.773946978021704e-06,
+      "loss": 0.113,
+      "step": 7225
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.6073929650853112,
+      "learning_rate": 2.7726275028242532e-06,
+      "loss": 0.0893,
+      "step": 7226
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.8179193342958866,
+      "learning_rate": 2.771308221117309e-06,
+      "loss": 0.1217,
+      "step": 7227
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.8956889407110802,
+      "learning_rate": 2.7699891330154825e-06,
+      "loss": 0.1259,
+      "step": 7228
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.460337773586239,
+      "learning_rate": 2.7686702386333584e-06,
+      "loss": 0.0934,
+      "step": 7229
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.6424100393049181,
+      "learning_rate": 2.767351538085512e-06,
+      "loss": 0.0964,
+      "step": 7230
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.6913486566528593,
+      "learning_rate": 2.7660330314864937e-06,
+      "loss": 0.0873,
+      "step": 7231
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.701191391234638,
+      "learning_rate": 2.7647147189508485e-06,
+      "loss": 0.1144,
+      "step": 7232
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.6999291881361889,
+      "learning_rate": 2.763396600593088e-06,
+      "loss": 0.1094,
+      "step": 7233
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.7532582731084572,
+      "learning_rate": 2.762078676527723e-06,
+      "loss": 0.1127,
+      "step": 7234
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.7298472905038729,
+      "learning_rate": 2.760760946869237e-06,
+      "loss": 0.1039,
+      "step": 7235
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.5900557445657748,
+      "learning_rate": 2.7594434117321044e-06,
+      "loss": 0.0987,
+      "step": 7236
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.4951148741114944,
+      "learning_rate": 2.758126071230776e-06,
+      "loss": 0.0864,
+      "step": 7237
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.5728734917060558,
+      "learning_rate": 2.7568089254796893e-06,
+      "loss": 0.0885,
+      "step": 7238
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6464266458570087,
+      "learning_rate": 2.755491974593261e-06,
+      "loss": 0.092,
+      "step": 7239
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.5961460745524656,
+      "learning_rate": 2.754175218685899e-06,
+      "loss": 0.095,
+      "step": 7240
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.5216670022738257,
+      "learning_rate": 2.752858657871984e-06,
+      "loss": 0.0891,
+      "step": 7241
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.7064685439725689,
+      "learning_rate": 2.7515422922658895e-06,
+      "loss": 0.1132,
+      "step": 7242
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6062990593945012,
+      "learning_rate": 2.750226121981965e-06,
+      "loss": 0.1035,
+      "step": 7243
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.3464676455187288,
+      "learning_rate": 2.748910147134546e-06,
+      "loss": 0.0757,
+      "step": 7244
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.8028790340029268,
+      "learning_rate": 2.7475943678379474e-06,
+      "loss": 0.1146,
+      "step": 7245
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.486442315788568,
+      "learning_rate": 2.7462787842064753e-06,
+      "loss": 0.0904,
+      "step": 7246
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.3316406424539249,
+      "learning_rate": 2.7449633963544085e-06,
+      "loss": 0.0765,
+      "step": 7247
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6877050824710027,
+      "learning_rate": 2.743648204396019e-06,
+      "loss": 0.1097,
+      "step": 7248
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.4288566936345768,
+      "learning_rate": 2.7423332084455543e-06,
+      "loss": 0.0989,
+      "step": 7249
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.651835742041863,
+      "learning_rate": 2.7410184086172477e-06,
+      "loss": 0.1033,
+      "step": 7250
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.5573793818989963,
+      "learning_rate": 2.7397038050253122e-06,
+      "loss": 0.0933,
+      "step": 7251
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6267708221021036,
+      "learning_rate": 2.7383893977839513e-06,
+      "loss": 0.1049,
+      "step": 7252
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.5069900939669867,
+      "learning_rate": 2.7370751870073433e-06,
+      "loss": 0.1052,
+      "step": 7253
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.67892337249066,
+      "learning_rate": 2.7357611728096554e-06,
+      "loss": 0.1056,
+      "step": 7254
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6107600017331722,
+      "learning_rate": 2.7344473553050343e-06,
+      "loss": 0.0906,
+      "step": 7255
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6241641450450923,
+      "learning_rate": 2.7331337346076105e-06,
+      "loss": 0.1056,
+      "step": 7256
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.785500657253143,
+      "learning_rate": 2.7318203108314946e-06,
+      "loss": 0.1091,
+      "step": 7257
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.54328979219966,
+      "learning_rate": 2.7305070840907878e-06,
+      "loss": 0.0999,
+      "step": 7258
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.7276050688772684,
+      "learning_rate": 2.7291940544995655e-06,
+      "loss": 0.096,
+      "step": 7259
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6506816192588067,
+      "learning_rate": 2.7278812221718927e-06,
+      "loss": 0.1107,
+      "step": 7260
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.631150693250893,
+      "learning_rate": 2.7265685872218133e-06,
+      "loss": 0.1026,
+      "step": 7261
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.009749878469168,
+      "learning_rate": 2.7252561497633546e-06,
+      "loss": 0.1256,
+      "step": 7262
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.7850088002472462,
+      "learning_rate": 2.723943909910526e-06,
+      "loss": 0.1281,
+      "step": 7263
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6519711705112885,
+      "learning_rate": 2.7226318677773243e-06,
+      "loss": 0.0953,
+      "step": 7264
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.7023963356303091,
+      "learning_rate": 2.7213200234777215e-06,
+      "loss": 0.1058,
+      "step": 7265
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6359390387739592,
+      "learning_rate": 2.720008377125682e-06,
+      "loss": 0.1044,
+      "step": 7266
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.5748952679884474,
+      "learning_rate": 2.7186969288351438e-06,
+      "loss": 0.0974,
+      "step": 7267
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.5528687813587494,
+      "learning_rate": 2.717385678720034e-06,
+      "loss": 0.0956,
+      "step": 7268
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.552913459266542,
+      "learning_rate": 2.716074626894256e-06,
+      "loss": 0.085,
+      "step": 7269
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6228902919900716,
+      "learning_rate": 2.714763773471706e-06,
+      "loss": 0.1017,
+      "step": 7270
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.7527929220348546,
+      "learning_rate": 2.7134531185662503e-06,
+      "loss": 0.1131,
+      "step": 7271
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.3411667009347046,
+      "learning_rate": 2.712142662291752e-06,
+      "loss": 0.0756,
+      "step": 7272
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7774831166574165,
+      "learning_rate": 2.710832404762045e-06,
+      "loss": 0.1089,
+      "step": 7273
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.5581469711660068,
+      "learning_rate": 2.7095223460909527e-06,
+      "loss": 0.091,
+      "step": 7274
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7317721214944743,
+      "learning_rate": 2.7082124863922753e-06,
+      "loss": 0.1099,
+      "step": 7275
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.3953330923989558,
+      "learning_rate": 2.706902825779804e-06,
+      "loss": 0.0862,
+      "step": 7276
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.5237344771688994,
+      "learning_rate": 2.705593364367305e-06,
+      "loss": 0.0965,
+      "step": 7277
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.8694709926866722,
+      "learning_rate": 2.704284102268534e-06,
+      "loss": 0.1121,
+      "step": 7278
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7153002321741082,
+      "learning_rate": 2.702975039597223e-06,
+      "loss": 0.1029,
+      "step": 7279
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.6831223233200074,
+      "learning_rate": 2.7016661764670917e-06,
+      "loss": 0.0984,
+      "step": 7280
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.6794019385659853,
+      "learning_rate": 2.700357512991836e-06,
+      "loss": 0.1023,
+      "step": 7281
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.5072118646135964,
+      "learning_rate": 2.6990490492851408e-06,
+      "loss": 0.084,
+      "step": 7282
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7878137781908692,
+      "learning_rate": 2.697740785460675e-06,
+      "loss": 0.1213,
+      "step": 7283
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7805820895861189,
+      "learning_rate": 2.696432721632082e-06,
+      "loss": 0.1245,
+      "step": 7284
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.4809674169756688,
+      "learning_rate": 2.695124857912998e-06,
+      "loss": 0.0829,
+      "step": 7285
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.6097491242758988,
+      "learning_rate": 2.693817194417029e-06,
+      "loss": 0.0996,
+      "step": 7286
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.5983240073255753,
+      "learning_rate": 2.6925097312577766e-06,
+      "loss": 0.089,
+      "step": 7287
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7929460376267201,
+      "learning_rate": 2.6912024685488157e-06,
+      "loss": 0.106,
+      "step": 7288
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.5870922776565113,
+      "learning_rate": 2.6898954064037107e-06,
+      "loss": 0.0813,
+      "step": 7289
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7451435979680616,
+      "learning_rate": 2.6885885449360027e-06,
+      "loss": 0.1053,
+      "step": 7290
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.6579836738803748,
+      "learning_rate": 2.687281884259223e-06,
+      "loss": 0.103,
+      "step": 7291
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7626971308146393,
+      "learning_rate": 2.685975424486872e-06,
+      "loss": 0.0952,
+      "step": 7292
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.610609072312392,
+      "learning_rate": 2.6846691657324473e-06,
+      "loss": 0.1025,
+      "step": 7293
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.5128506597178897,
+      "learning_rate": 2.6833631081094197e-06,
+      "loss": 0.0804,
+      "step": 7294
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7390467726294139,
+      "learning_rate": 2.682057251731249e-06,
+      "loss": 0.0992,
+      "step": 7295
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.8714125602694576,
+      "learning_rate": 2.68075159671137e-06,
+      "loss": 0.1181,
+      "step": 7296
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7170208327461447,
+      "learning_rate": 2.67944614316321e-06,
+      "loss": 0.1028,
+      "step": 7297
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.747307017217943,
+      "learning_rate": 2.678140891200166e-06,
+      "loss": 0.103,
+      "step": 7298
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7633026272952508,
+      "learning_rate": 2.67683584093563e-06,
+      "loss": 0.1136,
+      "step": 7299
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.4899012096744266,
+      "learning_rate": 2.6755309924829657e-06,
+      "loss": 0.0832,
+      "step": 7300
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.0455858058549876,
+      "learning_rate": 2.67422634595553e-06,
+      "loss": 0.1199,
+      "step": 7301
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7978889943576057,
+      "learning_rate": 2.6729219014666525e-06,
+      "loss": 0.1008,
+      "step": 7302
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.6481999397348024,
+      "learning_rate": 2.671617659129655e-06,
+      "loss": 0.0862,
+      "step": 7303
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.9390660617091382,
+      "learning_rate": 2.6703136190578287e-06,
+      "loss": 0.1134,
+      "step": 7304
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.617260294334028,
+      "learning_rate": 2.6690097813644605e-06,
+      "loss": 0.1002,
+      "step": 7305
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7063717681422734,
+      "learning_rate": 2.6677061461628107e-06,
+      "loss": 0.107,
+      "step": 7306
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7514628430293397,
+      "learning_rate": 2.6664027135661276e-06,
+      "loss": 0.1077,
+      "step": 7307
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6180866602492607,
+      "learning_rate": 2.6650994836876375e-06,
+      "loss": 0.0987,
+      "step": 7308
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.5463802326109355,
+      "learning_rate": 2.663796456640556e-06,
+      "loss": 0.1033,
+      "step": 7309
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6570079090517802,
+      "learning_rate": 2.662493632538069e-06,
+      "loss": 0.0999,
+      "step": 7310
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.8068238154761658,
+      "learning_rate": 2.6611910114933574e-06,
+      "loss": 0.1106,
+      "step": 7311
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.605067896762438,
+      "learning_rate": 2.6598885936195764e-06,
+      "loss": 0.0979,
+      "step": 7312
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6824570821489555,
+      "learning_rate": 2.658586379029868e-06,
+      "loss": 0.108,
+      "step": 7313
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4339261810768502,
+      "learning_rate": 2.657284367837355e-06,
+      "loss": 0.0803,
+      "step": 7314
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6340442260029877,
+      "learning_rate": 2.6559825601551408e-06,
+      "loss": 0.0981,
+      "step": 7315
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.730102882418146,
+      "learning_rate": 2.6546809560963116e-06,
+      "loss": 0.0862,
+      "step": 7316
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.1579266581057506,
+      "learning_rate": 2.6533795557739407e-06,
+      "loss": 0.1177,
+      "step": 7317
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.9444952685341077,
+      "learning_rate": 2.6520783593010757e-06,
+      "loss": 0.1419,
+      "step": 7318
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.2848759468870636,
+      "learning_rate": 2.6507773667907556e-06,
+      "loss": 0.0662,
+      "step": 7319
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.5879438239999113,
+      "learning_rate": 2.6494765783559933e-06,
+      "loss": 0.0893,
+      "step": 7320
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.5386976159788832,
+      "learning_rate": 2.648175994109789e-06,
+      "loss": 0.0934,
+      "step": 7321
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6614816690175895,
+      "learning_rate": 2.646875614165121e-06,
+      "loss": 0.1043,
+      "step": 7322
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.619953320479506,
+      "learning_rate": 2.6455754386349564e-06,
+      "loss": 0.0925,
+      "step": 7323
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4722674062384284,
+      "learning_rate": 2.6442754676322367e-06,
+      "loss": 0.0805,
+      "step": 7324
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4799403335139003,
+      "learning_rate": 2.642975701269894e-06,
+      "loss": 0.082,
+      "step": 7325
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.435261160281723,
+      "learning_rate": 2.6416761396608365e-06,
+      "loss": 0.0748,
+      "step": 7326
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.3568963328331998,
+      "learning_rate": 2.6403767829179554e-06,
+      "loss": 0.0637,
+      "step": 7327
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.2740169651542217,
+      "learning_rate": 2.6390776311541233e-06,
+      "loss": 0.0563,
+      "step": 7328
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.2567758384707397,
+      "learning_rate": 2.6377786844822016e-06,
+      "loss": 0.0563,
+      "step": 7329
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.3120807912587755,
+      "learning_rate": 2.6364799430150233e-06,
+      "loss": 0.065,
+      "step": 7330
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.217007382454917,
+      "learning_rate": 2.635181406865415e-06,
+      "loss": 0.0527,
+      "step": 7331
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.3872972378576565,
+      "learning_rate": 2.6338830761461775e-06,
+      "loss": 0.0723,
+      "step": 7332
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4179324463064191,
+      "learning_rate": 2.6325849509700952e-06,
+      "loss": 0.0591,
+      "step": 7333
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.2798805250707819,
+      "learning_rate": 2.6312870314499335e-06,
+      "loss": 0.0588,
+      "step": 7334
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1434480405905167,
+      "learning_rate": 2.629989317698446e-06,
+      "loss": 0.0476,
+      "step": 7335
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1577792024372717,
+      "learning_rate": 2.628691809828361e-06,
+      "loss": 0.0471,
+      "step": 7336
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4374254104606945,
+      "learning_rate": 2.6273945079523955e-06,
+      "loss": 0.0514,
+      "step": 7337
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.372018751991338,
+      "learning_rate": 2.626097412183244e-06,
+      "loss": 0.0537,
+      "step": 7338
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4467715580539955,
+      "learning_rate": 2.624800522633584e-06,
+      "loss": 0.0585,
+      "step": 7339
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.387471906589724,
+      "learning_rate": 2.623503839416073e-06,
+      "loss": 0.0508,
+      "step": 7340
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4256426230446908,
+      "learning_rate": 2.6222073626433587e-06,
+      "loss": 0.0522,
+      "step": 7341
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6542905353343873,
+      "learning_rate": 2.620911092428059e-06,
+      "loss": 0.0556,
+      "step": 7342
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6773813074210058,
+      "learning_rate": 2.619615028882786e-06,
+      "loss": 0.0514,
+      "step": 7343
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.408097651400215,
+      "learning_rate": 2.618319172120125e-06,
+      "loss": 0.0433,
+      "step": 7344
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.3240745528452251,
+      "learning_rate": 2.6170235222526467e-06,
+      "loss": 0.0438,
+      "step": 7345
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4711598680949536,
+      "learning_rate": 2.615728079392902e-06,
+      "loss": 0.0449,
+      "step": 7346
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.8441296659576043,
+      "learning_rate": 2.614432843653427e-06,
+      "loss": 0.0701,
+      "step": 7347
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.6548224370969575,
+      "learning_rate": 2.6131378151467367e-06,
+      "loss": 0.0522,
+      "step": 7348
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.7068503164507207,
+      "learning_rate": 2.6118429939853324e-06,
+      "loss": 0.0565,
+      "step": 7349
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.6759990171934467,
+      "learning_rate": 2.6105483802816922e-06,
+      "loss": 0.0518,
+      "step": 7350
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.2740941025993895,
+      "learning_rate": 2.609253974148278e-06,
+      "loss": 0.0341,
+      "step": 7351
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4751201330416444,
+      "learning_rate": 2.6079597756975335e-06,
+      "loss": 0.0536,
+      "step": 7352
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.688395100770885,
+      "learning_rate": 2.6066657850418873e-06,
+      "loss": 0.0605,
+      "step": 7353
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.591832562245072,
+      "learning_rate": 2.6053720022937455e-06,
+      "loss": 0.0475,
+      "step": 7354
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.5753955779859328,
+      "learning_rate": 2.6040784275655008e-06,
+      "loss": 0.0472,
+      "step": 7355
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.5256774696345954,
+      "learning_rate": 2.6027850609695227e-06,
+      "loss": 0.0523,
+      "step": 7356
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.617633273581408,
+      "learning_rate": 2.601491902618167e-06,
+      "loss": 0.0491,
+      "step": 7357
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.558444597630549,
+      "learning_rate": 2.6001989526237658e-06,
+      "loss": 0.0549,
+      "step": 7358
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4077567773627562,
+      "learning_rate": 2.598906211098643e-06,
+      "loss": 0.0462,
+      "step": 7359
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4776447362686629,
+      "learning_rate": 2.597613678155092e-06,
+      "loss": 0.0439,
+      "step": 7360
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4333624272881615,
+      "learning_rate": 2.5963213539054e-06,
+      "loss": 0.0475,
+      "step": 7361
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4189783014139743,
+      "learning_rate": 2.595029238461827e-06,
+      "loss": 0.0431,
+      "step": 7362
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.3922648000311357,
+      "learning_rate": 2.59373733193662e-06,
+      "loss": 0.0524,
+      "step": 7363
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.5123864265977387,
+      "learning_rate": 2.592445634442003e-06,
+      "loss": 0.0493,
+      "step": 7364
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.6550714555230197,
+      "learning_rate": 2.591154146090189e-06,
+      "loss": 0.0511,
+      "step": 7365
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.446265961660988,
+      "learning_rate": 2.5898628669933657e-06,
+      "loss": 0.0627,
+      "step": 7366
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4225404011018752,
+      "learning_rate": 2.588571797263708e-06,
+      "loss": 0.0481,
+      "step": 7367
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.6611383821426202,
+      "learning_rate": 2.5872809370133704e-06,
+      "loss": 0.0529,
+      "step": 7368
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.3008208415512186,
+      "learning_rate": 2.5859902863544884e-06,
+      "loss": 0.0414,
+      "step": 7369
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4238148948160205,
+      "learning_rate": 2.5846998453991767e-06,
+      "loss": 0.0513,
+      "step": 7370
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4617546079497878,
+      "learning_rate": 2.583409614259541e-06,
+      "loss": 0.0476,
+      "step": 7371
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.0786328938246228,
+      "learning_rate": 2.5821195930476584e-06,
+      "loss": 0.0408,
+      "step": 7372
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.5300018109719198,
+      "learning_rate": 2.5808297818755956e-06,
+      "loss": 0.0494,
+      "step": 7373
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.1106590652643458,
+      "learning_rate": 2.5795401808553966e-06,
+      "loss": 0.0326,
+      "step": 7374
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.3793240569270968,
+      "learning_rate": 2.5782507900990863e-06,
+      "loss": 0.0535,
+      "step": 7375
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4542917979982788,
+      "learning_rate": 2.5769616097186757e-06,
+      "loss": 0.0489,
+      "step": 7376
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.6487980627515233,
+      "learning_rate": 2.575672639826153e-06,
+      "loss": 0.0517,
+      "step": 7377
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.433688994609693,
+      "learning_rate": 2.574383880533493e-06,
+      "loss": 0.0438,
+      "step": 7378
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.5210258353656534,
+      "learning_rate": 2.573095331952646e-06,
+      "loss": 0.0485,
+      "step": 7379
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4086045465835808,
+      "learning_rate": 2.5718069941955535e-06,
+      "loss": 0.046,
+      "step": 7380
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.2875572746444996,
+      "learning_rate": 2.5705188673741253e-06,
+      "loss": 0.057,
+      "step": 7381
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.4561133844636518,
+      "learning_rate": 2.5692309516002643e-06,
+      "loss": 0.0475,
+      "step": 7382
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.2815168270673225,
+      "learning_rate": 2.567943246985849e-06,
+      "loss": 0.0428,
+      "step": 7383
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.823361314780626,
+      "learning_rate": 2.5666557536427445e-06,
+      "loss": 0.0397,
+      "step": 7384
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.468079471269978,
+      "learning_rate": 2.5653684716827904e-06,
+      "loss": 0.0483,
+      "step": 7385
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.4127813118974655,
+      "learning_rate": 2.5640814012178182e-06,
+      "loss": 0.051,
+      "step": 7386
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.1636121526870367,
+      "learning_rate": 2.562794542359628e-06,
+      "loss": 0.0437,
+      "step": 7387
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.7615040124789034,
+      "learning_rate": 2.5615078952200125e-06,
+      "loss": 0.0605,
+      "step": 7388
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6512237153234959,
+      "learning_rate": 2.56022145991074e-06,
+      "loss": 0.0523,
+      "step": 7389
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6503977102223706,
+      "learning_rate": 2.558935236543565e-06,
+      "loss": 0.051,
+      "step": 7390
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.3603800327432807,
+      "learning_rate": 2.557649225230219e-06,
+      "loss": 0.0471,
+      "step": 7391
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.558341517709461,
+      "learning_rate": 2.556363426082418e-06,
+      "loss": 0.0518,
+      "step": 7392
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.41757345379561,
+      "learning_rate": 2.5550778392118557e-06,
+      "loss": 0.046,
+      "step": 7393
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.5505715332213015,
+      "learning_rate": 2.5537924647302146e-06,
+      "loss": 0.0597,
+      "step": 7394
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6234831606488078,
+      "learning_rate": 2.5525073027491504e-06,
+      "loss": 0.0462,
+      "step": 7395
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.5554681912972133,
+      "learning_rate": 2.5512223533803084e-06,
+      "loss": 0.0442,
+      "step": 7396
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6043505882955307,
+      "learning_rate": 2.5499376167353097e-06,
+      "loss": 0.0563,
+      "step": 7397
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.410214041468284,
+      "learning_rate": 2.5486530929257574e-06,
+      "loss": 0.0479,
+      "step": 7398
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.5013271592069959,
+      "learning_rate": 2.5473687820632365e-06,
+      "loss": 0.053,
+      "step": 7399
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.4422673286773653,
+      "learning_rate": 2.546084684259318e-06,
+      "loss": 0.0488,
+      "step": 7400
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.4239287133665368,
+      "learning_rate": 2.5448007996255463e-06,
+      "loss": 0.0504,
+      "step": 7401
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.383135698870463,
+      "learning_rate": 2.5435171282734563e-06,
+      "loss": 0.0419,
+      "step": 7402
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.4730217323645354,
+      "learning_rate": 2.542233670314558e-06,
+      "loss": 0.0547,
+      "step": 7403
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.4428264350338156,
+      "learning_rate": 2.5409504258603436e-06,
+      "loss": 0.0482,
+      "step": 7404
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.4679914670242349,
+      "learning_rate": 2.5396673950222863e-06,
+      "loss": 0.0509,
+      "step": 7405
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6079100410897562,
+      "learning_rate": 2.5383845779118453e-06,
+      "loss": 0.0533,
+      "step": 7406
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.5992690830740246,
+      "learning_rate": 2.5371019746404564e-06,
+      "loss": 0.0535,
+      "step": 7407
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.509529496556637,
+      "learning_rate": 2.535819585319541e-06,
+      "loss": 0.0513,
+      "step": 7408
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.3454993188122852,
+      "learning_rate": 2.534537410060497e-06,
+      "loss": 0.0445,
+      "step": 7409
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.396632980280583,
+      "learning_rate": 2.5332554489747076e-06,
+      "loss": 0.0507,
+      "step": 7410
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.5057522271667847,
+      "learning_rate": 2.531973702173533e-06,
+      "loss": 0.0521,
+      "step": 7411
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.549303596754514,
+      "learning_rate": 2.5306921697683216e-06,
+      "loss": 0.0484,
+      "step": 7412
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.2432104638500188,
+      "learning_rate": 2.529410851870397e-06,
+      "loss": 0.0393,
+      "step": 7413
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.610559354259884,
+      "learning_rate": 2.5281297485910684e-06,
+      "loss": 0.0577,
+      "step": 7414
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.7158197403596251,
+      "learning_rate": 2.526848860041624e-06,
+      "loss": 0.0498,
+      "step": 7415
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.4251336752988741,
+      "learning_rate": 2.5255681863333325e-06,
+      "loss": 0.0469,
+      "step": 7416
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6794591194943982,
+      "learning_rate": 2.5242877275774446e-06,
+      "loss": 0.0609,
+      "step": 7417
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.5752059402759766,
+      "learning_rate": 2.523007483885196e-06,
+      "loss": 0.0539,
+      "step": 7418
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.5811898924718548,
+      "learning_rate": 2.5217274553677975e-06,
+      "loss": 0.0482,
+      "step": 7419
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.4837178677679705,
+      "learning_rate": 2.5204476421364475e-06,
+      "loss": 0.0454,
+      "step": 7420
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6456402158089958,
+      "learning_rate": 2.5191680443023214e-06,
+      "loss": 0.0581,
+      "step": 7421
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.4849872239333415,
+      "learning_rate": 2.5178886619765764e-06,
+      "loss": 0.0484,
+      "step": 7422
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.5699893075348668,
+      "learning_rate": 2.516609495270351e-06,
+      "loss": 0.0577,
+      "step": 7423
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.5577087552874045,
+      "learning_rate": 2.515330544294768e-06,
+      "loss": 0.053,
+      "step": 7424
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.4895696556941862,
+      "learning_rate": 2.5140518091609254e-06,
+      "loss": 0.0417,
+      "step": 7425
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.390785552920445,
+      "learning_rate": 2.512773289979911e-06,
+      "loss": 0.0425,
+      "step": 7426
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.4305359561950748,
+      "learning_rate": 2.5114949868627867e-06,
+      "loss": 0.0452,
+      "step": 7427
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.3933955787765868,
+      "learning_rate": 2.510216899920598e-06,
+      "loss": 0.0467,
+      "step": 7428
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.232316743659909,
+      "learning_rate": 2.5089390292643686e-06,
+      "loss": 0.0525,
+      "step": 7429
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6024451789703997,
+      "learning_rate": 2.5076613750051113e-06,
+      "loss": 0.0496,
+      "step": 7430
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.5597350045047362,
+      "learning_rate": 2.5063839372538112e-06,
+      "loss": 0.0503,
+      "step": 7431
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.458295372823315,
+      "learning_rate": 2.5051067161214414e-06,
+      "loss": 0.0515,
+      "step": 7432
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.559836412126493,
+      "learning_rate": 2.5038297117189535e-06,
+      "loss": 0.052,
+      "step": 7433
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.5286929700263967,
+      "learning_rate": 2.502552924157278e-06,
+      "loss": 0.0468,
+      "step": 7434
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.5490301309509895,
+      "learning_rate": 2.501276353547327e-06,
+      "loss": 0.0479,
+      "step": 7435
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.344690950141608,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.0446,
+      "step": 7436
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6782629746954085,
+      "learning_rate": 2.4987238636261705e-06,
+      "loss": 0.0566,
+      "step": 7437
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.7506567297002045,
+      "learning_rate": 2.4974479445366973e-06,
+      "loss": 0.0535,
+      "step": 7438
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.595440966679077,
+      "learning_rate": 2.4961722428424177e-06,
+      "loss": 0.04,
+      "step": 7439
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.648204419915552,
+      "learning_rate": 2.4948967586541508e-06,
+      "loss": 0.0525,
+      "step": 7440
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6446411292681524,
+      "learning_rate": 2.4936214920826956e-06,
+      "loss": 0.0484,
+      "step": 7441
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.3760623284832816,
+      "learning_rate": 2.4923464432388373e-06,
+      "loss": 0.0448,
+      "step": 7442
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.643020447031329,
+      "learning_rate": 2.4910716122333352e-06,
+      "loss": 0.0496,
+      "step": 7443
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.3621315461022658,
+      "learning_rate": 2.489796999176936e-06,
+      "loss": 0.0428,
+      "step": 7444
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.474105265027352,
+      "learning_rate": 2.488522604180364e-06,
+      "loss": 0.0462,
+      "step": 7445
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.478295408482525,
+      "learning_rate": 2.487248427354324e-06,
+      "loss": 0.0517,
+      "step": 7446
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.535593749311935,
+      "learning_rate": 2.4859744688095015e-06,
+      "loss": 0.0492,
+      "step": 7447
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.601177095502935,
+      "learning_rate": 2.484700728656569e-06,
+      "loss": 0.0477,
+      "step": 7448
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.669743354640075,
+      "learning_rate": 2.4834272070061706e-06,
+      "loss": 0.0462,
+      "step": 7449
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.4411696460978156,
+      "learning_rate": 2.4821539039689404e-06,
+      "loss": 0.0427,
+      "step": 7450
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6604570465520747,
+      "learning_rate": 2.4808808196554877e-06,
+      "loss": 0.0532,
+      "step": 7451
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6137018233754035,
+      "learning_rate": 2.479607954176406e-06,
+      "loss": 0.0576,
+      "step": 7452
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.565220772210761,
+      "learning_rate": 2.478335307642264e-06,
+      "loss": 0.0485,
+      "step": 7453
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.4392445272742047,
+      "learning_rate": 2.4770628801636205e-06,
+      "loss": 0.0428,
+      "step": 7454
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.8259654542720174,
+      "learning_rate": 2.475790671851007e-06,
+      "loss": 0.0605,
+      "step": 7455
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.343402316311559,
+      "learning_rate": 2.4745186828149435e-06,
+      "loss": 0.0485,
+      "step": 7456
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.6412224725789413,
+      "learning_rate": 2.473246913165925e-06,
+      "loss": 0.0472,
+      "step": 7457
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.394987341594411,
+      "learning_rate": 2.4719753630144283e-06,
+      "loss": 0.0538,
+      "step": 7458
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.4981573299034243,
+      "learning_rate": 2.4707040324709115e-06,
+      "loss": 0.0503,
+      "step": 7459
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.3451634809157227,
+      "learning_rate": 2.469432921645818e-06,
+      "loss": 0.0442,
+      "step": 7460
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.4369916909061546,
+      "learning_rate": 2.4681620306495634e-06,
+      "loss": 0.0426,
+      "step": 7461
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.4846313308218781,
+      "learning_rate": 2.4668913595925548e-06,
+      "loss": 0.0521,
+      "step": 7462
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.277636252275021,
+      "learning_rate": 2.4656209085851712e-06,
+      "loss": 0.0428,
+      "step": 7463
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.3098100017505718,
+      "learning_rate": 2.464350677737777e-06,
+      "loss": 0.0391,
+      "step": 7464
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.427784442839232,
+      "learning_rate": 2.463080667160714e-06,
+      "loss": 0.0516,
+      "step": 7465
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.696646045226093,
+      "learning_rate": 2.4618108769643105e-06,
+      "loss": 0.054,
+      "step": 7466
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.5385584582936893,
+      "learning_rate": 2.4605413072588702e-06,
+      "loss": 0.0439,
+      "step": 7467
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.6374315347955912,
+      "learning_rate": 2.4592719581546826e-06,
+      "loss": 0.0505,
+      "step": 7468
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.6202139145784817,
+      "learning_rate": 2.458002829762013e-06,
+      "loss": 0.0502,
+      "step": 7469
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.6485678714691454,
+      "learning_rate": 2.4567339221911086e-06,
+      "loss": 0.0524,
+      "step": 7470
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.7735600327315095,
+      "learning_rate": 2.455465235552202e-06,
+      "loss": 0.0625,
+      "step": 7471
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.492149330739676,
+      "learning_rate": 2.4541967699555004e-06,
+      "loss": 0.0392,
+      "step": 7472
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.4075805997822857,
+      "learning_rate": 2.4529285255111974e-06,
+      "loss": 0.0462,
+      "step": 7473
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.5386038381665181,
+      "learning_rate": 2.4516605023294626e-06,
+      "loss": 0.0466,
+      "step": 7474
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.367320301506333,
+      "learning_rate": 2.4503927005204497e-06,
+      "loss": 0.045,
+      "step": 7475
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.659796747824848,
+      "learning_rate": 2.4491251201942882e-06,
+      "loss": 0.0494,
+      "step": 7476
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.527278945743993,
+      "learning_rate": 2.4478577614610975e-06,
+      "loss": 0.0508,
+      "step": 7477
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.4836560247064783,
+      "learning_rate": 2.4465906244309677e-06,
+      "loss": 0.0416,
+      "step": 7478
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.4239991068320188,
+      "learning_rate": 2.445323709213978e-06,
+      "loss": 0.041,
+      "step": 7479
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.226835209340665,
+      "learning_rate": 2.444057015920183e-06,
+      "loss": 0.039,
+      "step": 7480
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.5524655882130292,
+      "learning_rate": 2.4427905446596194e-06,
+      "loss": 0.0438,
+      "step": 7481
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.3302898912936039,
+      "learning_rate": 2.441524295542303e-06,
+      "loss": 0.0455,
+      "step": 7482
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.4749669997387318,
+      "learning_rate": 2.4402582686782354e-06,
+      "loss": 0.0528,
+      "step": 7483
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.6239266235656735,
+      "learning_rate": 2.4389924641773925e-06,
+      "loss": 0.0541,
+      "step": 7484
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.7063224063434792,
+      "learning_rate": 2.4377268821497375e-06,
+      "loss": 0.0518,
+      "step": 7485
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.5313408510321915,
+      "learning_rate": 2.4364615227052086e-06,
+      "loss": 0.0488,
+      "step": 7486
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.4511069205582734,
+      "learning_rate": 2.435196385953727e-06,
+      "loss": 0.043,
+      "step": 7487
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.3635326906107355,
+      "learning_rate": 2.4339314720051927e-06,
+      "loss": 0.0445,
+      "step": 7488
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.63884359841843,
+      "learning_rate": 2.432666780969491e-06,
+      "loss": 0.0471,
+      "step": 7489
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.5528716654981882,
+      "learning_rate": 2.4314023129564824e-06,
+      "loss": 0.0459,
+      "step": 7490
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.5044901073500467,
+      "learning_rate": 2.430138068076013e-06,
+      "loss": 0.0505,
+      "step": 7491
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.3804834922620526,
+      "learning_rate": 2.4288740464379057e-06,
+      "loss": 0.055,
+      "step": 7492
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.5485970606536095,
+      "learning_rate": 2.4276102481519655e-06,
+      "loss": 0.055,
+      "step": 7493
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.8198067453404883,
+      "learning_rate": 2.4263466733279756e-06,
+      "loss": 0.0538,
+      "step": 7494
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.248492221541558,
+      "learning_rate": 2.4250833220757054e-06,
+      "loss": 0.0402,
+      "step": 7495
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.393824966633775,
+      "learning_rate": 2.4238201945048983e-06,
+      "loss": 0.0433,
+      "step": 7496
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.8301016736329818,
+      "learning_rate": 2.4225572907252853e-06,
+      "loss": 0.0654,
+      "step": 7497
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.6364757796590972,
+      "learning_rate": 2.421294610846571e-06,
+      "loss": 0.0634,
+      "step": 7498
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.4973883996622943,
+      "learning_rate": 2.4200321549784455e-06,
+      "loss": 0.0544,
+      "step": 7499
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.4207472298863366,
+      "learning_rate": 2.4187699232305745e-06,
+      "loss": 0.0505,
+      "step": 7500
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.639412770186421,
+      "learning_rate": 2.4175079157126115e-06,
+      "loss": 0.0556,
+      "step": 7501
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.5859266317595415,
+      "learning_rate": 2.4162461325341816e-06,
+      "loss": 0.0483,
+      "step": 7502
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.688275537229775,
+      "learning_rate": 2.4149845738049007e-06,
+      "loss": 0.0532,
+      "step": 7503
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.462156957982611,
+      "learning_rate": 2.413723239634356e-06,
+      "loss": 0.0446,
+      "step": 7504
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.3320981125435771,
+      "learning_rate": 2.41246213013212e-06,
+      "loss": 0.0477,
+      "step": 7505
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.496831387504434,
+      "learning_rate": 2.4112012454077422e-06,
+      "loss": 0.0528,
+      "step": 7506
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.4519157371185114,
+      "learning_rate": 2.4099405855707585e-06,
+      "loss": 0.0537,
+      "step": 7507
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.519177619896096,
+      "learning_rate": 2.4086801507306783e-06,
+      "loss": 0.0439,
+      "step": 7508
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.4687177155402915,
+      "learning_rate": 2.4074199409969984e-06,
+      "loss": 0.0471,
+      "step": 7509
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.5134619785816281,
+      "learning_rate": 2.4061599564791906e-06,
+      "loss": 0.0478,
+      "step": 7510
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.3224667379931698,
+      "learning_rate": 2.4049001972867086e-06,
+      "loss": 0.0421,
+      "step": 7511
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.6547161636983905,
+      "learning_rate": 2.403640663528986e-06,
+      "loss": 0.062,
+      "step": 7512
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.279436450402271,
+      "learning_rate": 2.402381355315441e-06,
+      "loss": 0.0402,
+      "step": 7513
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.6079417964693181,
+      "learning_rate": 2.401122272755464e-06,
+      "loss": 0.053,
+      "step": 7514
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.604471758190186,
+      "learning_rate": 2.3998634159584365e-06,
+      "loss": 0.0456,
+      "step": 7515
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.4442918534026248,
+      "learning_rate": 2.398604785033712e-06,
+      "loss": 0.0543,
+      "step": 7516
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.2814003626851682,
+      "learning_rate": 2.397346380090626e-06,
+      "loss": 0.0443,
+      "step": 7517
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.3672364848158798,
+      "learning_rate": 2.396088201238495e-06,
+      "loss": 0.0494,
+      "step": 7518
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.2890424896285348,
+      "learning_rate": 2.3948302485866194e-06,
+      "loss": 0.0399,
+      "step": 7519
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.623200219523274,
+      "learning_rate": 2.3935725222442728e-06,
+      "loss": 0.0478,
+      "step": 7520
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.4165671444426533,
+      "learning_rate": 2.3923150223207176e-06,
+      "loss": 0.0396,
+      "step": 7521
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.5192968090708392,
+      "learning_rate": 2.391057748925189e-06,
+      "loss": 0.0501,
+      "step": 7522
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.6285446009097,
+      "learning_rate": 2.3898007021669068e-06,
+      "loss": 0.0498,
+      "step": 7523
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.627294506939516,
+      "learning_rate": 2.388543882155067e-06,
+      "loss": 0.0551,
+      "step": 7524
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.079491390926197,
+      "learning_rate": 2.3872872889988535e-06,
+      "loss": 0.0694,
+      "step": 7525
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.8523640841048559,
+      "learning_rate": 2.3860309228074213e-06,
+      "loss": 0.0613,
+      "step": 7526
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.4936739729095836,
+      "learning_rate": 2.3847747836899144e-06,
+      "loss": 0.0455,
+      "step": 7527
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.387453667191235,
+      "learning_rate": 2.383518871755451e-06,
+      "loss": 0.0471,
+      "step": 7528
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.5305233749771534,
+      "learning_rate": 2.3822631871131306e-06,
+      "loss": 0.0566,
+      "step": 7529
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.4275702727064148,
+      "learning_rate": 2.381007729872033e-06,
+      "loss": 0.0433,
+      "step": 7530
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.2058282063974046,
+      "learning_rate": 2.379752500141222e-06,
+      "loss": 0.0417,
+      "step": 7531
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.5051223681557075,
+      "learning_rate": 2.378497498029735e-06,
+      "loss": 0.0462,
+      "step": 7532
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.399590767431314,
+      "learning_rate": 2.3772427236465974e-06,
+      "loss": 0.0518,
+      "step": 7533
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.4756301228571063,
+      "learning_rate": 2.3759881771008088e-06,
+      "loss": 0.0474,
+      "step": 7534
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.425308484059849,
+      "learning_rate": 2.37473385850135e-06,
+      "loss": 0.0497,
+      "step": 7535
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.6695545738068702,
+      "learning_rate": 2.3734797679571826e-06,
+      "loss": 0.0515,
+      "step": 7536
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.56954603327496,
+      "learning_rate": 2.372225905577251e-06,
+      "loss": 0.048,
+      "step": 7537
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.4803808616726128,
+      "learning_rate": 2.370972271470475e-06,
+      "loss": 0.0467,
+      "step": 7538
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.624043540279434,
+      "learning_rate": 2.3697188657457592e-06,
+      "loss": 0.061,
+      "step": 7539
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.028672661928012,
+      "learning_rate": 2.3684656885119856e-06,
+      "loss": 0.0586,
+      "step": 7540
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.195633559454792,
+      "learning_rate": 2.367212739878017e-06,
+      "loss": 0.0416,
+      "step": 7541
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.5847414277539829,
+      "learning_rate": 2.3659600199526933e-06,
+      "loss": 0.0484,
+      "step": 7542
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.4604825290204053,
+      "learning_rate": 2.3647075288448423e-06,
+      "loss": 0.0499,
+      "step": 7543
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.3517427438559508,
+      "learning_rate": 2.3634552666632633e-06,
+      "loss": 0.0442,
+      "step": 7544
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.634010794142163,
+      "learning_rate": 2.362203233516743e-06,
+      "loss": 0.0635,
+      "step": 7545
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.395742698915889,
+      "learning_rate": 2.360951429514043e-06,
+      "loss": 0.0481,
+      "step": 7546
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.4699971577633688,
+      "learning_rate": 2.3596998547639066e-06,
+      "loss": 0.0493,
+      "step": 7547
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.5654270394048593,
+      "learning_rate": 2.3584485093750554e-06,
+      "loss": 0.0429,
+      "step": 7548
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.5530620315914359,
+      "learning_rate": 2.3571973934561978e-06,
+      "loss": 0.0551,
+      "step": 7549
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.547121438041144,
+      "learning_rate": 2.355946507116012e-06,
+      "loss": 0.0509,
+      "step": 7550
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.3625585103568778,
+      "learning_rate": 2.3546958504631666e-06,
+      "loss": 0.0326,
+      "step": 7551
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.596851983142129,
+      "learning_rate": 2.3534454236063036e-06,
+      "loss": 0.0529,
+      "step": 7552
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.457607478887334,
+      "learning_rate": 2.3521952266540466e-06,
+      "loss": 0.0431,
+      "step": 7553
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.5848549870324236,
+      "learning_rate": 2.3509452597149972e-06,
+      "loss": 0.0514,
+      "step": 7554
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.5093190023146514,
+      "learning_rate": 2.3496955228977437e-06,
+      "loss": 0.0512,
+      "step": 7555
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.4306186650122474,
+      "learning_rate": 2.3484460163108457e-06,
+      "loss": 0.0518,
+      "step": 7556
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.6820046827272304,
+      "learning_rate": 2.3471967400628513e-06,
+      "loss": 0.0543,
+      "step": 7557
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.8762232605375888,
+      "learning_rate": 2.3459476942622823e-06,
+      "loss": 0.0578,
+      "step": 7558
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.4650275461258537,
+      "learning_rate": 2.3446988790176425e-06,
+      "loss": 0.0432,
+      "step": 7559
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.279400585314781,
+      "learning_rate": 2.3434502944374137e-06,
+      "loss": 0.0413,
+      "step": 7560
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.401065587262593,
+      "learning_rate": 2.3422019406300617e-06,
+      "loss": 0.0431,
+      "step": 7561
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.7711106598792907,
+      "learning_rate": 2.3409538177040324e-06,
+      "loss": 0.0426,
+      "step": 7562
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.5488316766119747,
+      "learning_rate": 2.339705925767747e-06,
+      "loss": 0.0441,
+      "step": 7563
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.4892965854736837,
+      "learning_rate": 2.3384582649296093e-06,
+      "loss": 0.0492,
+      "step": 7564
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.3209549733764223,
+      "learning_rate": 2.337210835298002e-06,
+      "loss": 0.0411,
+      "step": 7565
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.506860795917665,
+      "learning_rate": 2.335963636981291e-06,
+      "loss": 0.0535,
+      "step": 7566
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.4824930722399863,
+      "learning_rate": 2.3347166700878165e-06,
+      "loss": 0.0435,
+      "step": 7567
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.5624416421719172,
+      "learning_rate": 2.3334699347259053e-06,
+      "loss": 0.0478,
+      "step": 7568
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.469282830944571,
+      "learning_rate": 2.332223431003859e-06,
+      "loss": 0.0348,
+      "step": 7569
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.6469403463427092,
+      "learning_rate": 2.33097715902996e-06,
+      "loss": 0.0562,
+      "step": 7570
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.6192635478897062,
+      "learning_rate": 2.32973111891247e-06,
+      "loss": 0.0451,
+      "step": 7571
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.4604251326654094,
+      "learning_rate": 2.328485310759635e-06,
+      "loss": 0.0466,
+      "step": 7572
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.3994967606709425,
+      "learning_rate": 2.3272397346796743e-06,
+      "loss": 0.0422,
+      "step": 7573
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.4450697974333713,
+      "learning_rate": 2.325994390780794e-06,
+      "loss": 0.051,
+      "step": 7574
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.346727976837846,
+      "learning_rate": 2.3247492791711744e-06,
+      "loss": 0.0439,
+      "step": 7575
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.748942646475676,
+      "learning_rate": 2.323504399958978e-06,
+      "loss": 0.0541,
+      "step": 7576
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.3682559555317957,
+      "learning_rate": 2.322259753252344e-06,
+      "loss": 0.0444,
+      "step": 7577
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.571304413620297,
+      "learning_rate": 2.3210153391593978e-06,
+      "loss": 0.0428,
+      "step": 7578
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.4523939815908802,
+      "learning_rate": 2.319771157788238e-06,
+      "loss": 0.0552,
+      "step": 7579
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.943102107899454,
+      "learning_rate": 2.3185272092469497e-06,
+      "loss": 0.057,
+      "step": 7580
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.450829826275882,
+      "learning_rate": 2.3172834936435913e-06,
+      "loss": 0.0492,
+      "step": 7581
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.5851933402831677,
+      "learning_rate": 2.316040011086204e-06,
+      "loss": 0.0517,
+      "step": 7582
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.589527550328945,
+      "learning_rate": 2.3147967616828067e-06,
+      "loss": 0.045,
+      "step": 7583
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.2777718755646204,
+      "learning_rate": 2.313553745541403e-06,
+      "loss": 0.037,
+      "step": 7584
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.4591973466812223,
+      "learning_rate": 2.3123109627699695e-06,
+      "loss": 0.0506,
+      "step": 7585
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.502324905492929,
+      "learning_rate": 2.31106841347647e-06,
+      "loss": 0.0501,
+      "step": 7586
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.5512376524106535,
+      "learning_rate": 2.3098260977688412e-06,
+      "loss": 0.0511,
+      "step": 7587
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.5601547479680107,
+      "learning_rate": 2.3085840157550036e-06,
+      "loss": 0.0558,
+      "step": 7588
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.7306740814267172,
+      "learning_rate": 2.307342167542854e-06,
+      "loss": 0.0551,
+      "step": 7589
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.492895469026317,
+      "learning_rate": 2.306100553240274e-06,
+      "loss": 0.0408,
+      "step": 7590
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.6068578787089192,
+      "learning_rate": 2.3048591729551184e-06,
+      "loss": 0.0489,
+      "step": 7591
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.5158334598622607,
+      "learning_rate": 2.303618026795229e-06,
+      "loss": 0.0439,
+      "step": 7592
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.5983196453549129,
+      "learning_rate": 2.302377114868422e-06,
+      "loss": 0.04,
+      "step": 7593
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.7551625050848039,
+      "learning_rate": 2.301136437282494e-06,
+      "loss": 0.0456,
+      "step": 7594
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.6857404180284448,
+      "learning_rate": 2.2998959941452203e-06,
+      "loss": 0.0612,
+      "step": 7595
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.400021033123697,
+      "learning_rate": 2.2986557855643617e-06,
+      "loss": 0.0443,
+      "step": 7596
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.8589435909544976,
+      "learning_rate": 2.297415811647649e-06,
+      "loss": 0.064,
+      "step": 7597
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.4273731790347688,
+      "learning_rate": 2.2961760725028036e-06,
+      "loss": 0.0459,
+      "step": 7598
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.5788197167286022,
+      "learning_rate": 2.2949365682375185e-06,
+      "loss": 0.0499,
+      "step": 7599
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.472154848246478,
+      "learning_rate": 2.2936972989594684e-06,
+      "loss": 0.0549,
+      "step": 7600
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.3366290195493562,
+      "learning_rate": 2.292458264776306e-06,
+      "loss": 0.0383,
+      "step": 7601
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.675983968477484,
+      "learning_rate": 2.291219465795669e-06,
+      "loss": 0.0511,
+      "step": 7602
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.544732135812631,
+      "learning_rate": 2.289980902125168e-06,
+      "loss": 0.0524,
+      "step": 7603
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.376764389317635,
+      "learning_rate": 2.2887425738723994e-06,
+      "loss": 0.0487,
+      "step": 7604
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.7467665709834814,
+      "learning_rate": 2.2875044811449347e-06,
+      "loss": 0.0565,
+      "step": 7605
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.3727162744176389,
+      "learning_rate": 2.286266624050326e-06,
+      "loss": 0.0439,
+      "step": 7606
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.7540345273310114,
+      "learning_rate": 2.2850290026961032e-06,
+      "loss": 0.0544,
+      "step": 7607
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.4530033901467136,
+      "learning_rate": 2.2837916171897816e-06,
+      "loss": 0.0465,
+      "step": 7608
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.518431323614025,
+      "learning_rate": 2.282554467638849e-06,
+      "loss": 0.0463,
+      "step": 7609
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.4532104231630625,
+      "learning_rate": 2.2813175541507782e-06,
+      "loss": 0.0459,
+      "step": 7610
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.4851585783861172,
+      "learning_rate": 2.2800808768330184e-06,
+      "loss": 0.0438,
+      "step": 7611
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.8190002962686183,
+      "learning_rate": 2.278844435792998e-06,
+      "loss": 0.0524,
+      "step": 7612
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.42588456768309,
+      "learning_rate": 2.277608231138126e-06,
+      "loss": 0.046,
+      "step": 7613
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.2059015319087816,
+      "learning_rate": 2.2763722629757924e-06,
+      "loss": 0.0373,
+      "step": 7614
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.364858285271689,
+      "learning_rate": 2.2751365314133623e-06,
+      "loss": 0.0408,
+      "step": 7615
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.289317735072688,
+      "learning_rate": 2.2739010365581866e-06,
+      "loss": 0.0334,
+      "step": 7616
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.4027579949343196,
+      "learning_rate": 2.2726657785175892e-06,
+      "loss": 0.0493,
+      "step": 7617
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.3856619102350474,
+      "learning_rate": 2.2714307573988776e-06,
+      "loss": 0.0415,
+      "step": 7618
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.617443408700295,
+      "learning_rate": 2.2701959733093347e-06,
+      "loss": 0.0497,
+      "step": 7619
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.5006837865932632,
+      "learning_rate": 2.2689614263562297e-06,
+      "loss": 0.0435,
+      "step": 7620
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.460853769798232,
+      "learning_rate": 2.2677271166468024e-06,
+      "loss": 0.0432,
+      "step": 7621
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.7579801425646928,
+      "learning_rate": 2.266493044288281e-06,
+      "loss": 0.0617,
+      "step": 7622
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.599145519630225,
+      "learning_rate": 2.265259209387867e-06,
+      "loss": 0.0482,
+      "step": 7623
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.46630507774899,
+      "learning_rate": 2.2640256120527413e-06,
+      "loss": 0.0502,
+      "step": 7624
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.3967425420194994,
+      "learning_rate": 2.262792252390066e-06,
+      "loss": 0.0418,
+      "step": 7625
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.775530077724054,
+      "learning_rate": 2.2615591305069846e-06,
+      "loss": 0.0568,
+      "step": 7626
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.3429775183850952,
+      "learning_rate": 2.2603262465106147e-06,
+      "loss": 0.0361,
+      "step": 7627
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.592797036798222,
+      "learning_rate": 2.2590936005080594e-06,
+      "loss": 0.0573,
+      "step": 7628
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.4405335971063797,
+      "learning_rate": 2.257861192606396e-06,
+      "loss": 0.0481,
+      "step": 7629
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.3926611469973187,
+      "learning_rate": 2.2566290229126837e-06,
+      "loss": 0.0403,
+      "step": 7630
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.4336549257731726,
+      "learning_rate": 2.255397091533958e-06,
+      "loss": 0.0453,
+      "step": 7631
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.4193505176959316,
+      "learning_rate": 2.2541653985772394e-06,
+      "loss": 0.0486,
+      "step": 7632
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.4782932694553221,
+      "learning_rate": 2.252933944149522e-06,
+      "loss": 0.0379,
+      "step": 7633
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.343700709155917,
+      "learning_rate": 2.251702728357783e-06,
+      "loss": 0.0437,
+      "step": 7634
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.293332829306894,
+      "learning_rate": 2.2504717513089773e-06,
+      "loss": 0.0367,
+      "step": 7635
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.3029671763377195,
+      "learning_rate": 2.249241013110039e-06,
+      "loss": 0.0425,
+      "step": 7636
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.7098572515457942,
+      "learning_rate": 2.248010513867879e-06,
+      "loss": 0.0481,
+      "step": 7637
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.403924983273616,
+      "learning_rate": 2.246780253689394e-06,
+      "loss": 0.0471,
+      "step": 7638
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.781435672744397,
+      "learning_rate": 2.245550232681453e-06,
+      "loss": 0.0529,
+      "step": 7639
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.3709736956694176,
+      "learning_rate": 2.2443204509509094e-06,
+      "loss": 0.0411,
+      "step": 7640
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.6308684174630381,
+      "learning_rate": 2.243090908604593e-06,
+      "loss": 0.0447,
+      "step": 7641
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.3301624465146882,
+      "learning_rate": 2.2418616057493125e-06,
+      "loss": 0.0463,
+      "step": 7642
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.8268856860764175,
+      "learning_rate": 2.2406325424918562e-06,
+      "loss": 0.0662,
+      "step": 7643
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.3916864802324946,
+      "learning_rate": 2.2394037189389943e-06,
+      "loss": 0.0463,
+      "step": 7644
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.3999831368461508,
+      "learning_rate": 2.238175135197471e-06,
+      "loss": 0.0406,
+      "step": 7645
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.2398709726568975,
+      "learning_rate": 2.236946791374016e-06,
+      "loss": 0.0407,
+      "step": 7646
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.7159626967387376,
+      "learning_rate": 2.2357186875753333e-06,
+      "loss": 0.0534,
+      "step": 7647
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.4948736675191263,
+      "learning_rate": 2.2344908239081076e-06,
+      "loss": 0.0442,
+      "step": 7648
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.5138630148447723,
+      "learning_rate": 2.2332632004790007e-06,
+      "loss": 0.0416,
+      "step": 7649
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.6280263184553732,
+      "learning_rate": 2.2320358173946587e-06,
+      "loss": 0.0481,
+      "step": 7650
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.575910483837429,
+      "learning_rate": 2.230808674761701e-06,
+      "loss": 0.0559,
+      "step": 7651
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.4842978247482534,
+      "learning_rate": 2.2295817726867313e-06,
+      "loss": 0.0536,
+      "step": 7652
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.514600048320214,
+      "learning_rate": 2.2283551112763284e-06,
+      "loss": 0.0483,
+      "step": 7653
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.8199166667140514,
+      "learning_rate": 2.2271286906370504e-06,
+      "loss": 0.0489,
+      "step": 7654
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.4253198151145172,
+      "learning_rate": 2.2259025108754388e-06,
+      "loss": 0.0453,
+      "step": 7655
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.2171111099162017,
+      "learning_rate": 2.2246765720980074e-06,
+      "loss": 0.0419,
+      "step": 7656
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.3174041181723641,
+      "learning_rate": 2.2234508744112564e-06,
+      "loss": 0.0424,
+      "step": 7657
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.4188081701404573,
+      "learning_rate": 2.2222254179216602e-06,
+      "loss": 0.0487,
+      "step": 7658
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.3457789391986241,
+      "learning_rate": 2.2210002027356723e-06,
+      "loss": 0.0432,
+      "step": 7659
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.686292024139762,
+      "learning_rate": 2.219775228959726e-06,
+      "loss": 0.0507,
+      "step": 7660
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.6161021225646754,
+      "learning_rate": 2.218550496700237e-06,
+      "loss": 0.0554,
+      "step": 7661
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.3862607310245512,
+      "learning_rate": 2.2173260060635927e-06,
+      "loss": 0.0401,
+      "step": 7662
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.667351863916119,
+      "learning_rate": 2.216101757156169e-06,
+      "loss": 0.0568,
+      "step": 7663
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.58665205752601,
+      "learning_rate": 2.2148777500843125e-06,
+      "loss": 0.0472,
+      "step": 7664
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.5140741642804514,
+      "learning_rate": 2.2136539849543525e-06,
+      "loss": 0.0489,
+      "step": 7665
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.520828894147409,
+      "learning_rate": 2.2124304618725956e-06,
+      "loss": 0.049,
+      "step": 7666
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.6303484004422177,
+      "learning_rate": 2.2112071809453306e-06,
+      "loss": 0.0514,
+      "step": 7667
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.4937625396977685,
+      "learning_rate": 2.209984142278821e-06,
+      "loss": 0.0487,
+      "step": 7668
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.8526561129813044,
+      "learning_rate": 2.2087613459793143e-06,
+      "loss": 0.0536,
+      "step": 7669
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.4570860009307929,
+      "learning_rate": 2.2075387921530327e-06,
+      "loss": 0.0486,
+      "step": 7670
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.6369724323436896,
+      "learning_rate": 2.2063164809061783e-06,
+      "loss": 0.0545,
+      "step": 7671
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.490502897413803,
+      "learning_rate": 2.205094412344931e-06,
+      "loss": 0.0448,
+      "step": 7672
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.4971969392969942,
+      "learning_rate": 2.2038725865754543e-06,
+      "loss": 0.0468,
+      "step": 7673
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.5807379655128104,
+      "learning_rate": 2.202651003703885e-06,
+      "loss": 0.0585,
+      "step": 7674
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.6362436732041103,
+      "learning_rate": 2.2014296638363437e-06,
+      "loss": 0.0409,
+      "step": 7675
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.4379636652063073,
+      "learning_rate": 2.2002085670789257e-06,
+      "loss": 0.0431,
+      "step": 7676
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.6130344090275623,
+      "learning_rate": 2.198987713537708e-06,
+      "loss": 0.0488,
+      "step": 7677
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.4453948439224231,
+      "learning_rate": 2.1977671033187425e-06,
+      "loss": 0.0474,
+      "step": 7678
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.8024263482546783,
+      "learning_rate": 2.196546736528067e-06,
+      "loss": 0.0453,
+      "step": 7679
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.5864496000804666,
+      "learning_rate": 2.1953266132716903e-06,
+      "loss": 0.0532,
+      "step": 7680
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.3410641497346425,
+      "learning_rate": 2.1941067336556082e-06,
+      "loss": 0.0393,
+      "step": 7681
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.039115095323855,
+      "learning_rate": 2.1928870977857873e-06,
+      "loss": 0.0525,
+      "step": 7682
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.7243659912974534,
+      "learning_rate": 2.1916677057681786e-06,
+      "loss": 0.0491,
+      "step": 7683
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.659263106049305,
+      "learning_rate": 2.1904485577087066e-06,
+      "loss": 0.0584,
+      "step": 7684
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.6009967663664904,
+      "learning_rate": 2.1892296537132822e-06,
+      "loss": 0.0481,
+      "step": 7685
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.6696131274065544,
+      "learning_rate": 2.188010993887787e-06,
+      "loss": 0.0539,
+      "step": 7686
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.444899124482223,
+      "learning_rate": 2.1867925783380893e-06,
+      "loss": 0.0453,
+      "step": 7687
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.8349081832420642,
+      "learning_rate": 2.1855744071700303e-06,
+      "loss": 0.0587,
+      "step": 7688
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.480256626876591,
+      "learning_rate": 2.1843564804894316e-06,
+      "loss": 0.0462,
+      "step": 7689
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.5649200161013583,
+      "learning_rate": 2.183138798402092e-06,
+      "loss": 0.0487,
+      "step": 7690
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.5257571402479257,
+      "learning_rate": 2.181921361013794e-06,
+      "loss": 0.0508,
+      "step": 7691
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.205779753780435,
+      "learning_rate": 2.1807041684302928e-06,
+      "loss": 0.0407,
+      "step": 7692
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.7846379075064627,
+      "learning_rate": 2.1794872207573286e-06,
+      "loss": 0.0546,
+      "step": 7693
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.5997319416493758,
+      "learning_rate": 2.1782705181006148e-06,
+      "loss": 0.0554,
+      "step": 7694
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.2957778245233489,
+      "learning_rate": 2.177054060565845e-06,
+      "loss": 0.0376,
+      "step": 7695
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.4009549130388859,
+      "learning_rate": 2.1758378482586924e-06,
+      "loss": 0.0413,
+      "step": 7696
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.5925360488429634,
+      "learning_rate": 2.1746218812848097e-06,
+      "loss": 0.0519,
+      "step": 7697
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.3701076520046696,
+      "learning_rate": 2.1734061597498256e-06,
+      "loss": 0.0499,
+      "step": 7698
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.5069849078690918,
+      "learning_rate": 2.1721906837593514e-06,
+      "loss": 0.0444,
+      "step": 7699
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.456057254387466,
+      "learning_rate": 2.170975453418974e-06,
+      "loss": 0.0511,
+      "step": 7700
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.4658315923399063,
+      "learning_rate": 2.1697604688342594e-06,
+      "loss": 0.0452,
+      "step": 7701
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.6552396112015717,
+      "learning_rate": 2.1685457301107506e-06,
+      "loss": 0.0471,
+      "step": 7702
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.3671635513037255,
+      "learning_rate": 2.167331237353974e-06,
+      "loss": 0.0441,
+      "step": 7703
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.4439055345538996,
+      "learning_rate": 2.16611699066943e-06,
+      "loss": 0.0424,
+      "step": 7704
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.7250617868294074,
+      "learning_rate": 2.164902990162602e-06,
+      "loss": 0.0534,
+      "step": 7705
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.7564844538457465,
+      "learning_rate": 2.1636892359389476e-06,
+      "loss": 0.0545,
+      "step": 7706
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.1110422181675603,
+      "learning_rate": 2.1624757281039056e-06,
+      "loss": 0.0599,
+      "step": 7707
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.7453026628414732,
+      "learning_rate": 2.16126246676289e-06,
+      "loss": 0.0547,
+      "step": 7708
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.515852528891318,
+      "learning_rate": 2.1600494520213006e-06,
+      "loss": 0.0482,
+      "step": 7709
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.624721070154593,
+      "learning_rate": 2.158836683984507e-06,
+      "loss": 0.0542,
+      "step": 7710
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.4601312872432688,
+      "learning_rate": 2.1576241627578654e-06,
+      "loss": 0.0398,
+      "step": 7711
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5589654476108326,
+      "learning_rate": 2.156411888446705e-06,
+      "loss": 0.0466,
+      "step": 7712
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5921808041888277,
+      "learning_rate": 2.1551998611563355e-06,
+      "loss": 0.0663,
+      "step": 7713
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5978527651273398,
+      "learning_rate": 2.1539880809920433e-06,
+      "loss": 0.0483,
+      "step": 7714
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.4437560235513083,
+      "learning_rate": 2.152776548059098e-06,
+      "loss": 0.0471,
+      "step": 7715
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.70637716803343,
+      "learning_rate": 2.151565262462742e-06,
+      "loss": 0.0627,
+      "step": 7716
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.2347661224502224,
+      "learning_rate": 2.1503542243082016e-06,
+      "loss": 0.0391,
+      "step": 7717
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5886820747654424,
+      "learning_rate": 2.1491434337006777e-06,
+      "loss": 0.0464,
+      "step": 7718
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.515478311199961,
+      "learning_rate": 2.147932890745351e-06,
+      "loss": 0.0574,
+      "step": 7719
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.3630286522709556,
+      "learning_rate": 2.1467225955473786e-06,
+      "loss": 0.0424,
+      "step": 7720
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.729905560066306,
+      "learning_rate": 2.145512548211902e-06,
+      "loss": 0.0517,
+      "step": 7721
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.671694412506426,
+      "learning_rate": 2.1443027488440338e-06,
+      "loss": 0.0478,
+      "step": 7722
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.4671785264596322,
+      "learning_rate": 2.1430931975488715e-06,
+      "loss": 0.0528,
+      "step": 7723
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.3996389864197183,
+      "learning_rate": 2.1418838944314866e-06,
+      "loss": 0.046,
+      "step": 7724
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.3042824222804872,
+      "learning_rate": 2.140674839596931e-06,
+      "loss": 0.0403,
+      "step": 7725
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.9163178994240455,
+      "learning_rate": 2.1394660331502322e-06,
+      "loss": 0.0585,
+      "step": 7726
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5980340307750944,
+      "learning_rate": 2.138257475196402e-06,
+      "loss": 0.0496,
+      "step": 7727
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.3834564319937144,
+      "learning_rate": 2.1370491658404235e-06,
+      "loss": 0.0441,
+      "step": 7728
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.311981403660786,
+      "learning_rate": 2.135841105187266e-06,
+      "loss": 0.0476,
+      "step": 7729
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.4578929951399202,
+      "learning_rate": 2.134633293341871e-06,
+      "loss": 0.0453,
+      "step": 7730
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.7951757284219738,
+      "learning_rate": 2.1334257304091603e-06,
+      "loss": 0.0566,
+      "step": 7731
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.4829455331152133,
+      "learning_rate": 2.1322184164940324e-06,
+      "loss": 0.0489,
+      "step": 7732
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5165166913305823,
+      "learning_rate": 2.1310113517013693e-06,
+      "loss": 0.0489,
+      "step": 7733
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.7203524176482812,
+      "learning_rate": 2.129804536136025e-06,
+      "loss": 0.054,
+      "step": 7734
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5739682140469393,
+      "learning_rate": 2.1285979699028376e-06,
+      "loss": 0.05,
+      "step": 7735
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5223370492182904,
+      "learning_rate": 2.1273916531066193e-06,
+      "loss": 0.0528,
+      "step": 7736
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.3697488144628571,
+      "learning_rate": 2.126185585852162e-06,
+      "loss": 0.0454,
+      "step": 7737
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.4414655254685338,
+      "learning_rate": 2.1249797682442346e-06,
+      "loss": 0.0409,
+      "step": 7738
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.4609203083422198,
+      "learning_rate": 2.1237742003875895e-06,
+      "loss": 0.0462,
+      "step": 7739
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5035723275710244,
+      "learning_rate": 2.1225688823869494e-06,
+      "loss": 0.0499,
+      "step": 7740
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.3718125797347978,
+      "learning_rate": 2.1213638143470234e-06,
+      "loss": 0.0455,
+      "step": 7741
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.616421031835314,
+      "learning_rate": 2.1201589963724933e-06,
+      "loss": 0.0585,
+      "step": 7742
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.4159906043808597,
+      "learning_rate": 2.1189544285680214e-06,
+      "loss": 0.0477,
+      "step": 7743
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.6069130153010018,
+      "learning_rate": 2.1177501110382455e-06,
+      "loss": 0.0506,
+      "step": 7744
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.766691361378118,
+      "learning_rate": 2.1165460438877856e-06,
+      "loss": 0.0549,
+      "step": 7745
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5530794788415467,
+      "learning_rate": 2.1153422272212398e-06,
+      "loss": 0.0506,
+      "step": 7746
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.6474714942926358,
+      "learning_rate": 2.1141386611431818e-06,
+      "loss": 0.0469,
+      "step": 7747
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.5125926852891596,
+      "learning_rate": 2.1129353457581647e-06,
+      "loss": 0.0472,
+      "step": 7748
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.4729865465977838,
+      "learning_rate": 2.111732281170718e-06,
+      "loss": 0.0512,
+      "step": 7749
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.5229022782093555,
+      "learning_rate": 2.1105294674853543e-06,
+      "loss": 0.0504,
+      "step": 7750
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.8257730028337207,
+      "learning_rate": 2.109326904806558e-06,
+      "loss": 0.0659,
+      "step": 7751
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.5871578368332793,
+      "learning_rate": 2.108124593238798e-06,
+      "loss": 0.0551,
+      "step": 7752
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.4547307346475309,
+      "learning_rate": 2.106922532886517e-06,
+      "loss": 0.0498,
+      "step": 7753
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.6077294998445286,
+      "learning_rate": 2.105720723854138e-06,
+      "loss": 0.0493,
+      "step": 7754
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.3829068844835544,
+      "learning_rate": 2.104519166246059e-06,
+      "loss": 0.0435,
+      "step": 7755
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.5420311823340107,
+      "learning_rate": 2.103317860166662e-06,
+      "loss": 0.045,
+      "step": 7756
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.5249571409645588,
+      "learning_rate": 2.1021168057203008e-06,
+      "loss": 0.0517,
+      "step": 7757
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.9595829931747413,
+      "learning_rate": 2.1009160030113128e-06,
+      "loss": 0.0501,
+      "step": 7758
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.7430004946930417,
+      "learning_rate": 2.09971545214401e-06,
+      "loss": 0.0608,
+      "step": 7759
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.515632271760031,
+      "learning_rate": 2.0985151532226834e-06,
+      "loss": 0.0483,
+      "step": 7760
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.4754426486601246,
+      "learning_rate": 2.0973151063516e-06,
+      "loss": 0.0557,
+      "step": 7761
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.264308021842779,
+      "learning_rate": 2.096115311635011e-06,
+      "loss": 0.0413,
+      "step": 7762
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.3739104782719076,
+      "learning_rate": 2.0949157691771395e-06,
+      "loss": 0.0422,
+      "step": 7763
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.9306220316413207,
+      "learning_rate": 2.0937164790821907e-06,
+      "loss": 0.0548,
+      "step": 7764
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.354528985374528,
+      "learning_rate": 2.0925174414543454e-06,
+      "loss": 0.0413,
+      "step": 7765
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.4123816436259278,
+      "learning_rate": 2.0913186563977634e-06,
+      "loss": 0.0436,
+      "step": 7766
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.537961081431161,
+      "learning_rate": 2.0901201240165797e-06,
+      "loss": 0.0511,
+      "step": 7767
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.4646898787400453,
+      "learning_rate": 2.0889218444149145e-06,
+      "loss": 0.0454,
+      "step": 7768
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.6327447718797505,
+      "learning_rate": 2.0877238176968585e-06,
+      "loss": 0.0502,
+      "step": 7769
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.356354250775867,
+      "learning_rate": 2.0865260439664857e-06,
+      "loss": 0.0448,
+      "step": 7770
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.5118410978571668,
+      "learning_rate": 2.0853285233278454e-06,
+      "loss": 0.0548,
+      "step": 7771
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.5133616736592606,
+      "learning_rate": 2.0841312558849653e-06,
+      "loss": 0.0476,
+      "step": 7772
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.389973799756541,
+      "learning_rate": 2.0829342417418493e-06,
+      "loss": 0.0506,
+      "step": 7773
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.3037531408099183,
+      "learning_rate": 2.081737481002484e-06,
+      "loss": 0.0408,
+      "step": 7774
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.6352657510066801,
+      "learning_rate": 2.0805409737708297e-06,
+      "loss": 0.0543,
+      "step": 7775
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.4295526286713385,
+      "learning_rate": 2.0793447201508288e-06,
+      "loss": 0.0486,
+      "step": 7776
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.5172547434495156,
+      "learning_rate": 2.078148720246397e-06,
+      "loss": 0.0445,
+      "step": 7777
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.4740608359245537,
+      "learning_rate": 2.0769529741614297e-06,
+      "loss": 0.05,
+      "step": 7778
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.6067616543549852,
+      "learning_rate": 2.0757574819998e-06,
+      "loss": 0.0527,
+      "step": 7779
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.5781559215022924,
+      "learning_rate": 2.0745622438653627e-06,
+      "loss": 0.052,
+      "step": 7780
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.559826462112527,
+      "learning_rate": 2.0733672598619444e-06,
+      "loss": 0.0517,
+      "step": 7781
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.3888181006178777,
+      "learning_rate": 2.0721725300933552e-06,
+      "loss": 0.0468,
+      "step": 7782
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.5638577420089863,
+      "learning_rate": 2.070978054663379e-06,
+      "loss": 0.0476,
+      "step": 7783
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.6050613637259945,
+      "learning_rate": 2.0697838336757796e-06,
+      "loss": 0.0576,
+      "step": 7784
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.4089142071152132,
+      "learning_rate": 2.0685898672342967e-06,
+      "loss": 0.0456,
+      "step": 7785
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.5300530579668767,
+      "learning_rate": 2.067396155442652e-06,
+      "loss": 0.0612,
+      "step": 7786
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.6484293749905639,
+      "learning_rate": 2.0662026984045396e-06,
+      "loss": 0.0563,
+      "step": 7787
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.4691297601126798,
+      "learning_rate": 2.065009496223638e-06,
+      "loss": 0.0491,
+      "step": 7788
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.6824805799177214,
+      "learning_rate": 2.063816549003599e-06,
+      "loss": 0.0519,
+      "step": 7789
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.5403406514618005,
+      "learning_rate": 2.062623856848051e-06,
+      "loss": 0.0449,
+      "step": 7790
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.4618512018526246,
+      "learning_rate": 2.061431419860603e-06,
+      "loss": 0.0508,
+      "step": 7791
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.5768809500425693,
+      "learning_rate": 2.0602392381448427e-06,
+      "loss": 0.053,
+      "step": 7792
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.5707215816601925,
+      "learning_rate": 2.0590473118043326e-06,
+      "loss": 0.0544,
+      "step": 7793
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.626813443045718,
+      "learning_rate": 2.057855640942617e-06,
+      "loss": 0.0482,
+      "step": 7794
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.4206355601913683,
+      "learning_rate": 2.056664225663214e-06,
+      "loss": 0.0477,
+      "step": 7795
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.5760286190845691,
+      "learning_rate": 2.0554730660696214e-06,
+      "loss": 0.0498,
+      "step": 7796
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.3561025048514879,
+      "learning_rate": 2.054282162265313e-06,
+      "loss": 0.0434,
+      "step": 7797
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.3167705952037672,
+      "learning_rate": 2.053091514353745e-06,
+      "loss": 0.0443,
+      "step": 7798
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.4423705057792942,
+      "learning_rate": 2.051901122438345e-06,
+      "loss": 0.0478,
+      "step": 7799
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.8799137959968626,
+      "learning_rate": 2.0507109866225243e-06,
+      "loss": 0.0497,
+      "step": 7800
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.750496124531781,
+      "learning_rate": 2.049521107009669e-06,
+      "loss": 0.0491,
+      "step": 7801
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.4535862961989705,
+      "learning_rate": 2.048331483703142e-06,
+      "loss": 0.0478,
+      "step": 7802
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.7139688425136053,
+      "learning_rate": 2.0471421168062845e-06,
+      "loss": 0.0399,
+      "step": 7803
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.6477892250932662,
+      "learning_rate": 2.0459530064224183e-06,
+      "loss": 0.0505,
+      "step": 7804
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.7039773118983894,
+      "learning_rate": 2.0447641526548377e-06,
+      "loss": 0.0531,
+      "step": 7805
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.449287930773047,
+      "learning_rate": 2.043575555606822e-06,
+      "loss": 0.0423,
+      "step": 7806
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.494215798853032,
+      "learning_rate": 2.042387215381621e-06,
+      "loss": 0.0455,
+      "step": 7807
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.5124702792621303,
+      "learning_rate": 2.0411991320824657e-06,
+      "loss": 0.0559,
+      "step": 7808
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.4620068170302547,
+      "learning_rate": 2.040011305812563e-06,
+      "loss": 0.0425,
+      "step": 7809
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.6498792186300844,
+      "learning_rate": 2.0388237366751005e-06,
+      "loss": 0.0449,
+      "step": 7810
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.9403859103290324,
+      "learning_rate": 2.03763642477324e-06,
+      "loss": 0.0573,
+      "step": 7811
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.5908742165498022,
+      "learning_rate": 2.036449370210125e-06,
+      "loss": 0.046,
+      "step": 7812
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.7260610038266784,
+      "learning_rate": 2.0352625730888727e-06,
+      "loss": 0.0478,
+      "step": 7813
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.5962290717880836,
+      "learning_rate": 2.0340760335125794e-06,
+      "loss": 0.0551,
+      "step": 7814
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.4312057243207117,
+      "learning_rate": 2.032889751584317e-06,
+      "loss": 0.0496,
+      "step": 7815
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.4944692181821482,
+      "learning_rate": 2.0317037274071412e-06,
+      "loss": 0.0429,
+      "step": 7816
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.5121052407610622,
+      "learning_rate": 2.0305179610840775e-06,
+      "loss": 0.0514,
+      "step": 7817
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.7338597665570605,
+      "learning_rate": 2.0293324527181363e-06,
+      "loss": 0.0538,
+      "step": 7818
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.7575409884749362,
+      "learning_rate": 2.0281472024122992e-06,
+      "loss": 0.058,
+      "step": 7819
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.2665488462710315,
+      "learning_rate": 2.0269622102695303e-06,
+      "loss": 0.043,
+      "step": 7820
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.5591025757795025,
+      "learning_rate": 2.0257774763927656e-06,
+      "loss": 0.0468,
+      "step": 7821
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.1625389945961262,
+      "learning_rate": 2.0245930008849267e-06,
+      "loss": 0.0368,
+      "step": 7822
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.5968746556318127,
+      "learning_rate": 2.0234087838489042e-06,
+      "loss": 0.0502,
+      "step": 7823
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.5542149784664891,
+      "learning_rate": 2.0222248253875735e-06,
+      "loss": 0.0567,
+      "step": 7824
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.3231833573843619,
+      "learning_rate": 2.0210411256037844e-06,
+      "loss": 0.0402,
+      "step": 7825
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.3403769738104137,
+      "learning_rate": 2.019857684600362e-06,
+      "loss": 0.0442,
+      "step": 7826
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.4480927657289338,
+      "learning_rate": 2.01867450248011e-06,
+      "loss": 0.0516,
+      "step": 7827
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.563754124969961,
+      "learning_rate": 2.0174915793458154e-06,
+      "loss": 0.0564,
+      "step": 7828
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.355365020456432,
+      "learning_rate": 2.016308915300233e-06,
+      "loss": 0.0391,
+      "step": 7829
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.8248091909783004,
+      "learning_rate": 2.015126510446104e-06,
+      "loss": 0.0585,
+      "step": 7830
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.7451387463920665,
+      "learning_rate": 2.013944364886143e-06,
+      "loss": 0.0512,
+      "step": 7831
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.3700698483102016,
+      "learning_rate": 2.0127624787230397e-06,
+      "loss": 0.046,
+      "step": 7832
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.4535732672335389,
+      "learning_rate": 2.0115808520594638e-06,
+      "loss": 0.0434,
+      "step": 7833
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.4497472982744464,
+      "learning_rate": 2.010399484998065e-06,
+      "loss": 0.0564,
+      "step": 7834
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.7040216777047015,
+      "learning_rate": 2.009218377641466e-06,
+      "loss": 0.0587,
+      "step": 7835
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.5892149292824322,
+      "learning_rate": 2.0080375300922703e-06,
+      "loss": 0.0493,
+      "step": 7836
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.421009217921668,
+      "learning_rate": 2.0068569424530577e-06,
+      "loss": 0.0442,
+      "step": 7837
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.694277776676992,
+      "learning_rate": 2.0056766148263825e-06,
+      "loss": 0.0524,
+      "step": 7838
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.6856894760815495,
+      "learning_rate": 2.0044965473147815e-06,
+      "loss": 0.0577,
+      "step": 7839
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.3632547154858765,
+      "learning_rate": 2.0033167400207647e-06,
+      "loss": 0.0422,
+      "step": 7840
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.3626975326963566,
+      "learning_rate": 2.0021371930468235e-06,
+      "loss": 0.0404,
+      "step": 7841
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.5505909845709056,
+      "learning_rate": 2.0009579064954236e-06,
+      "loss": 0.0457,
+      "step": 7842
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.3973529997189955,
+      "learning_rate": 1.999778880469009e-06,
+      "loss": 0.0531,
+      "step": 7843
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.5258542477377819,
+      "learning_rate": 1.998600115069998e-06,
+      "loss": 0.0412,
+      "step": 7844
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.4302891056326903,
+      "learning_rate": 1.997421610400793e-06,
+      "loss": 0.045,
+      "step": 7845
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.5669878738284893,
+      "learning_rate": 1.996243366563768e-06,
+      "loss": 0.0493,
+      "step": 7846
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.3760944038427028,
+      "learning_rate": 1.9950653836612783e-06,
+      "loss": 0.046,
+      "step": 7847
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.6146863147395114,
+      "learning_rate": 1.9938876617956533e-06,
+      "loss": 0.0548,
+      "step": 7848
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.8327212882209223,
+      "learning_rate": 1.9927102010692014e-06,
+      "loss": 0.0511,
+      "step": 7849
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.651219090180509,
+      "learning_rate": 1.9915330015842055e-06,
+      "loss": 0.0464,
+      "step": 7850
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.6362518885381698,
+      "learning_rate": 1.990356063442932e-06,
+      "loss": 0.0516,
+      "step": 7851
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.454400970406789,
+      "learning_rate": 1.989179386747617e-06,
+      "loss": 0.0412,
+      "step": 7852
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.3803465715619918,
+      "learning_rate": 1.9880029716004817e-06,
+      "loss": 0.0452,
+      "step": 7853
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.442993687750572,
+      "learning_rate": 1.9868268181037186e-06,
+      "loss": 0.0508,
+      "step": 7854
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.3319003427694345,
+      "learning_rate": 1.9856509263595e-06,
+      "loss": 0.0396,
+      "step": 7855
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.5554488904825834,
+      "learning_rate": 1.984475296469972e-06,
+      "loss": 0.0465,
+      "step": 7856
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.4927437563339783,
+      "learning_rate": 1.9832999285372653e-06,
+      "loss": 0.0488,
+      "step": 7857
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.265028351196914,
+      "learning_rate": 1.9821248226634793e-06,
+      "loss": 0.036,
+      "step": 7858
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.4579311169951723,
+      "learning_rate": 1.9809499789506985e-06,
+      "loss": 0.0538,
+      "step": 7859
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.4340642524253164,
+      "learning_rate": 1.9797753975009794e-06,
+      "loss": 0.0401,
+      "step": 7860
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.6209045204191102,
+      "learning_rate": 1.978601078416357e-06,
+      "loss": 0.044,
+      "step": 7861
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.6675981489799814,
+      "learning_rate": 1.977427021798841e-06,
+      "loss": 0.0567,
+      "step": 7862
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5695406567711878,
+      "learning_rate": 1.9762532277504266e-06,
+      "loss": 0.041,
+      "step": 7863
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.4371224960782363,
+      "learning_rate": 1.9750796963730752e-06,
+      "loss": 0.0429,
+      "step": 7864
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.4728631341312466,
+      "learning_rate": 1.973906427768735e-06,
+      "loss": 0.041,
+      "step": 7865
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5418999644244622,
+      "learning_rate": 1.9727334220393253e-06,
+      "loss": 0.0498,
+      "step": 7866
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5038065838690255,
+      "learning_rate": 1.971560679286744e-06,
+      "loss": 0.0522,
+      "step": 7867
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.8844241518084501,
+      "learning_rate": 1.970388199612866e-06,
+      "loss": 0.0486,
+      "step": 7868
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5750638815612796,
+      "learning_rate": 1.969215983119546e-06,
+      "loss": 0.0472,
+      "step": 7869
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.6603770252210994,
+      "learning_rate": 1.9680440299086114e-06,
+      "loss": 0.0495,
+      "step": 7870
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5777041981175985,
+      "learning_rate": 1.966872340081872e-06,
+      "loss": 0.0526,
+      "step": 7871
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.4258247151912036,
+      "learning_rate": 1.9657009137411097e-06,
+      "loss": 0.051,
+      "step": 7872
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.8414140369359633,
+      "learning_rate": 1.964529750988086e-06,
+      "loss": 0.0546,
+      "step": 7873
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.7254335481130056,
+      "learning_rate": 1.9633588519245378e-06,
+      "loss": 0.0588,
+      "step": 7874
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.735786107561333,
+      "learning_rate": 1.962188216652183e-06,
+      "loss": 0.0534,
+      "step": 7875
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.6700067904609694,
+      "learning_rate": 1.961017845272711e-06,
+      "loss": 0.0484,
+      "step": 7876
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5637940603191256,
+      "learning_rate": 1.9598477378877944e-06,
+      "loss": 0.05,
+      "step": 7877
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.6698178826815329,
+      "learning_rate": 1.9586778945990785e-06,
+      "loss": 0.0495,
+      "step": 7878
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.4925687346383951,
+      "learning_rate": 1.957508315508187e-06,
+      "loss": 0.0487,
+      "step": 7879
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.6600549662376056,
+      "learning_rate": 1.956339000716718e-06,
+      "loss": 0.0506,
+      "step": 7880
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.4897150270189374,
+      "learning_rate": 1.9551699503262534e-06,
+      "loss": 0.0538,
+      "step": 7881
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5569128098914728,
+      "learning_rate": 1.954001164438344e-06,
+      "loss": 0.0438,
+      "step": 7882
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5739700993582202,
+      "learning_rate": 1.9528326431545248e-06,
+      "loss": 0.0531,
+      "step": 7883
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5970191869259367,
+      "learning_rate": 1.951664386576303e-06,
+      "loss": 0.0464,
+      "step": 7884
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.604324173765507,
+      "learning_rate": 1.9504963948051646e-06,
+      "loss": 0.0565,
+      "step": 7885
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.663927712548092,
+      "learning_rate": 1.949328667942571e-06,
+      "loss": 0.0464,
+      "step": 7886
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.4923735181733429,
+      "learning_rate": 1.9481612060899646e-06,
+      "loss": 0.0425,
+      "step": 7887
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5539395918729548,
+      "learning_rate": 1.946994009348759e-06,
+      "loss": 0.0461,
+      "step": 7888
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.3165604951095617,
+      "learning_rate": 1.945827077820351e-06,
+      "loss": 0.0441,
+      "step": 7889
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.628446485530381,
+      "learning_rate": 1.9446604116061095e-06,
+      "loss": 0.045,
+      "step": 7890
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5397673096694424,
+      "learning_rate": 1.943494010807383e-06,
+      "loss": 0.0583,
+      "step": 7891
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5181087651565,
+      "learning_rate": 1.9423278755254933e-06,
+      "loss": 0.0466,
+      "step": 7892
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5163892077282617,
+      "learning_rate": 1.9411620058617458e-06,
+      "loss": 0.0508,
+      "step": 7893
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.4829450339473196,
+      "learning_rate": 1.939996401917415e-06,
+      "loss": 0.048,
+      "step": 7894
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.5366090046248948,
+      "learning_rate": 1.9388310637937606e-06,
+      "loss": 0.0557,
+      "step": 7895
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.513697605954505,
+      "learning_rate": 1.937665991592012e-06,
+      "loss": 0.0558,
+      "step": 7896
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.396043539619283,
+      "learning_rate": 1.936501185413379e-06,
+      "loss": 0.0453,
+      "step": 7897
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.2863188736265565,
+      "learning_rate": 1.935336645359046e-06,
+      "loss": 0.0448,
+      "step": 7898
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.5870620095832648,
+      "learning_rate": 1.9341723715301786e-06,
+      "loss": 0.0498,
+      "step": 7899
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.399188406110023,
+      "learning_rate": 1.933008364027914e-06,
+      "loss": 0.0456,
+      "step": 7900
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.5052417179934954,
+      "learning_rate": 1.9318446229533717e-06,
+      "loss": 0.0476,
+      "step": 7901
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.3205901716175221,
+      "learning_rate": 1.9306811484076433e-06,
+      "loss": 0.04,
+      "step": 7902
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.509832493319649,
+      "learning_rate": 1.9295179404918e-06,
+      "loss": 0.0501,
+      "step": 7903
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.4907555968252728,
+      "learning_rate": 1.9283549993068863e-06,
+      "loss": 0.0449,
+      "step": 7904
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.3975844515662,
+      "learning_rate": 1.92719232495393e-06,
+      "loss": 0.0405,
+      "step": 7905
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.5680883691912584,
+      "learning_rate": 1.9260299175339288e-06,
+      "loss": 0.0536,
+      "step": 7906
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.4729180938036717,
+      "learning_rate": 1.924867777147863e-06,
+      "loss": 0.0412,
+      "step": 7907
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.430743135315728,
+      "learning_rate": 1.9237059038966867e-06,
+      "loss": 0.0479,
+      "step": 7908
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.3856068120074836,
+      "learning_rate": 1.9225442978813296e-06,
+      "loss": 0.046,
+      "step": 7909
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.2903083566513454,
+      "learning_rate": 1.921382959202699e-06,
+      "loss": 0.0405,
+      "step": 7910
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.559217148370068,
+      "learning_rate": 1.9202218879616824e-06,
+      "loss": 0.0532,
+      "step": 7911
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.7755910772841519,
+      "learning_rate": 1.9190610842591386e-06,
+      "loss": 0.0508,
+      "step": 7912
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.8211453612759008,
+      "learning_rate": 1.917900548195909e-06,
+      "loss": 0.0543,
+      "step": 7913
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.5043877453022898,
+      "learning_rate": 1.9167402798728068e-06,
+      "loss": 0.0496,
+      "step": 7914
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.762457313373451,
+      "learning_rate": 1.915580279390624e-06,
+      "loss": 0.0465,
+      "step": 7915
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.4527783345581387,
+      "learning_rate": 1.914420546850128e-06,
+      "loss": 0.0433,
+      "step": 7916
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.36241044729645,
+      "learning_rate": 1.9132610823520663e-06,
+      "loss": 0.0464,
+      "step": 7917
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.4772735313052772,
+      "learning_rate": 1.9121018859971584e-06,
+      "loss": 0.0498,
+      "step": 7918
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.616506194829487,
+      "learning_rate": 1.9109429578861066e-06,
+      "loss": 0.0553,
+      "step": 7919
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.6082722954537647,
+      "learning_rate": 1.9097842981195836e-06,
+      "loss": 0.0412,
+      "step": 7920
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.5191249937291298,
+      "learning_rate": 1.908625906798242e-06,
+      "loss": 0.0511,
+      "step": 7921
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.4618685177366575,
+      "learning_rate": 1.907467784022709e-06,
+      "loss": 0.0493,
+      "step": 7922
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.4637727809802143,
+      "learning_rate": 1.9063099298935933e-06,
+      "loss": 0.0496,
+      "step": 7923
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.479710842556938,
+      "learning_rate": 1.9051523445114733e-06,
+      "loss": 0.0481,
+      "step": 7924
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.4820086439070792,
+      "learning_rate": 1.9039950279769114e-06,
+      "loss": 0.0468,
+      "step": 7925
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.3682367134027762,
+      "learning_rate": 1.9028379803904417e-06,
+      "loss": 0.0422,
+      "step": 7926
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.4440082212492422,
+      "learning_rate": 1.9016812018525753e-06,
+      "loss": 0.0556,
+      "step": 7927
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.2781817672722595,
+      "learning_rate": 1.9005246924638e-06,
+      "loss": 0.0371,
+      "step": 7928
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.7713988279791533,
+      "learning_rate": 1.8993684523245842e-06,
+      "loss": 0.0564,
+      "step": 7929
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.467250007441262,
+      "learning_rate": 1.8982124815353665e-06,
+      "loss": 0.0493,
+      "step": 7930
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.7129036898423464,
+      "learning_rate": 1.8970567801965683e-06,
+      "loss": 0.048,
+      "step": 7931
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.5738389476471533,
+      "learning_rate": 1.8959013484085836e-06,
+      "loss": 0.0557,
+      "step": 7932
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.9010371420348786,
+      "learning_rate": 1.894746186271782e-06,
+      "loss": 0.0663,
+      "step": 7933
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.4948607626610058,
+      "learning_rate": 1.8935912938865147e-06,
+      "loss": 0.042,
+      "step": 7934
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.4713419730197863,
+      "learning_rate": 1.8924366713531045e-06,
+      "loss": 0.0412,
+      "step": 7935
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.600376300169827,
+      "learning_rate": 1.8912823187718548e-06,
+      "loss": 0.0507,
+      "step": 7936
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.6508937238407522,
+      "learning_rate": 1.8901282362430424e-06,
+      "loss": 0.06,
+      "step": 7937
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.4163073160036752,
+      "learning_rate": 1.8889744238669216e-06,
+      "loss": 0.0385,
+      "step": 7938
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.3955973018967271,
+      "learning_rate": 1.8878208817437216e-06,
+      "loss": 0.0412,
+      "step": 7939
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.4066356353652896,
+      "learning_rate": 1.8866676099736536e-06,
+      "loss": 0.0457,
+      "step": 7940
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.637933505667924,
+      "learning_rate": 1.8855146086568982e-06,
+      "loss": 0.062,
+      "step": 7941
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.7075314762120908,
+      "learning_rate": 1.8843618778936195e-06,
+      "loss": 0.0583,
+      "step": 7942
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.58256969676848,
+      "learning_rate": 1.883209417783952e-06,
+      "loss": 0.0576,
+      "step": 7943
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.5541818941354404,
+      "learning_rate": 1.8820572284280102e-06,
+      "loss": 0.0509,
+      "step": 7944
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.4799399075083148,
+      "learning_rate": 1.8809053099258817e-06,
+      "loss": 0.0431,
+      "step": 7945
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.4743103841131193,
+      "learning_rate": 1.879753662377637e-06,
+      "loss": 0.0578,
+      "step": 7946
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.4398616584650685,
+      "learning_rate": 1.8786022858833148e-06,
+      "loss": 0.0477,
+      "step": 7947
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.3442030241696408,
+      "learning_rate": 1.8774511805429385e-06,
+      "loss": 0.043,
+      "step": 7948
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.4179464346123711,
+      "learning_rate": 1.8763003464565022e-06,
+      "loss": 0.0387,
+      "step": 7949
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.5208001600803147,
+      "learning_rate": 1.875149783723978e-06,
+      "loss": 0.0416,
+      "step": 7950
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.8263714613786792,
+      "learning_rate": 1.873999492445313e-06,
+      "loss": 0.0669,
+      "step": 7951
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.7793321224033813,
+      "learning_rate": 1.8728494727204354e-06,
+      "loss": 0.0461,
+      "step": 7952
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.8735140163478394,
+      "learning_rate": 1.871699724649244e-06,
+      "loss": 0.0607,
+      "step": 7953
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.3845715251761794,
+      "learning_rate": 1.8705502483316196e-06,
+      "loss": 0.0397,
+      "step": 7954
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.4686588841829997,
+      "learning_rate": 1.8694010438674144e-06,
+      "loss": 0.0507,
+      "step": 7955
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.5587442667979023,
+      "learning_rate": 1.86825211135646e-06,
+      "loss": 0.0516,
+      "step": 7956
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.4229870764587589,
+      "learning_rate": 1.8671034508985615e-06,
+      "loss": 0.0485,
+      "step": 7957
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.3384009962772436,
+      "learning_rate": 1.8659550625935052e-06,
+      "loss": 0.0434,
+      "step": 7958
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.8612363626812651,
+      "learning_rate": 1.8648069465410483e-06,
+      "loss": 0.0579,
+      "step": 7959
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.8425298960201364,
+      "learning_rate": 1.8636591028409302e-06,
+      "loss": 0.0599,
+      "step": 7960
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.3945461370950045,
+      "learning_rate": 1.862511531592861e-06,
+      "loss": 0.038,
+      "step": 7961
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.3389043576546187,
+      "learning_rate": 1.8613642328965303e-06,
+      "loss": 0.0414,
+      "step": 7962
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.1975136829967095,
+      "learning_rate": 1.8602172068516011e-06,
+      "loss": 0.0338,
+      "step": 7963
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.378548529035634,
+      "learning_rate": 1.8590704535577187e-06,
+      "loss": 0.0449,
+      "step": 7964
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.5011630520910686,
+      "learning_rate": 1.8579239731144971e-06,
+      "loss": 0.0545,
+      "step": 7965
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.531846314230187,
+      "learning_rate": 1.8567777656215336e-06,
+      "loss": 0.0509,
+      "step": 7966
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.4693219926895753,
+      "learning_rate": 1.8556318311783977e-06,
+      "loss": 0.0419,
+      "step": 7967
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.2588706748102763,
+      "learning_rate": 1.854486169884635e-06,
+      "loss": 0.0388,
+      "step": 7968
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.4164366033679199,
+      "learning_rate": 1.853340781839767e-06,
+      "loss": 0.0439,
+      "step": 7969
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.4050453327497048,
+      "learning_rate": 1.8521956671432967e-06,
+      "loss": 0.0356,
+      "step": 7970
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.8175427152650618,
+      "learning_rate": 1.8510508258946957e-06,
+      "loss": 0.0513,
+      "step": 7971
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.3251804820540158,
+      "learning_rate": 1.8499062581934197e-06,
+      "loss": 0.0462,
+      "step": 7972
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.6045768540629455,
+      "learning_rate": 1.8487619641388938e-06,
+      "loss": 0.0458,
+      "step": 7973
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.8535817236846683,
+      "learning_rate": 1.847617943830523e-06,
+      "loss": 0.0577,
+      "step": 7974
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.481519681921305,
+      "learning_rate": 1.846474197367686e-06,
+      "loss": 0.0448,
+      "step": 7975
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.3772340241703847,
+      "learning_rate": 1.845330724849742e-06,
+      "loss": 0.0408,
+      "step": 7976
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.3416299516981818,
+      "learning_rate": 1.8441875263760211e-06,
+      "loss": 0.0421,
+      "step": 7977
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.49517404288257,
+      "learning_rate": 1.8430446020458353e-06,
+      "loss": 0.049,
+      "step": 7978
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.8490889058601157,
+      "learning_rate": 1.8419019519584685e-06,
+      "loss": 0.0562,
+      "step": 7979
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.5075018156490743,
+      "learning_rate": 1.8407595762131814e-06,
+      "loss": 0.0489,
+      "step": 7980
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.6817974447433879,
+      "learning_rate": 1.8396174749092105e-06,
+      "loss": 0.0546,
+      "step": 7981
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.4796280241751474,
+      "learning_rate": 1.8384756481457723e-06,
+      "loss": 0.0488,
+      "step": 7982
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.5436936192131094,
+      "learning_rate": 1.8373340960220531e-06,
+      "loss": 0.0489,
+      "step": 7983
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.36442829898563,
+      "learning_rate": 1.836192818637223e-06,
+      "loss": 0.0426,
+      "step": 7984
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.2747600440766713,
+      "learning_rate": 1.8350518160904213e-06,
+      "loss": 0.0409,
+      "step": 7985
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.6958963639081952,
+      "learning_rate": 1.8339110884807671e-06,
+      "loss": 0.0529,
+      "step": 7986
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.6459501166549269,
+      "learning_rate": 1.8327706359073526e-06,
+      "loss": 0.051,
+      "step": 7987
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.332145295643568,
+      "learning_rate": 1.8316304584692517e-06,
+      "loss": 0.0398,
+      "step": 7988
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.6092169476961729,
+      "learning_rate": 1.830490556265508e-06,
+      "loss": 0.0409,
+      "step": 7989
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.4238509726761988,
+      "learning_rate": 1.8293509293951468e-06,
+      "loss": 0.0407,
+      "step": 7990
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.7838877695225555,
+      "learning_rate": 1.8282115779571651e-06,
+      "loss": 0.0478,
+      "step": 7991
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.5751210671199023,
+      "learning_rate": 1.8270725020505387e-06,
+      "loss": 0.0491,
+      "step": 7992
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.4135249575881417,
+      "learning_rate": 1.8259337017742158e-06,
+      "loss": 0.0507,
+      "step": 7993
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.527191930554032,
+      "learning_rate": 1.8247951772271267e-06,
+      "loss": 0.0416,
+      "step": 7994
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.7271732494130831,
+      "learning_rate": 1.8236569285081707e-06,
+      "loss": 0.0494,
+      "step": 7995
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.5132401768208266,
+      "learning_rate": 1.8225189557162315e-06,
+      "loss": 0.047,
+      "step": 7996
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.380208601564516,
+      "learning_rate": 1.8213812589501611e-06,
+      "loss": 0.0416,
+      "step": 7997
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.7298877135663888,
+      "learning_rate": 1.820243838308791e-06,
+      "loss": 0.0464,
+      "step": 7998
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.7337897090316465,
+      "learning_rate": 1.8191066938909263e-06,
+      "loss": 0.0535,
+      "step": 7999
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.5534026481394936,
+      "learning_rate": 1.8179698257953543e-06,
+      "loss": 0.0563,
+      "step": 8000
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.559369432043881,
+      "learning_rate": 1.8168332341208294e-06,
+      "loss": 0.0459,
+      "step": 8001
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.4050147857745554,
+      "learning_rate": 1.8156969189660911e-06,
+      "loss": 0.0389,
+      "step": 8002
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.554806608506223,
+      "learning_rate": 1.8145608804298482e-06,
+      "loss": 0.0483,
+      "step": 8003
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.5440321275716231,
+      "learning_rate": 1.8134251186107875e-06,
+      "loss": 0.0442,
+      "step": 8004
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.7955266330060053,
+      "learning_rate": 1.8122896336075708e-06,
+      "loss": 0.0562,
+      "step": 8005
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.9204853015330836,
+      "learning_rate": 1.8111544255188402e-06,
+      "loss": 0.0512,
+      "step": 8006
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.6436730648768774,
+      "learning_rate": 1.8100194944432064e-06,
+      "loss": 0.0479,
+      "step": 8007
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.6989422880228322,
+      "learning_rate": 1.8088848404792652e-06,
+      "loss": 0.0555,
+      "step": 8008
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.9347168220677684,
+      "learning_rate": 1.80775046372558e-06,
+      "loss": 0.0609,
+      "step": 8009
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.691600062440382,
+      "learning_rate": 1.8066163642806945e-06,
+      "loss": 0.0487,
+      "step": 8010
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.4340196154343636,
+      "learning_rate": 1.8054825422431248e-06,
+      "loss": 0.0444,
+      "step": 8011
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.4120811621061868,
+      "learning_rate": 1.8043489977113688e-06,
+      "loss": 0.044,
+      "step": 8012
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.9203459576589121,
+      "learning_rate": 1.8032157307838943e-06,
+      "loss": 0.0568,
+      "step": 8013
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.799035326438951,
+      "learning_rate": 1.8020827415591496e-06,
+      "loss": 0.0618,
+      "step": 8014
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.4528087066170114,
+      "learning_rate": 1.8009500301355564e-06,
+      "loss": 0.0444,
+      "step": 8015
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.589583956397223,
+      "learning_rate": 1.7998175966115116e-06,
+      "loss": 0.0503,
+      "step": 8016
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.4324950618079342,
+      "learning_rate": 1.7986854410853882e-06,
+      "loss": 0.0435,
+      "step": 8017
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.4062335053653605,
+      "learning_rate": 1.7975535636555387e-06,
+      "loss": 0.0419,
+      "step": 8018
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.7419709273642674,
+      "learning_rate": 1.7964219644202852e-06,
+      "loss": 0.0526,
+      "step": 8019
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.3341372470147885,
+      "learning_rate": 1.7952906434779327e-06,
+      "loss": 0.0369,
+      "step": 8020
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.4258158119073165,
+      "learning_rate": 1.794159600926757e-06,
+      "loss": 0.0438,
+      "step": 8021
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.379169774591084,
+      "learning_rate": 1.79302883686501e-06,
+      "loss": 0.0463,
+      "step": 8022
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.540008647033397,
+      "learning_rate": 1.7918983513909199e-06,
+      "loss": 0.0491,
+      "step": 8023
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.5660911562346658,
+      "learning_rate": 1.790768144602692e-06,
+      "loss": 0.0537,
+      "step": 8024
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.3889209327573362,
+      "learning_rate": 1.7896382165985094e-06,
+      "loss": 0.0448,
+      "step": 8025
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.3516853927826242,
+      "learning_rate": 1.7885085674765263e-06,
+      "loss": 0.0416,
+      "step": 8026
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.3760715658806681,
+      "learning_rate": 1.7873791973348737e-06,
+      "loss": 0.0397,
+      "step": 8027
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.4267088077724834,
+      "learning_rate": 1.7862501062716591e-06,
+      "loss": 0.0414,
+      "step": 8028
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.4855078173588327,
+      "learning_rate": 1.7851212943849682e-06,
+      "loss": 0.051,
+      "step": 8029
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.5198410609474282,
+      "learning_rate": 1.7839927617728569e-06,
+      "loss": 0.0499,
+      "step": 8030
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.6076867577005096,
+      "learning_rate": 1.7828645085333645e-06,
+      "loss": 0.0451,
+      "step": 8031
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.5671940949394,
+      "learning_rate": 1.7817365347644993e-06,
+      "loss": 0.0479,
+      "step": 8032
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.7189734338329012,
+      "learning_rate": 1.7806088405642474e-06,
+      "loss": 0.0502,
+      "step": 8033
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.4450312993810916,
+      "learning_rate": 1.7794814260305699e-06,
+      "loss": 0.0452,
+      "step": 8034
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.3382518455599572,
+      "learning_rate": 1.7783542912614076e-06,
+      "loss": 0.0466,
+      "step": 8035
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.614169787987761,
+      "learning_rate": 1.7772274363546704e-06,
+      "loss": 0.0475,
+      "step": 8036
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.4421716431037035,
+      "learning_rate": 1.7761008614082515e-06,
+      "loss": 0.0471,
+      "step": 8037
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.0413040133287192,
+      "learning_rate": 1.774974566520014e-06,
+      "loss": 0.0522,
+      "step": 8038
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.566075586598455,
+      "learning_rate": 1.773848551787798e-06,
+      "loss": 0.0498,
+      "step": 8039
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.6986095896473912,
+      "learning_rate": 1.7727228173094184e-06,
+      "loss": 0.0456,
+      "step": 8040
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.7825050305559207,
+      "learning_rate": 1.7715973631826705e-06,
+      "loss": 0.052,
+      "step": 8041
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.272601867544646,
+      "learning_rate": 1.7704721895053179e-06,
+      "loss": 0.041,
+      "step": 8042
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.514402068915755,
+      "learning_rate": 1.7693472963751079e-06,
+      "loss": 0.0488,
+      "step": 8043
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.6702403608819327,
+      "learning_rate": 1.768222683889757e-06,
+      "loss": 0.0486,
+      "step": 8044
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.6935623722822029,
+      "learning_rate": 1.7670983521469597e-06,
+      "loss": 0.0547,
+      "step": 8045
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.431338935557627,
+      "learning_rate": 1.7659743012443853e-06,
+      "loss": 0.0438,
+      "step": 8046
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.4563254395396996,
+      "learning_rate": 1.7648505312796814e-06,
+      "loss": 0.0447,
+      "step": 8047
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.3035631970500714,
+      "learning_rate": 1.7637270423504664e-06,
+      "loss": 0.0426,
+      "step": 8048
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.3488227248215918,
+      "learning_rate": 1.7626038345543405e-06,
+      "loss": 0.0434,
+      "step": 8049
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.2698785890593605,
+      "learning_rate": 1.7614809079888744e-06,
+      "loss": 0.0343,
+      "step": 8050
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.5533410238420473,
+      "learning_rate": 1.7603582627516163e-06,
+      "loss": 0.0479,
+      "step": 8051
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.9021485521373942,
+      "learning_rate": 1.7592358989400882e-06,
+      "loss": 0.0538,
+      "step": 8052
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.303909725115401,
+      "learning_rate": 1.7581138166517913e-06,
+      "loss": 0.044,
+      "step": 8053
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.6831880049600092,
+      "learning_rate": 1.7569920159841985e-06,
+      "loss": 0.0575,
+      "step": 8054
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.7407240845738021,
+      "learning_rate": 1.7558704970347622e-06,
+      "loss": 0.0516,
+      "step": 8055
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.3835252675587282,
+      "learning_rate": 1.7547492599009063e-06,
+      "loss": 0.0426,
+      "step": 8056
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.4277012234796869,
+      "learning_rate": 1.7536283046800328e-06,
+      "loss": 0.0505,
+      "step": 8057
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.4210503861735482,
+      "learning_rate": 1.7525076314695167e-06,
+      "loss": 0.0374,
+      "step": 8058
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.8054495669733548,
+      "learning_rate": 1.7513872403667125e-06,
+      "loss": 0.0528,
+      "step": 8059
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.3772449148379393,
+      "learning_rate": 1.7502671314689457e-06,
+      "loss": 0.046,
+      "step": 8060
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.6692236331777652,
+      "learning_rate": 1.749147304873522e-06,
+      "loss": 0.0518,
+      "step": 8061
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.412111684924484,
+      "learning_rate": 1.748027760677719e-06,
+      "loss": 0.0457,
+      "step": 8062
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.718977405971166,
+      "learning_rate": 1.746908498978791e-06,
+      "loss": 0.0597,
+      "step": 8063
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.7404234800898917,
+      "learning_rate": 1.7457895198739649e-06,
+      "loss": 0.0512,
+      "step": 8064
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.5089479098161227,
+      "learning_rate": 1.7446708234604498e-06,
+      "loss": 0.0537,
+      "step": 8065
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.7481606711683564,
+      "learning_rate": 1.7435524098354228e-06,
+      "loss": 0.0634,
+      "step": 8066
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.363526497440536,
+      "learning_rate": 1.7424342790960436e-06,
+      "loss": 0.0402,
+      "step": 8067
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.7307837232743941,
+      "learning_rate": 1.741316431339441e-06,
+      "loss": 0.0544,
+      "step": 8068
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.714394427871754,
+      "learning_rate": 1.7401988666627217e-06,
+      "loss": 0.0507,
+      "step": 8069
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.5831542380686199,
+      "learning_rate": 1.7390815851629672e-06,
+      "loss": 0.056,
+      "step": 8070
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.41649346975632,
+      "learning_rate": 1.737964586937238e-06,
+      "loss": 0.048,
+      "step": 8071
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.3478573344921505,
+      "learning_rate": 1.7368478720825633e-06,
+      "loss": 0.0461,
+      "step": 8072
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.3444617830925452,
+      "learning_rate": 1.7357314406959552e-06,
+      "loss": 0.0435,
+      "step": 8073
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.3615280247243722,
+      "learning_rate": 1.7346152928743958e-06,
+      "loss": 0.0531,
+      "step": 8074
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.4176401270277832,
+      "learning_rate": 1.7334994287148438e-06,
+      "loss": 0.0454,
+      "step": 8075
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.542296395951026,
+      "learning_rate": 1.7323838483142318e-06,
+      "loss": 0.0509,
+      "step": 8076
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.3237752219958212,
+      "learning_rate": 1.7312685517694737e-06,
+      "loss": 0.0415,
+      "step": 8077
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.611921170291769,
+      "learning_rate": 1.7301535391774516e-06,
+      "loss": 0.0549,
+      "step": 8078
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.575347973006268,
+      "learning_rate": 1.7290388106350276e-06,
+      "loss": 0.0522,
+      "step": 8079
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.45970055885112,
+      "learning_rate": 1.7279243662390377e-06,
+      "loss": 0.0446,
+      "step": 8080
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.7104706386011317,
+      "learning_rate": 1.7268102060862918e-06,
+      "loss": 0.0534,
+      "step": 8081
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.431100505675895,
+      "learning_rate": 1.7256963302735752e-06,
+      "loss": 0.0405,
+      "step": 8082
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.3900376877398863,
+      "learning_rate": 1.7245827388976527e-06,
+      "loss": 0.0433,
+      "step": 8083
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.2620913341087951,
+      "learning_rate": 1.723469432055258e-06,
+      "loss": 0.0452,
+      "step": 8084
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.7308685436396103,
+      "learning_rate": 1.7223564098431067e-06,
+      "loss": 0.0501,
+      "step": 8085
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.6840377202226817,
+      "learning_rate": 1.7212436723578851e-06,
+      "loss": 0.0516,
+      "step": 8086
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.4636601149365835,
+      "learning_rate": 1.7201312196962561e-06,
+      "loss": 0.0487,
+      "step": 8087
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.6426251357750212,
+      "learning_rate": 1.7190190519548555e-06,
+      "loss": 0.0454,
+      "step": 8088
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.7792510714521894,
+      "learning_rate": 1.7179071692303002e-06,
+      "loss": 0.0615,
+      "step": 8089
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.6803777796043524,
+      "learning_rate": 1.7167955716191753e-06,
+      "loss": 0.051,
+      "step": 8090
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.4625246032667736,
+      "learning_rate": 1.7156842592180484e-06,
+      "loss": 0.049,
+      "step": 8091
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.2250366791458176,
+      "learning_rate": 1.7145732321234565e-06,
+      "loss": 0.0379,
+      "step": 8092
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.6438603753343128,
+      "learning_rate": 1.7134624904319142e-06,
+      "loss": 0.0436,
+      "step": 8093
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.6209682337135403,
+      "learning_rate": 1.7123520342399091e-06,
+      "loss": 0.0499,
+      "step": 8094
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.9108591718720465,
+      "learning_rate": 1.7112418636439093e-06,
+      "loss": 0.0575,
+      "step": 8095
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.9517829343892141,
+      "learning_rate": 1.710131978740351e-06,
+      "loss": 0.0627,
+      "step": 8096
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.6434603121572873,
+      "learning_rate": 1.7090223796256527e-06,
+      "loss": 0.0461,
+      "step": 8097
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.343880307664543,
+      "learning_rate": 1.7079130663962034e-06,
+      "loss": 0.038,
+      "step": 8098
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.5921481396759396,
+      "learning_rate": 1.7068040391483676e-06,
+      "loss": 0.0482,
+      "step": 8099
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.625854252274159,
+      "learning_rate": 1.7056952979784853e-06,
+      "loss": 0.0502,
+      "step": 8100
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.3860285372928698,
+      "learning_rate": 1.7045868429828745e-06,
+      "loss": 0.0427,
+      "step": 8101
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.2755142136046578,
+      "learning_rate": 1.703478674257823e-06,
+      "loss": 0.0394,
+      "step": 8102
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.5144339967954465,
+      "learning_rate": 1.7023707918995996e-06,
+      "loss": 0.0473,
+      "step": 8103
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.4380683276855128,
+      "learning_rate": 1.701263196004445e-06,
+      "loss": 0.0508,
+      "step": 8104
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.5011324477205539,
+      "learning_rate": 1.7001558866685747e-06,
+      "loss": 0.0488,
+      "step": 8105
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.715964400843852,
+      "learning_rate": 1.699048863988178e-06,
+      "loss": 0.048,
+      "step": 8106
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.5905625264814254,
+      "learning_rate": 1.6979421280594249e-06,
+      "loss": 0.0471,
+      "step": 8107
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.3480203972939815,
+      "learning_rate": 1.6968356789784535e-06,
+      "loss": 0.0437,
+      "step": 8108
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.373952464173078,
+      "learning_rate": 1.695729516841384e-06,
+      "loss": 0.042,
+      "step": 8109
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.3873495148985726,
+      "learning_rate": 1.6946236417443062e-06,
+      "loss": 0.0478,
+      "step": 8110
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.3800542523265455,
+      "learning_rate": 1.6935180537832862e-06,
+      "loss": 0.0472,
+      "step": 8111
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.455749077453802,
+      "learning_rate": 1.692412753054365e-06,
+      "loss": 0.0459,
+      "step": 8112
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.3448691339260663,
+      "learning_rate": 1.6913077396535626e-06,
+      "loss": 0.0439,
+      "step": 8113
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.4869596539761307,
+      "learning_rate": 1.6902030136768665e-06,
+      "loss": 0.0444,
+      "step": 8114
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.4002155329355563,
+      "learning_rate": 1.6890985752202488e-06,
+      "loss": 0.0395,
+      "step": 8115
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.4232349895983019,
+      "learning_rate": 1.6879944243796477e-06,
+      "loss": 0.0484,
+      "step": 8116
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.4808223520786983,
+      "learning_rate": 1.68689056125098e-06,
+      "loss": 0.0452,
+      "step": 8117
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.7180578842373968,
+      "learning_rate": 1.6857869859301401e-06,
+      "loss": 0.0584,
+      "step": 8118
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.4331455416117003,
+      "learning_rate": 1.6846836985129916e-06,
+      "loss": 0.051,
+      "step": 8119
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.6463101528931443,
+      "learning_rate": 1.6835806990953802e-06,
+      "loss": 0.0521,
+      "step": 8120
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.5398514608393399,
+      "learning_rate": 1.6824779877731211e-06,
+      "loss": 0.052,
+      "step": 8121
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.429756491684159,
+      "learning_rate": 1.681375564642006e-06,
+      "loss": 0.0429,
+      "step": 8122
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.3680878711697797,
+      "learning_rate": 1.6802734297977997e-06,
+      "loss": 0.0417,
+      "step": 8123
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.3931442380185792,
+      "learning_rate": 1.6791715833362482e-06,
+      "loss": 0.0471,
+      "step": 8124
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.649261543585834,
+      "learning_rate": 1.6780700253530642e-06,
+      "loss": 0.0539,
+      "step": 8125
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.4619022813850429,
+      "learning_rate": 1.6769687559439425e-06,
+      "loss": 0.0486,
+      "step": 8126
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.7863664521790024,
+      "learning_rate": 1.6758677752045487e-06,
+      "loss": 0.0535,
+      "step": 8127
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.7445993245602909,
+      "learning_rate": 1.674767083230524e-06,
+      "loss": 0.0639,
+      "step": 8128
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.5023427726537997,
+      "learning_rate": 1.673666680117484e-06,
+      "loss": 0.0489,
+      "step": 8129
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.513167149987485,
+      "learning_rate": 1.6725665659610218e-06,
+      "loss": 0.0497,
+      "step": 8130
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.621633800008838,
+      "learning_rate": 1.6714667408567015e-06,
+      "loss": 0.0494,
+      "step": 8131
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.8745827472701766,
+      "learning_rate": 1.6703672049000673e-06,
+      "loss": 0.0662,
+      "step": 8132
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.6706845039157907,
+      "learning_rate": 1.6692679581866334e-06,
+      "loss": 0.059,
+      "step": 8133
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.3533243399441997,
+      "learning_rate": 1.6681690008118912e-06,
+      "loss": 0.0427,
+      "step": 8134
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.3262834352381931,
+      "learning_rate": 1.6670703328713039e-06,
+      "loss": 0.0389,
+      "step": 8135
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.6236684887114572,
+      "learning_rate": 1.665971954460316e-06,
+      "loss": 0.0567,
+      "step": 8136
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.4524135917482524,
+      "learning_rate": 1.6648738656743402e-06,
+      "loss": 0.0477,
+      "step": 8137
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.6256087440277407,
+      "learning_rate": 1.6637760666087688e-06,
+      "loss": 0.0548,
+      "step": 8138
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.5138650294760831,
+      "learning_rate": 1.6626785573589667e-06,
+      "loss": 0.0543,
+      "step": 8139
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.4041457464715152,
+      "learning_rate": 1.6615813380202728e-06,
+      "loss": 0.0415,
+      "step": 8140
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.6309638677545721,
+      "learning_rate": 1.6604844086880012e-06,
+      "loss": 0.052,
+      "step": 8141
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.3477925797770858,
+      "learning_rate": 1.6593877694574435e-06,
+      "loss": 0.0475,
+      "step": 8142
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.2577528534930584,
+      "learning_rate": 1.6582914204238621e-06,
+      "loss": 0.04,
+      "step": 8143
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.7834576210377315,
+      "learning_rate": 1.6571953616824987e-06,
+      "loss": 0.0532,
+      "step": 8144
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.4863588895234348,
+      "learning_rate": 1.6560995933285656e-06,
+      "loss": 0.0531,
+      "step": 8145
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.5233881321112273,
+      "learning_rate": 1.6550041154572521e-06,
+      "loss": 0.0517,
+      "step": 8146
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.4501349439502198,
+      "learning_rate": 1.65390892816372e-06,
+      "loss": 0.0453,
+      "step": 8147
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.4360532250591533,
+      "learning_rate": 1.6528140315431102e-06,
+      "loss": 0.0415,
+      "step": 8148
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.7072872659935638,
+      "learning_rate": 1.6517194256905329e-06,
+      "loss": 0.0482,
+      "step": 8149
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.2083911373634373,
+      "learning_rate": 1.650625110701079e-06,
+      "loss": 0.0357,
+      "step": 8150
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.6809449758243455,
+      "learning_rate": 1.6495310866698095e-06,
+      "loss": 0.0492,
+      "step": 8151
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.5753231230797256,
+      "learning_rate": 1.6484373536917615e-06,
+      "loss": 0.043,
+      "step": 8152
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.5035187177744316,
+      "learning_rate": 1.647343911861945e-06,
+      "loss": 0.0513,
+      "step": 8153
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.4157189322156665,
+      "learning_rate": 1.6462507612753503e-06,
+      "loss": 0.0513,
+      "step": 8154
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.6742848983734384,
+      "learning_rate": 1.6451579020269353e-06,
+      "loss": 0.0455,
+      "step": 8155
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.6456541930501847,
+      "learning_rate": 1.6440653342116398e-06,
+      "loss": 0.0427,
+      "step": 8156
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.4984074307545814,
+      "learning_rate": 1.642973057924372e-06,
+      "loss": 0.0507,
+      "step": 8157
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.4772195637859555,
+      "learning_rate": 1.6418810732600177e-06,
+      "loss": 0.0515,
+      "step": 8158
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.7196871922385537,
+      "learning_rate": 1.6407893803134357e-06,
+      "loss": 0.0567,
+      "step": 8159
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.48391662020529,
+      "learning_rate": 1.6396979791794631e-06,
+      "loss": 0.054,
+      "step": 8160
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.3107200014290368,
+      "learning_rate": 1.6386068699529067e-06,
+      "loss": 0.0434,
+      "step": 8161
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.4567064986577705,
+      "learning_rate": 1.6375160527285538e-06,
+      "loss": 0.0418,
+      "step": 8162
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.268562273018905,
+      "learning_rate": 1.636425527601161e-06,
+      "loss": 0.0395,
+      "step": 8163
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.4922041447965735,
+      "learning_rate": 1.635335294665462e-06,
+      "loss": 0.0436,
+      "step": 8164
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.0988693132261922,
+      "learning_rate": 1.6342453540161624e-06,
+      "loss": 0.0351,
+      "step": 8165
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.6733420531633307,
+      "learning_rate": 1.6331557057479485e-06,
+      "loss": 0.0567,
+      "step": 8166
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.3788757259888342,
+      "learning_rate": 1.632066349955474e-06,
+      "loss": 0.049,
+      "step": 8167
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.4908065032829123,
+      "learning_rate": 1.630977286733374e-06,
+      "loss": 0.0389,
+      "step": 8168
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.6533841740056816,
+      "learning_rate": 1.6298885161762528e-06,
+      "loss": 0.0574,
+      "step": 8169
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.5621180884770862,
+      "learning_rate": 1.6288000383786912e-06,
+      "loss": 0.0465,
+      "step": 8170
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.502062037079539,
+      "learning_rate": 1.6277118534352432e-06,
+      "loss": 0.0448,
+      "step": 8171
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.5945718235288897,
+      "learning_rate": 1.6266239614404421e-06,
+      "loss": 0.0588,
+      "step": 8172
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.6403079569615602,
+      "learning_rate": 1.6255363624887894e-06,
+      "loss": 0.0509,
+      "step": 8173
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.7523551739406498,
+      "learning_rate": 1.6244490566747667e-06,
+      "loss": 0.0559,
+      "step": 8174
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.2695197993651435,
+      "learning_rate": 1.6233620440928265e-06,
+      "loss": 0.0368,
+      "step": 8175
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.520931362145433,
+      "learning_rate": 1.6222753248373969e-06,
+      "loss": 0.0467,
+      "step": 8176
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.6231449694365343,
+      "learning_rate": 1.6211888990028785e-06,
+      "loss": 0.0583,
+      "step": 8177
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.3378397772371289,
+      "learning_rate": 1.6201027666836522e-06,
+      "loss": 0.0451,
+      "step": 8178
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.392382853265308,
+      "learning_rate": 1.6190169279740665e-06,
+      "loss": 0.0448,
+      "step": 8179
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.4290789217294706,
+      "learning_rate": 1.6179313829684506e-06,
+      "loss": 0.044,
+      "step": 8180
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.503374201286744,
+      "learning_rate": 1.6168461317611028e-06,
+      "loss": 0.0427,
+      "step": 8181
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.5968457455456138,
+      "learning_rate": 1.6157611744462998e-06,
+      "loss": 0.0497,
+      "step": 8182
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.371448316537398,
+      "learning_rate": 1.6146765111182877e-06,
+      "loss": 0.0445,
+      "step": 8183
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.1858457832977316,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.0383,
+      "step": 8184
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.6058401758467902,
+      "learning_rate": 1.6125080667995174e-06,
+      "loss": 0.0499,
+      "step": 8185
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.5202149913847696,
+      "learning_rate": 1.6114242859971302e-06,
+      "loss": 0.0458,
+      "step": 8186
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.4769161337097285,
+      "learning_rate": 1.6103407995582794e-06,
+      "loss": 0.0443,
+      "step": 8187
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.6533571102798723,
+      "learning_rate": 1.6092576075770861e-06,
+      "loss": 0.0522,
+      "step": 8188
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.5422006182545938,
+      "learning_rate": 1.6081747101476464e-06,
+      "loss": 0.0545,
+      "step": 8189
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.5070512459258634,
+      "learning_rate": 1.6070921073640328e-06,
+      "loss": 0.0515,
+      "step": 8190
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.4620442310669806,
+      "learning_rate": 1.6060097993202878e-06,
+      "loss": 0.0497,
+      "step": 8191
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.489696731931627,
+      "learning_rate": 1.6049277861104345e-06,
+      "loss": 0.0567,
+      "step": 8192
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.612398889856248,
+      "learning_rate": 1.6038460678284644e-06,
+      "loss": 0.0491,
+      "step": 8193
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.9378443665037954,
+      "learning_rate": 1.602764644568346e-06,
+      "loss": 0.0542,
+      "step": 8194
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.6721160058414224,
+      "learning_rate": 1.6016835164240196e-06,
+      "loss": 0.05,
+      "step": 8195
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.6975156677644365,
+      "learning_rate": 1.6006026834894068e-06,
+      "loss": 0.0518,
+      "step": 8196
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.4458671892588826,
+      "learning_rate": 1.5995221458583943e-06,
+      "loss": 0.0489,
+      "step": 8197
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.6811829650201813,
+      "learning_rate": 1.5984419036248516e-06,
+      "loss": 0.0518,
+      "step": 8198
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.449515878495356,
+      "learning_rate": 1.5973619568826177e-06,
+      "loss": 0.0416,
+      "step": 8199
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.3467642431847793,
+      "learning_rate": 1.5962823057255055e-06,
+      "loss": 0.0426,
+      "step": 8200
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.6081957757636973,
+      "learning_rate": 1.5952029502473032e-06,
+      "loss": 0.0463,
+      "step": 8201
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.4086829315300893,
+      "learning_rate": 1.594123890541776e-06,
+      "loss": 0.0412,
+      "step": 8202
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.6199228277316036,
+      "learning_rate": 1.5930451267026592e-06,
+      "loss": 0.0446,
+      "step": 8203
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.6248291539419761,
+      "learning_rate": 1.5919666588236666e-06,
+      "loss": 0.0546,
+      "step": 8204
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.4842107686924249,
+      "learning_rate": 1.5908884869984831e-06,
+      "loss": 0.0484,
+      "step": 8205
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.671322613966543,
+      "learning_rate": 1.5898106113207685e-06,
+      "loss": 0.0397,
+      "step": 8206
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.4622711130175166,
+      "learning_rate": 1.5887330318841548e-06,
+      "loss": 0.0408,
+      "step": 8207
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.6184641750276456,
+      "learning_rate": 1.5876557487822553e-06,
+      "loss": 0.0551,
+      "step": 8208
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.1953682265800787,
+      "learning_rate": 1.5865787621086491e-06,
+      "loss": 0.0394,
+      "step": 8209
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.4500851584533982,
+      "learning_rate": 1.585502071956897e-06,
+      "loss": 0.0462,
+      "step": 8210
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.5601892453149564,
+      "learning_rate": 1.5844256784205275e-06,
+      "loss": 0.0564,
+      "step": 8211
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.5545377472074746,
+      "learning_rate": 1.583349581593046e-06,
+      "loss": 0.0536,
+      "step": 8212
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.373363747993195,
+      "learning_rate": 1.5822737815679357e-06,
+      "loss": 0.0393,
+      "step": 8213
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.3016026986598959,
+      "learning_rate": 1.5811982784386465e-06,
+      "loss": 0.0398,
+      "step": 8214
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.585756995463118,
+      "learning_rate": 1.5801230722986104e-06,
+      "loss": 0.0465,
+      "step": 8215
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.2958485857896926,
+      "learning_rate": 1.5790481632412286e-06,
+      "loss": 0.0424,
+      "step": 8216
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.4837685635433642,
+      "learning_rate": 1.577973551359877e-06,
+      "loss": 0.0499,
+      "step": 8217
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.662840034025026,
+      "learning_rate": 1.5768992367479058e-06,
+      "loss": 0.0581,
+      "step": 8218
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.3661403932657985,
+      "learning_rate": 1.575825219498643e-06,
+      "loss": 0.039,
+      "step": 8219
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.9670123342118937,
+      "learning_rate": 1.5747514997053841e-06,
+      "loss": 0.0575,
+      "step": 8220
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.3005139170873676,
+      "learning_rate": 1.5736780774614064e-06,
+      "loss": 0.0407,
+      "step": 8221
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.6971504591094217,
+      "learning_rate": 1.5726049528599552e-06,
+      "loss": 0.0563,
+      "step": 8222
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.4422175737273988,
+      "learning_rate": 1.5715321259942529e-06,
+      "loss": 0.0424,
+      "step": 8223
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.397704009442535,
+      "learning_rate": 1.5704595969574933e-06,
+      "loss": 0.0434,
+      "step": 8224
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.7463487895342504,
+      "learning_rate": 1.5693873658428494e-06,
+      "loss": 0.0466,
+      "step": 8225
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.7809533676729927,
+      "learning_rate": 1.568315432743462e-06,
+      "loss": 0.0479,
+      "step": 8226
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.4753121900827932,
+      "learning_rate": 1.567243797752453e-06,
+      "loss": 0.046,
+      "step": 8227
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.827641416334775,
+      "learning_rate": 1.5661724609629132e-06,
+      "loss": 0.0519,
+      "step": 8228
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.4450154964479713,
+      "learning_rate": 1.5651014224679083e-06,
+      "loss": 0.0393,
+      "step": 8229
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.5660727273119208,
+      "learning_rate": 1.5640306823604778e-06,
+      "loss": 0.0518,
+      "step": 8230
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.38358427681166,
+      "learning_rate": 1.5629602407336386e-06,
+      "loss": 0.0453,
+      "step": 8231
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.5796561813012897,
+      "learning_rate": 1.5618900976803769e-06,
+      "loss": 0.0527,
+      "step": 8232
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.5153150126599382,
+      "learning_rate": 1.560820253293659e-06,
+      "loss": 0.0451,
+      "step": 8233
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.285072985482014,
+      "learning_rate": 1.5597507076664187e-06,
+      "loss": 0.0376,
+      "step": 8234
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.500505225742505,
+      "learning_rate": 1.5586814608915673e-06,
+      "loss": 0.0455,
+      "step": 8235
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.3908345798291033,
+      "learning_rate": 1.5576125130619885e-06,
+      "loss": 0.0428,
+      "step": 8236
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.7012003288897117,
+      "learning_rate": 1.5565438642705444e-06,
+      "loss": 0.0579,
+      "step": 8237
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.7326949587359681,
+      "learning_rate": 1.5554755146100641e-06,
+      "loss": 0.0442,
+      "step": 8238
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.6756464438744398,
+      "learning_rate": 1.5544074641733574e-06,
+      "loss": 0.0614,
+      "step": 8239
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.5122949431946533,
+      "learning_rate": 1.5533397130532053e-06,
+      "loss": 0.0462,
+      "step": 8240
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.652057648294175,
+      "learning_rate": 1.5522722613423608e-06,
+      "loss": 0.0499,
+      "step": 8241
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.9126688045031186,
+      "learning_rate": 1.5512051091335518e-06,
+      "loss": 0.0561,
+      "step": 8242
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.3140715771816118,
+      "learning_rate": 1.5501382565194845e-06,
+      "loss": 0.0427,
+      "step": 8243
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.5965372761091547,
+      "learning_rate": 1.5490717035928327e-06,
+      "loss": 0.0499,
+      "step": 8244
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.2378901385564058,
+      "learning_rate": 1.5480054504462505e-06,
+      "loss": 0.0447,
+      "step": 8245
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.4783177861200232,
+      "learning_rate": 1.54693949717236e-06,
+      "loss": 0.0492,
+      "step": 8246
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.734957528542089,
+      "learning_rate": 1.5458738438637616e-06,
+      "loss": 0.0556,
+      "step": 8247
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.3766987454261743,
+      "learning_rate": 1.5448084906130252e-06,
+      "loss": 0.0459,
+      "step": 8248
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.394131207218535,
+      "learning_rate": 1.5437434375127008e-06,
+      "loss": 0.0491,
+      "step": 8249
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.5074360486638545,
+      "learning_rate": 1.542678684655306e-06,
+      "loss": 0.0511,
+      "step": 8250
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.3399965064651054,
+      "learning_rate": 1.5416142321333382e-06,
+      "loss": 0.0389,
+      "step": 8251
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.2249441531394676,
+      "learning_rate": 1.5405500800392643e-06,
+      "loss": 0.0463,
+      "step": 8252
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.4473339032897732,
+      "learning_rate": 1.5394862284655266e-06,
+      "loss": 0.0475,
+      "step": 8253
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.4455559757662613,
+      "learning_rate": 1.5384226775045391e-06,
+      "loss": 0.0537,
+      "step": 8254
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.5672405166444547,
+      "learning_rate": 1.5373594272486958e-06,
+      "loss": 0.0513,
+      "step": 8255
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.5672297331037095,
+      "learning_rate": 1.5362964777903565e-06,
+      "loss": 0.0559,
+      "step": 8256
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.6241397004207176,
+      "learning_rate": 1.5352338292218633e-06,
+      "loss": 0.0546,
+      "step": 8257
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.3955630374610366,
+      "learning_rate": 1.5341714816355257e-06,
+      "loss": 0.0471,
+      "step": 8258
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.1999547123952439,
+      "learning_rate": 1.5331094351236287e-06,
+      "loss": 0.039,
+      "step": 8259
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.342496495178013,
+      "learning_rate": 1.5320476897784309e-06,
+      "loss": 0.0448,
+      "step": 8260
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.4455687678045765,
+      "learning_rate": 1.5309862456921682e-06,
+      "loss": 0.0517,
+      "step": 8261
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.39530499113681,
+      "learning_rate": 1.5299251029570445e-06,
+      "loss": 0.05,
+      "step": 8262
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.3220994274213405,
+      "learning_rate": 1.5288642616652437e-06,
+      "loss": 0.039,
+      "step": 8263
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.4951681318975283,
+      "learning_rate": 1.5278037219089191e-06,
+      "loss": 0.0442,
+      "step": 8264
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.6075466754630183,
+      "learning_rate": 1.5267434837801993e-06,
+      "loss": 0.0538,
+      "step": 8265
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.3379775521327348,
+      "learning_rate": 1.5256835473711844e-06,
+      "loss": 0.0437,
+      "step": 8266
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.4312589224188745,
+      "learning_rate": 1.5246239127739542e-06,
+      "loss": 0.0453,
+      "step": 8267
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.3663636985004097,
+      "learning_rate": 1.523564580080555e-06,
+      "loss": 0.0461,
+      "step": 8268
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.4176524878611945,
+      "learning_rate": 1.5225055493830132e-06,
+      "loss": 0.0425,
+      "step": 8269
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.732516355929053,
+      "learning_rate": 1.5214468207733258e-06,
+      "loss": 0.0504,
+      "step": 8270
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.224737498718848,
+      "learning_rate": 1.5203883943434622e-06,
+      "loss": 0.039,
+      "step": 8271
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.5288818835340634,
+      "learning_rate": 1.5193302701853674e-06,
+      "loss": 0.0476,
+      "step": 8272
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.4935261485706044,
+      "learning_rate": 1.5182724483909618e-06,
+      "loss": 0.0522,
+      "step": 8273
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.5565215827434007,
+      "learning_rate": 1.5172149290521354e-06,
+      "loss": 0.056,
+      "step": 8274
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.9884709849723492,
+      "learning_rate": 1.5161577122607573e-06,
+      "loss": 0.0402,
+      "step": 8275
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.3028556613529376,
+      "learning_rate": 1.5151007981086657e-06,
+      "loss": 0.0417,
+      "step": 8276
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.5348910211719016,
+      "learning_rate": 1.5140441866876737e-06,
+      "loss": 0.0442,
+      "step": 8277
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.3886215791758076,
+      "learning_rate": 1.5129878780895674e-06,
+      "loss": 0.0481,
+      "step": 8278
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.272547487762864,
+      "learning_rate": 1.5119318724061105e-06,
+      "loss": 0.0426,
+      "step": 8279
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.4731114345791578,
+      "learning_rate": 1.5108761697290348e-06,
+      "loss": 0.0497,
+      "step": 8280
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.3083630761844314,
+      "learning_rate": 1.5098207701500511e-06,
+      "loss": 0.0375,
+      "step": 8281
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.5699638342191817,
+      "learning_rate": 1.5087656737608403e-06,
+      "loss": 0.0483,
+      "step": 8282
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.498460961471617,
+      "learning_rate": 1.5077108806530582e-06,
+      "loss": 0.0433,
+      "step": 8283
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.5632110315905887,
+      "learning_rate": 1.5066563909183318e-06,
+      "loss": 0.0546,
+      "step": 8284
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.4392588062823115,
+      "learning_rate": 1.5056022046482678e-06,
+      "loss": 0.0486,
+      "step": 8285
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.4589594886235493,
+      "learning_rate": 1.5045483219344387e-06,
+      "loss": 0.04,
+      "step": 8286
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.5648431783414312,
+      "learning_rate": 1.5034947428683988e-06,
+      "loss": 0.0436,
+      "step": 8287
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.7609842689865909,
+      "learning_rate": 1.5024414675416693e-06,
+      "loss": 0.0459,
+      "step": 8288
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.350515146554619,
+      "learning_rate": 1.5013884960457486e-06,
+      "loss": 0.0397,
+      "step": 8289
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.3098330974189372,
+      "learning_rate": 1.5003358284721053e-06,
+      "loss": 0.0416,
+      "step": 8290
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.5036578497152129,
+      "learning_rate": 1.499283464912188e-06,
+      "loss": 0.0447,
+      "step": 8291
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.5651364818544082,
+      "learning_rate": 1.498231405457411e-06,
+      "loss": 0.0461,
+      "step": 8292
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.3121323365412014,
+      "learning_rate": 1.4971796501991698e-06,
+      "loss": 0.0386,
+      "step": 8293
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.5941825792596624,
+      "learning_rate": 1.4961281992288273e-06,
+      "loss": 0.0467,
+      "step": 8294
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.683937099982629,
+      "learning_rate": 1.4950770526377233e-06,
+      "loss": 0.0536,
+      "step": 8295
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.5409491534288338,
+      "learning_rate": 1.4940262105171683e-06,
+      "loss": 0.0424,
+      "step": 8296
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.5077052770456136,
+      "learning_rate": 1.4929756729584517e-06,
+      "loss": 0.0517,
+      "step": 8297
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.300047304475324,
+      "learning_rate": 1.4919254400528293e-06,
+      "loss": 0.0439,
+      "step": 8298
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 12.172105749993614,
+      "learning_rate": 1.490875511891538e-06,
+      "loss": 0.102,
+      "step": 8299
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.5024623483180286,
+      "learning_rate": 1.4898258885657829e-06,
+      "loss": 0.0479,
+      "step": 8300
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4931677611979917,
+      "learning_rate": 1.488776570166744e-06,
+      "loss": 0.0488,
+      "step": 8301
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4514742640496727,
+      "learning_rate": 1.4877275567855726e-06,
+      "loss": 0.047,
+      "step": 8302
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4019400305553669,
+      "learning_rate": 1.4866788485133988e-06,
+      "loss": 0.0485,
+      "step": 8303
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.6287996964540272,
+      "learning_rate": 1.4856304454413239e-06,
+      "loss": 0.0506,
+      "step": 8304
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4868805746551998,
+      "learning_rate": 1.484582347660421e-06,
+      "loss": 0.0531,
+      "step": 8305
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4888731329030027,
+      "learning_rate": 1.483534555261737e-06,
+      "loss": 0.0494,
+      "step": 8306
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4111030865181216,
+      "learning_rate": 1.4824870683362919e-06,
+      "loss": 0.0462,
+      "step": 8307
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.406357081249897,
+      "learning_rate": 1.4814398869750835e-06,
+      "loss": 0.0444,
+      "step": 8308
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.5717985211013366,
+      "learning_rate": 1.4803930112690767e-06,
+      "loss": 0.0522,
+      "step": 8309
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4446290558389596,
+      "learning_rate": 1.4793464413092161e-06,
+      "loss": 0.0516,
+      "step": 8310
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.5742204561616804,
+      "learning_rate": 1.4783001771864148e-06,
+      "loss": 0.0519,
+      "step": 8311
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.7160132763569504,
+      "learning_rate": 1.4772542189915607e-06,
+      "loss": 0.0617,
+      "step": 8312
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.6815000844088648,
+      "learning_rate": 1.4762085668155152e-06,
+      "loss": 0.0477,
+      "step": 8313
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.0618387408373655,
+      "learning_rate": 1.4751632207491156e-06,
+      "loss": 0.0415,
+      "step": 8314
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4048991861042903,
+      "learning_rate": 1.4741181808831679e-06,
+      "loss": 0.0479,
+      "step": 8315
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.2133614825597814,
+      "learning_rate": 1.4730734473084568e-06,
+      "loss": 0.0347,
+      "step": 8316
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.6134061828270194,
+      "learning_rate": 1.4720290201157361e-06,
+      "loss": 0.0554,
+      "step": 8317
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.3520837327594406,
+      "learning_rate": 1.4709848993957348e-06,
+      "loss": 0.0456,
+      "step": 8318
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.3233795649071949,
+      "learning_rate": 1.4699410852391538e-06,
+      "loss": 0.0458,
+      "step": 8319
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.5509257591050072,
+      "learning_rate": 1.4688975777366716e-06,
+      "loss": 0.0534,
+      "step": 8320
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.6362817497012556,
+      "learning_rate": 1.4678543769789334e-06,
+      "loss": 0.043,
+      "step": 8321
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4749488016933494,
+      "learning_rate": 1.4668114830565644e-06,
+      "loss": 0.0447,
+      "step": 8322
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4044337363804473,
+      "learning_rate": 1.4657688960601595e-06,
+      "loss": 0.0461,
+      "step": 8323
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.9998902185186296,
+      "learning_rate": 1.4647266160802876e-06,
+      "loss": 0.051,
+      "step": 8324
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.531686299491867,
+      "learning_rate": 1.4636846432074885e-06,
+      "loss": 0.0465,
+      "step": 8325
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4807920919029836,
+      "learning_rate": 1.4626429775322816e-06,
+      "loss": 0.0489,
+      "step": 8326
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.6165714203585164,
+      "learning_rate": 1.4616016191451522e-06,
+      "loss": 0.0483,
+      "step": 8327
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.6709100672648618,
+      "learning_rate": 1.4605605681365658e-06,
+      "loss": 0.0511,
+      "step": 8328
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.6902104631480812,
+      "learning_rate": 1.459519824596956e-06,
+      "loss": 0.052,
+      "step": 8329
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4895467800675666,
+      "learning_rate": 1.4584793886167326e-06,
+      "loss": 0.0427,
+      "step": 8330
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.9864664994658,
+      "learning_rate": 1.4574392602862746e-06,
+      "loss": 0.052,
+      "step": 8331
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.3695823961156617,
+      "learning_rate": 1.4563994396959419e-06,
+      "loss": 0.0469,
+      "step": 8332
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.4316052015152396,
+      "learning_rate": 1.455359926936059e-06,
+      "loss": 0.0437,
+      "step": 8333
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.8415366814078669,
+      "learning_rate": 1.4543207220969308e-06,
+      "loss": 0.0499,
+      "step": 8334
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.5702074325399997,
+      "learning_rate": 1.453281825268832e-06,
+      "loss": 0.0462,
+      "step": 8335
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.540584543189919,
+      "learning_rate": 1.4522432365420092e-06,
+      "loss": 0.0477,
+      "step": 8336
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.407507528222732,
+      "learning_rate": 1.4512049560066837e-06,
+      "loss": 0.0451,
+      "step": 8337
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.9855033825595423,
+      "learning_rate": 1.4501669837530535e-06,
+      "loss": 0.0646,
+      "step": 8338
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.4180086107954244,
+      "learning_rate": 1.4491293198712824e-06,
+      "loss": 0.0447,
+      "step": 8339
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.331053280706459,
+      "learning_rate": 1.4480919644515156e-06,
+      "loss": 0.0461,
+      "step": 8340
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.5193673526707658,
+      "learning_rate": 1.447054917583866e-06,
+      "loss": 0.0495,
+      "step": 8341
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.300157323420643,
+      "learning_rate": 1.4460181793584211e-06,
+      "loss": 0.0418,
+      "step": 8342
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.5097993982684867,
+      "learning_rate": 1.4449817498652402e-06,
+      "loss": 0.0402,
+      "step": 8343
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.6630864133225698,
+      "learning_rate": 1.4439456291943605e-06,
+      "loss": 0.0499,
+      "step": 8344
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.4299568979547996,
+      "learning_rate": 1.4429098174357852e-06,
+      "loss": 0.0552,
+      "step": 8345
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.7279825954082573,
+      "learning_rate": 1.4418743146794988e-06,
+      "loss": 0.0522,
+      "step": 8346
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.4154458776971695,
+      "learning_rate": 1.4408391210154532e-06,
+      "loss": 0.0397,
+      "step": 8347
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.8938103248443001,
+      "learning_rate": 1.4398042365335745e-06,
+      "loss": 0.0436,
+      "step": 8348
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.7226860280165233,
+      "learning_rate": 1.4387696613237612e-06,
+      "loss": 0.0554,
+      "step": 8349
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.3466427262747704,
+      "learning_rate": 1.4377353954758893e-06,
+      "loss": 0.0415,
+      "step": 8350
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.3161238505774129,
+      "learning_rate": 1.4367014390798023e-06,
+      "loss": 0.041,
+      "step": 8351
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.5900747019268302,
+      "learning_rate": 1.4356677922253215e-06,
+      "loss": 0.0477,
+      "step": 8352
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.32011524623867,
+      "learning_rate": 1.4346344550022384e-06,
+      "loss": 0.0442,
+      "step": 8353
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.3693774615991687,
+      "learning_rate": 1.433601427500318e-06,
+      "loss": 0.0434,
+      "step": 8354
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.7667166268277283,
+      "learning_rate": 1.4325687098092967e-06,
+      "loss": 0.0532,
+      "step": 8355
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.6198050635260666,
+      "learning_rate": 1.4315363020188905e-06,
+      "loss": 0.0535,
+      "step": 8356
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.5357686740989378,
+      "learning_rate": 1.43050420421878e-06,
+      "loss": 0.042,
+      "step": 8357
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.456401216956577,
+      "learning_rate": 1.4294724164986262e-06,
+      "loss": 0.0562,
+      "step": 8358
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.1297419015709385,
+      "learning_rate": 1.428440938948058e-06,
+      "loss": 0.0396,
+      "step": 8359
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.184290446734807,
+      "learning_rate": 1.4274097716566804e-06,
+      "loss": 0.0361,
+      "step": 8360
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.4144277719985296,
+      "learning_rate": 1.4263789147140672e-06,
+      "loss": 0.0471,
+      "step": 8361
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.352838176835097,
+      "learning_rate": 1.4253483682097724e-06,
+      "loss": 0.0464,
+      "step": 8362
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.5941458067979706,
+      "learning_rate": 1.424318132233316e-06,
+      "loss": 0.0463,
+      "step": 8363
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.238919348327796,
+      "learning_rate": 1.423288206874196e-06,
+      "loss": 0.0357,
+      "step": 8364
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.3419119425218575,
+      "learning_rate": 1.4222585922218812e-06,
+      "loss": 0.0426,
+      "step": 8365
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.4758226842481141,
+      "learning_rate": 1.4212292883658123e-06,
+      "loss": 0.0495,
+      "step": 8366
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.6322923517552461,
+      "learning_rate": 1.4202002953954042e-06,
+      "loss": 0.0507,
+      "step": 8367
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.4633052185354565,
+      "learning_rate": 1.4191716134000466e-06,
+      "loss": 0.0499,
+      "step": 8368
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.7650050568315068,
+      "learning_rate": 1.4181432424690978e-06,
+      "loss": 0.0483,
+      "step": 8369
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.6427714555071173,
+      "learning_rate": 1.4171151826918954e-06,
+      "loss": 0.0477,
+      "step": 8370
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4896222721922217,
+      "learning_rate": 1.4160874341577447e-06,
+      "loss": 0.0511,
+      "step": 8371
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.484449023956297,
+      "learning_rate": 1.4150599969559247e-06,
+      "loss": 0.0414,
+      "step": 8372
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.7047880114514589,
+      "learning_rate": 1.4140328711756878e-06,
+      "loss": 0.0544,
+      "step": 8373
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.5009777808506528,
+      "learning_rate": 1.4130060569062626e-06,
+      "loss": 0.0535,
+      "step": 8374
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.6048544200299082,
+      "learning_rate": 1.4119795542368441e-06,
+      "loss": 0.0481,
+      "step": 8375
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.5585570352010696,
+      "learning_rate": 1.410953363256608e-06,
+      "loss": 0.0508,
+      "step": 8376
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.3799036939663358,
+      "learning_rate": 1.409927484054696e-06,
+      "loss": 0.0441,
+      "step": 8377
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.5925558622322644,
+      "learning_rate": 1.4089019167202278e-06,
+      "loss": 0.0505,
+      "step": 8378
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.3611386914926655,
+      "learning_rate": 1.40787666134229e-06,
+      "loss": 0.0443,
+      "step": 8379
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.6997664744256886,
+      "learning_rate": 1.4068517180099505e-06,
+      "loss": 0.0539,
+      "step": 8380
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.5498846553070156,
+      "learning_rate": 1.4058270868122414e-06,
+      "loss": 0.048,
+      "step": 8381
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.7127464722234402,
+      "learning_rate": 1.404802767838176e-06,
+      "loss": 0.0512,
+      "step": 8382
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4403881440700712,
+      "learning_rate": 1.403778761176734e-06,
+      "loss": 0.0386,
+      "step": 8383
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.3583720807686566,
+      "learning_rate": 1.40275506691687e-06,
+      "loss": 0.0461,
+      "step": 8384
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.5222095081602156,
+      "learning_rate": 1.4017316851475105e-06,
+      "loss": 0.0436,
+      "step": 8385
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.521601829147279,
+      "learning_rate": 1.4007086159575595e-06,
+      "loss": 0.0436,
+      "step": 8386
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4458539856113788,
+      "learning_rate": 1.399685859435887e-06,
+      "loss": 0.046,
+      "step": 8387
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.3888603809176023,
+      "learning_rate": 1.3986634156713418e-06,
+      "loss": 0.0476,
+      "step": 8388
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4381609684188719,
+      "learning_rate": 1.3976412847527427e-06,
+      "loss": 0.0483,
+      "step": 8389
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.6570854948404339,
+      "learning_rate": 1.3966194667688804e-06,
+      "loss": 0.0446,
+      "step": 8390
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.5305075610106071,
+      "learning_rate": 1.3955979618085185e-06,
+      "loss": 0.0473,
+      "step": 8391
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4948790962246123,
+      "learning_rate": 1.394576769960398e-06,
+      "loss": 0.0526,
+      "step": 8392
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4374064108066564,
+      "learning_rate": 1.3935558913132252e-06,
+      "loss": 0.0446,
+      "step": 8393
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.5177200705284128,
+      "learning_rate": 1.3925353259556873e-06,
+      "loss": 0.0487,
+      "step": 8394
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.530365334909462,
+      "learning_rate": 1.3915150739764383e-06,
+      "loss": 0.0439,
+      "step": 8395
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.7696838517185172,
+      "learning_rate": 1.390495135464105e-06,
+      "loss": 0.0596,
+      "step": 8396
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4838624229345398,
+      "learning_rate": 1.3894755105072922e-06,
+      "loss": 0.0443,
+      "step": 8397
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.3905998379468776,
+      "learning_rate": 1.388456199194571e-06,
+      "loss": 0.049,
+      "step": 8398
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4161479350346862,
+      "learning_rate": 1.3874372016144915e-06,
+      "loss": 0.044,
+      "step": 8399
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.5687670441615165,
+      "learning_rate": 1.3864185178555722e-06,
+      "loss": 0.0645,
+      "step": 8400
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.3348445935435789,
+      "learning_rate": 1.3854001480063045e-06,
+      "loss": 0.0441,
+      "step": 8401
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4346253356636294,
+      "learning_rate": 1.3843820921551532e-06,
+      "loss": 0.0423,
+      "step": 8402
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.591177346691786,
+      "learning_rate": 1.3833643503905587e-06,
+      "loss": 0.0546,
+      "step": 8403
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.6276614540571444,
+      "learning_rate": 1.3823469228009284e-06,
+      "loss": 0.0461,
+      "step": 8404
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.4627172050713082,
+      "learning_rate": 1.3813298094746491e-06,
+      "loss": 0.0489,
+      "step": 8405
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.3288218497670228,
+      "learning_rate": 1.380313010500075e-06,
+      "loss": 0.043,
+      "step": 8406
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.617531211939985,
+      "learning_rate": 1.379296525965535e-06,
+      "loss": 0.0502,
+      "step": 8407
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.674245561824862,
+      "learning_rate": 1.3782803559593288e-06,
+      "loss": 0.0461,
+      "step": 8408
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.5522588077515165,
+      "learning_rate": 1.3772645005697337e-06,
+      "loss": 0.047,
+      "step": 8409
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.6345117720229083,
+      "learning_rate": 1.3762489598849937e-06,
+      "loss": 0.0498,
+      "step": 8410
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.4953551944731487,
+      "learning_rate": 1.3752337339933308e-06,
+      "loss": 0.0426,
+      "step": 8411
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.3703123107656219,
+      "learning_rate": 1.3742188229829351e-06,
+      "loss": 0.0461,
+      "step": 8412
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.6992285160849898,
+      "learning_rate": 1.3732042269419721e-06,
+      "loss": 0.0524,
+      "step": 8413
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.334320971273496,
+      "learning_rate": 1.3721899459585775e-06,
+      "loss": 0.0394,
+      "step": 8414
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.414424594508028,
+      "learning_rate": 1.371175980120864e-06,
+      "loss": 0.0428,
+      "step": 8415
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.4242672219815427,
+      "learning_rate": 1.3701623295169115e-06,
+      "loss": 0.047,
+      "step": 8416
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.457491914156786,
+      "learning_rate": 1.369148994234778e-06,
+      "loss": 0.0507,
+      "step": 8417
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.2359557515047375,
+      "learning_rate": 1.36813597436249e-06,
+      "loss": 0.0368,
+      "step": 8418
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.2911869016831212,
+      "learning_rate": 1.3671232699880477e-06,
+      "loss": 0.0406,
+      "step": 8419
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.484748242976053,
+      "learning_rate": 1.3661108811994228e-06,
+      "loss": 0.0524,
+      "step": 8420
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.3895683804827463,
+      "learning_rate": 1.365098808084564e-06,
+      "loss": 0.0446,
+      "step": 8421
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.4170835967344488,
+      "learning_rate": 1.3640870507313859e-06,
+      "loss": 0.0405,
+      "step": 8422
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.6071124760191486,
+      "learning_rate": 1.363075609227783e-06,
+      "loss": 0.0473,
+      "step": 8423
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.1428289128589673,
+      "learning_rate": 1.362064483661617e-06,
+      "loss": 0.0453,
+      "step": 8424
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.873356745272441,
+      "learning_rate": 1.3610536741207237e-06,
+      "loss": 0.0574,
+      "step": 8425
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.5425176026557368,
+      "learning_rate": 1.3600431806929092e-06,
+      "loss": 0.0459,
+      "step": 8426
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.5886468194550611,
+      "learning_rate": 1.3590330034659588e-06,
+      "loss": 0.0427,
+      "step": 8427
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.3648565176668856,
+      "learning_rate": 1.3580231425276224e-06,
+      "loss": 0.0433,
+      "step": 8428
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.7523541361556187,
+      "learning_rate": 1.3570135979656285e-06,
+      "loss": 0.0458,
+      "step": 8429
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.6013968146494437,
+      "learning_rate": 1.356004369867675e-06,
+      "loss": 0.0541,
+      "step": 8430
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.5824063498560141,
+      "learning_rate": 1.354995458321432e-06,
+      "loss": 0.0445,
+      "step": 8431
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.3884894606982499,
+      "learning_rate": 1.3539868634145425e-06,
+      "loss": 0.0413,
+      "step": 8432
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.5887128111742896,
+      "learning_rate": 1.352978585234625e-06,
+      "loss": 0.0471,
+      "step": 8433
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.5234331807270978,
+      "learning_rate": 1.3519706238692654e-06,
+      "loss": 0.0424,
+      "step": 8434
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.5269440177301963,
+      "learning_rate": 1.3509629794060269e-06,
+      "loss": 0.0528,
+      "step": 8435
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.6203135430157731,
+      "learning_rate": 1.3499556519324424e-06,
+      "loss": 0.0462,
+      "step": 8436
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.4982311276058105,
+      "learning_rate": 1.3489486415360175e-06,
+      "loss": 0.05,
+      "step": 8437
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.6381407859014134,
+      "learning_rate": 1.3479419483042288e-06,
+      "loss": 0.0444,
+      "step": 8438
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.6729275348263937,
+      "learning_rate": 1.3469355723245303e-06,
+      "loss": 0.0602,
+      "step": 8439
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.312937306709134,
+      "learning_rate": 1.3459295136843426e-06,
+      "loss": 0.0425,
+      "step": 8440
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.4151699155857165,
+      "learning_rate": 1.344923772471064e-06,
+      "loss": 0.0375,
+      "step": 8441
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.6076921729086373,
+      "learning_rate": 1.3439183487720608e-06,
+      "loss": 0.0389,
+      "step": 8442
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.7502260076523841,
+      "learning_rate": 1.3429132426746743e-06,
+      "loss": 0.0542,
+      "step": 8443
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.3759592163463548,
+      "learning_rate": 1.3419084542662159e-06,
+      "loss": 0.0436,
+      "step": 8444
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.3667868670914456,
+      "learning_rate": 1.3409039836339738e-06,
+      "loss": 0.0413,
+      "step": 8445
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.430006305991176,
+      "learning_rate": 1.3398998308652027e-06,
+      "loss": 0.0464,
+      "step": 8446
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.4000544687170169,
+      "learning_rate": 1.3388959960471354e-06,
+      "loss": 0.0471,
+      "step": 8447
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.5205588386122524,
+      "learning_rate": 1.337892479266974e-06,
+      "loss": 0.0485,
+      "step": 8448
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.547599164751151,
+      "learning_rate": 1.336889280611892e-06,
+      "loss": 0.0461,
+      "step": 8449
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.4917334250993264,
+      "learning_rate": 1.3358864001690358e-06,
+      "loss": 0.0483,
+      "step": 8450
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.3226379052393897,
+      "learning_rate": 1.3348838380255287e-06,
+      "loss": 0.0351,
+      "step": 8451
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.4735105183595614,
+      "learning_rate": 1.3338815942684586e-06,
+      "loss": 0.0393,
+      "step": 8452
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.3411039447895312,
+      "learning_rate": 1.3328796689848932e-06,
+      "loss": 0.0429,
+      "step": 8453
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.467414501587144,
+      "learning_rate": 1.3318780622618682e-06,
+      "loss": 0.039,
+      "step": 8454
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.5796853165072808,
+      "learning_rate": 1.3308767741863916e-06,
+      "loss": 0.0468,
+      "step": 8455
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.4478500287379872,
+      "learning_rate": 1.3298758048454436e-06,
+      "loss": 0.04,
+      "step": 8456
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.7639759433878386,
+      "learning_rate": 1.3288751543259814e-06,
+      "loss": 0.0506,
+      "step": 8457
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.7069931454481615,
+      "learning_rate": 1.327874822714927e-06,
+      "loss": 0.0562,
+      "step": 8458
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.4785677746262984,
+      "learning_rate": 1.3268748100991819e-06,
+      "loss": 0.0471,
+      "step": 8459
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.6479370001976672,
+      "learning_rate": 1.3258751165656154e-06,
+      "loss": 0.0491,
+      "step": 8460
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.907082005923064,
+      "learning_rate": 1.32487574220107e-06,
+      "loss": 0.0522,
+      "step": 8461
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.317951193676135,
+      "learning_rate": 1.3238766870923592e-06,
+      "loss": 0.041,
+      "step": 8462
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.5725227842357739,
+      "learning_rate": 1.3228779513262735e-06,
+      "loss": 0.0514,
+      "step": 8463
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.3940749983392537,
+      "learning_rate": 1.3218795349895696e-06,
+      "loss": 0.0428,
+      "step": 8464
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.7230539461782544,
+      "learning_rate": 1.3208814381689822e-06,
+      "loss": 0.0508,
+      "step": 8465
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.4873764065491668,
+      "learning_rate": 1.3198836609512134e-06,
+      "loss": 0.0466,
+      "step": 8466
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.5750390945745487,
+      "learning_rate": 1.3188862034229405e-06,
+      "loss": 0.0504,
+      "step": 8467
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.0003054516575274,
+      "learning_rate": 1.3178890656708094e-06,
+      "loss": 0.0564,
+      "step": 8468
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.5316559571103365,
+      "learning_rate": 1.3168922477814444e-06,
+      "loss": 0.0545,
+      "step": 8469
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.4546673889450077,
+      "learning_rate": 1.315895749841436e-06,
+      "loss": 0.0367,
+      "step": 8470
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.2558656215709567,
+      "learning_rate": 1.3148995719373514e-06,
+      "loss": 0.0384,
+      "step": 8471
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.3404345812948024,
+      "learning_rate": 1.313903714155727e-06,
+      "loss": 0.0485,
+      "step": 8472
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.3450725133900352,
+      "learning_rate": 1.3129081765830725e-06,
+      "loss": 0.0396,
+      "step": 8473
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.663221008325269,
+      "learning_rate": 1.3119129593058676e-06,
+      "loss": 0.0492,
+      "step": 8474
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.5002123392750324,
+      "learning_rate": 1.3109180624105699e-06,
+      "loss": 0.0494,
+      "step": 8475
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.2717798296710185,
+      "learning_rate": 1.3099234859836019e-06,
+      "loss": 0.0389,
+      "step": 8476
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.4234441191604637,
+      "learning_rate": 1.3089292301113654e-06,
+      "loss": 0.0455,
+      "step": 8477
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.3760897869864137,
+      "learning_rate": 1.3079352948802294e-06,
+      "loss": 0.0426,
+      "step": 8478
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.4137062615995946,
+      "learning_rate": 1.3069416803765355e-06,
+      "loss": 0.0391,
+      "step": 8479
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.4897742397366038,
+      "learning_rate": 1.3059483866865973e-06,
+      "loss": 0.0502,
+      "step": 8480
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.494539919238236,
+      "learning_rate": 1.3049554138967052e-06,
+      "loss": 0.0439,
+      "step": 8481
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.395305316581814,
+      "learning_rate": 1.303962762093115e-06,
+      "loss": 0.0474,
+      "step": 8482
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.3996936503578319,
+      "learning_rate": 1.30297043136206e-06,
+      "loss": 0.0472,
+      "step": 8483
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.4870434914698234,
+      "learning_rate": 1.3019784217897423e-06,
+      "loss": 0.0482,
+      "step": 8484
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.5743919203152352,
+      "learning_rate": 1.3009867334623383e-06,
+      "loss": 0.0519,
+      "step": 8485
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.3549515463752877,
+      "learning_rate": 1.299995366465992e-06,
+      "loss": 0.0441,
+      "step": 8486
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6221175382632504,
+      "learning_rate": 1.2990043208868253e-06,
+      "loss": 0.0555,
+      "step": 8487
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.3421737233564357,
+      "learning_rate": 1.2980135968109314e-06,
+      "loss": 0.0423,
+      "step": 8488
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.2774057437083652,
+      "learning_rate": 1.2970231943243716e-06,
+      "loss": 0.042,
+      "step": 8489
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.3371086568778108,
+      "learning_rate": 1.2960331135131826e-06,
+      "loss": 0.0477,
+      "step": 8490
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6058614752109275,
+      "learning_rate": 1.29504335446337e-06,
+      "loss": 0.0545,
+      "step": 8491
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6995573501756085,
+      "learning_rate": 1.2940539172609167e-06,
+      "loss": 0.0519,
+      "step": 8492
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6265437216382979,
+      "learning_rate": 1.2930648019917719e-06,
+      "loss": 0.0484,
+      "step": 8493
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6344399301009649,
+      "learning_rate": 1.2920760087418616e-06,
+      "loss": 0.0498,
+      "step": 8494
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.5905380987075208,
+      "learning_rate": 1.291087537597081e-06,
+      "loss": 0.0535,
+      "step": 8495
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.522136909110268,
+      "learning_rate": 1.2900993886432972e-06,
+      "loss": 0.0527,
+      "step": 8496
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6061318414981354,
+      "learning_rate": 1.2891115619663496e-06,
+      "loss": 0.0607,
+      "step": 8497
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.4407859489344852,
+      "learning_rate": 1.288124057652052e-06,
+      "loss": 0.0465,
+      "step": 8498
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.5940558056846568,
+      "learning_rate": 1.2871368757861863e-06,
+      "loss": 0.0522,
+      "step": 8499
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.4064833018887997,
+      "learning_rate": 1.286150016454511e-06,
+      "loss": 0.0426,
+      "step": 8500
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.4171605114573003,
+      "learning_rate": 1.285163479742752e-06,
+      "loss": 0.0481,
+      "step": 8501
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.3178897549524387,
+      "learning_rate": 1.2841772657366103e-06,
+      "loss": 0.0405,
+      "step": 8502
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.4766507787415275,
+      "learning_rate": 1.283191374521755e-06,
+      "loss": 0.0457,
+      "step": 8503
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.7000557818821727,
+      "learning_rate": 1.2822058061838333e-06,
+      "loss": 0.0515,
+      "step": 8504
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.5439167060505428,
+      "learning_rate": 1.2812205608084582e-06,
+      "loss": 0.0522,
+      "step": 8505
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.5870895114604513,
+      "learning_rate": 1.2802356384812203e-06,
+      "loss": 0.0508,
+      "step": 8506
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6451203205736327,
+      "learning_rate": 1.2792510392876777e-06,
+      "loss": 0.0504,
+      "step": 8507
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.2243738082045,
+      "learning_rate": 1.2782667633133617e-06,
+      "loss": 0.0443,
+      "step": 8508
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.418898319882461,
+      "learning_rate": 1.277282810643774e-06,
+      "loss": 0.042,
+      "step": 8509
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.1158275519841194,
+      "learning_rate": 1.2762991813643938e-06,
+      "loss": 0.0349,
+      "step": 8510
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.3366414155435924,
+      "learning_rate": 1.2753158755606649e-06,
+      "loss": 0.0456,
+      "step": 8511
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.5914447543491548,
+      "learning_rate": 1.2743328933180099e-06,
+      "loss": 0.0475,
+      "step": 8512
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.1600521540835773,
+      "learning_rate": 1.2733502347218174e-06,
+      "loss": 0.0348,
+      "step": 8513
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.5148091858485222,
+      "learning_rate": 1.2723678998574512e-06,
+      "loss": 0.0521,
+      "step": 8514
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.3473752828762064,
+      "learning_rate": 1.271385888810245e-06,
+      "loss": 0.0487,
+      "step": 8515
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.6546302300001483,
+      "learning_rate": 1.270404201665507e-06,
+      "loss": 0.0484,
+      "step": 8516
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.3633386331234514,
+      "learning_rate": 1.2694228385085144e-06,
+      "loss": 0.0422,
+      "step": 8517
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.5336970750911085,
+      "learning_rate": 1.2684417994245197e-06,
+      "loss": 0.0432,
+      "step": 8518
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.5603455240822794,
+      "learning_rate": 1.267461084498744e-06,
+      "loss": 0.0451,
+      "step": 8519
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.8762383812560237,
+      "learning_rate": 1.2664806938163816e-06,
+      "loss": 0.0554,
+      "step": 8520
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.28453381011712,
+      "learning_rate": 1.2655006274625959e-06,
+      "loss": 0.0389,
+      "step": 8521
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.428045596548661,
+      "learning_rate": 1.2645208855225289e-06,
+      "loss": 0.0406,
+      "step": 8522
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.9319783469996132,
+      "learning_rate": 1.263541468081287e-06,
+      "loss": 0.0572,
+      "step": 8523
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.7249688951400397,
+      "learning_rate": 1.262562375223954e-06,
+      "loss": 0.0549,
+      "step": 8524
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.4468872876757757,
+      "learning_rate": 1.2615836070355824e-06,
+      "loss": 0.0486,
+      "step": 8525
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.8849498713579076,
+      "learning_rate": 1.2606051636011963e-06,
+      "loss": 0.0551,
+      "step": 8526
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.6715171523833798,
+      "learning_rate": 1.2596270450057917e-06,
+      "loss": 0.0445,
+      "step": 8527
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.375410485003809,
+      "learning_rate": 1.2586492513343395e-06,
+      "loss": 0.0444,
+      "step": 8528
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.4248116454668638,
+      "learning_rate": 1.2576717826717782e-06,
+      "loss": 0.0519,
+      "step": 8529
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.5661791644275904,
+      "learning_rate": 1.2566946391030222e-06,
+      "loss": 0.0483,
+      "step": 8530
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.4570841787472775,
+      "learning_rate": 1.2557178207129533e-06,
+      "loss": 0.045,
+      "step": 8531
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.4426313862710176,
+      "learning_rate": 1.254741327586428e-06,
+      "loss": 0.0442,
+      "step": 8532
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.2683220277439153,
+      "learning_rate": 1.2537651598082718e-06,
+      "loss": 0.039,
+      "step": 8533
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.6782206136657742,
+      "learning_rate": 1.2527893174632872e-06,
+      "loss": 0.0546,
+      "step": 8534
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.734121312462512,
+      "learning_rate": 1.2518138006362413e-06,
+      "loss": 0.0551,
+      "step": 8535
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.501804140901601,
+      "learning_rate": 1.25083860941188e-06,
+      "loss": 0.0459,
+      "step": 8536
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.6752448574634988,
+      "learning_rate": 1.2498637438749162e-06,
+      "loss": 0.0499,
+      "step": 8537
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.5116942502817403,
+      "learning_rate": 1.2488892041100364e-06,
+      "loss": 0.047,
+      "step": 8538
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.6083312456738155,
+      "learning_rate": 1.2479149902018955e-06,
+      "loss": 0.0395,
+      "step": 8539
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.6663518167753921,
+      "learning_rate": 1.2469411022351273e-06,
+      "loss": 0.0488,
+      "step": 8540
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.50799782759602,
+      "learning_rate": 1.245967540294329e-06,
+      "loss": 0.0478,
+      "step": 8541
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.3063215056198543,
+      "learning_rate": 1.244994304464076e-06,
+      "loss": 0.0429,
+      "step": 8542
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.2989047598794272,
+      "learning_rate": 1.2440213948289121e-06,
+      "loss": 0.0421,
+      "step": 8543
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.7475331138053989,
+      "learning_rate": 1.243048811473353e-06,
+      "loss": 0.0524,
+      "step": 8544
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.7630693986406565,
+      "learning_rate": 1.2420765544818847e-06,
+      "loss": 0.0583,
+      "step": 8545
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.579089601300963,
+      "learning_rate": 1.2411046239389701e-06,
+      "loss": 0.0464,
+      "step": 8546
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.540150314642911,
+      "learning_rate": 1.2401330199290368e-06,
+      "loss": 0.0444,
+      "step": 8547
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.1953884400127959,
+      "learning_rate": 1.2391617425364904e-06,
+      "loss": 0.0355,
+      "step": 8548
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.37445388901159,
+      "learning_rate": 1.2381907918457042e-06,
+      "loss": 0.0483,
+      "step": 8549
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.754158845305628,
+      "learning_rate": 1.2372201679410233e-06,
+      "loss": 0.0436,
+      "step": 8550
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.796609485108871,
+      "learning_rate": 1.236249870906765e-06,
+      "loss": 0.0429,
+      "step": 8551
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.2562714141270679,
+      "learning_rate": 1.2352799008272198e-06,
+      "loss": 0.0386,
+      "step": 8552
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.4240908051816248,
+      "learning_rate": 1.2343102577866467e-06,
+      "loss": 0.0401,
+      "step": 8553
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.5004670950701247,
+      "learning_rate": 1.2333409418692804e-06,
+      "loss": 0.0441,
+      "step": 8554
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.4212615492423462,
+      "learning_rate": 1.2323719531593236e-06,
+      "loss": 0.0433,
+      "step": 8555
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.5654421741205267,
+      "learning_rate": 1.2314032917409513e-06,
+      "loss": 0.0556,
+      "step": 8556
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.6395034598891283,
+      "learning_rate": 1.2304349576983094e-06,
+      "loss": 0.0599,
+      "step": 8557
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.386590966855871,
+      "learning_rate": 1.2294669511155193e-06,
+      "loss": 0.0428,
+      "step": 8558
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.3621744843099224,
+      "learning_rate": 1.2284992720766686e-06,
+      "loss": 0.0401,
+      "step": 8559
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.2418250408879725,
+      "learning_rate": 1.2275319206658215e-06,
+      "loss": 0.0387,
+      "step": 8560
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.6521369990603894,
+      "learning_rate": 1.2265648969670096e-06,
+      "loss": 0.0508,
+      "step": 8561
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.596360838352744,
+      "learning_rate": 1.2255982010642387e-06,
+      "loss": 0.0518,
+      "step": 8562
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.9350441098558337,
+      "learning_rate": 1.2246318330414824e-06,
+      "loss": 0.0468,
+      "step": 8563
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.4171459940900415,
+      "learning_rate": 1.2236657929826917e-06,
+      "loss": 0.0411,
+      "step": 8564
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.60396847608048,
+      "learning_rate": 1.2227000809717838e-06,
+      "loss": 0.05,
+      "step": 8565
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.3460343642192503,
+      "learning_rate": 1.221734697092652e-06,
+      "loss": 0.0376,
+      "step": 8566
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.3750708227058803,
+      "learning_rate": 1.2207696414291563e-06,
+      "loss": 0.0408,
+      "step": 8567
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.6324929934378343,
+      "learning_rate": 1.219804914065132e-06,
+      "loss": 0.0638,
+      "step": 8568
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.353178696400372,
+      "learning_rate": 1.2188405150843812e-06,
+      "loss": 0.0405,
+      "step": 8569
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.2432653278241774,
+      "learning_rate": 1.2178764445706854e-06,
+      "loss": 0.0329,
+      "step": 8570
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.6605459794269863,
+      "learning_rate": 1.2169127026077888e-06,
+      "loss": 0.0497,
+      "step": 8571
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.338599510254725,
+      "learning_rate": 1.2159492892794144e-06,
+      "loss": 0.0398,
+      "step": 8572
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.3337499688084775,
+      "learning_rate": 1.2149862046692513e-06,
+      "loss": 0.041,
+      "step": 8573
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.6871546847299972,
+      "learning_rate": 1.2140234488609631e-06,
+      "loss": 0.0595,
+      "step": 8574
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.3621560394321242,
+      "learning_rate": 1.2130610219381811e-06,
+      "loss": 0.0461,
+      "step": 8575
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.431846331056706,
+      "learning_rate": 1.2120989239845149e-06,
+      "loss": 0.0458,
+      "step": 8576
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.330883434438782,
+      "learning_rate": 1.2111371550835377e-06,
+      "loss": 0.0379,
+      "step": 8577
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.4324720737034748,
+      "learning_rate": 1.210175715318801e-06,
+      "loss": 0.0473,
+      "step": 8578
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.386022627075222,
+      "learning_rate": 1.2092146047738229e-06,
+      "loss": 0.0459,
+      "step": 8579
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.5147285559876726,
+      "learning_rate": 1.2082538235320928e-06,
+      "loss": 0.051,
+      "step": 8580
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.340001570993958,
+      "learning_rate": 1.207293371677077e-06,
+      "loss": 0.0478,
+      "step": 8581
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.3771317047921547,
+      "learning_rate": 1.2063332492922052e-06,
+      "loss": 0.0426,
+      "step": 8582
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.515285126580002,
+      "learning_rate": 1.2053734564608865e-06,
+      "loss": 0.0462,
+      "step": 8583
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.6173835816289572,
+      "learning_rate": 1.2044139932664955e-06,
+      "loss": 0.0519,
+      "step": 8584
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.5601534038519072,
+      "learning_rate": 1.2034548597923812e-06,
+      "loss": 0.0516,
+      "step": 8585
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.2423781192261574,
+      "learning_rate": 1.20249605612186e-06,
+      "loss": 0.0426,
+      "step": 8586
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.396407086050511,
+      "learning_rate": 1.2015375823382264e-06,
+      "loss": 0.0429,
+      "step": 8587
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.5993913145005045,
+      "learning_rate": 1.2005794385247398e-06,
+      "loss": 0.0503,
+      "step": 8588
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.5227206796595838,
+      "learning_rate": 1.199621624764636e-06,
+      "loss": 0.0515,
+      "step": 8589
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.6243251975685995,
+      "learning_rate": 1.1986641411411181e-06,
+      "loss": 0.0524,
+      "step": 8590
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.3810038748179703,
+      "learning_rate": 1.1977069877373625e-06,
+      "loss": 0.042,
+      "step": 8591
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.4217777920582164,
+      "learning_rate": 1.1967501646365147e-06,
+      "loss": 0.0502,
+      "step": 8592
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.6167066473991174,
+      "learning_rate": 1.1957936719216966e-06,
+      "loss": 0.0512,
+      "step": 8593
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.5120967403357835,
+      "learning_rate": 1.1948375096759956e-06,
+      "loss": 0.0517,
+      "step": 8594
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.5946784231182125,
+      "learning_rate": 1.1938816779824753e-06,
+      "loss": 0.0473,
+      "step": 8595
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.7880335970379435,
+      "learning_rate": 1.1929261769241662e-06,
+      "loss": 0.0469,
+      "step": 8596
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.7452235234168656,
+      "learning_rate": 1.1919710065840733e-06,
+      "loss": 0.0496,
+      "step": 8597
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.4659878584255868,
+      "learning_rate": 1.1910161670451697e-06,
+      "loss": 0.0465,
+      "step": 8598
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.652531048565458,
+      "learning_rate": 1.1900616583904046e-06,
+      "loss": 0.0459,
+      "step": 8599
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.5988455220923825,
+      "learning_rate": 1.1891074807026926e-06,
+      "loss": 0.051,
+      "step": 8600
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.3012670161704276,
+      "learning_rate": 1.1881536340649258e-06,
+      "loss": 0.0369,
+      "step": 8601
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.4469568776144042,
+      "learning_rate": 1.1872001185599625e-06,
+      "loss": 0.0459,
+      "step": 8602
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.316046676489979,
+      "learning_rate": 1.186246934270634e-06,
+      "loss": 0.0349,
+      "step": 8603
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.6045879043557556,
+      "learning_rate": 1.185294081279742e-06,
+      "loss": 0.0543,
+      "step": 8604
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.610926294484075,
+      "learning_rate": 1.1843415596700618e-06,
+      "loss": 0.05,
+      "step": 8605
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.4180162785189552,
+      "learning_rate": 1.183389369524337e-06,
+      "loss": 0.0453,
+      "step": 8606
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.4907686049086637,
+      "learning_rate": 1.182437510925286e-06,
+      "loss": 0.0456,
+      "step": 8607
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.5878190486874169,
+      "learning_rate": 1.1814859839555947e-06,
+      "loss": 0.0544,
+      "step": 8608
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.3623270645943937,
+      "learning_rate": 1.1805347886979219e-06,
+      "loss": 0.0442,
+      "step": 8609
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.599371467693969,
+      "learning_rate": 1.1795839252348957e-06,
+      "loss": 0.0453,
+      "step": 8610
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.4999803437332517,
+      "learning_rate": 1.17863339364912e-06,
+      "loss": 0.0509,
+      "step": 8611
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.59768194251578,
+      "learning_rate": 1.1776831940231642e-06,
+      "loss": 0.0597,
+      "step": 8612
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.6869592713976012,
+      "learning_rate": 1.1767333264395735e-06,
+      "loss": 0.0438,
+      "step": 8613
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.449916767455108,
+      "learning_rate": 1.1757837909808628e-06,
+      "loss": 0.043,
+      "step": 8614
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.576697244086845,
+      "learning_rate": 1.1748345877295158e-06,
+      "loss": 0.0621,
+      "step": 8615
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.6121585874276614,
+      "learning_rate": 1.1738857167679884e-06,
+      "loss": 0.0486,
+      "step": 8616
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.610681863397244,
+      "learning_rate": 1.1729371781787119e-06,
+      "loss": 0.0442,
+      "step": 8617
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.589676945449953,
+      "learning_rate": 1.171988972044082e-06,
+      "loss": 0.0526,
+      "step": 8618
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.4328779255876392,
+      "learning_rate": 1.1710410984464716e-06,
+      "loss": 0.0466,
+      "step": 8619
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.5986272448433725,
+      "learning_rate": 1.1700935574682204e-06,
+      "loss": 0.0476,
+      "step": 8620
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.3945821234217615,
+      "learning_rate": 1.1691463491916404e-06,
+      "loss": 0.0424,
+      "step": 8621
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.3581316199233786,
+      "learning_rate": 1.1681994736990143e-06,
+      "loss": 0.0421,
+      "step": 8622
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.616051125250449,
+      "learning_rate": 1.1672529310725995e-06,
+      "loss": 0.0449,
+      "step": 8623
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.513672825527011,
+      "learning_rate": 1.1663067213946177e-06,
+      "loss": 0.0478,
+      "step": 8624
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.581738660379025,
+      "learning_rate": 1.1653608447472698e-06,
+      "loss": 0.0452,
+      "step": 8625
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.6175716375464169,
+      "learning_rate": 1.1644153012127208e-06,
+      "loss": 0.0499,
+      "step": 8626
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.6567081946869795,
+      "learning_rate": 1.1634700908731106e-06,
+      "loss": 0.0563,
+      "step": 8627
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.5909292535527835,
+      "learning_rate": 1.162525213810547e-06,
+      "loss": 0.042,
+      "step": 8628
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.3541146481365631,
+      "learning_rate": 1.1615806701071137e-06,
+      "loss": 0.0427,
+      "step": 8629
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.372837942420161,
+      "learning_rate": 1.1606364598448605e-06,
+      "loss": 0.0444,
+      "step": 8630
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.5627315361978833,
+      "learning_rate": 1.159692583105812e-06,
+      "loss": 0.053,
+      "step": 8631
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.5419208587621287,
+      "learning_rate": 1.158749039971962e-06,
+      "loss": 0.049,
+      "step": 8632
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.2840754490400847,
+      "learning_rate": 1.157805830525275e-06,
+      "loss": 0.0346,
+      "step": 8633
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.4843195866467578,
+      "learning_rate": 1.1568629548476856e-06,
+      "loss": 0.0396,
+      "step": 8634
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.5814382541359278,
+      "learning_rate": 1.1559204130211039e-06,
+      "loss": 0.0528,
+      "step": 8635
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.7951448828343117,
+      "learning_rate": 1.1549782051274045e-06,
+      "loss": 0.0465,
+      "step": 8636
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.6467678756415827,
+      "learning_rate": 1.15403633124844e-06,
+      "loss": 0.0452,
+      "step": 8637
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.2946089299435162,
+      "learning_rate": 1.1530947914660285e-06,
+      "loss": 0.0456,
+      "step": 8638
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.84400729011203,
+      "learning_rate": 1.1521535858619615e-06,
+      "loss": 0.0394,
+      "step": 8639
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.724159872037726,
+      "learning_rate": 1.151212714517999e-06,
+      "loss": 0.0523,
+      "step": 8640
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.6081840643448688,
+      "learning_rate": 1.1502721775158772e-06,
+      "loss": 0.0494,
+      "step": 8641
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.282742011640965,
+      "learning_rate": 1.1493319749372967e-06,
+      "loss": 0.0413,
+      "step": 8642
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.4778855809345055,
+      "learning_rate": 1.1483921068639353e-06,
+      "loss": 0.0445,
+      "step": 8643
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.3695715773883415,
+      "learning_rate": 1.1474525733774377e-06,
+      "loss": 0.0463,
+      "step": 8644
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.576472202366201,
+      "learning_rate": 1.1465133745594203e-06,
+      "loss": 0.0561,
+      "step": 8645
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.5425734133063451,
+      "learning_rate": 1.14557451049147e-06,
+      "loss": 0.0472,
+      "step": 8646
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.5060810439061327,
+      "learning_rate": 1.1446359812551473e-06,
+      "loss": 0.0499,
+      "step": 8647
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.6021134994409165,
+      "learning_rate": 1.1436977869319787e-06,
+      "loss": 0.0472,
+      "step": 8648
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.4865952915573266,
+      "learning_rate": 1.1427599276034685e-06,
+      "loss": 0.0431,
+      "step": 8649
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.3829558053150885,
+      "learning_rate": 1.1418224033510855e-06,
+      "loss": 0.0402,
+      "step": 8650
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.535414289528933,
+      "learning_rate": 1.140885214256272e-06,
+      "loss": 0.0577,
+      "step": 8651
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.7243132175440012,
+      "learning_rate": 1.1399483604004403e-06,
+      "loss": 0.0543,
+      "step": 8652
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.5194590495364049,
+      "learning_rate": 1.139011841864977e-06,
+      "loss": 0.0491,
+      "step": 8653
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.717303260322061,
+      "learning_rate": 1.1380756587312335e-06,
+      "loss": 0.0567,
+      "step": 8654
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.4285278869234561,
+      "learning_rate": 1.1371398110805386e-06,
+      "loss": 0.0443,
+      "step": 8655
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.5087968700629493,
+      "learning_rate": 1.136204298994188e-06,
+      "loss": 0.0465,
+      "step": 8656
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.3287636887802292,
+      "learning_rate": 1.135269122553448e-06,
+      "loss": 0.0374,
+      "step": 8657
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.5752362278404695,
+      "learning_rate": 1.1343342818395558e-06,
+      "loss": 0.0466,
+      "step": 8658
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.486605507161374,
+      "learning_rate": 1.133399776933724e-06,
+      "loss": 0.0518,
+      "step": 8659
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.7163191113180185,
+      "learning_rate": 1.1324656079171288e-06,
+      "loss": 0.0513,
+      "step": 8660
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.493976505178587,
+      "learning_rate": 1.1315317748709237e-06,
+      "loss": 0.0457,
+      "step": 8661
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.2073708834407713,
+      "learning_rate": 1.1305982778762291e-06,
+      "loss": 0.0371,
+      "step": 8662
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.3983621879023187,
+      "learning_rate": 1.1296651170141376e-06,
+      "loss": 0.0424,
+      "step": 8663
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4080438562072919,
+      "learning_rate": 1.1287322923657106e-06,
+      "loss": 0.0457,
+      "step": 8664
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5453861729578757,
+      "learning_rate": 1.1277998040119853e-06,
+      "loss": 0.0388,
+      "step": 8665
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4268554747247901,
+      "learning_rate": 1.1268676520339628e-06,
+      "loss": 0.0431,
+      "step": 8666
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5741436458767701,
+      "learning_rate": 1.1259358365126217e-06,
+      "loss": 0.047,
+      "step": 8667
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.2473785471110066,
+      "learning_rate": 1.1250043575289065e-06,
+      "loss": 0.0355,
+      "step": 8668
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.6203638563143967,
+      "learning_rate": 1.1240732151637352e-06,
+      "loss": 0.0471,
+      "step": 8669
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.627235085690202,
+      "learning_rate": 1.1231424094979932e-06,
+      "loss": 0.0512,
+      "step": 8670
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4865486633142886,
+      "learning_rate": 1.1222119406125426e-06,
+      "loss": 0.0497,
+      "step": 8671
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.9003270357968307,
+      "learning_rate": 1.1212818085882094e-06,
+      "loss": 0.0495,
+      "step": 8672
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.6082930753624527,
+      "learning_rate": 1.120352013505796e-06,
+      "loss": 0.0532,
+      "step": 8673
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5603501555908568,
+      "learning_rate": 1.1194225554460725e-06,
+      "loss": 0.0441,
+      "step": 8674
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.6501065297966977,
+      "learning_rate": 1.118493434489779e-06,
+      "loss": 0.0511,
+      "step": 8675
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5138720737797267,
+      "learning_rate": 1.1175646507176302e-06,
+      "loss": 0.0559,
+      "step": 8676
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5372216024952963,
+      "learning_rate": 1.1166362042103056e-06,
+      "loss": 0.0524,
+      "step": 8677
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.3272649858523338,
+      "learning_rate": 1.1157080950484628e-06,
+      "loss": 0.0441,
+      "step": 8678
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4882585394961207,
+      "learning_rate": 1.1147803233127241e-06,
+      "loss": 0.0474,
+      "step": 8679
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4134773122234536,
+      "learning_rate": 1.1138528890836842e-06,
+      "loss": 0.0459,
+      "step": 8680
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4280170020970242,
+      "learning_rate": 1.1129257924419074e-06,
+      "loss": 0.0513,
+      "step": 8681
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.3179120664920447,
+      "learning_rate": 1.111999033467933e-06,
+      "loss": 0.047,
+      "step": 8682
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5697454101463022,
+      "learning_rate": 1.1110726122422654e-06,
+      "loss": 0.0484,
+      "step": 8683
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.3311403900360779,
+      "learning_rate": 1.110146528845385e-06,
+      "loss": 0.0335,
+      "step": 8684
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4230840686567787,
+      "learning_rate": 1.1092207833577384e-06,
+      "loss": 0.0349,
+      "step": 8685
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.3892815891291426,
+      "learning_rate": 1.1082953758597447e-06,
+      "loss": 0.0392,
+      "step": 8686
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.307990506217155,
+      "learning_rate": 1.107370306431792e-06,
+      "loss": 0.04,
+      "step": 8687
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5814182920239135,
+      "learning_rate": 1.1064455751542436e-06,
+      "loss": 0.0438,
+      "step": 8688
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.343865160788658,
+      "learning_rate": 1.1055211821074275e-06,
+      "loss": 0.0349,
+      "step": 8689
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.2723597488772327,
+      "learning_rate": 1.1045971273716476e-06,
+      "loss": 0.0412,
+      "step": 8690
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4912736167483633,
+      "learning_rate": 1.1036734110271753e-06,
+      "loss": 0.0484,
+      "step": 8691
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.7384963574201484,
+      "learning_rate": 1.1027500331542523e-06,
+      "loss": 0.0504,
+      "step": 8692
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5853992787740907,
+      "learning_rate": 1.1018269938330912e-06,
+      "loss": 0.0522,
+      "step": 8693
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4642178130049124,
+      "learning_rate": 1.1009042931438784e-06,
+      "loss": 0.0471,
+      "step": 8694
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4127145072286977,
+      "learning_rate": 1.0999819311667658e-06,
+      "loss": 0.0495,
+      "step": 8695
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5683912330819498,
+      "learning_rate": 1.099059907981881e-06,
+      "loss": 0.0473,
+      "step": 8696
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.4596551246417173,
+      "learning_rate": 1.0981382236693184e-06,
+      "loss": 0.0461,
+      "step": 8697
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.635664357936403,
+      "learning_rate": 1.0972168783091436e-06,
+      "loss": 0.0566,
+      "step": 8698
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.3264279600670759,
+      "learning_rate": 1.0962958719813926e-06,
+      "loss": 0.0471,
+      "step": 8699
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.257315618850486,
+      "learning_rate": 1.0953752047660754e-06,
+      "loss": 0.0393,
+      "step": 8700
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.341497022512968,
+      "learning_rate": 1.0944548767431667e-06,
+      "loss": 0.039,
+      "step": 8701
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.4645846401110711,
+      "learning_rate": 1.0935348879926178e-06,
+      "loss": 0.0419,
+      "step": 8702
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.4038228855487405,
+      "learning_rate": 1.0926152385943456e-06,
+      "loss": 0.0367,
+      "step": 8703
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.8403982762513762,
+      "learning_rate": 1.0916959286282409e-06,
+      "loss": 0.0513,
+      "step": 8704
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.4067659706810443,
+      "learning_rate": 1.0907769581741606e-06,
+      "loss": 0.0458,
+      "step": 8705
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.354428472313196,
+      "learning_rate": 1.089858327311939e-06,
+      "loss": 0.0408,
+      "step": 8706
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.5481662604829547,
+      "learning_rate": 1.0889400361213737e-06,
+      "loss": 0.0607,
+      "step": 8707
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.744281711896534,
+      "learning_rate": 1.0880220846822392e-06,
+      "loss": 0.0537,
+      "step": 8708
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.4002141535611814,
+      "learning_rate": 1.0871044730742752e-06,
+      "loss": 0.0502,
+      "step": 8709
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.5578376821595767,
+      "learning_rate": 1.0861872013771958e-06,
+      "loss": 0.0511,
+      "step": 8710
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.7716819235603667,
+      "learning_rate": 1.0852702696706807e-06,
+      "loss": 0.0537,
+      "step": 8711
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.6727194201491118,
+      "learning_rate": 1.0843536780343866e-06,
+      "loss": 0.0456,
+      "step": 8712
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.49711488513233,
+      "learning_rate": 1.0834374265479347e-06,
+      "loss": 0.0458,
+      "step": 8713
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.6697889951619662,
+      "learning_rate": 1.082521515290922e-06,
+      "loss": 0.0463,
+      "step": 8714
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.1893122752854486,
+      "learning_rate": 1.081605944342911e-06,
+      "loss": 0.032,
+      "step": 8715
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.3788904308151453,
+      "learning_rate": 1.0806907137834377e-06,
+      "loss": 0.0423,
+      "step": 8716
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.6836252583564824,
+      "learning_rate": 1.0797758236920063e-06,
+      "loss": 0.0557,
+      "step": 8717
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.7030975644250759,
+      "learning_rate": 1.0788612741480947e-06,
+      "loss": 0.0438,
+      "step": 8718
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.7254683561152628,
+      "learning_rate": 1.0779470652311475e-06,
+      "loss": 0.0475,
+      "step": 8719
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.3649290956955344,
+      "learning_rate": 1.0770331970205834e-06,
+      "loss": 0.043,
+      "step": 8720
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.718566869492126,
+      "learning_rate": 1.0761196695957882e-06,
+      "loss": 0.0514,
+      "step": 8721
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.4067620604936915,
+      "learning_rate": 1.0752064830361202e-06,
+      "loss": 0.0402,
+      "step": 8722
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.5341310704741666,
+      "learning_rate": 1.0742936374209056e-06,
+      "loss": 0.046,
+      "step": 8723
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.3092533755309306,
+      "learning_rate": 1.0733811328294453e-06,
+      "loss": 0.0414,
+      "step": 8724
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.3943252796019423,
+      "learning_rate": 1.0724689693410052e-06,
+      "loss": 0.0463,
+      "step": 8725
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.3947905734558317,
+      "learning_rate": 1.071557147034828e-06,
+      "loss": 0.0546,
+      "step": 8726
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.7754939271621943,
+      "learning_rate": 1.0706456659901204e-06,
+      "loss": 0.0586,
+      "step": 8727
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.537108118182416,
+      "learning_rate": 1.0697345262860638e-06,
+      "loss": 0.0379,
+      "step": 8728
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.6417435736150496,
+      "learning_rate": 1.068823728001806e-06,
+      "loss": 0.0504,
+      "step": 8729
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.4142194029666615,
+      "learning_rate": 1.0679132712164702e-06,
+      "loss": 0.0426,
+      "step": 8730
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.448040446353482,
+      "learning_rate": 1.067003156009145e-06,
+      "loss": 0.0515,
+      "step": 8731
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.4846537490367535,
+      "learning_rate": 1.0660933824588932e-06,
+      "loss": 0.0485,
+      "step": 8732
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.5657631118437902,
+      "learning_rate": 1.0651839506447464e-06,
+      "loss": 0.0498,
+      "step": 8733
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.4579179848584627,
+      "learning_rate": 1.064274860645706e-06,
+      "loss": 0.0483,
+      "step": 8734
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.589541583739505,
+      "learning_rate": 1.0633661125407418e-06,
+      "loss": 0.0506,
+      "step": 8735
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.506847849878915,
+      "learning_rate": 1.0624577064087998e-06,
+      "loss": 0.0417,
+      "step": 8736
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.3345558905313566,
+      "learning_rate": 1.0615496423287896e-06,
+      "loss": 0.0456,
+      "step": 8737
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.5341032669021522,
+      "learning_rate": 1.0606419203795975e-06,
+      "loss": 0.0484,
+      "step": 8738
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.7359820339919008,
+      "learning_rate": 1.059734540640075e-06,
+      "loss": 0.0505,
+      "step": 8739
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.2955877374309666,
+      "learning_rate": 1.0588275031890455e-06,
+      "loss": 0.0378,
+      "step": 8740
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4003891859511555,
+      "learning_rate": 1.057920808105301e-06,
+      "loss": 0.0448,
+      "step": 8741
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.3820908158222436,
+      "learning_rate": 1.0570144554676092e-06,
+      "loss": 0.0418,
+      "step": 8742
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4152866756912188,
+      "learning_rate": 1.0561084453547016e-06,
+      "loss": 0.0462,
+      "step": 8743
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4233780016520368,
+      "learning_rate": 1.055202777845285e-06,
+      "loss": 0.0403,
+      "step": 8744
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.5129650197065576,
+      "learning_rate": 1.0542974530180327e-06,
+      "loss": 0.05,
+      "step": 8745
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.3690041538815456,
+      "learning_rate": 1.0533924709515902e-06,
+      "loss": 0.0455,
+      "step": 8746
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4596056520261682,
+      "learning_rate": 1.0524878317245713e-06,
+      "loss": 0.0504,
+      "step": 8747
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4479850841049984,
+      "learning_rate": 1.051583535415564e-06,
+      "loss": 0.0365,
+      "step": 8748
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.373508659862148,
+      "learning_rate": 1.0506795821031212e-06,
+      "loss": 0.0464,
+      "step": 8749
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.6301214825816965,
+      "learning_rate": 1.049775971865772e-06,
+      "loss": 0.0498,
+      "step": 8750
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.3193886075016072,
+      "learning_rate": 1.0488727047820108e-06,
+      "loss": 0.033,
+      "step": 8751
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.613290170834081,
+      "learning_rate": 1.0479697809303035e-06,
+      "loss": 0.0516,
+      "step": 8752
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.5584302344052667,
+      "learning_rate": 1.0470672003890858e-06,
+      "loss": 0.0433,
+      "step": 8753
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4777825509915168,
+      "learning_rate": 1.046164963236767e-06,
+      "loss": 0.0501,
+      "step": 8754
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.5072193656964288,
+      "learning_rate": 1.0452630695517208e-06,
+      "loss": 0.0427,
+      "step": 8755
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4850964624208396,
+      "learning_rate": 1.0443615194122969e-06,
+      "loss": 0.0463,
+      "step": 8756
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4646351833393905,
+      "learning_rate": 1.0434603128968112e-06,
+      "loss": 0.0469,
+      "step": 8757
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.2784578683891326,
+      "learning_rate": 1.0425594500835512e-06,
+      "loss": 0.0404,
+      "step": 8758
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.6368838695048964,
+      "learning_rate": 1.0416589310507723e-06,
+      "loss": 0.0489,
+      "step": 8759
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.509262635574599,
+      "learning_rate": 1.0407587558767056e-06,
+      "loss": 0.0494,
+      "step": 8760
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.6624956555128414,
+      "learning_rate": 1.0398589246395457e-06,
+      "loss": 0.0536,
+      "step": 8761
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4471894983615914,
+      "learning_rate": 1.0389594374174628e-06,
+      "loss": 0.0434,
+      "step": 8762
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.5689389000444693,
+      "learning_rate": 1.0380602942885937e-06,
+      "loss": 0.0486,
+      "step": 8763
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.59127909641896,
+      "learning_rate": 1.0371614953310465e-06,
+      "loss": 0.053,
+      "step": 8764
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.5117878643022244,
+      "learning_rate": 1.0362630406228986e-06,
+      "loss": 0.0434,
+      "step": 8765
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.430700174163597,
+      "learning_rate": 1.0353649302421982e-06,
+      "loss": 0.0474,
+      "step": 8766
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.8450598099289466,
+      "learning_rate": 1.0344671642669656e-06,
+      "loss": 0.0497,
+      "step": 8767
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.2708096092054417,
+      "learning_rate": 1.033569742775188e-06,
+      "loss": 0.0379,
+      "step": 8768
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.632131534412719,
+      "learning_rate": 1.0326726658448238e-06,
+      "loss": 0.0532,
+      "step": 8769
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.245203290786845,
+      "learning_rate": 1.0317759335538002e-06,
+      "loss": 0.0378,
+      "step": 8770
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.3530264251542963,
+      "learning_rate": 1.0308795459800186e-06,
+      "loss": 0.0441,
+      "step": 8771
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4332085647035122,
+      "learning_rate": 1.029983503201345e-06,
+      "loss": 0.0444,
+      "step": 8772
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.400410541932451,
+      "learning_rate": 1.02908780529562e-06,
+      "loss": 0.0425,
+      "step": 8773
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.6666841004742612,
+      "learning_rate": 1.0281924523406518e-06,
+      "loss": 0.0504,
+      "step": 8774
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.4666928689361818,
+      "learning_rate": 1.0272974444142192e-06,
+      "loss": 0.0481,
+      "step": 8775
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.28456884729643,
+      "learning_rate": 1.0264027815940692e-06,
+      "loss": 0.0376,
+      "step": 8776
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.8066625105372123,
+      "learning_rate": 1.0255084639579232e-06,
+      "loss": 0.0537,
+      "step": 8777
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.5567682959897602,
+      "learning_rate": 1.0246144915834683e-06,
+      "loss": 0.0512,
+      "step": 8778
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.7222915824930451,
+      "learning_rate": 1.0237208645483648e-06,
+      "loss": 0.0506,
+      "step": 8779
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.607556983171655,
+      "learning_rate": 1.0228275829302415e-06,
+      "loss": 0.0509,
+      "step": 8780
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.8842353193484842,
+      "learning_rate": 1.021934646806696e-06,
+      "loss": 0.059,
+      "step": 8781
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.5680425666035738,
+      "learning_rate": 1.0210420562552963e-06,
+      "loss": 0.0432,
+      "step": 8782
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.6600658814451268,
+      "learning_rate": 1.020149811353584e-06,
+      "loss": 0.0407,
+      "step": 8783
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.5964090690810402,
+      "learning_rate": 1.0192579121790652e-06,
+      "loss": 0.0457,
+      "step": 8784
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.4714120993328248,
+      "learning_rate": 1.0183663588092214e-06,
+      "loss": 0.0503,
+      "step": 8785
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.5384554420353587,
+      "learning_rate": 1.0174751513214992e-06,
+      "loss": 0.0483,
+      "step": 8786
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.4258340331164856,
+      "learning_rate": 1.0165842897933188e-06,
+      "loss": 0.0462,
+      "step": 8787
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.0253356486129674,
+      "learning_rate": 1.0156937743020657e-06,
+      "loss": 0.0515,
+      "step": 8788
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.3714356621762054,
+      "learning_rate": 1.014803604925102e-06,
+      "loss": 0.0444,
+      "step": 8789
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.542973732715464,
+      "learning_rate": 1.0139137817397537e-06,
+      "loss": 0.0418,
+      "step": 8790
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.5581835248773026,
+      "learning_rate": 1.013024304823322e-06,
+      "loss": 0.0527,
+      "step": 8791
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.7057100004497252,
+      "learning_rate": 1.0121351742530728e-06,
+      "loss": 0.0464,
+      "step": 8792
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.5767212206346282,
+      "learning_rate": 1.0112463901062453e-06,
+      "loss": 0.0478,
+      "step": 8793
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.5295288425758895,
+      "learning_rate": 1.010357952460046e-06,
+      "loss": 0.051,
+      "step": 8794
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.613798376132376,
+      "learning_rate": 1.0094698613916558e-06,
+      "loss": 0.0457,
+      "step": 8795
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.6117486954109663,
+      "learning_rate": 1.00858211697822e-06,
+      "loss": 0.0461,
+      "step": 8796
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.5788641001584318,
+      "learning_rate": 1.007694719296859e-06,
+      "loss": 0.0496,
+      "step": 8797
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.6846278568899848,
+      "learning_rate": 1.0068076684246586e-06,
+      "loss": 0.0584,
+      "step": 8798
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.5396276638518092,
+      "learning_rate": 1.0059209644386775e-06,
+      "loss": 0.044,
+      "step": 8799
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.6523128589519536,
+      "learning_rate": 1.0050346074159406e-06,
+      "loss": 0.0472,
+      "step": 8800
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.5279275546385982,
+      "learning_rate": 1.0041485974334493e-06,
+      "loss": 0.0511,
+      "step": 8801
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.4378350064755228,
+      "learning_rate": 1.0032629345681666e-06,
+      "loss": 0.0539,
+      "step": 8802
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.603667699107373,
+      "learning_rate": 1.0023776188970325e-06,
+      "loss": 0.0544,
+      "step": 8803
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.5081346375230482,
+      "learning_rate": 1.0014926504969535e-06,
+      "loss": 0.0515,
+      "step": 8804
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.214810196934005,
+      "learning_rate": 1.000608029444805e-06,
+      "loss": 0.0412,
+      "step": 8805
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.4006742833955323,
+      "learning_rate": 9.997237558174334e-07,
+      "loss": 0.04,
+      "step": 8806
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.4129550328153881,
+      "learning_rate": 9.988398296916569e-07,
+      "loss": 0.0458,
+      "step": 8807
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.341890443918045,
+      "learning_rate": 9.979562511442586e-07,
+      "loss": 0.0484,
+      "step": 8808
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.4135832754950768,
+      "learning_rate": 9.970730202519986e-07,
+      "loss": 0.0423,
+      "step": 8809
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.373944204233886,
+      "learning_rate": 9.961901370915994e-07,
+      "loss": 0.0488,
+      "step": 8810
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.3851600214535538,
+      "learning_rate": 9.953076017397579e-07,
+      "loss": 0.041,
+      "step": 8811
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.2682180285044817,
+      "learning_rate": 9.944254142731375e-07,
+      "loss": 0.0408,
+      "step": 8812
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.3586509882649467,
+      "learning_rate": 9.935435747683758e-07,
+      "loss": 0.0475,
+      "step": 8813
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.5952142618839091,
+      "learning_rate": 9.926620833020755e-07,
+      "loss": 0.0555,
+      "step": 8814
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.4946890099049241,
+      "learning_rate": 9.917809399508144e-07,
+      "loss": 0.0453,
+      "step": 8815
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.2764898861252318,
+      "learning_rate": 9.909001447911336e-07,
+      "loss": 0.0419,
+      "step": 8816
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.6023593000761078,
+      "learning_rate": 9.900196978995497e-07,
+      "loss": 0.0479,
+      "step": 8817
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.4781385659474782,
+      "learning_rate": 9.891395993525433e-07,
+      "loss": 0.0491,
+      "step": 8818
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.5359673023133742,
+      "learning_rate": 9.882598492265716e-07,
+      "loss": 0.0489,
+      "step": 8819
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.2640913978210047,
+      "learning_rate": 9.873804475980552e-07,
+      "loss": 0.0361,
+      "step": 8820
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.4363770850360897,
+      "learning_rate": 9.865013945433905e-07,
+      "loss": 0.0451,
+      "step": 8821
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.3441206019644,
+      "learning_rate": 9.856226901389376e-07,
+      "loss": 0.0402,
+      "step": 8822
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.4839445448950859,
+      "learning_rate": 9.847443344610296e-07,
+      "loss": 0.0417,
+      "step": 8823
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.558399843938944,
+      "learning_rate": 9.838663275859678e-07,
+      "loss": 0.0507,
+      "step": 8824
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.5097129476885902,
+      "learning_rate": 9.829886695900265e-07,
+      "loss": 0.0467,
+      "step": 8825
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.4783404184696236,
+      "learning_rate": 9.821113605494449e-07,
+      "loss": 0.0451,
+      "step": 8826
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.3442201264781266,
+      "learning_rate": 9.812344005404361e-07,
+      "loss": 0.0397,
+      "step": 8827
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.3151017061671793,
+      "learning_rate": 9.803577896391809e-07,
+      "loss": 0.0369,
+      "step": 8828
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.7710901547375577,
+      "learning_rate": 9.794815279218288e-07,
+      "loss": 0.0521,
+      "step": 8829
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.483713453038063,
+      "learning_rate": 9.786056154645001e-07,
+      "loss": 0.052,
+      "step": 8830
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.4099590464477867,
+      "learning_rate": 9.77730052343287e-07,
+      "loss": 0.0489,
+      "step": 8831
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.6979591030207262,
+      "learning_rate": 9.768548386342458e-07,
+      "loss": 0.0527,
+      "step": 8832
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.6792590347500782,
+      "learning_rate": 9.75979974413409e-07,
+      "loss": 0.0555,
+      "step": 8833
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.076762898293552,
+      "learning_rate": 9.751054597567744e-07,
+      "loss": 0.0322,
+      "step": 8834
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.4642910314927935,
+      "learning_rate": 9.742312947403103e-07,
+      "loss": 0.049,
+      "step": 8835
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.4940258544132559,
+      "learning_rate": 9.733574794399537e-07,
+      "loss": 0.0444,
+      "step": 8836
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.4037131497400552,
+      "learning_rate": 9.724840139316144e-07,
+      "loss": 0.0421,
+      "step": 8837
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.5103223092410143,
+      "learning_rate": 9.71610898291168e-07,
+      "loss": 0.0479,
+      "step": 8838
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.2864339785832624,
+      "learning_rate": 9.707381325944642e-07,
+      "loss": 0.0416,
+      "step": 8839
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.3260510759892015,
+      "learning_rate": 9.698657169173176e-07,
+      "loss": 0.042,
+      "step": 8840
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.5442615736113157,
+      "learning_rate": 9.689936513355147e-07,
+      "loss": 0.0496,
+      "step": 8841
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.3643606418963607,
+      "learning_rate": 9.681219359248106e-07,
+      "loss": 0.0406,
+      "step": 8842
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.5220865572806,
+      "learning_rate": 9.672505707609326e-07,
+      "loss": 0.0537,
+      "step": 8843
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.182433106239767,
+      "learning_rate": 9.663795559195733e-07,
+      "loss": 0.0317,
+      "step": 8844
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.5795583322988773,
+      "learning_rate": 9.655088914763994e-07,
+      "loss": 0.0515,
+      "step": 8845
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.4861466924809996,
+      "learning_rate": 9.646385775070444e-07,
+      "loss": 0.0432,
+      "step": 8846
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.1290982155382336,
+      "learning_rate": 9.637686140871121e-07,
+      "loss": 0.0391,
+      "step": 8847
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4827467735414195,
+      "learning_rate": 9.628990012921734e-07,
+      "loss": 0.0632,
+      "step": 8848
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.7924657188638018,
+      "learning_rate": 9.620297391977746e-07,
+      "loss": 0.0545,
+      "step": 8849
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.022742935537372,
+      "learning_rate": 9.611608278794249e-07,
+      "loss": 0.0553,
+      "step": 8850
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4422068554644836,
+      "learning_rate": 9.602922674126085e-07,
+      "loss": 0.0394,
+      "step": 8851
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.6369553880332486,
+      "learning_rate": 9.59424057872776e-07,
+      "loss": 0.049,
+      "step": 8852
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.451703924426917,
+      "learning_rate": 9.585561993353482e-07,
+      "loss": 0.0425,
+      "step": 8853
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.3786391330358592,
+      "learning_rate": 9.576886918757134e-07,
+      "loss": 0.0425,
+      "step": 8854
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.483401282370702,
+      "learning_rate": 9.568215355692351e-07,
+      "loss": 0.0538,
+      "step": 8855
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4605764005482178,
+      "learning_rate": 9.559547304912392e-07,
+      "loss": 0.0481,
+      "step": 8856
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.6277322929927311,
+      "learning_rate": 9.550882767170278e-07,
+      "loss": 0.0536,
+      "step": 8857
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.5463550556685373,
+      "learning_rate": 9.54222174321867e-07,
+      "loss": 0.0469,
+      "step": 8858
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4353914517934903,
+      "learning_rate": 9.533564233809939e-07,
+      "loss": 0.0449,
+      "step": 8859
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.6869508507798658,
+      "learning_rate": 9.524910239696189e-07,
+      "loss": 0.0581,
+      "step": 8860
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.5780154260581238,
+      "learning_rate": 9.516259761629148e-07,
+      "loss": 0.0427,
+      "step": 8861
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.51502926948865,
+      "learning_rate": 9.507612800360316e-07,
+      "loss": 0.0464,
+      "step": 8862
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.5584709207979923,
+      "learning_rate": 9.498969356640836e-07,
+      "loss": 0.0492,
+      "step": 8863
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4185636705077296,
+      "learning_rate": 9.490329431221545e-07,
+      "loss": 0.0489,
+      "step": 8864
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4515765777261733,
+      "learning_rate": 9.48169302485299e-07,
+      "loss": 0.0475,
+      "step": 8865
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.7595135351921316,
+      "learning_rate": 9.473060138285434e-07,
+      "loss": 0.0643,
+      "step": 8866
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4003256745507489,
+      "learning_rate": 9.464430772268779e-07,
+      "loss": 0.0437,
+      "step": 8867
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.388042495068433,
+      "learning_rate": 9.455804927552681e-07,
+      "loss": 0.0375,
+      "step": 8868
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.3559387656083617,
+      "learning_rate": 9.447182604886446e-07,
+      "loss": 0.0442,
+      "step": 8869
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.416570404498866,
+      "learning_rate": 9.438563805019096e-07,
+      "loss": 0.0386,
+      "step": 8870
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.705767330847675,
+      "learning_rate": 9.429948528699329e-07,
+      "loss": 0.0568,
+      "step": 8871
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4819682440463875,
+      "learning_rate": 9.421336776675565e-07,
+      "loss": 0.0499,
+      "step": 8872
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.5690218538326075,
+      "learning_rate": 9.412728549695888e-07,
+      "loss": 0.0524,
+      "step": 8873
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.2435858016717076,
+      "learning_rate": 9.404123848508107e-07,
+      "loss": 0.0368,
+      "step": 8874
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.5186243940801072,
+      "learning_rate": 9.395522673859698e-07,
+      "loss": 0.0442,
+      "step": 8875
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.5427425080659716,
+      "learning_rate": 9.386925026497835e-07,
+      "loss": 0.0468,
+      "step": 8876
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4019306326656193,
+      "learning_rate": 9.378330907169387e-07,
+      "loss": 0.0452,
+      "step": 8877
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.7366399206142773,
+      "learning_rate": 9.369740316620935e-07,
+      "loss": 0.056,
+      "step": 8878
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.1102651671172517,
+      "learning_rate": 9.361153255598721e-07,
+      "loss": 0.0349,
+      "step": 8879
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.7912278041817948,
+      "learning_rate": 9.352569724848715e-07,
+      "loss": 0.0573,
+      "step": 8880
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4831430918563488,
+      "learning_rate": 9.34398972511656e-07,
+      "loss": 0.0582,
+      "step": 8881
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.4082439020614645,
+      "learning_rate": 9.33541325714759e-07,
+      "loss": 0.0406,
+      "step": 8882
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.5344946218336881,
+      "learning_rate": 9.326840321686826e-07,
+      "loss": 0.0437,
+      "step": 8883
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.557982352612514,
+      "learning_rate": 9.318270919479022e-07,
+      "loss": 0.0477,
+      "step": 8884
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.2931425737049431,
+      "learning_rate": 9.309705051268564e-07,
+      "loss": 0.0437,
+      "step": 8885
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.3837348866786165,
+      "learning_rate": 9.301142717799594e-07,
+      "loss": 0.0424,
+      "step": 8886
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.6136291831205603,
+      "learning_rate": 9.292583919815906e-07,
+      "loss": 0.0536,
+      "step": 8887
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.2269054288742103,
+      "learning_rate": 9.284028658060995e-07,
+      "loss": 0.0365,
+      "step": 8888
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.7875405393852057,
+      "learning_rate": 9.275476933278038e-07,
+      "loss": 0.058,
+      "step": 8889
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.405820008470654,
+      "learning_rate": 9.266928746209946e-07,
+      "loss": 0.0438,
+      "step": 8890
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.3503676716642974,
+      "learning_rate": 9.258384097599266e-07,
+      "loss": 0.0328,
+      "step": 8891
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.66583858820306,
+      "learning_rate": 9.249842988188295e-07,
+      "loss": 0.0446,
+      "step": 8892
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.2711118931266623,
+      "learning_rate": 9.241305418718982e-07,
+      "loss": 0.0378,
+      "step": 8893
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.555801336235142,
+      "learning_rate": 9.232771389932976e-07,
+      "loss": 0.0459,
+      "step": 8894
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.480189942479658,
+      "learning_rate": 9.224240902571618e-07,
+      "loss": 0.043,
+      "step": 8895
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.4656811459885832,
+      "learning_rate": 9.215713957375961e-07,
+      "loss": 0.0454,
+      "step": 8896
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.4912856411713014,
+      "learning_rate": 9.20719055508672e-07,
+      "loss": 0.0451,
+      "step": 8897
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.5684205187235856,
+      "learning_rate": 9.198670696444339e-07,
+      "loss": 0.0474,
+      "step": 8898
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.3387273689649501,
+      "learning_rate": 9.190154382188921e-07,
+      "loss": 0.0445,
+      "step": 8899
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.7424405804316927,
+      "learning_rate": 9.181641613060271e-07,
+      "loss": 0.0591,
+      "step": 8900
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.2953173454594806,
+      "learning_rate": 9.173132389797878e-07,
+      "loss": 0.0313,
+      "step": 8901
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.5458891993022181,
+      "learning_rate": 9.164626713140956e-07,
+      "loss": 0.0477,
+      "step": 8902
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.958214775298957,
+      "learning_rate": 9.156124583828368e-07,
+      "loss": 0.0561,
+      "step": 8903
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.5415768556482872,
+      "learning_rate": 9.147626002598708e-07,
+      "loss": 0.0568,
+      "step": 8904
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.7458338192573024,
+      "learning_rate": 9.139130970190235e-07,
+      "loss": 0.0502,
+      "step": 8905
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.33948263990128,
+      "learning_rate": 9.130639487340903e-07,
+      "loss": 0.0429,
+      "step": 8906
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.4869310372449824,
+      "learning_rate": 9.12215155478835e-07,
+      "loss": 0.0444,
+      "step": 8907
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.6133436769444875,
+      "learning_rate": 9.113667173269947e-07,
+      "loss": 0.0486,
+      "step": 8908
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.4643045993720862,
+      "learning_rate": 9.105186343522698e-07,
+      "loss": 0.0412,
+      "step": 8909
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.3937916003743358,
+      "learning_rate": 9.096709066283355e-07,
+      "loss": 0.0449,
+      "step": 8910
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.5561828186758775,
+      "learning_rate": 9.088235342288315e-07,
+      "loss": 0.047,
+      "step": 8911
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.6725011260253135,
+      "learning_rate": 9.079765172273697e-07,
+      "loss": 0.0494,
+      "step": 8912
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.676783555269748,
+      "learning_rate": 9.071298556975278e-07,
+      "loss": 0.0474,
+      "step": 8913
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.398297046029956,
+      "learning_rate": 9.062835497128575e-07,
+      "loss": 0.0438,
+      "step": 8914
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.3314652328600187,
+      "learning_rate": 9.054375993468745e-07,
+      "loss": 0.0369,
+      "step": 8915
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.7447067324934413,
+      "learning_rate": 9.045920046730683e-07,
+      "loss": 0.0576,
+      "step": 8916
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.3778030108000554,
+      "learning_rate": 9.037467657648941e-07,
+      "loss": 0.033,
+      "step": 8917
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.682401234750559,
+      "learning_rate": 9.029018826957775e-07,
+      "loss": 0.0468,
+      "step": 8918
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.97998012338404,
+      "learning_rate": 9.020573555391116e-07,
+      "loss": 0.058,
+      "step": 8919
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.5185950448877157,
+      "learning_rate": 9.01213184368262e-07,
+      "loss": 0.0477,
+      "step": 8920
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.3077831073047794,
+      "learning_rate": 9.00369369256559e-07,
+      "loss": 0.0444,
+      "step": 8921
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.258866649604549,
+      "learning_rate": 8.99525910277308e-07,
+      "loss": 0.0378,
+      "step": 8922
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.496496291836476,
+      "learning_rate": 8.986828075037768e-07,
+      "loss": 0.048,
+      "step": 8923
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.743012011117733,
+      "learning_rate": 8.978400610092058e-07,
+      "loss": 0.0555,
+      "step": 8924
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.7607407885442508,
+      "learning_rate": 8.969976708668032e-07,
+      "loss": 0.0469,
+      "step": 8925
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.606959196856989,
+      "learning_rate": 8.961556371497493e-07,
+      "loss": 0.0464,
+      "step": 8926
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.4554689227465931,
+      "learning_rate": 8.953139599311883e-07,
+      "loss": 0.0414,
+      "step": 8927
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.27526330763992,
+      "learning_rate": 8.944726392842385e-07,
+      "loss": 0.047,
+      "step": 8928
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.6510390643667272,
+      "learning_rate": 8.936316752819834e-07,
+      "loss": 0.0507,
+      "step": 8929
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.4372799593863448,
+      "learning_rate": 8.927910679974783e-07,
+      "loss": 0.0496,
+      "step": 8930
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.455611894118685,
+      "learning_rate": 8.919508175037439e-07,
+      "loss": 0.0409,
+      "step": 8931
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.5521467655265355,
+      "learning_rate": 8.911109238737748e-07,
+      "loss": 0.0536,
+      "step": 8932
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.5023983141121513,
+      "learning_rate": 8.902713871805302e-07,
+      "loss": 0.0467,
+      "step": 8933
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.7411531547120003,
+      "learning_rate": 8.894322074969419e-07,
+      "loss": 0.0536,
+      "step": 8934
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.3954188702441621,
+      "learning_rate": 8.885933848959083e-07,
+      "loss": 0.0461,
+      "step": 8935
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.2754190422906153,
+      "learning_rate": 8.877549194502972e-07,
+      "loss": 0.0377,
+      "step": 8936
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.5800340367118,
+      "learning_rate": 8.86916811232944e-07,
+      "loss": 0.0551,
+      "step": 8937
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.467730972761269,
+      "learning_rate": 8.86079060316658e-07,
+      "loss": 0.0557,
+      "step": 8938
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.4660268590982395,
+      "learning_rate": 8.852416667742108e-07,
+      "loss": 0.0429,
+      "step": 8939
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.5712180374033295,
+      "learning_rate": 8.844046306783488e-07,
+      "loss": 0.0412,
+      "step": 8940
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.4608688824960407,
+      "learning_rate": 8.835679521017842e-07,
+      "loss": 0.0433,
+      "step": 8941
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.497105914919798,
+      "learning_rate": 8.827316311171986e-07,
+      "loss": 0.0477,
+      "step": 8942
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.5381419947103372,
+      "learning_rate": 8.818956677972407e-07,
+      "loss": 0.0498,
+      "step": 8943
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.5855770823760886,
+      "learning_rate": 8.810600622145337e-07,
+      "loss": 0.054,
+      "step": 8944
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.4341688561111856,
+      "learning_rate": 8.802248144416625e-07,
+      "loss": 0.0416,
+      "step": 8945
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.6459914327100782,
+      "learning_rate": 8.793899245511884e-07,
+      "loss": 0.0467,
+      "step": 8946
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.3821434203234158,
+      "learning_rate": 8.785553926156354e-07,
+      "loss": 0.047,
+      "step": 8947
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.4548495232975707,
+      "learning_rate": 8.777212187074996e-07,
+      "loss": 0.0447,
+      "step": 8948
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.551450707363934,
+      "learning_rate": 8.768874028992431e-07,
+      "loss": 0.0438,
+      "step": 8949
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.513944080432506,
+      "learning_rate": 8.76053945263301e-07,
+      "loss": 0.0489,
+      "step": 8950
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.3817290415720977,
+      "learning_rate": 8.752208458720762e-07,
+      "loss": 0.0432,
+      "step": 8951
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.346948102230618,
+      "learning_rate": 8.743881047979381e-07,
+      "loss": 0.0413,
+      "step": 8952
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.513365042625213,
+      "learning_rate": 8.735557221132268e-07,
+      "loss": 0.0474,
+      "step": 8953
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.440045238603853,
+      "learning_rate": 8.727236978902492e-07,
+      "loss": 0.0438,
+      "step": 8954
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.4428019679273554,
+      "learning_rate": 8.718920322012858e-07,
+      "loss": 0.0468,
+      "step": 8955
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.479941510035178,
+      "learning_rate": 8.710607251185799e-07,
+      "loss": 0.0432,
+      "step": 8956
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4818894274913803,
+      "learning_rate": 8.702297767143497e-07,
+      "loss": 0.0361,
+      "step": 8957
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4974284264178659,
+      "learning_rate": 8.693991870607771e-07,
+      "loss": 0.0475,
+      "step": 8958
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4934142023669315,
+      "learning_rate": 8.685689562300159e-07,
+      "loss": 0.0525,
+      "step": 8959
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.6622345692125744,
+      "learning_rate": 8.677390842941857e-07,
+      "loss": 0.0502,
+      "step": 8960
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.606256589133991,
+      "learning_rate": 8.669095713253795e-07,
+      "loss": 0.0444,
+      "step": 8961
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.5078717472467353,
+      "learning_rate": 8.66080417395655e-07,
+      "loss": 0.0546,
+      "step": 8962
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4019987152525188,
+      "learning_rate": 8.652516225770419e-07,
+      "loss": 0.0435,
+      "step": 8963
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.3791596253621285,
+      "learning_rate": 8.64423186941536e-07,
+      "loss": 0.0432,
+      "step": 8964
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4422979317113884,
+      "learning_rate": 8.635951105611035e-07,
+      "loss": 0.041,
+      "step": 8965
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.3701412877269576,
+      "learning_rate": 8.627673935076769e-07,
+      "loss": 0.0449,
+      "step": 8966
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.3285394799292558,
+      "learning_rate": 8.619400358531626e-07,
+      "loss": 0.0422,
+      "step": 8967
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.241747993182049,
+      "learning_rate": 8.611130376694299e-07,
+      "loss": 0.0376,
+      "step": 8968
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.662961027116381,
+      "learning_rate": 8.602863990283217e-07,
+      "loss": 0.0432,
+      "step": 8969
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4828059178672184,
+      "learning_rate": 8.594601200016472e-07,
+      "loss": 0.0463,
+      "step": 8970
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.5003132376295374,
+      "learning_rate": 8.586342006611847e-07,
+      "loss": 0.0482,
+      "step": 8971
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4228374227419227,
+      "learning_rate": 8.578086410786796e-07,
+      "loss": 0.0496,
+      "step": 8972
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4029162320759927,
+      "learning_rate": 8.569834413258505e-07,
+      "loss": 0.0418,
+      "step": 8973
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.3007415715774118,
+      "learning_rate": 8.561586014743789e-07,
+      "loss": 0.0433,
+      "step": 8974
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4906295592309138,
+      "learning_rate": 8.553341215959215e-07,
+      "loss": 0.0408,
+      "step": 8975
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.5469424546303432,
+      "learning_rate": 8.545100017620988e-07,
+      "loss": 0.0528,
+      "step": 8976
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4790081689765433,
+      "learning_rate": 8.536862420445019e-07,
+      "loss": 0.0411,
+      "step": 8977
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.325325073211631,
+      "learning_rate": 8.528628425146885e-07,
+      "loss": 0.0438,
+      "step": 8978
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.6423657308090662,
+      "learning_rate": 8.520398032441896e-07,
+      "loss": 0.0475,
+      "step": 8979
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.8370377242915266,
+      "learning_rate": 8.512171243044992e-07,
+      "loss": 0.0464,
+      "step": 8980
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.3428002091129605,
+      "learning_rate": 8.503948057670863e-07,
+      "loss": 0.0363,
+      "step": 8981
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.5992235106442614,
+      "learning_rate": 8.495728477033832e-07,
+      "loss": 0.0529,
+      "step": 8982
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.5251309877657038,
+      "learning_rate": 8.487512501847933e-07,
+      "loss": 0.049,
+      "step": 8983
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.3934953433798714,
+      "learning_rate": 8.479300132826873e-07,
+      "loss": 0.0387,
+      "step": 8984
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.656173642326572,
+      "learning_rate": 8.47109137068407e-07,
+      "loss": 0.0524,
+      "step": 8985
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.600286517758158,
+      "learning_rate": 8.462886216132604e-07,
+      "loss": 0.0581,
+      "step": 8986
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.5065598208541373,
+      "learning_rate": 8.45468466988526e-07,
+      "loss": 0.0485,
+      "step": 8987
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.5658449333804347,
+      "learning_rate": 8.446486732654508e-07,
+      "loss": 0.0455,
+      "step": 8988
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.3885582390485607,
+      "learning_rate": 8.438292405152477e-07,
+      "loss": 0.0517,
+      "step": 8989
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.338702883809766,
+      "learning_rate": 8.430101688091009e-07,
+      "loss": 0.0377,
+      "step": 8990
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4501031663604222,
+      "learning_rate": 8.421914582181639e-07,
+      "loss": 0.0484,
+      "step": 8991
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.606599166203397,
+      "learning_rate": 8.413731088135563e-07,
+      "loss": 0.049,
+      "step": 8992
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.6099786011135102,
+      "learning_rate": 8.405551206663686e-07,
+      "loss": 0.0443,
+      "step": 8993
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.4472703398053017,
+      "learning_rate": 8.397374938476594e-07,
+      "loss": 0.0442,
+      "step": 8994
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.4943388990749984,
+      "learning_rate": 8.389202284284536e-07,
+      "loss": 0.0476,
+      "step": 8995
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.511444428661888,
+      "learning_rate": 8.38103324479747e-07,
+      "loss": 0.0504,
+      "step": 8996
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.363267847532087,
+      "learning_rate": 8.37286782072505e-07,
+      "loss": 0.0358,
+      "step": 8997
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.4133418350198572,
+      "learning_rate": 8.36470601277658e-07,
+      "loss": 0.0435,
+      "step": 8998
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.6288271493226334,
+      "learning_rate": 8.356547821661098e-07,
+      "loss": 0.0548,
+      "step": 8999
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.6517041938394745,
+      "learning_rate": 8.348393248087289e-07,
+      "loss": 0.0503,
+      "step": 9000
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.2574762738780376,
+      "learning_rate": 8.340242292763529e-07,
+      "loss": 0.0397,
+      "step": 9001
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.3575275716684383,
+      "learning_rate": 8.33209495639788e-07,
+      "loss": 0.039,
+      "step": 9002
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.4252422645237195,
+      "learning_rate": 8.323951239698119e-07,
+      "loss": 0.0489,
+      "step": 9003
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.5413535320784262,
+      "learning_rate": 8.315811143371666e-07,
+      "loss": 0.0487,
+      "step": 9004
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.541034198856851,
+      "learning_rate": 8.307674668125665e-07,
+      "loss": 0.0444,
+      "step": 9005
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.2813976530781592,
+      "learning_rate": 8.299541814666917e-07,
+      "loss": 0.0371,
+      "step": 9006
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.6958676468289544,
+      "learning_rate": 8.291412583701913e-07,
+      "loss": 0.0501,
+      "step": 9007
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.5297521809715184,
+      "learning_rate": 8.283286975936833e-07,
+      "loss": 0.0476,
+      "step": 9008
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.4493495446458038,
+      "learning_rate": 8.275164992077555e-07,
+      "loss": 0.0458,
+      "step": 9009
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.4896130776669245,
+      "learning_rate": 8.267046632829618e-07,
+      "loss": 0.0491,
+      "step": 9010
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.6503397423052593,
+      "learning_rate": 8.258931898898276e-07,
+      "loss": 0.0546,
+      "step": 9011
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.5549274654388119,
+      "learning_rate": 8.250820790988446e-07,
+      "loss": 0.0501,
+      "step": 9012
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.358183941349689,
+      "learning_rate": 8.242713309804729e-07,
+      "loss": 0.0404,
+      "step": 9013
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.3940337228855835,
+      "learning_rate": 8.234609456051402e-07,
+      "loss": 0.0439,
+      "step": 9014
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.5995225619911047,
+      "learning_rate": 8.226509230432472e-07,
+      "loss": 0.0459,
+      "step": 9015
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.4005265088750665,
+      "learning_rate": 8.218412633651579e-07,
+      "loss": 0.0383,
+      "step": 9016
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.231710447697946,
+      "learning_rate": 8.210319666412087e-07,
+      "loss": 0.0409,
+      "step": 9017
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.5731805999632344,
+      "learning_rate": 8.202230329417016e-07,
+      "loss": 0.0511,
+      "step": 9018
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.6128783761907122,
+      "learning_rate": 8.194144623369083e-07,
+      "loss": 0.0462,
+      "step": 9019
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.6316539814389979,
+      "learning_rate": 8.18606254897068e-07,
+      "loss": 0.0526,
+      "step": 9020
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.4191188583693426,
+      "learning_rate": 8.177984106923914e-07,
+      "loss": 0.0456,
+      "step": 9021
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.5408719925757557,
+      "learning_rate": 8.169909297930528e-07,
+      "loss": 0.0546,
+      "step": 9022
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.3613786770416876,
+      "learning_rate": 8.161838122692e-07,
+      "loss": 0.0379,
+      "step": 9023
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.4965932979154073,
+      "learning_rate": 8.15377058190946e-07,
+      "loss": 0.0443,
+      "step": 9024
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.5226486809817725,
+      "learning_rate": 8.145706676283727e-07,
+      "loss": 0.0459,
+      "step": 9025
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.5170960817361812,
+      "learning_rate": 8.137646406515293e-07,
+      "loss": 0.0515,
+      "step": 9026
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.215493483329827,
+      "learning_rate": 8.129589773304381e-07,
+      "loss": 0.0393,
+      "step": 9027
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.5544798075202735,
+      "learning_rate": 8.121536777350836e-07,
+      "loss": 0.0444,
+      "step": 9028
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.3790009615106615,
+      "learning_rate": 8.113487419354244e-07,
+      "loss": 0.0421,
+      "step": 9029
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.6083808501873749,
+      "learning_rate": 8.105441700013827e-07,
+      "loss": 0.046,
+      "step": 9030
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.4936871187777547,
+      "learning_rate": 8.097399620028523e-07,
+      "loss": 0.0445,
+      "step": 9031
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.591663116355617,
+      "learning_rate": 8.089361180096927e-07,
+      "loss": 0.0495,
+      "step": 9032
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.5017803024250382,
+      "learning_rate": 8.08132638091736e-07,
+      "loss": 0.0482,
+      "step": 9033
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.4127755181942563,
+      "learning_rate": 8.073295223187766e-07,
+      "loss": 0.0349,
+      "step": 9034
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.1853239417764114,
+      "learning_rate": 8.06526770760584e-07,
+      "loss": 0.0384,
+      "step": 9035
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.3324437910311249,
+      "learning_rate": 8.057243834868916e-07,
+      "loss": 0.0436,
+      "step": 9036
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.1707072770807363,
+      "learning_rate": 8.049223605674023e-07,
+      "loss": 0.0361,
+      "step": 9037
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.451944514938382,
+      "learning_rate": 8.041207020717851e-07,
+      "loss": 0.0526,
+      "step": 9038
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.287628025223005,
+      "learning_rate": 8.033194080696833e-07,
+      "loss": 0.0408,
+      "step": 9039
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.5210471715688836,
+      "learning_rate": 8.025184786307016e-07,
+      "loss": 0.0422,
+      "step": 9040
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.2261789493315818,
+      "learning_rate": 8.017179138244191e-07,
+      "loss": 0.0347,
+      "step": 9041
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.4218751493565884,
+      "learning_rate": 8.009177137203794e-07,
+      "loss": 0.0407,
+      "step": 9042
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.4753775062765253,
+      "learning_rate": 8.001178783880936e-07,
+      "loss": 0.0479,
+      "step": 9043
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.3520247096842597,
+      "learning_rate": 7.99318407897045e-07,
+      "loss": 0.0423,
+      "step": 9044
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.758429802604417,
+      "learning_rate": 7.985193023166821e-07,
+      "loss": 0.0454,
+      "step": 9045
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.1879696122580803,
+      "learning_rate": 7.977205617164241e-07,
+      "loss": 0.0358,
+      "step": 9046
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.5109030996365918,
+      "learning_rate": 7.969221861656557e-07,
+      "loss": 0.0445,
+      "step": 9047
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.4234089690038207,
+      "learning_rate": 7.961241757337324e-07,
+      "loss": 0.0469,
+      "step": 9048
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.6390994822687233,
+      "learning_rate": 7.953265304899743e-07,
+      "loss": 0.0455,
+      "step": 9049
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.3827650030811074,
+      "learning_rate": 7.945292505036762e-07,
+      "loss": 0.0401,
+      "step": 9050
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.5179390231928465,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.038,
+      "step": 9051
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.4298489588422658,
+      "learning_rate": 7.929357865804571e-07,
+      "loss": 0.0457,
+      "step": 9052
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.5887061713564796,
+      "learning_rate": 7.921396027819616e-07,
+      "loss": 0.0511,
+      "step": 9053
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.6842889479197347,
+      "learning_rate": 7.913437845177701e-07,
+      "loss": 0.0569,
+      "step": 9054
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.4826350381534015,
+      "learning_rate": 7.905483318570145e-07,
+      "loss": 0.0501,
+      "step": 9055
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.4449163465246155,
+      "learning_rate": 7.897532448687978e-07,
+      "loss": 0.0468,
+      "step": 9056
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.364153784443725,
+      "learning_rate": 7.889585236221853e-07,
+      "loss": 0.0422,
+      "step": 9057
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.1839617974893517,
+      "learning_rate": 7.881641681862173e-07,
+      "loss": 0.0372,
+      "step": 9058
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.4551291759469345,
+      "learning_rate": 7.873701786298976e-07,
+      "loss": 0.0468,
+      "step": 9059
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.7694456129801226,
+      "learning_rate": 7.865765550221993e-07,
+      "loss": 0.0532,
+      "step": 9060
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.595986510898096,
+      "learning_rate": 7.857832974320634e-07,
+      "loss": 0.0485,
+      "step": 9061
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.333406306805132,
+      "learning_rate": 7.849904059284014e-07,
+      "loss": 0.038,
+      "step": 9062
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.4310867823210012,
+      "learning_rate": 7.841978805800887e-07,
+      "loss": 0.0435,
+      "step": 9063
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.5610426955813441,
+      "learning_rate": 7.834057214559749e-07,
+      "loss": 0.0438,
+      "step": 9064
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.3302655459612682,
+      "learning_rate": 7.82613928624873e-07,
+      "loss": 0.0368,
+      "step": 9065
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.3312887321039264,
+      "learning_rate": 7.818225021555648e-07,
+      "loss": 0.0439,
+      "step": 9066
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.4530976531464472,
+      "learning_rate": 7.810314421168003e-07,
+      "loss": 0.047,
+      "step": 9067
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.4705509609400045,
+      "learning_rate": 7.802407485773011e-07,
+      "loss": 0.0456,
+      "step": 9068
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.290940710341443,
+      "learning_rate": 7.794504216057513e-07,
+      "loss": 0.0404,
+      "step": 9069
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.5746991313637482,
+      "learning_rate": 7.786604612708093e-07,
+      "loss": 0.0461,
+      "step": 9070
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.317041374842992,
+      "learning_rate": 7.778708676410962e-07,
+      "loss": 0.0435,
+      "step": 9071
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.5982556350044739,
+      "learning_rate": 7.770816407852045e-07,
+      "loss": 0.05,
+      "step": 9072
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.4054662972230303,
+      "learning_rate": 7.762927807716925e-07,
+      "loss": 0.0498,
+      "step": 9073
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.6816352853804117,
+      "learning_rate": 7.755042876690893e-07,
+      "loss": 0.0434,
+      "step": 9074
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.5184365092681416,
+      "learning_rate": 7.747161615458903e-07,
+      "loss": 0.049,
+      "step": 9075
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.5790024943273662,
+      "learning_rate": 7.739284024705601e-07,
+      "loss": 0.0481,
+      "step": 9076
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.5601657919042449,
+      "learning_rate": 7.731410105115311e-07,
+      "loss": 0.0438,
+      "step": 9077
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.4016034601639242,
+      "learning_rate": 7.723539857372026e-07,
+      "loss": 0.043,
+      "step": 9078
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.2513766785266978,
+      "learning_rate": 7.715673282159425e-07,
+      "loss": 0.042,
+      "step": 9079
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.3304815997111177,
+      "learning_rate": 7.707810380160891e-07,
+      "loss": 0.0428,
+      "step": 9080
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.4750189006759722,
+      "learning_rate": 7.699951152059448e-07,
+      "loss": 0.0465,
+      "step": 9081
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.2760917233304985,
+      "learning_rate": 7.692095598537847e-07,
+      "loss": 0.0418,
+      "step": 9082
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.5261198510865022,
+      "learning_rate": 7.684243720278478e-07,
+      "loss": 0.0432,
+      "step": 9083
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.7639380720526652,
+      "learning_rate": 7.676395517963436e-07,
+      "loss": 0.0427,
+      "step": 9084
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.433297945762989,
+      "learning_rate": 7.668550992274476e-07,
+      "loss": 0.0453,
+      "step": 9085
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.522719045783516,
+      "learning_rate": 7.660710143893069e-07,
+      "loss": 0.0463,
+      "step": 9086
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.6697006447180247,
+      "learning_rate": 7.652872973500325e-07,
+      "loss": 0.0496,
+      "step": 9087
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.6728096140500466,
+      "learning_rate": 7.645039481777073e-07,
+      "loss": 0.055,
+      "step": 9088
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.5890631151395276,
+      "learning_rate": 7.637209669403789e-07,
+      "loss": 0.0534,
+      "step": 9089
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.655969852936737,
+      "learning_rate": 7.629383537060653e-07,
+      "loss": 0.0412,
+      "step": 9090
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.7967731667136608,
+      "learning_rate": 7.621561085427503e-07,
+      "loss": 0.05,
+      "step": 9091
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.7604609885702558,
+      "learning_rate": 7.613742315183887e-07,
+      "loss": 0.0458,
+      "step": 9092
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.8146974337203476,
+      "learning_rate": 7.605927227009002e-07,
+      "loss": 0.0504,
+      "step": 9093
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.6356878186592496,
+      "learning_rate": 7.598115821581759e-07,
+      "loss": 0.0497,
+      "step": 9094
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.493286672819636,
+      "learning_rate": 7.590308099580718e-07,
+      "loss": 0.0519,
+      "step": 9095
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.6021212748673392,
+      "learning_rate": 7.582504061684131e-07,
+      "loss": 0.0583,
+      "step": 9096
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.6090582071614288,
+      "learning_rate": 7.57470370856992e-07,
+      "loss": 0.0551,
+      "step": 9097
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.3090115400577307,
+      "learning_rate": 7.566907040915721e-07,
+      "loss": 0.0458,
+      "step": 9098
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.3645653663675927,
+      "learning_rate": 7.559114059398804e-07,
+      "loss": 0.0478,
+      "step": 9099
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.6990392092582034,
+      "learning_rate": 7.551324764696155e-07,
+      "loss": 0.0519,
+      "step": 9100
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.477471738769153,
+      "learning_rate": 7.543539157484425e-07,
+      "loss": 0.0479,
+      "step": 9101
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.3260412189248787,
+      "learning_rate": 7.535757238439939e-07,
+      "loss": 0.0385,
+      "step": 9102
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.4482554309030768,
+      "learning_rate": 7.527979008238695e-07,
+      "loss": 0.049,
+      "step": 9103
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.507983976866288,
+      "learning_rate": 7.520204467556407e-07,
+      "loss": 0.0417,
+      "step": 9104
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.7837008751581924,
+      "learning_rate": 7.512433617068426e-07,
+      "loss": 0.0512,
+      "step": 9105
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.5918365709689872,
+      "learning_rate": 7.504666457449822e-07,
+      "loss": 0.0539,
+      "step": 9106
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.9112005434083585,
+      "learning_rate": 7.49690298937531e-07,
+      "loss": 0.0569,
+      "step": 9107
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.7623075250190148,
+      "learning_rate": 7.489143213519301e-07,
+      "loss": 0.0487,
+      "step": 9108
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.2710173598255867,
+      "learning_rate": 7.481387130555868e-07,
+      "loss": 0.0388,
+      "step": 9109
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.4368405095243892,
+      "learning_rate": 7.473634741158797e-07,
+      "loss": 0.0479,
+      "step": 9110
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.586905853660764,
+      "learning_rate": 7.465886046001519e-07,
+      "loss": 0.0482,
+      "step": 9111
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.5129254124410179,
+      "learning_rate": 7.458141045757172e-07,
+      "loss": 0.0442,
+      "step": 9112
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.5185101930640894,
+      "learning_rate": 7.450399741098557e-07,
+      "loss": 0.0433,
+      "step": 9113
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.3830462799794978,
+      "learning_rate": 7.442662132698148e-07,
+      "loss": 0.0435,
+      "step": 9114
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.3768448813877903,
+      "learning_rate": 7.434928221228105e-07,
+      "loss": 0.0417,
+      "step": 9115
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.6428633489229871,
+      "learning_rate": 7.427198007360282e-07,
+      "loss": 0.0553,
+      "step": 9116
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.2559999935875805,
+      "learning_rate": 7.419471491766173e-07,
+      "loss": 0.0392,
+      "step": 9117
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.7471362532604113,
+      "learning_rate": 7.411748675117008e-07,
+      "loss": 0.0592,
+      "step": 9118
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.5774018878719787,
+      "learning_rate": 7.404029558083653e-07,
+      "loss": 0.0592,
+      "step": 9119
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.4492076342546396,
+      "learning_rate": 7.396314141336652e-07,
+      "loss": 0.0456,
+      "step": 9120
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.4668121524540976,
+      "learning_rate": 7.388602425546237e-07,
+      "loss": 0.0508,
+      "step": 9121
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.7264432790167168,
+      "learning_rate": 7.380894411382339e-07,
+      "loss": 0.0565,
+      "step": 9122
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.3763246314609863,
+      "learning_rate": 7.373190099514521e-07,
+      "loss": 0.0513,
+      "step": 9123
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.5509895088181258,
+      "learning_rate": 7.365489490612083e-07,
+      "loss": 0.0462,
+      "step": 9124
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.4387997706783684,
+      "learning_rate": 7.357792585343959e-07,
+      "loss": 0.0483,
+      "step": 9125
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.116156077270865,
+      "learning_rate": 7.350099384378773e-07,
+      "loss": 0.0336,
+      "step": 9126
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.4659906791550648,
+      "learning_rate": 7.342409888384816e-07,
+      "loss": 0.0422,
+      "step": 9127
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.252971903235181,
+      "learning_rate": 7.334724098030094e-07,
+      "loss": 0.0415,
+      "step": 9128
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.0192814914859376,
+      "learning_rate": 7.32704201398225e-07,
+      "loss": 0.0664,
+      "step": 9129
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.4548849854891277,
+      "learning_rate": 7.319363636908633e-07,
+      "loss": 0.0382,
+      "step": 9130
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.3015757993164256,
+      "learning_rate": 7.311688967476255e-07,
+      "loss": 0.0427,
+      "step": 9131
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.3161088599106257,
+      "learning_rate": 7.30401800635181e-07,
+      "loss": 0.0394,
+      "step": 9132
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.3149310551921567,
+      "learning_rate": 7.296350754201653e-07,
+      "loss": 0.0433,
+      "step": 9133
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.5344309965861216,
+      "learning_rate": 7.288687211691864e-07,
+      "loss": 0.0463,
+      "step": 9134
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.4192949624323892,
+      "learning_rate": 7.281027379488143e-07,
+      "loss": 0.0415,
+      "step": 9135
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.628533774123234,
+      "learning_rate": 7.273371258255923e-07,
+      "loss": 0.0528,
+      "step": 9136
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.435250830705924,
+      "learning_rate": 7.26571884866027e-07,
+      "loss": 0.0342,
+      "step": 9137
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.6009934184279602,
+      "learning_rate": 7.258070151365931e-07,
+      "loss": 0.0489,
+      "step": 9138
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.5112383965783929,
+      "learning_rate": 7.250425167037367e-07,
+      "loss": 0.0478,
+      "step": 9139
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.6967305849472853,
+      "learning_rate": 7.242783896338678e-07,
+      "loss": 0.0481,
+      "step": 9140
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5222964590595602,
+      "learning_rate": 7.235146339933674e-07,
+      "loss": 0.0472,
+      "step": 9141
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5473918720573918,
+      "learning_rate": 7.227512498485812e-07,
+      "loss": 0.0449,
+      "step": 9142
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.3845105852340565,
+      "learning_rate": 7.219882372658237e-07,
+      "loss": 0.0425,
+      "step": 9143
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.2379720794844282,
+      "learning_rate": 7.212255963113773e-07,
+      "loss": 0.0405,
+      "step": 9144
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.7886078232425295,
+      "learning_rate": 7.204633270514932e-07,
+      "loss": 0.0593,
+      "step": 9145
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.7484915588835566,
+      "learning_rate": 7.197014295523879e-07,
+      "loss": 0.0558,
+      "step": 9146
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5321798563832638,
+      "learning_rate": 7.189399038802492e-07,
+      "loss": 0.0525,
+      "step": 9147
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.583366697512015,
+      "learning_rate": 7.181787501012283e-07,
+      "loss": 0.0478,
+      "step": 9148
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.6110618663636844,
+      "learning_rate": 7.17417968281447e-07,
+      "loss": 0.0454,
+      "step": 9149
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5776490410012918,
+      "learning_rate": 7.166575584869929e-07,
+      "loss": 0.0458,
+      "step": 9150
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.2747137327793674,
+      "learning_rate": 7.158975207839241e-07,
+      "loss": 0.0416,
+      "step": 9151
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.4365939432555175,
+      "learning_rate": 7.151378552382627e-07,
+      "loss": 0.0485,
+      "step": 9152
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.6321785953028134,
+      "learning_rate": 7.143785619160026e-07,
+      "loss": 0.0441,
+      "step": 9153
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.3846858880309934,
+      "learning_rate": 7.136196408831014e-07,
+      "loss": 0.0423,
+      "step": 9154
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.6919957307652642,
+      "learning_rate": 7.128610922054874e-07,
+      "loss": 0.0427,
+      "step": 9155
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.7780764373082132,
+      "learning_rate": 7.121029159490533e-07,
+      "loss": 0.0629,
+      "step": 9156
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.637754464819675,
+      "learning_rate": 7.113451121796632e-07,
+      "loss": 0.0458,
+      "step": 9157
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.6384577532866529,
+      "learning_rate": 7.105876809631462e-07,
+      "loss": 0.041,
+      "step": 9158
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.382256613876428,
+      "learning_rate": 7.098306223653013e-07,
+      "loss": 0.0366,
+      "step": 9159
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.4383202923067384,
+      "learning_rate": 7.090739364518923e-07,
+      "loss": 0.04,
+      "step": 9160
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.3692997221121066,
+      "learning_rate": 7.083176232886524e-07,
+      "loss": 0.0394,
+      "step": 9161
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5784348692505208,
+      "learning_rate": 7.075616829412806e-07,
+      "loss": 0.041,
+      "step": 9162
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.4999909944735648,
+      "learning_rate": 7.068061154754485e-07,
+      "loss": 0.0499,
+      "step": 9163
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5382113976134686,
+      "learning_rate": 7.060509209567878e-07,
+      "loss": 0.0365,
+      "step": 9164
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5916825069678298,
+      "learning_rate": 7.052960994509056e-07,
+      "loss": 0.0445,
+      "step": 9165
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.3037641116882095,
+      "learning_rate": 7.045416510233705e-07,
+      "loss": 0.0349,
+      "step": 9166
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.6036895352373282,
+      "learning_rate": 7.037875757397211e-07,
+      "loss": 0.0461,
+      "step": 9167
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.6689980251471799,
+      "learning_rate": 7.030338736654629e-07,
+      "loss": 0.0381,
+      "step": 9168
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.7461677607971402,
+      "learning_rate": 7.022805448660719e-07,
+      "loss": 0.0538,
+      "step": 9169
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5844878706328465,
+      "learning_rate": 7.015275894069862e-07,
+      "loss": 0.0526,
+      "step": 9170
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5224405119550164,
+      "learning_rate": 7.007750073536179e-07,
+      "loss": 0.043,
+      "step": 9171
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5387973710966365,
+      "learning_rate": 7.000227987713415e-07,
+      "loss": 0.0445,
+      "step": 9172
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5559841441470788,
+      "learning_rate": 6.992709637255007e-07,
+      "loss": 0.0391,
+      "step": 9173
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.6610742721382772,
+      "learning_rate": 6.985195022814068e-07,
+      "loss": 0.0472,
+      "step": 9174
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5973456885303319,
+      "learning_rate": 6.9776841450434e-07,
+      "loss": 0.0461,
+      "step": 9175
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.4824059611412461,
+      "learning_rate": 6.970177004595452e-07,
+      "loss": 0.0458,
+      "step": 9176
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.7082853378616765,
+      "learning_rate": 6.962673602122388e-07,
+      "loss": 0.0515,
+      "step": 9177
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.196760984309167,
+      "learning_rate": 6.955173938276011e-07,
+      "loss": 0.0372,
+      "step": 9178
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.5315850464093408,
+      "learning_rate": 6.947678013707809e-07,
+      "loss": 0.0453,
+      "step": 9179
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.5257737735715897,
+      "learning_rate": 6.940185829068946e-07,
+      "loss": 0.0516,
+      "step": 9180
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.2746727349338245,
+      "learning_rate": 6.932697385010273e-07,
+      "loss": 0.0348,
+      "step": 9181
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.4676712665266491,
+      "learning_rate": 6.925212682182298e-07,
+      "loss": 0.0389,
+      "step": 9182
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.382390721292978,
+      "learning_rate": 6.917731721235227e-07,
+      "loss": 0.0341,
+      "step": 9183
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.7156762319224332,
+      "learning_rate": 6.910254502818914e-07,
+      "loss": 0.0501,
+      "step": 9184
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.5315022220843613,
+      "learning_rate": 6.902781027582905e-07,
+      "loss": 0.0474,
+      "step": 9185
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.3430577575572498,
+      "learning_rate": 6.895311296176404e-07,
+      "loss": 0.0448,
+      "step": 9186
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.7919947656392654,
+      "learning_rate": 6.887845309248326e-07,
+      "loss": 0.046,
+      "step": 9187
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.5762793180089436,
+      "learning_rate": 6.880383067447211e-07,
+      "loss": 0.0524,
+      "step": 9188
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.7235385464822204,
+      "learning_rate": 6.872924571421318e-07,
+      "loss": 0.0476,
+      "step": 9189
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.7512384186112453,
+      "learning_rate": 6.865469821818566e-07,
+      "loss": 0.0474,
+      "step": 9190
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.367431366014104,
+      "learning_rate": 6.858018819286527e-07,
+      "loss": 0.0421,
+      "step": 9191
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.482470364816719,
+      "learning_rate": 6.850571564472463e-07,
+      "loss": 0.0461,
+      "step": 9192
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.3204672103503778,
+      "learning_rate": 6.843128058023335e-07,
+      "loss": 0.0392,
+      "step": 9193
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.4637631717892052,
+      "learning_rate": 6.835688300585735e-07,
+      "loss": 0.0403,
+      "step": 9194
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.8680475887151076,
+      "learning_rate": 6.828252292805965e-07,
+      "loss": 0.0481,
+      "step": 9195
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.572716331843182,
+      "learning_rate": 6.820820035329984e-07,
+      "loss": 0.0491,
+      "step": 9196
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.2657270178321114,
+      "learning_rate": 6.813391528803426e-07,
+      "loss": 0.0436,
+      "step": 9197
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.5149542145923707,
+      "learning_rate": 6.80596677387158e-07,
+      "loss": 0.05,
+      "step": 9198
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.3473478657768,
+      "learning_rate": 6.798545771179466e-07,
+      "loss": 0.039,
+      "step": 9199
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.391550161682129,
+      "learning_rate": 6.791128521371715e-07,
+      "loss": 0.041,
+      "step": 9200
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.325188866237484,
+      "learning_rate": 6.783715025092674e-07,
+      "loss": 0.0434,
+      "step": 9201
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.238757577990556,
+      "learning_rate": 6.776305282986346e-07,
+      "loss": 0.0388,
+      "step": 9202
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.6936175802651203,
+      "learning_rate": 6.768899295696413e-07,
+      "loss": 0.0486,
+      "step": 9203
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.5645034524621013,
+      "learning_rate": 6.761497063866207e-07,
+      "loss": 0.0512,
+      "step": 9204
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.5128111396100308,
+      "learning_rate": 6.754098588138791e-07,
+      "loss": 0.0398,
+      "step": 9205
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.55964305812144,
+      "learning_rate": 6.746703869156829e-07,
+      "loss": 0.0531,
+      "step": 9206
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.865922968543829,
+      "learning_rate": 6.739312907562734e-07,
+      "loss": 0.0549,
+      "step": 9207
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.730120597173305,
+      "learning_rate": 6.731925703998526e-07,
+      "loss": 0.0452,
+      "step": 9208
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.526950796569622,
+      "learning_rate": 6.724542259105943e-07,
+      "loss": 0.0418,
+      "step": 9209
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.6833136013791101,
+      "learning_rate": 6.717162573526359e-07,
+      "loss": 0.0512,
+      "step": 9210
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.676693408619882,
+      "learning_rate": 6.709786647900874e-07,
+      "loss": 0.0501,
+      "step": 9211
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.3358527261887407,
+      "learning_rate": 6.702414482870195e-07,
+      "loss": 0.0401,
+      "step": 9212
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.445444937305827,
+      "learning_rate": 6.695046079074774e-07,
+      "loss": 0.0385,
+      "step": 9213
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.383730121091557,
+      "learning_rate": 6.687681437154681e-07,
+      "loss": 0.0418,
+      "step": 9214
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.4009520205237163,
+      "learning_rate": 6.680320557749675e-07,
+      "loss": 0.0444,
+      "step": 9215
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.4730571523622686,
+      "learning_rate": 6.672963441499186e-07,
+      "loss": 0.0496,
+      "step": 9216
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.4193557421202503,
+      "learning_rate": 6.66561008904234e-07,
+      "loss": 0.0416,
+      "step": 9217
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.4044115948524016,
+      "learning_rate": 6.658260501017905e-07,
+      "loss": 0.0412,
+      "step": 9218
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.7636711477410567,
+      "learning_rate": 6.650914678064346e-07,
+      "loss": 0.0557,
+      "step": 9219
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.3283509774025268,
+      "learning_rate": 6.643572620819783e-07,
+      "loss": 0.0301,
+      "step": 9220
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.4811753756869708,
+      "learning_rate": 6.63623432992202e-07,
+      "loss": 0.0462,
+      "step": 9221
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.2914863031586483,
+      "learning_rate": 6.628899806008515e-07,
+      "loss": 0.0432,
+      "step": 9222
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.6641015925117764,
+      "learning_rate": 6.621569049716442e-07,
+      "loss": 0.0557,
+      "step": 9223
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.4158388268219357,
+      "learning_rate": 6.614242061682585e-07,
+      "loss": 0.042,
+      "step": 9224
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.4025117825778293,
+      "learning_rate": 6.606918842543481e-07,
+      "loss": 0.0355,
+      "step": 9225
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.5946008237703155,
+      "learning_rate": 6.599599392935241e-07,
+      "loss": 0.0461,
+      "step": 9226
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.6414571574034247,
+      "learning_rate": 6.592283713493741e-07,
+      "loss": 0.0549,
+      "step": 9227
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.5732034354358095,
+      "learning_rate": 6.584971804854457e-07,
+      "loss": 0.0515,
+      "step": 9228
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.8049634188996078,
+      "learning_rate": 6.577663667652595e-07,
+      "loss": 0.0586,
+      "step": 9229
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.7082091171672416,
+      "learning_rate": 6.570359302523011e-07,
+      "loss": 0.0478,
+      "step": 9230
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.5798190716279514,
+      "learning_rate": 6.563058710100218e-07,
+      "loss": 0.0521,
+      "step": 9231
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.6792697881846939,
+      "learning_rate": 6.55576189101842e-07,
+      "loss": 0.0522,
+      "step": 9232
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.38889217632605,
+      "learning_rate": 6.548468845911471e-07,
+      "loss": 0.0417,
+      "step": 9233
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.4232526639331131,
+      "learning_rate": 6.541179575412942e-07,
+      "loss": 0.0457,
+      "step": 9234
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.6222448425794103,
+      "learning_rate": 6.533894080156017e-07,
+      "loss": 0.0463,
+      "step": 9235
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.5032123485024722,
+      "learning_rate": 6.526612360773615e-07,
+      "loss": 0.0408,
+      "step": 9236
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.4639932929462964,
+      "learning_rate": 6.519334417898277e-07,
+      "loss": 0.0421,
+      "step": 9237
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.4222474806463148,
+      "learning_rate": 6.512060252162228e-07,
+      "loss": 0.0444,
+      "step": 9238
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.5655759177579105,
+      "learning_rate": 6.504789864197375e-07,
+      "loss": 0.0552,
+      "step": 9239
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.3414952635130217,
+      "learning_rate": 6.497523254635296e-07,
+      "loss": 0.0424,
+      "step": 9240
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.3160366227969231,
+      "learning_rate": 6.490260424107231e-07,
+      "loss": 0.0371,
+      "step": 9241
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.4885130425785391,
+      "learning_rate": 6.483001373244107e-07,
+      "loss": 0.0459,
+      "step": 9242
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.8104844383498992,
+      "learning_rate": 6.475746102676517e-07,
+      "loss": 0.0512,
+      "step": 9243
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.7218507461531034,
+      "learning_rate": 6.468494613034704e-07,
+      "loss": 0.0526,
+      "step": 9244
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.6974153564100232,
+      "learning_rate": 6.461246904948604e-07,
+      "loss": 0.0516,
+      "step": 9245
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.362255119743437,
+      "learning_rate": 6.454002979047836e-07,
+      "loss": 0.0382,
+      "step": 9246
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.3405947808462468,
+      "learning_rate": 6.446762835961656e-07,
+      "loss": 0.0438,
+      "step": 9247
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.6080820401130123,
+      "learning_rate": 6.439526476319031e-07,
+      "loss": 0.0419,
+      "step": 9248
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.6322231813534775,
+      "learning_rate": 6.432293900748571e-07,
+      "loss": 0.0423,
+      "step": 9249
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.7618590367801918,
+      "learning_rate": 6.425065109878559e-07,
+      "loss": 0.0541,
+      "step": 9250
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.5443724106379078,
+      "learning_rate": 6.417840104336953e-07,
+      "loss": 0.0424,
+      "step": 9251
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.361147940565434,
+      "learning_rate": 6.410618884751407e-07,
+      "loss": 0.036,
+      "step": 9252
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.6279273750122605,
+      "learning_rate": 6.403401451749197e-07,
+      "loss": 0.0541,
+      "step": 9253
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.4105709248009561,
+      "learning_rate": 6.396187805957315e-07,
+      "loss": 0.0406,
+      "step": 9254
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.4051043991359757,
+      "learning_rate": 6.388977948002406e-07,
+      "loss": 0.0413,
+      "step": 9255
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.4966326023433527,
+      "learning_rate": 6.381771878510779e-07,
+      "loss": 0.0502,
+      "step": 9256
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.1587844564237084,
+      "learning_rate": 6.37456959810841e-07,
+      "loss": 0.031,
+      "step": 9257
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.73880482175389,
+      "learning_rate": 6.367371107420983e-07,
+      "loss": 0.0542,
+      "step": 9258
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.355296823024841,
+      "learning_rate": 6.360176407073798e-07,
+      "loss": 0.0446,
+      "step": 9259
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.3758192502550173,
+      "learning_rate": 6.352985497691883e-07,
+      "loss": 0.042,
+      "step": 9260
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.470427061121933,
+      "learning_rate": 6.345798379899898e-07,
+      "loss": 0.0453,
+      "step": 9261
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.4650421763447183,
+      "learning_rate": 6.338615054322173e-07,
+      "loss": 0.0513,
+      "step": 9262
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.3292632706227707,
+      "learning_rate": 6.331435521582718e-07,
+      "loss": 0.0362,
+      "step": 9263
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.5117734241372645,
+      "learning_rate": 6.324259782305237e-07,
+      "loss": 0.041,
+      "step": 9264
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.3517403952211193,
+      "learning_rate": 6.31708783711305e-07,
+      "loss": 0.0404,
+      "step": 9265
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.4673009304547475,
+      "learning_rate": 6.309919686629212e-07,
+      "loss": 0.0468,
+      "step": 9266
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.3759096290853892,
+      "learning_rate": 6.302755331476401e-07,
+      "loss": 0.0435,
+      "step": 9267
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.4391997785213244,
+      "learning_rate": 6.295594772276981e-07,
+      "loss": 0.0444,
+      "step": 9268
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.436912540357196,
+      "learning_rate": 6.288438009652969e-07,
+      "loss": 0.0493,
+      "step": 9269
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.6853775413334768,
+      "learning_rate": 6.281285044226104e-07,
+      "loss": 0.054,
+      "step": 9270
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.3539939606525964,
+      "learning_rate": 6.274135876617726e-07,
+      "loss": 0.0408,
+      "step": 9271
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.1652654143620704,
+      "learning_rate": 6.2669905074489e-07,
+      "loss": 0.0334,
+      "step": 9272
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.6040188450962098,
+      "learning_rate": 6.25984893734034e-07,
+      "loss": 0.0532,
+      "step": 9273
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.6332789250665911,
+      "learning_rate": 6.252711166912418e-07,
+      "loss": 0.0524,
+      "step": 9274
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.2154998608562746,
+      "learning_rate": 6.245577196785186e-07,
+      "loss": 0.0335,
+      "step": 9275
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.6350521779976859,
+      "learning_rate": 6.238447027578387e-07,
+      "loss": 0.0418,
+      "step": 9276
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.615776139702605,
+      "learning_rate": 6.231320659911388e-07,
+      "loss": 0.0486,
+      "step": 9277
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.312790196202859,
+      "learning_rate": 6.224198094403278e-07,
+      "loss": 0.046,
+      "step": 9278
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.347429532863126,
+      "learning_rate": 6.217079331672777e-07,
+      "loss": 0.0373,
+      "step": 9279
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.425931296115451,
+      "learning_rate": 6.20996437233829e-07,
+      "loss": 0.043,
+      "step": 9280
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.841467047033049,
+      "learning_rate": 6.202853217017879e-07,
+      "loss": 0.0569,
+      "step": 9281
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.3220236107840497,
+      "learning_rate": 6.195745866329305e-07,
+      "loss": 0.0416,
+      "step": 9282
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.707855717149118,
+      "learning_rate": 6.188642320889959e-07,
+      "loss": 0.0548,
+      "step": 9283
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.3117725998502034,
+      "learning_rate": 6.181542581316941e-07,
+      "loss": 0.0409,
+      "step": 9284
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.1135003004366497,
+      "learning_rate": 6.174446648226995e-07,
+      "loss": 0.0359,
+      "step": 9285
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.2542941203560147,
+      "learning_rate": 6.167354522236535e-07,
+      "loss": 0.0377,
+      "step": 9286
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.525791939327361,
+      "learning_rate": 6.160266203961645e-07,
+      "loss": 0.0381,
+      "step": 9287
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.4542004719795047,
+      "learning_rate": 6.153181694018101e-07,
+      "loss": 0.0391,
+      "step": 9288
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.4369041304255807,
+      "learning_rate": 6.146100993021308e-07,
+      "loss": 0.0457,
+      "step": 9289
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.5730652059783,
+      "learning_rate": 6.139024101586383e-07,
+      "loss": 0.051,
+      "step": 9290
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.5706678971258587,
+      "learning_rate": 6.131951020328081e-07,
+      "loss": 0.0477,
+      "step": 9291
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.4507090190316347,
+      "learning_rate": 6.124881749860839e-07,
+      "loss": 0.0513,
+      "step": 9292
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.5157359426230064,
+      "learning_rate": 6.117816290798751e-07,
+      "loss": 0.0461,
+      "step": 9293
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.4491017462610927,
+      "learning_rate": 6.110754643755606e-07,
+      "loss": 0.0537,
+      "step": 9294
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.511588796433832,
+      "learning_rate": 6.103696809344823e-07,
+      "loss": 0.049,
+      "step": 9295
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.6418665109222186,
+      "learning_rate": 6.096642788179535e-07,
+      "loss": 0.0449,
+      "step": 9296
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.5615796065450795,
+      "learning_rate": 6.089592580872511e-07,
+      "loss": 0.0483,
+      "step": 9297
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.646715306849953,
+      "learning_rate": 6.082546188036204e-07,
+      "loss": 0.043,
+      "step": 9298
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.409845531805855,
+      "learning_rate": 6.075503610282707e-07,
+      "loss": 0.0455,
+      "step": 9299
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.5084798445332748,
+      "learning_rate": 6.068464848223831e-07,
+      "loss": 0.0482,
+      "step": 9300
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.5154741647755088,
+      "learning_rate": 6.061429902471011e-07,
+      "loss": 0.0366,
+      "step": 9301
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.7248905250114255,
+      "learning_rate": 6.054398773635395e-07,
+      "loss": 0.0518,
+      "step": 9302
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.3060936472282296,
+      "learning_rate": 6.047371462327733e-07,
+      "loss": 0.0389,
+      "step": 9303
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.3389706740125993,
+      "learning_rate": 6.040347969158517e-07,
+      "loss": 0.0397,
+      "step": 9304
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.4878187323089656,
+      "learning_rate": 6.03332829473785e-07,
+      "loss": 0.0489,
+      "step": 9305
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.5757453983418181,
+      "learning_rate": 6.026312439675553e-07,
+      "loss": 0.0507,
+      "step": 9306
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.4519569305653428,
+      "learning_rate": 6.019300404581057e-07,
+      "loss": 0.0466,
+      "step": 9307
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.4549603392332116,
+      "learning_rate": 6.012292190063535e-07,
+      "loss": 0.0432,
+      "step": 9308
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.3277910538830735,
+      "learning_rate": 6.005287796731746e-07,
+      "loss": 0.0408,
+      "step": 9309
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.5773560520113425,
+      "learning_rate": 5.998287225194177e-07,
+      "loss": 0.0548,
+      "step": 9310
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.3295425083017247,
+      "learning_rate": 5.991290476058953e-07,
+      "loss": 0.0442,
+      "step": 9311
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.331608052580573,
+      "learning_rate": 5.984297549933893e-07,
+      "loss": 0.0427,
+      "step": 9312
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.497564092423009,
+      "learning_rate": 5.97730844742645e-07,
+      "loss": 0.0468,
+      "step": 9313
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.6906800811012426,
+      "learning_rate": 5.970323169143793e-07,
+      "loss": 0.0522,
+      "step": 9314
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.4527454842122562,
+      "learning_rate": 5.963341715692689e-07,
+      "loss": 0.0469,
+      "step": 9315
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.4022621882997055,
+      "learning_rate": 5.956364087679644e-07,
+      "loss": 0.0436,
+      "step": 9316
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.3522228791772575,
+      "learning_rate": 5.949390285710777e-07,
+      "loss": 0.0459,
+      "step": 9317
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.5420115392691391,
+      "learning_rate": 5.942420310391916e-07,
+      "loss": 0.0456,
+      "step": 9318
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.2810661070643923,
+      "learning_rate": 5.93545416232853e-07,
+      "loss": 0.0406,
+      "step": 9319
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.6907657552588,
+      "learning_rate": 5.928491842125783e-07,
+      "loss": 0.0418,
+      "step": 9320
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.4718772456581073,
+      "learning_rate": 5.921533350388448e-07,
+      "loss": 0.042,
+      "step": 9321
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.2634409351652187,
+      "learning_rate": 5.914578687721034e-07,
+      "loss": 0.0333,
+      "step": 9322
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.355904576735825,
+      "learning_rate": 5.907627854727688e-07,
+      "loss": 0.0447,
+      "step": 9323
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.2270687372438003,
+      "learning_rate": 5.900680852012209e-07,
+      "loss": 0.0379,
+      "step": 9324
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.7310683579335753,
+      "learning_rate": 5.893737680178102e-07,
+      "loss": 0.0478,
+      "step": 9325
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.3703101561172066,
+      "learning_rate": 5.886798339828498e-07,
+      "loss": 0.0417,
+      "step": 9326
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.5655299570881718,
+      "learning_rate": 5.879862831566225e-07,
+      "loss": 0.0449,
+      "step": 9327
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.6846351679118978,
+      "learning_rate": 5.872931155993744e-07,
+      "loss": 0.0409,
+      "step": 9328
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.475778780504766,
+      "learning_rate": 5.866003313713231e-07,
+      "loss": 0.0452,
+      "step": 9329
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.3535298060701344,
+      "learning_rate": 5.859079305326487e-07,
+      "loss": 0.04,
+      "step": 9330
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.4156496614101144,
+      "learning_rate": 5.852159131435015e-07,
+      "loss": 0.045,
+      "step": 9331
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.4696627664881157,
+      "learning_rate": 5.845242792639955e-07,
+      "loss": 0.0459,
+      "step": 9332
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.5132987026955178,
+      "learning_rate": 5.838330289542121e-07,
+      "loss": 0.0421,
+      "step": 9333
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.5638255493013575,
+      "learning_rate": 5.831421622741995e-07,
+      "loss": 0.0485,
+      "step": 9334
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.4464948469440047,
+      "learning_rate": 5.82451679283974e-07,
+      "loss": 0.0426,
+      "step": 9335
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.6653777623951753,
+      "learning_rate": 5.817615800435167e-07,
+      "loss": 0.0482,
+      "step": 9336
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.5987814745554494,
+      "learning_rate": 5.810718646127772e-07,
+      "loss": 0.0471,
+      "step": 9337
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.6159677273389355,
+      "learning_rate": 5.803825330516699e-07,
+      "loss": 0.0415,
+      "step": 9338
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.7725605347204223,
+      "learning_rate": 5.796935854200764e-07,
+      "loss": 0.0486,
+      "step": 9339
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.2116651530771918,
+      "learning_rate": 5.790050217778442e-07,
+      "loss": 0.0365,
+      "step": 9340
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.3366514082309722,
+      "learning_rate": 5.783168421847912e-07,
+      "loss": 0.0392,
+      "step": 9341
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.5357472584251142,
+      "learning_rate": 5.776290467006961e-07,
+      "loss": 0.0441,
+      "step": 9342
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.4931475573569182,
+      "learning_rate": 5.769416353853097e-07,
+      "loss": 0.0431,
+      "step": 9343
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.5973699024862298,
+      "learning_rate": 5.762546082983462e-07,
+      "loss": 0.0427,
+      "step": 9344
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.4727593058720514,
+      "learning_rate": 5.755679654994866e-07,
+      "loss": 0.0371,
+      "step": 9345
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.3102635345845686,
+      "learning_rate": 5.748817070483792e-07,
+      "loss": 0.0352,
+      "step": 9346
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.5098352210174155,
+      "learning_rate": 5.741958330046399e-07,
+      "loss": 0.0448,
+      "step": 9347
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.7112577625926164,
+      "learning_rate": 5.735103434278482e-07,
+      "loss": 0.058,
+      "step": 9348
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.3312174687743756,
+      "learning_rate": 5.728252383775551e-07,
+      "loss": 0.037,
+      "step": 9349
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.6262182635753129,
+      "learning_rate": 5.721405179132733e-07,
+      "loss": 0.0478,
+      "step": 9350
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.5288705711280295,
+      "learning_rate": 5.714561820944848e-07,
+      "loss": 0.0498,
+      "step": 9351
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.3130130647280651,
+      "learning_rate": 5.707722309806352e-07,
+      "loss": 0.0398,
+      "step": 9352
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.5272330078855576,
+      "learning_rate": 5.700886646311427e-07,
+      "loss": 0.0489,
+      "step": 9353
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.8284411040755089,
+      "learning_rate": 5.694054831053847e-07,
+      "loss": 0.0469,
+      "step": 9354
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.2921777293746586,
+      "learning_rate": 5.687226864627115e-07,
+      "loss": 0.0401,
+      "step": 9355
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.5249134050999804,
+      "learning_rate": 5.680402747624364e-07,
+      "loss": 0.0462,
+      "step": 9356
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.5415871212250776,
+      "learning_rate": 5.673582480638395e-07,
+      "loss": 0.0486,
+      "step": 9357
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.4828366130767072,
+      "learning_rate": 5.666766064261681e-07,
+      "loss": 0.0495,
+      "step": 9358
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.7910842337444195,
+      "learning_rate": 5.659953499086368e-07,
+      "loss": 0.0488,
+      "step": 9359
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.3899166506067544,
+      "learning_rate": 5.653144785704245e-07,
+      "loss": 0.0422,
+      "step": 9360
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.8409343123085,
+      "learning_rate": 5.6463399247068e-07,
+      "loss": 0.055,
+      "step": 9361
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.7177966708246812,
+      "learning_rate": 5.639538916685161e-07,
+      "loss": 0.0448,
+      "step": 9362
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5371485866944967,
+      "learning_rate": 5.63274176223012e-07,
+      "loss": 0.0453,
+      "step": 9363
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.500343283868266,
+      "learning_rate": 5.625948461932135e-07,
+      "loss": 0.0503,
+      "step": 9364
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.3823709689841568,
+      "learning_rate": 5.619159016381359e-07,
+      "loss": 0.0418,
+      "step": 9365
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.3323515608385956,
+      "learning_rate": 5.612373426167566e-07,
+      "loss": 0.0393,
+      "step": 9366
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.358664271691672,
+      "learning_rate": 5.60559169188023e-07,
+      "loss": 0.0488,
+      "step": 9367
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.3337144003115673,
+      "learning_rate": 5.598813814108478e-07,
+      "loss": 0.0423,
+      "step": 9368
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.307399947489894,
+      "learning_rate": 5.59203979344109e-07,
+      "loss": 0.0391,
+      "step": 9369
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.3910824184274528,
+      "learning_rate": 5.585269630466511e-07,
+      "loss": 0.045,
+      "step": 9370
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.6143070674202626,
+      "learning_rate": 5.578503325772889e-07,
+      "loss": 0.048,
+      "step": 9371
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5642765953222224,
+      "learning_rate": 5.571740879947979e-07,
+      "loss": 0.0484,
+      "step": 9372
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5730907122242064,
+      "learning_rate": 5.564982293579258e-07,
+      "loss": 0.0479,
+      "step": 9373
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.2930179538546491,
+      "learning_rate": 5.558227567253832e-07,
+      "loss": 0.0394,
+      "step": 9374
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5825501312858514,
+      "learning_rate": 5.551476701558473e-07,
+      "loss": 0.0514,
+      "step": 9375
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.4867669702295414,
+      "learning_rate": 5.544729697079615e-07,
+      "loss": 0.0407,
+      "step": 9376
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.4824601417061116,
+      "learning_rate": 5.537986554403391e-07,
+      "loss": 0.0503,
+      "step": 9377
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5718637874099572,
+      "learning_rate": 5.531247274115553e-07,
+      "loss": 0.0458,
+      "step": 9378
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.7632920201416338,
+      "learning_rate": 5.524511856801567e-07,
+      "loss": 0.0521,
+      "step": 9379
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.53203699301683,
+      "learning_rate": 5.517780303046494e-07,
+      "loss": 0.0427,
+      "step": 9380
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.4663945088993786,
+      "learning_rate": 5.511052613435131e-07,
+      "loss": 0.0435,
+      "step": 9381
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.4457285922026681,
+      "learning_rate": 5.504328788551888e-07,
+      "loss": 0.0453,
+      "step": 9382
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.45961542746993,
+      "learning_rate": 5.497608828980877e-07,
+      "loss": 0.043,
+      "step": 9383
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5230644950006442,
+      "learning_rate": 5.490892735305842e-07,
+      "loss": 0.0459,
+      "step": 9384
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.523004316136607,
+      "learning_rate": 5.484180508110232e-07,
+      "loss": 0.0487,
+      "step": 9385
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.8684622774376212,
+      "learning_rate": 5.477472147977097e-07,
+      "loss": 0.0538,
+      "step": 9386
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.4833514447720368,
+      "learning_rate": 5.470767655489217e-07,
+      "loss": 0.0415,
+      "step": 9387
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.593617356309045,
+      "learning_rate": 5.464067031228987e-07,
+      "loss": 0.0516,
+      "step": 9388
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.3943362612278334,
+      "learning_rate": 5.457370275778506e-07,
+      "loss": 0.0418,
+      "step": 9389
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5165033016974703,
+      "learning_rate": 5.450677389719494e-07,
+      "loss": 0.0425,
+      "step": 9390
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.4436019870015415,
+      "learning_rate": 5.443988373633397e-07,
+      "loss": 0.0406,
+      "step": 9391
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.6795210602622075,
+      "learning_rate": 5.437303228101238e-07,
+      "loss": 0.0514,
+      "step": 9392
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.4252441487840701,
+      "learning_rate": 5.430621953703785e-07,
+      "loss": 0.0435,
+      "step": 9393
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.6414920293094153,
+      "learning_rate": 5.423944551021409e-07,
+      "loss": 0.0559,
+      "step": 9394
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5078702152771486,
+      "learning_rate": 5.417271020634207e-07,
+      "loss": 0.0412,
+      "step": 9395
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.6842539671184469,
+      "learning_rate": 5.41060136312187e-07,
+      "loss": 0.0519,
+      "step": 9396
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4328445366183313,
+      "learning_rate": 5.403935579063824e-07,
+      "loss": 0.0423,
+      "step": 9397
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4699558178617385,
+      "learning_rate": 5.397273669039083e-07,
+      "loss": 0.0448,
+      "step": 9398
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.5151817262613314,
+      "learning_rate": 5.390615633626384e-07,
+      "loss": 0.0371,
+      "step": 9399
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.5243240678260968,
+      "learning_rate": 5.383961473404098e-07,
+      "loss": 0.0421,
+      "step": 9400
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.5187390755136203,
+      "learning_rate": 5.377311188950279e-07,
+      "loss": 0.0436,
+      "step": 9401
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.3205795879981002,
+      "learning_rate": 5.370664780842622e-07,
+      "loss": 0.0449,
+      "step": 9402
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4413842704868112,
+      "learning_rate": 5.364022249658519e-07,
+      "loss": 0.0404,
+      "step": 9403
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.3966146927325749,
+      "learning_rate": 5.357383595974969e-07,
+      "loss": 0.0424,
+      "step": 9404
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.7454430441998872,
+      "learning_rate": 5.350748820368689e-07,
+      "loss": 0.0437,
+      "step": 9405
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.6705080782759565,
+      "learning_rate": 5.344117923416026e-07,
+      "loss": 0.055,
+      "step": 9406
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.327100049820239,
+      "learning_rate": 5.337490905693016e-07,
+      "loss": 0.0516,
+      "step": 9407
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.3836378415133572,
+      "learning_rate": 5.330867767775333e-07,
+      "loss": 0.0388,
+      "step": 9408
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.3747909813151689,
+      "learning_rate": 5.324248510238345e-07,
+      "loss": 0.0413,
+      "step": 9409
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.3015105217796543,
+      "learning_rate": 5.317633133657029e-07,
+      "loss": 0.0426,
+      "step": 9410
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.5816899056895841,
+      "learning_rate": 5.311021638606084e-07,
+      "loss": 0.0467,
+      "step": 9411
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4776297669941274,
+      "learning_rate": 5.304414025659832e-07,
+      "loss": 0.0458,
+      "step": 9412
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.6886019805220323,
+      "learning_rate": 5.297810295392291e-07,
+      "loss": 0.0512,
+      "step": 9413
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4024395631991442,
+      "learning_rate": 5.291210448377099e-07,
+      "loss": 0.038,
+      "step": 9414
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.496374015976713,
+      "learning_rate": 5.284614485187606e-07,
+      "loss": 0.0479,
+      "step": 9415
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4703070558785776,
+      "learning_rate": 5.278022406396788e-07,
+      "loss": 0.0496,
+      "step": 9416
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.610789699533333,
+      "learning_rate": 5.271434212577281e-07,
+      "loss": 0.0492,
+      "step": 9417
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4449983962305737,
+      "learning_rate": 5.264849904301422e-07,
+      "loss": 0.0425,
+      "step": 9418
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.328663111603859,
+      "learning_rate": 5.258269482141165e-07,
+      "loss": 0.0432,
+      "step": 9419
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.409689085100629,
+      "learning_rate": 5.251692946668169e-07,
+      "loss": 0.0476,
+      "step": 9420
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.5474513427162788,
+      "learning_rate": 5.245120298453715e-07,
+      "loss": 0.0481,
+      "step": 9421
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.766975422799182,
+      "learning_rate": 5.238551538068776e-07,
+      "loss": 0.054,
+      "step": 9422
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4026953004612066,
+      "learning_rate": 5.23198666608396e-07,
+      "loss": 0.0454,
+      "step": 9423
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.295197058555472,
+      "learning_rate": 5.22542568306958e-07,
+      "loss": 0.0403,
+      "step": 9424
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.5130454643863047,
+      "learning_rate": 5.218868589595555e-07,
+      "loss": 0.0405,
+      "step": 9425
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.5970329732191668,
+      "learning_rate": 5.21231538623152e-07,
+      "loss": 0.0548,
+      "step": 9426
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.5037455687960726,
+      "learning_rate": 5.205766073546742e-07,
+      "loss": 0.0468,
+      "step": 9427
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.5931211400825729,
+      "learning_rate": 5.199220652110148e-07,
+      "loss": 0.0417,
+      "step": 9428
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.5350254098100875,
+      "learning_rate": 5.19267912249033e-07,
+      "loss": 0.0489,
+      "step": 9429
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.544681422183019,
+      "learning_rate": 5.186141485255569e-07,
+      "loss": 0.0436,
+      "step": 9430
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.5045752686857166,
+      "learning_rate": 5.179607740973764e-07,
+      "loss": 0.0534,
+      "step": 9431
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.4016989356848832,
+      "learning_rate": 5.173077890212508e-07,
+      "loss": 0.0449,
+      "step": 9432
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.3679013200555543,
+      "learning_rate": 5.166551933539049e-07,
+      "loss": 0.0393,
+      "step": 9433
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.412741537200752,
+      "learning_rate": 5.160029871520284e-07,
+      "loss": 0.0473,
+      "step": 9434
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.285126378454,
+      "learning_rate": 5.153511704722775e-07,
+      "loss": 0.0432,
+      "step": 9435
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.5473949157744449,
+      "learning_rate": 5.146997433712769e-07,
+      "loss": 0.0438,
+      "step": 9436
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.3398707130295597,
+      "learning_rate": 5.140487059056143e-07,
+      "loss": 0.0431,
+      "step": 9437
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.2907777540719125,
+      "learning_rate": 5.133980581318459e-07,
+      "loss": 0.0394,
+      "step": 9438
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.5461921453787888,
+      "learning_rate": 5.127478001064928e-07,
+      "loss": 0.0476,
+      "step": 9439
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.2466253673084657,
+      "learning_rate": 5.120979318860419e-07,
+      "loss": 0.0349,
+      "step": 9440
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.422079879613309,
+      "learning_rate": 5.114484535269465e-07,
+      "loss": 0.0421,
+      "step": 9441
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.4591550594869038,
+      "learning_rate": 5.107993650856285e-07,
+      "loss": 0.0406,
+      "step": 9442
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.4511048983545245,
+      "learning_rate": 5.101506666184708e-07,
+      "loss": 0.038,
+      "step": 9443
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.3487680519418748,
+      "learning_rate": 5.095023581818287e-07,
+      "loss": 0.0414,
+      "step": 9444
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.4867755700430791,
+      "learning_rate": 5.088544398320189e-07,
+      "loss": 0.0423,
+      "step": 9445
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.439062274577456,
+      "learning_rate": 5.082069116253252e-07,
+      "loss": 0.0412,
+      "step": 9446
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.247696186418509,
+      "learning_rate": 5.075597736179977e-07,
+      "loss": 0.0378,
+      "step": 9447
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.3988958666516418,
+      "learning_rate": 5.069130258662541e-07,
+      "loss": 0.0439,
+      "step": 9448
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.3810304775337894,
+      "learning_rate": 5.062666684262757e-07,
+      "loss": 0.0436,
+      "step": 9449
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.2471137972676791,
+      "learning_rate": 5.056207013542131e-07,
+      "loss": 0.0362,
+      "step": 9450
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.7717503129411514,
+      "learning_rate": 5.049751247061796e-07,
+      "loss": 0.0561,
+      "step": 9451
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.3781118365836218,
+      "learning_rate": 5.04329938538256e-07,
+      "loss": 0.0339,
+      "step": 9452
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.4630232072172364,
+      "learning_rate": 5.036851429064893e-07,
+      "loss": 0.0432,
+      "step": 9453
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.4949059316421671,
+      "learning_rate": 5.030407378668939e-07,
+      "loss": 0.039,
+      "step": 9454
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.5491035998803881,
+      "learning_rate": 5.023967234754462e-07,
+      "loss": 0.0467,
+      "step": 9455
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.7369319243384285,
+      "learning_rate": 5.017530997880948e-07,
+      "loss": 0.0429,
+      "step": 9456
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.2036899260130764,
+      "learning_rate": 5.011098668607478e-07,
+      "loss": 0.0328,
+      "step": 9457
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.6144115463247422,
+      "learning_rate": 5.004670247492838e-07,
+      "loss": 0.0465,
+      "step": 9458
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.6980740804591137,
+      "learning_rate": 4.998245735095459e-07,
+      "loss": 0.0448,
+      "step": 9459
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.5715629139801757,
+      "learning_rate": 4.991825131973438e-07,
+      "loss": 0.0427,
+      "step": 9460
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.5939347609083596,
+      "learning_rate": 4.985408438684519e-07,
+      "loss": 0.0451,
+      "step": 9461
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.4050774036195004,
+      "learning_rate": 4.978995655786145e-07,
+      "loss": 0.0409,
+      "step": 9462
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.758837100701156,
+      "learning_rate": 4.972586783835348e-07,
+      "loss": 0.052,
+      "step": 9463
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.5576675345820696,
+      "learning_rate": 4.966181823388893e-07,
+      "loss": 0.0486,
+      "step": 9464
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.4273600991779292,
+      "learning_rate": 4.959780775003153e-07,
+      "loss": 0.0418,
+      "step": 9465
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.9990415096647731,
+      "learning_rate": 4.953383639234216e-07,
+      "loss": 0.0573,
+      "step": 9466
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.4494208550684164,
+      "learning_rate": 4.946990416637759e-07,
+      "loss": 0.0403,
+      "step": 9467
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.4911959701832518,
+      "learning_rate": 4.9406011077692e-07,
+      "loss": 0.0466,
+      "step": 9468
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.4539111239228553,
+      "learning_rate": 4.934215713183527e-07,
+      "loss": 0.0341,
+      "step": 9469
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.5458588281804682,
+      "learning_rate": 4.927834233435474e-07,
+      "loss": 0.0368,
+      "step": 9470
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.4138629052321512,
+      "learning_rate": 4.921456669079366e-07,
+      "loss": 0.0379,
+      "step": 9471
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.323973871698604,
+      "learning_rate": 4.915083020669248e-07,
+      "loss": 0.041,
+      "step": 9472
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.5093223602738592,
+      "learning_rate": 4.908713288758771e-07,
+      "loss": 0.0437,
+      "step": 9473
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.5445211578531475,
+      "learning_rate": 4.902347473901297e-07,
+      "loss": 0.0407,
+      "step": 9474
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.4153931419642454,
+      "learning_rate": 4.895985576649781e-07,
+      "loss": 0.0407,
+      "step": 9475
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.6210309251156636,
+      "learning_rate": 4.889627597556911e-07,
+      "loss": 0.0518,
+      "step": 9476
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.6978262220180413,
+      "learning_rate": 4.883273537174976e-07,
+      "loss": 0.0562,
+      "step": 9477
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.5321382138704425,
+      "learning_rate": 4.876923396055977e-07,
+      "loss": 0.0441,
+      "step": 9478
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.2016520276614269,
+      "learning_rate": 4.870577174751517e-07,
+      "loss": 0.034,
+      "step": 9479
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.6641765635057333,
+      "learning_rate": 4.864234873812928e-07,
+      "loss": 0.0419,
+      "step": 9480
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.585848376751937,
+      "learning_rate": 4.857896493791114e-07,
+      "loss": 0.047,
+      "step": 9481
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.4335185339832035,
+      "learning_rate": 4.851562035236723e-07,
+      "loss": 0.0395,
+      "step": 9482
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.7254863440110364,
+      "learning_rate": 4.845231498699998e-07,
+      "loss": 0.0536,
+      "step": 9483
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.7456862602568157,
+      "learning_rate": 4.8389048847309e-07,
+      "loss": 0.0567,
+      "step": 9484
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.44295099503431,
+      "learning_rate": 4.832582193878988e-07,
+      "loss": 0.0465,
+      "step": 9485
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.6516397546748454,
+      "learning_rate": 4.826263426693539e-07,
+      "loss": 0.0448,
+      "step": 9486
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.42135690675478,
+      "learning_rate": 4.819948583723427e-07,
+      "loss": 0.0455,
+      "step": 9487
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.370516455354655,
+      "learning_rate": 4.813637665517251e-07,
+      "loss": 0.0458,
+      "step": 9488
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.4810766280952907,
+      "learning_rate": 4.807330672623211e-07,
+      "loss": 0.0396,
+      "step": 9489
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.368160412066376,
+      "learning_rate": 4.80102760558921e-07,
+      "loss": 0.042,
+      "step": 9490
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.5253654707617381,
+      "learning_rate": 4.794728464962778e-07,
+      "loss": 0.0413,
+      "step": 9491
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.56328917780785,
+      "learning_rate": 4.788433251291141e-07,
+      "loss": 0.0487,
+      "step": 9492
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.4845320735215057,
+      "learning_rate": 4.782141965121129e-07,
+      "loss": 0.0428,
+      "step": 9493
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.4400576891437717,
+      "learning_rate": 4.775854606999286e-07,
+      "loss": 0.0427,
+      "step": 9494
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.5139143426400339,
+      "learning_rate": 4.769571177471771e-07,
+      "loss": 0.0424,
+      "step": 9495
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.255037020951712,
+      "learning_rate": 4.763291677084442e-07,
+      "loss": 0.033,
+      "step": 9496
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.5055425892205259,
+      "learning_rate": 4.757016106382778e-07,
+      "loss": 0.0481,
+      "step": 9497
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.6102096304762528,
+      "learning_rate": 4.750744465911955e-07,
+      "loss": 0.0483,
+      "step": 9498
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.28193087951188,
+      "learning_rate": 4.744476756216765e-07,
+      "loss": 0.0427,
+      "step": 9499
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.6470718244890163,
+      "learning_rate": 4.7382129778416885e-07,
+      "loss": 0.0436,
+      "step": 9500
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.3114637889482257,
+      "learning_rate": 4.7319531313308573e-07,
+      "loss": 0.0363,
+      "step": 9501
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.4572003532748765,
+      "learning_rate": 4.7256972172280646e-07,
+      "loss": 0.0437,
+      "step": 9502
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.7873252348252229,
+      "learning_rate": 4.7194452360767417e-07,
+      "loss": 0.0474,
+      "step": 9503
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.3694109220424995,
+      "learning_rate": 4.713197188420027e-07,
+      "loss": 0.043,
+      "step": 9504
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.501250763209713,
+      "learning_rate": 4.7069530748006463e-07,
+      "loss": 0.0448,
+      "step": 9505
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.4919803786136137,
+      "learning_rate": 4.7007128957610447e-07,
+      "loss": 0.0437,
+      "step": 9506
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.4569439756482014,
+      "learning_rate": 4.6944766518432936e-07,
+      "loss": 0.0392,
+      "step": 9507
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.3668100599845112,
+      "learning_rate": 4.6882443435891325e-07,
+      "loss": 0.0381,
+      "step": 9508
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5930763697735733,
+      "learning_rate": 4.6820159715399715e-07,
+      "loss": 0.0502,
+      "step": 9509
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.4620540062493423,
+      "learning_rate": 4.6757915362368567e-07,
+      "loss": 0.0488,
+      "step": 9510
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5094373079756083,
+      "learning_rate": 4.669571038220494e-07,
+      "loss": 0.0415,
+      "step": 9511
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.3819065961209056,
+      "learning_rate": 4.6633544780312565e-07,
+      "loss": 0.04,
+      "step": 9512
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.4430010230852144,
+      "learning_rate": 4.657141856209185e-07,
+      "loss": 0.0468,
+      "step": 9513
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5329100104241211,
+      "learning_rate": 4.6509331732939476e-07,
+      "loss": 0.0352,
+      "step": 9514
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.6388111210819711,
+      "learning_rate": 4.6447284298249127e-07,
+      "loss": 0.0461,
+      "step": 9515
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.6384250207853872,
+      "learning_rate": 4.6385276263410604e-07,
+      "loss": 0.0479,
+      "step": 9516
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.4805767987830838,
+      "learning_rate": 4.6323307633810653e-07,
+      "loss": 0.0461,
+      "step": 9517
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5445254517366718,
+      "learning_rate": 4.6261378414832304e-07,
+      "loss": 0.0469,
+      "step": 9518
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.6100393614242214,
+      "learning_rate": 4.619948861185547e-07,
+      "loss": 0.0362,
+      "step": 9519
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.4934181876837194,
+      "learning_rate": 4.6137638230256353e-07,
+      "loss": 0.0409,
+      "step": 9520
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.3688147581514603,
+      "learning_rate": 4.607582727540799e-07,
+      "loss": 0.0375,
+      "step": 9521
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.350878062334665,
+      "learning_rate": 4.601405575267981e-07,
+      "loss": 0.0405,
+      "step": 9522
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.8483355241968689,
+      "learning_rate": 4.5952323667437795e-07,
+      "loss": 0.0516,
+      "step": 9523
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5055110586907121,
+      "learning_rate": 4.589063102504454e-07,
+      "loss": 0.0453,
+      "step": 9524
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.291934826872168,
+      "learning_rate": 4.582897783085949e-07,
+      "loss": 0.0427,
+      "step": 9525
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.6701623552675646,
+      "learning_rate": 4.576736409023813e-07,
+      "loss": 0.0536,
+      "step": 9526
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.6483259650702111,
+      "learning_rate": 4.570578980853302e-07,
+      "loss": 0.0533,
+      "step": 9527
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.4582374512795266,
+      "learning_rate": 4.564425499109304e-07,
+      "loss": 0.0421,
+      "step": 9528
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.4289541497975753,
+      "learning_rate": 4.5582759643263583e-07,
+      "loss": 0.0364,
+      "step": 9529
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.4426341299269536,
+      "learning_rate": 4.552130377038677e-07,
+      "loss": 0.0427,
+      "step": 9530
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5100662040164692,
+      "learning_rate": 4.5459887377801273e-07,
+      "loss": 0.0502,
+      "step": 9531
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.2817017970217974,
+      "learning_rate": 4.5398510470842207e-07,
+      "loss": 0.0428,
+      "step": 9532
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5375537563431756,
+      "learning_rate": 4.533717305484153e-07,
+      "loss": 0.048,
+      "step": 9533
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.4918268622991384,
+      "learning_rate": 4.5275875135127325e-07,
+      "loss": 0.0489,
+      "step": 9534
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5373561181011297,
+      "learning_rate": 4.5214616717024764e-07,
+      "loss": 0.0414,
+      "step": 9535
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.3558061623370292,
+      "learning_rate": 4.5153397805855094e-07,
+      "loss": 0.0424,
+      "step": 9536
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.3004160844422055,
+      "learning_rate": 4.509221840693656e-07,
+      "loss": 0.035,
+      "step": 9537
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5876561997230576,
+      "learning_rate": 4.503107852558358e-07,
+      "loss": 0.0477,
+      "step": 9538
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.2031334073626196,
+      "learning_rate": 4.4969978167107684e-07,
+      "loss": 0.0396,
+      "step": 9539
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5658193588597842,
+      "learning_rate": 4.4908917336816237e-07,
+      "loss": 0.0462,
+      "step": 9540
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.2056230739369738,
+      "learning_rate": 4.484789604001377e-07,
+      "loss": 0.0366,
+      "step": 9541
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.6164753919974024,
+      "learning_rate": 4.478691428200099e-07,
+      "loss": 0.0432,
+      "step": 9542
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.5479220592543506,
+      "learning_rate": 4.472597206807561e-07,
+      "loss": 0.0502,
+      "step": 9543
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.6689289070238862,
+      "learning_rate": 4.466506940353138e-07,
+      "loss": 0.0506,
+      "step": 9544
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.4197064964887574,
+      "learning_rate": 4.460420629365919e-07,
+      "loss": 0.0422,
+      "step": 9545
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.384649335660778,
+      "learning_rate": 4.454338274374587e-07,
+      "loss": 0.0416,
+      "step": 9546
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.3782718443648383,
+      "learning_rate": 4.448259875907523e-07,
+      "loss": 0.0378,
+      "step": 9547
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.614217968265468,
+      "learning_rate": 4.4421854344927575e-07,
+      "loss": 0.0506,
+      "step": 9548
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.290495904666864,
+      "learning_rate": 4.4361149506579716e-07,
+      "loss": 0.0382,
+      "step": 9549
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.7277424937807064,
+      "learning_rate": 4.4300484249304996e-07,
+      "loss": 0.0503,
+      "step": 9550
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.335265947730024,
+      "learning_rate": 4.4239858578373597e-07,
+      "loss": 0.044,
+      "step": 9551
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.699859481302849,
+      "learning_rate": 4.4179272499051686e-07,
+      "loss": 0.0493,
+      "step": 9552
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.4255126742501394,
+      "learning_rate": 4.411872601660261e-07,
+      "loss": 0.0471,
+      "step": 9553
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.6536891684380957,
+      "learning_rate": 4.4058219136285774e-07,
+      "loss": 0.0526,
+      "step": 9554
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.4532082337850223,
+      "learning_rate": 4.399775186335764e-07,
+      "loss": 0.0425,
+      "step": 9555
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.469013044530932,
+      "learning_rate": 4.393732420307073e-07,
+      "loss": 0.0435,
+      "step": 9556
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.5723628394817593,
+      "learning_rate": 4.3876936160674623e-07,
+      "loss": 0.0455,
+      "step": 9557
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.5266400676304106,
+      "learning_rate": 4.381658774141484e-07,
+      "loss": 0.0478,
+      "step": 9558
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.3870519485012351,
+      "learning_rate": 4.375627895053408e-07,
+      "loss": 0.0474,
+      "step": 9559
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.4271746261598763,
+      "learning_rate": 4.3696009793271213e-07,
+      "loss": 0.0457,
+      "step": 9560
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.5415089083127158,
+      "learning_rate": 4.363578027486187e-07,
+      "loss": 0.0413,
+      "step": 9561
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.5119009632183478,
+      "learning_rate": 4.3575590400538046e-07,
+      "loss": 0.0456,
+      "step": 9562
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.6478816027790337,
+      "learning_rate": 4.351544017552861e-07,
+      "loss": 0.0572,
+      "step": 9563
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.4489145353676427,
+      "learning_rate": 4.3455329605058436e-07,
+      "loss": 0.0449,
+      "step": 9564
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.6815556982251532,
+      "learning_rate": 4.339525869434963e-07,
+      "loss": 0.0465,
+      "step": 9565
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.2571063108565679,
+      "learning_rate": 4.333522744862023e-07,
+      "loss": 0.0369,
+      "step": 9566
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.3178132071690807,
+      "learning_rate": 4.327523587308535e-07,
+      "loss": 0.0425,
+      "step": 9567
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.7437975616228967,
+      "learning_rate": 4.321528397295621e-07,
+      "loss": 0.0476,
+      "step": 9568
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.4169624417733622,
+      "learning_rate": 4.3155371753441146e-07,
+      "loss": 0.0447,
+      "step": 9569
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.5134645657278236,
+      "learning_rate": 4.309549921974421e-07,
+      "loss": 0.0451,
+      "step": 9570
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.3311828437239435,
+      "learning_rate": 4.3035666377066855e-07,
+      "loss": 0.0404,
+      "step": 9571
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.322036452766648,
+      "learning_rate": 4.2975873230606536e-07,
+      "loss": 0.04,
+      "step": 9572
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.501500500641906,
+      "learning_rate": 4.291611978555765e-07,
+      "loss": 0.0453,
+      "step": 9573
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.3516762879611213,
+      "learning_rate": 4.285640604711067e-07,
+      "loss": 0.0436,
+      "step": 9574
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.350929535121936,
+      "learning_rate": 4.279673202045326e-07,
+      "loss": 0.0411,
+      "step": 9575
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.6777444177088288,
+      "learning_rate": 4.2737097710768837e-07,
+      "loss": 0.054,
+      "step": 9576
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.4376696621562164,
+      "learning_rate": 4.2677503123238094e-07,
+      "loss": 0.0451,
+      "step": 9577
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.632608926454742,
+      "learning_rate": 4.261794826303783e-07,
+      "loss": 0.0493,
+      "step": 9578
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4420829343145307,
+      "learning_rate": 4.2558433135341694e-07,
+      "loss": 0.0455,
+      "step": 9579
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.6448636713443647,
+      "learning_rate": 4.249895774531948e-07,
+      "loss": 0.0518,
+      "step": 9580
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4376438838887593,
+      "learning_rate": 4.2439522098138173e-07,
+      "loss": 0.0434,
+      "step": 9581
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.477456807762911,
+      "learning_rate": 4.238012619896048e-07,
+      "loss": 0.0474,
+      "step": 9582
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.3350933413206312,
+      "learning_rate": 4.232077005294638e-07,
+      "loss": 0.0352,
+      "step": 9583
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.3643065807432677,
+      "learning_rate": 4.226145366525192e-07,
+      "loss": 0.0383,
+      "step": 9584
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4225632854187988,
+      "learning_rate": 4.2202177041030025e-07,
+      "loss": 0.0382,
+      "step": 9585
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4153227719599126,
+      "learning_rate": 4.2142940185429915e-07,
+      "loss": 0.0391,
+      "step": 9586
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.6851671514111863,
+      "learning_rate": 4.208374310359764e-07,
+      "loss": 0.0497,
+      "step": 9587
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.3968557618100055,
+      "learning_rate": 4.202458580067531e-07,
+      "loss": 0.0445,
+      "step": 9588
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.5098432942304885,
+      "learning_rate": 4.1965468281802145e-07,
+      "loss": 0.0444,
+      "step": 9589
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4841087959216106,
+      "learning_rate": 4.190639055211349e-07,
+      "loss": 0.0403,
+      "step": 9590
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.5362823199359597,
+      "learning_rate": 4.18473526167415e-07,
+      "loss": 0.0438,
+      "step": 9591
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.2829272993133471,
+      "learning_rate": 4.1788354480814696e-07,
+      "loss": 0.0398,
+      "step": 9592
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.441937812614088,
+      "learning_rate": 4.1729396149458367e-07,
+      "loss": 0.0453,
+      "step": 9593
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.3965150439936667,
+      "learning_rate": 4.167047762779391e-07,
+      "loss": 0.0422,
+      "step": 9594
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4641677120384513,
+      "learning_rate": 4.16115989209398e-07,
+      "loss": 0.0471,
+      "step": 9595
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4624805999265817,
+      "learning_rate": 4.155276003401054e-07,
+      "loss": 0.0413,
+      "step": 9596
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.3503470474501447,
+      "learning_rate": 4.149396097211772e-07,
+      "loss": 0.0386,
+      "step": 9597
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.7506135868904837,
+      "learning_rate": 4.1435201740368914e-07,
+      "loss": 0.0489,
+      "step": 9598
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.2793660070393602,
+      "learning_rate": 4.137648234386871e-07,
+      "loss": 0.038,
+      "step": 9599
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.5572386084438272,
+      "learning_rate": 4.1317802787717963e-07,
+      "loss": 0.0485,
+      "step": 9600
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4277414060374767,
+      "learning_rate": 4.1259163077014e-07,
+      "loss": 0.0427,
+      "step": 9601
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4858533183587261,
+      "learning_rate": 4.120056321685101e-07,
+      "loss": 0.044,
+      "step": 9602
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.384292873930322,
+      "learning_rate": 4.114200321231937e-07,
+      "loss": 0.0465,
+      "step": 9603
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.5467212615041195,
+      "learning_rate": 4.108348306850629e-07,
+      "loss": 0.0472,
+      "step": 9604
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4713975399676553,
+      "learning_rate": 4.1025002790495317e-07,
+      "loss": 0.0371,
+      "step": 9605
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.622189308365567,
+      "learning_rate": 4.09665623833666e-07,
+      "loss": 0.0456,
+      "step": 9606
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4034371176090872,
+      "learning_rate": 4.0908161852196706e-07,
+      "loss": 0.0465,
+      "step": 9607
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.3138294371510686,
+      "learning_rate": 4.0849801202059113e-07,
+      "loss": 0.039,
+      "step": 9608
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.833006646548874,
+      "learning_rate": 4.079148043802328e-07,
+      "loss": 0.0493,
+      "step": 9609
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.318008375752975,
+      "learning_rate": 4.0733199565155814e-07,
+      "loss": 0.0403,
+      "step": 9610
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.3939835669538023,
+      "learning_rate": 4.067495858851922e-07,
+      "loss": 0.0467,
+      "step": 9611
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.5090937273351503,
+      "learning_rate": 4.0616757513173123e-07,
+      "loss": 0.0385,
+      "step": 9612
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.234917698747976,
+      "learning_rate": 4.055859634417314e-07,
+      "loss": 0.0374,
+      "step": 9613
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.3393417058968564,
+      "learning_rate": 4.050047508657201e-07,
+      "loss": 0.0403,
+      "step": 9614
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.2797919141688916,
+      "learning_rate": 4.0442393745418415e-07,
+      "loss": 0.0368,
+      "step": 9615
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.5212888060663954,
+      "learning_rate": 4.0384352325758104e-07,
+      "loss": 0.0462,
+      "step": 9616
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.3996560028828897,
+      "learning_rate": 4.0326350832632865e-07,
+      "loss": 0.0373,
+      "step": 9617
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.5057362796375684,
+      "learning_rate": 4.02683892710814e-07,
+      "loss": 0.0376,
+      "step": 9618
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.4863508628178341,
+      "learning_rate": 4.0210467646138674e-07,
+      "loss": 0.0452,
+      "step": 9619
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.5403708190458993,
+      "learning_rate": 4.0152585962836444e-07,
+      "loss": 0.0488,
+      "step": 9620
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.3672349842028424,
+      "learning_rate": 4.009474422620269e-07,
+      "loss": 0.0448,
+      "step": 9621
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.3738632875135273,
+      "learning_rate": 4.0036942441262385e-07,
+      "loss": 0.0392,
+      "step": 9622
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.4511639709781454,
+      "learning_rate": 3.997918061303635e-07,
+      "loss": 0.0438,
+      "step": 9623
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.658930113970053,
+      "learning_rate": 3.992145874654263e-07,
+      "loss": 0.0407,
+      "step": 9624
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.400134270401821,
+      "learning_rate": 3.9863776846795265e-07,
+      "loss": 0.0387,
+      "step": 9625
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.235493170328634,
+      "learning_rate": 3.980613491880525e-07,
+      "loss": 0.0339,
+      "step": 9626
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.3468224840469483,
+      "learning_rate": 3.974853296757969e-07,
+      "loss": 0.0406,
+      "step": 9627
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.7401560436294927,
+      "learning_rate": 3.9690970998122745e-07,
+      "loss": 0.0432,
+      "step": 9628
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.5139068090801995,
+      "learning_rate": 3.963344901543437e-07,
+      "loss": 0.0491,
+      "step": 9629
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.2413803350643091,
+      "learning_rate": 3.957596702451183e-07,
+      "loss": 0.0421,
+      "step": 9630
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.2187718116857942,
+      "learning_rate": 3.9518525030348307e-07,
+      "loss": 0.0357,
+      "step": 9631
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.3373550799879357,
+      "learning_rate": 3.9461123037933923e-07,
+      "loss": 0.0393,
+      "step": 9632
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.6875917024728762,
+      "learning_rate": 3.940376105225496e-07,
+      "loss": 0.0437,
+      "step": 9633
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.595828548429279,
+      "learning_rate": 3.934643907829477e-07,
+      "loss": 0.0434,
+      "step": 9634
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.3621943665747567,
+      "learning_rate": 3.9289157121032485e-07,
+      "loss": 0.0391,
+      "step": 9635
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.5098424185350363,
+      "learning_rate": 3.9231915185444337e-07,
+      "loss": 0.0432,
+      "step": 9636
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.6312294134192091,
+      "learning_rate": 3.9174713276502853e-07,
+      "loss": 0.0409,
+      "step": 9637
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.3645249435977806,
+      "learning_rate": 3.911755139917722e-07,
+      "loss": 0.0454,
+      "step": 9638
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.6431027718609474,
+      "learning_rate": 3.906042955843298e-07,
+      "loss": 0.0478,
+      "step": 9639
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.4013439191510617,
+      "learning_rate": 3.900334775923237e-07,
+      "loss": 0.0454,
+      "step": 9640
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.5890009361789863,
+      "learning_rate": 3.894630600653382e-07,
+      "loss": 0.0446,
+      "step": 9641
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.589326027576601,
+      "learning_rate": 3.888930430529275e-07,
+      "loss": 0.0438,
+      "step": 9642
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.4258426411561043,
+      "learning_rate": 3.883234266046071e-07,
+      "loss": 0.0386,
+      "step": 9643
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.39026434914808,
+      "learning_rate": 3.8775421076986066e-07,
+      "loss": 0.0403,
+      "step": 9644
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.3744058080541643,
+      "learning_rate": 3.871853955981336e-07,
+      "loss": 0.0394,
+      "step": 9645
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.3062424475989316,
+      "learning_rate": 3.866169811388415e-07,
+      "loss": 0.039,
+      "step": 9646
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.482074727659546,
+      "learning_rate": 3.8604896744135923e-07,
+      "loss": 0.0504,
+      "step": 9647
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.5515775600029753,
+      "learning_rate": 3.8548135455503176e-07,
+      "loss": 0.0503,
+      "step": 9648
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.570176912303616,
+      "learning_rate": 3.849141425291658e-07,
+      "loss": 0.0486,
+      "step": 9649
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.546208018880304,
+      "learning_rate": 3.843473314130358e-07,
+      "loss": 0.0433,
+      "step": 9650
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.3709802341085098,
+      "learning_rate": 3.837809212558796e-07,
+      "loss": 0.0367,
+      "step": 9651
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.2039069547037646,
+      "learning_rate": 3.832149121069029e-07,
+      "loss": 0.0345,
+      "step": 9652
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.6244670252580329,
+      "learning_rate": 3.8264930401527123e-07,
+      "loss": 0.045,
+      "step": 9653
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.5845169308905884,
+      "learning_rate": 3.8208409703012153e-07,
+      "loss": 0.0491,
+      "step": 9654
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.2980211178456464,
+      "learning_rate": 3.815192912005505e-07,
+      "loss": 0.0386,
+      "step": 9655
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.5798748475729993,
+      "learning_rate": 3.809548865756246e-07,
+      "loss": 0.0486,
+      "step": 9656
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.6081215345888544,
+      "learning_rate": 3.803908832043718e-07,
+      "loss": 0.0444,
+      "step": 9657
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.688638157637924,
+      "learning_rate": 3.7982728113578946e-07,
+      "loss": 0.0526,
+      "step": 9658
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.470612772243646,
+      "learning_rate": 3.7926408041883355e-07,
+      "loss": 0.0481,
+      "step": 9659
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.4298932739818198,
+      "learning_rate": 3.7870128110243155e-07,
+      "loss": 0.0417,
+      "step": 9660
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.1776982384922046,
+      "learning_rate": 3.7813888323547155e-07,
+      "loss": 0.0311,
+      "step": 9661
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.5898361754061843,
+      "learning_rate": 3.7757688686681117e-07,
+      "loss": 0.0469,
+      "step": 9662
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.2990397451577733,
+      "learning_rate": 3.7701529204526856e-07,
+      "loss": 0.0351,
+      "step": 9663
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.4558233502207352,
+      "learning_rate": 3.7645409881963133e-07,
+      "loss": 0.042,
+      "step": 9664
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.3717147224539585,
+      "learning_rate": 3.7589330723864724e-07,
+      "loss": 0.0392,
+      "step": 9665
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.3737318122822282,
+      "learning_rate": 3.753329173510345e-07,
+      "loss": 0.0432,
+      "step": 9666
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.5664873477638628,
+      "learning_rate": 3.7477292920547134e-07,
+      "loss": 0.0431,
+      "step": 9667
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.505345058074906,
+      "learning_rate": 3.7421334285060617e-07,
+      "loss": 0.0443,
+      "step": 9668
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.7109639389418407,
+      "learning_rate": 3.736541583350473e-07,
+      "loss": 0.0555,
+      "step": 9669
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.5779923672220566,
+      "learning_rate": 3.730953757073741e-07,
+      "loss": 0.0471,
+      "step": 9670
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.625205631063717,
+      "learning_rate": 3.7253699501612394e-07,
+      "loss": 0.0457,
+      "step": 9671
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.512805026027645,
+      "learning_rate": 3.719790163098058e-07,
+      "loss": 0.046,
+      "step": 9672
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.4187021956642465,
+      "learning_rate": 3.7142143963688927e-07,
+      "loss": 0.0447,
+      "step": 9673
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.6555273882899562,
+      "learning_rate": 3.7086426504581166e-07,
+      "loss": 0.0489,
+      "step": 9674
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.6604519275628469,
+      "learning_rate": 3.7030749258497365e-07,
+      "loss": 0.0583,
+      "step": 9675
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.513595617640187,
+      "learning_rate": 3.697511223027439e-07,
+      "loss": 0.0433,
+      "step": 9676
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.3531490921404383,
+      "learning_rate": 3.6919515424745035e-07,
+      "loss": 0.0415,
+      "step": 9677
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.5663841629182775,
+      "learning_rate": 3.6863958846739213e-07,
+      "loss": 0.046,
+      "step": 9678
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.3475148826851961,
+      "learning_rate": 3.6808442501083007e-07,
+      "loss": 0.0356,
+      "step": 9679
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.4235683745953958,
+      "learning_rate": 3.675296639259912e-07,
+      "loss": 0.0439,
+      "step": 9680
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.2864579277767414,
+      "learning_rate": 3.6697530526106697e-07,
+      "loss": 0.0417,
+      "step": 9681
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.5248006513046146,
+      "learning_rate": 3.66421349064216e-07,
+      "loss": 0.0474,
+      "step": 9682
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.6962991151385718,
+      "learning_rate": 3.6586779538355656e-07,
+      "loss": 0.0423,
+      "step": 9683
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.4313109310666983,
+      "learning_rate": 3.6531464426717843e-07,
+      "loss": 0.0497,
+      "step": 9684
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.5021430581608235,
+      "learning_rate": 3.6476189576313215e-07,
+      "loss": 0.0399,
+      "step": 9685
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.5188652925885526,
+      "learning_rate": 3.6420954991943537e-07,
+      "loss": 0.0476,
+      "step": 9686
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.5099403700176421,
+      "learning_rate": 3.636576067840697e-07,
+      "loss": 0.0435,
+      "step": 9687
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.2953869764399193,
+      "learning_rate": 3.631060664049824e-07,
+      "loss": 0.036,
+      "step": 9688
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.135637780544938,
+      "learning_rate": 3.6255492883008446e-07,
+      "loss": 0.0839,
+      "step": 9689
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.297224895146072,
+      "learning_rate": 3.620041941072544e-07,
+      "loss": 0.0333,
+      "step": 9690
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.403688915026389,
+      "learning_rate": 3.614538622843328e-07,
+      "loss": 0.04,
+      "step": 9691
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.428840921297277,
+      "learning_rate": 3.609039334091269e-07,
+      "loss": 0.0446,
+      "step": 9692
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.6866592808076206,
+      "learning_rate": 3.6035440752941075e-07,
+      "loss": 0.0441,
+      "step": 9693
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.6774085384434785,
+      "learning_rate": 3.598052846929184e-07,
+      "loss": 0.0495,
+      "step": 9694
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.6489876335898097,
+      "learning_rate": 3.592565649473534e-07,
+      "loss": 0.0521,
+      "step": 9695
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.4852908212474134,
+      "learning_rate": 3.58708248340382e-07,
+      "loss": 0.0481,
+      "step": 9696
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.7915707820764906,
+      "learning_rate": 3.581603349196372e-07,
+      "loss": 0.0474,
+      "step": 9697
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.3782621004698132,
+      "learning_rate": 3.576128247327143e-07,
+      "loss": 0.0368,
+      "step": 9698
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.3292731299290403,
+      "learning_rate": 3.57065717827178e-07,
+      "loss": 0.0393,
+      "step": 9699
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.6146172252114868,
+      "learning_rate": 3.565190142505515e-07,
+      "loss": 0.0494,
+      "step": 9700
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.2775572420920787,
+      "learning_rate": 3.5597271405032887e-07,
+      "loss": 0.0325,
+      "step": 9701
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.5191835708494437,
+      "learning_rate": 3.5542681727396613e-07,
+      "loss": 0.0412,
+      "step": 9702
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.510335564061284,
+      "learning_rate": 3.548813239688853e-07,
+      "loss": 0.044,
+      "step": 9703
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.854392936934028,
+      "learning_rate": 3.54336234182473e-07,
+      "loss": 0.0517,
+      "step": 9704
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.4958541756591652,
+      "learning_rate": 3.537915479620818e-07,
+      "loss": 0.0469,
+      "step": 9705
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.3512280113677804,
+      "learning_rate": 3.532472653550262e-07,
+      "loss": 0.0423,
+      "step": 9706
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.5478550426148945,
+      "learning_rate": 3.5270338640858993e-07,
+      "loss": 0.0472,
+      "step": 9707
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.4953008899616966,
+      "learning_rate": 3.521599111700169e-07,
+      "loss": 0.039,
+      "step": 9708
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.7281674031853826,
+      "learning_rate": 3.5161683968652104e-07,
+      "loss": 0.053,
+      "step": 9709
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.6259853940180455,
+      "learning_rate": 3.5107417200527625e-07,
+      "loss": 0.0421,
+      "step": 9710
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.6828368280114945,
+      "learning_rate": 3.5053190817342707e-07,
+      "loss": 0.0535,
+      "step": 9711
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.3286215114643205,
+      "learning_rate": 3.499900482380758e-07,
+      "loss": 0.0414,
+      "step": 9712
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.539048145541395,
+      "learning_rate": 3.4944859224629645e-07,
+      "loss": 0.0471,
+      "step": 9713
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.4901233155701168,
+      "learning_rate": 3.4890754024512254e-07,
+      "loss": 0.043,
+      "step": 9714
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.5051511939728763,
+      "learning_rate": 3.4836689228155697e-07,
+      "loss": 0.0462,
+      "step": 9715
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.377644116671599,
+      "learning_rate": 3.4782664840256387e-07,
+      "loss": 0.0368,
+      "step": 9716
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.2006219765145378,
+      "learning_rate": 3.472868086550768e-07,
+      "loss": 0.0358,
+      "step": 9717
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.2678294845884381,
+      "learning_rate": 3.4674737308598714e-07,
+      "loss": 0.0376,
+      "step": 9718
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.3929636370664398,
+      "learning_rate": 3.4620834174215856e-07,
+      "loss": 0.0412,
+      "step": 9719
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.4893405556902184,
+      "learning_rate": 3.4566971467041463e-07,
+      "loss": 0.047,
+      "step": 9720
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.5853225786233056,
+      "learning_rate": 3.4513149191754635e-07,
+      "loss": 0.0524,
+      "step": 9721
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.6986716196288256,
+      "learning_rate": 3.4459367353030846e-07,
+      "loss": 0.0506,
+      "step": 9722
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.5638335667719534,
+      "learning_rate": 3.4405625955542254e-07,
+      "loss": 0.0375,
+      "step": 9723
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.411961688956543,
+      "learning_rate": 3.4351925003957065e-07,
+      "loss": 0.0413,
+      "step": 9724
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.3626148803904228,
+      "learning_rate": 3.4298264502940436e-07,
+      "loss": 0.047,
+      "step": 9725
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.5640031429974925,
+      "learning_rate": 3.42446444571537e-07,
+      "loss": 0.0526,
+      "step": 9726
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.3401156168510278,
+      "learning_rate": 3.419106487125495e-07,
+      "loss": 0.0389,
+      "step": 9727
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.4997626914865505,
+      "learning_rate": 3.4137525749898425e-07,
+      "loss": 0.044,
+      "step": 9728
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.3883680349965608,
+      "learning_rate": 3.408402709773534e-07,
+      "loss": 0.0438,
+      "step": 9729
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.2925281710706147,
+      "learning_rate": 3.4030568919412697e-07,
+      "loss": 0.0425,
+      "step": 9730
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.272416215838726,
+      "learning_rate": 3.397715121957468e-07,
+      "loss": 0.0377,
+      "step": 9731
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.363483812621177,
+      "learning_rate": 3.3923774002861454e-07,
+      "loss": 0.0393,
+      "step": 9732
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.3283657593865004,
+      "learning_rate": 3.387043727391004e-07,
+      "loss": 0.0362,
+      "step": 9733
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.2920712417065263,
+      "learning_rate": 3.3817141037353565e-07,
+      "loss": 0.0378,
+      "step": 9734
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.6409444840690268,
+      "learning_rate": 3.3763885297822153e-07,
+      "loss": 0.0557,
+      "step": 9735
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.8048520824904202,
+      "learning_rate": 3.3710670059941777e-07,
+      "loss": 0.0484,
+      "step": 9736
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.476179626598445,
+      "learning_rate": 3.36574953283354e-07,
+      "loss": 0.0486,
+      "step": 9737
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.4594338579927015,
+      "learning_rate": 3.3604361107622106e-07,
+      "loss": 0.0431,
+      "step": 9738
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.2122888322147463,
+      "learning_rate": 3.3551267402417874e-07,
+      "loss": 0.0351,
+      "step": 9739
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.5503853545996096,
+      "learning_rate": 3.349821421733468e-07,
+      "loss": 0.0387,
+      "step": 9740
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.342768497659534,
+      "learning_rate": 3.34452015569815e-07,
+      "loss": 0.0421,
+      "step": 9741
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.3232370565035567,
+      "learning_rate": 3.339222942596321e-07,
+      "loss": 0.0371,
+      "step": 9742
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.9085039908039205,
+      "learning_rate": 3.333929782888168e-07,
+      "loss": 0.05,
+      "step": 9743
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.4828854416236226,
+      "learning_rate": 3.3286406770334843e-07,
+      "loss": 0.0423,
+      "step": 9744
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.5442392845080655,
+      "learning_rate": 3.323355625491759e-07,
+      "loss": 0.0513,
+      "step": 9745
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.3520004766054956,
+      "learning_rate": 3.318074628722079e-07,
+      "loss": 0.0383,
+      "step": 9746
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.5964540815010628,
+      "learning_rate": 3.312797687183217e-07,
+      "loss": 0.0582,
+      "step": 9747
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.56785560375489,
+      "learning_rate": 3.3075248013335614e-07,
+      "loss": 0.0508,
+      "step": 9748
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.614260796169554,
+      "learning_rate": 3.30225597163118e-07,
+      "loss": 0.0518,
+      "step": 9749
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.5694730490528126,
+      "learning_rate": 3.2969911985337556e-07,
+      "loss": 0.0451,
+      "step": 9750
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.7683488225591102,
+      "learning_rate": 3.2917304824986505e-07,
+      "loss": 0.0561,
+      "step": 9751
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.4454028560418148,
+      "learning_rate": 3.2864738239828553e-07,
+      "loss": 0.0418,
+      "step": 9752
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.5873321412313446,
+      "learning_rate": 3.281221223443026e-07,
+      "loss": 0.0452,
+      "step": 9753
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.533996345376737,
+      "learning_rate": 3.275972681335421e-07,
+      "loss": 0.0366,
+      "step": 9754
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.4070084597921413,
+      "learning_rate": 3.2707281981160075e-07,
+      "loss": 0.0414,
+      "step": 9755
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.1728929179470517,
+      "learning_rate": 3.26548777424035e-07,
+      "loss": 0.0315,
+      "step": 9756
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.392700248420635,
+      "learning_rate": 3.2602514101637004e-07,
+      "loss": 0.0431,
+      "step": 9757
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.410957420236946,
+      "learning_rate": 3.255019106340923e-07,
+      "loss": 0.0426,
+      "step": 9758
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.4396989430568352,
+      "learning_rate": 3.249790863226565e-07,
+      "loss": 0.0507,
+      "step": 9759
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.6143241200088134,
+      "learning_rate": 3.244566681274769e-07,
+      "loss": 0.0507,
+      "step": 9760
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.7198136483955655,
+      "learning_rate": 3.2393465609393825e-07,
+      "loss": 0.0536,
+      "step": 9761
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.398777095462013,
+      "learning_rate": 3.23413050267386e-07,
+      "loss": 0.0429,
+      "step": 9762
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.3250820426437029,
+      "learning_rate": 3.2289185069313277e-07,
+      "loss": 0.0392,
+      "step": 9763
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.3451370940852598,
+      "learning_rate": 3.2237105741645456e-07,
+      "loss": 0.0423,
+      "step": 9764
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.599758765473879,
+      "learning_rate": 3.2185067048259245e-07,
+      "loss": 0.0429,
+      "step": 9765
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.6036227106949597,
+      "learning_rate": 3.213306899367508e-07,
+      "loss": 0.0483,
+      "step": 9766
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.482535618187445,
+      "learning_rate": 3.208111158241023e-07,
+      "loss": 0.0426,
+      "step": 9767
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.4123398585691431,
+      "learning_rate": 3.2029194818977984e-07,
+      "loss": 0.0404,
+      "step": 9768
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.421548355183196,
+      "learning_rate": 3.1977318707888506e-07,
+      "loss": 0.04,
+      "step": 9769
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.6064524386653583,
+      "learning_rate": 3.1925483253648135e-07,
+      "loss": 0.0441,
+      "step": 9770
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.383278161948324,
+      "learning_rate": 3.187368846075983e-07,
+      "loss": 0.039,
+      "step": 9771
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.391997503570171,
+      "learning_rate": 3.182193433372288e-07,
+      "loss": 0.0437,
+      "step": 9772
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.633697649344028,
+      "learning_rate": 3.1770220877033243e-07,
+      "loss": 0.0527,
+      "step": 9773
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.5719644713419079,
+      "learning_rate": 3.1718548095183153e-07,
+      "loss": 0.0447,
+      "step": 9774
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.494395449206361,
+      "learning_rate": 3.166691599266153e-07,
+      "loss": 0.048,
+      "step": 9775
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.3734912339847678,
+      "learning_rate": 3.161532457395355e-07,
+      "loss": 0.0433,
+      "step": 9776
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.463169521929121,
+      "learning_rate": 3.156377384354087e-07,
+      "loss": 0.0453,
+      "step": 9777
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.4940975046453457,
+      "learning_rate": 3.1512263805901667e-07,
+      "loss": 0.0482,
+      "step": 9778
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.678748396347184,
+      "learning_rate": 3.14607944655107e-07,
+      "loss": 0.0549,
+      "step": 9779
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.2637093962025079,
+      "learning_rate": 3.14093658268389e-07,
+      "loss": 0.0344,
+      "step": 9780
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.6270888878882568,
+      "learning_rate": 3.135797789435407e-07,
+      "loss": 0.0462,
+      "step": 9781
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.21991156245253,
+      "learning_rate": 3.1306630672520153e-07,
+      "loss": 0.0319,
+      "step": 9782
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.4038695739936458,
+      "learning_rate": 3.125532416579763e-07,
+      "loss": 0.0406,
+      "step": 9783
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.3508710337807508,
+      "learning_rate": 3.1204058378643375e-07,
+      "loss": 0.0432,
+      "step": 9784
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.6502976609643316,
+      "learning_rate": 3.115283331551089e-07,
+      "loss": 0.0419,
+      "step": 9785
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.4305730020335734,
+      "learning_rate": 3.1101648980850217e-07,
+      "loss": 0.0452,
+      "step": 9786
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.5634464708578433,
+      "learning_rate": 3.105050537910742e-07,
+      "loss": 0.0461,
+      "step": 9787
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.4910660517319974,
+      "learning_rate": 3.099940251472572e-07,
+      "loss": 0.0403,
+      "step": 9788
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.6738891191621585,
+      "learning_rate": 3.0948340392143897e-07,
+      "loss": 0.0513,
+      "step": 9789
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.3986875170971205,
+      "learning_rate": 3.0897319015798067e-07,
+      "loss": 0.0378,
+      "step": 9790
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.5293050996106932,
+      "learning_rate": 3.084633839012019e-07,
+      "loss": 0.0484,
+      "step": 9791
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.5281477245803312,
+      "learning_rate": 3.0795398519539113e-07,
+      "loss": 0.043,
+      "step": 9792
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.6091043511257537,
+      "learning_rate": 3.074449940847979e-07,
+      "loss": 0.0533,
+      "step": 9793
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.5523465910460366,
+      "learning_rate": 3.069364106136402e-07,
+      "loss": 0.0475,
+      "step": 9794
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.699985764421929,
+      "learning_rate": 3.0642823482609495e-07,
+      "loss": 0.0489,
+      "step": 9795
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.6321301043488206,
+      "learning_rate": 3.0592046676631015e-07,
+      "loss": 0.0487,
+      "step": 9796
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.2115564049325827,
+      "learning_rate": 3.054131064783933e-07,
+      "loss": 0.04,
+      "step": 9797
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.3469935338261332,
+      "learning_rate": 3.049061540064202e-07,
+      "loss": 0.0427,
+      "step": 9798
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.3876714390383034,
+      "learning_rate": 3.0439960939442794e-07,
+      "loss": 0.0397,
+      "step": 9799
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.4282545431976708,
+      "learning_rate": 3.038934726864218e-07,
+      "loss": 0.034,
+      "step": 9800
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.6714890567533012,
+      "learning_rate": 3.033877439263666e-07,
+      "loss": 0.0488,
+      "step": 9801
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.4031503952263187,
+      "learning_rate": 3.0288242315819724e-07,
+      "loss": 0.0471,
+      "step": 9802
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.3463164412475201,
+      "learning_rate": 3.0237751042580866e-07,
+      "loss": 0.0388,
+      "step": 9803
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.6023658979704292,
+      "learning_rate": 3.0187300577306456e-07,
+      "loss": 0.0413,
+      "step": 9804
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.5198271549143991,
+      "learning_rate": 3.013689092437888e-07,
+      "loss": 0.0472,
+      "step": 9805
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.6933446982158806,
+      "learning_rate": 3.0086522088177415e-07,
+      "loss": 0.0474,
+      "step": 9806
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.465241027092162,
+      "learning_rate": 3.003619407307734e-07,
+      "loss": 0.042,
+      "step": 9807
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.2992516944884374,
+      "learning_rate": 2.9985906883450765e-07,
+      "loss": 0.0419,
+      "step": 9808
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.4912614738979613,
+      "learning_rate": 2.9935660523665976e-07,
+      "loss": 0.0436,
+      "step": 9809
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.383811868812435,
+      "learning_rate": 2.988545499808804e-07,
+      "loss": 0.0432,
+      "step": 9810
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.3341734929850606,
+      "learning_rate": 2.9835290311078123e-07,
+      "loss": 0.0387,
+      "step": 9811
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.3192136579361862,
+      "learning_rate": 2.9785166466994195e-07,
+      "loss": 0.0361,
+      "step": 9812
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.553888799852158,
+      "learning_rate": 2.9735083470190164e-07,
+      "loss": 0.0527,
+      "step": 9813
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.652181334845828,
+      "learning_rate": 2.9685041325016983e-07,
+      "loss": 0.0471,
+      "step": 9814
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.443994193885354,
+      "learning_rate": 2.9635040035821627e-07,
+      "loss": 0.0437,
+      "step": 9815
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.4326933168637934,
+      "learning_rate": 2.9585079606947843e-07,
+      "loss": 0.0422,
+      "step": 9816
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.6185615132232907,
+      "learning_rate": 2.953516004273543e-07,
+      "loss": 0.0471,
+      "step": 9817
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.3645734980620279,
+      "learning_rate": 2.948528134752121e-07,
+      "loss": 0.035,
+      "step": 9818
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.5018381755880716,
+      "learning_rate": 2.943544352563771e-07,
+      "loss": 0.0447,
+      "step": 9819
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.3158266951887627,
+      "learning_rate": 2.938564658141463e-07,
+      "loss": 0.0356,
+      "step": 9820
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.5306089743479394,
+      "learning_rate": 2.933589051917757e-07,
+      "loss": 0.046,
+      "step": 9821
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.5657788120774125,
+      "learning_rate": 2.9286175343249015e-07,
+      "loss": 0.0446,
+      "step": 9822
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.5756581173013304,
+      "learning_rate": 2.9236501057947506e-07,
+      "loss": 0.0426,
+      "step": 9823
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.545620882655996,
+      "learning_rate": 2.918686766758844e-07,
+      "loss": 0.0446,
+      "step": 9824
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.4705908053138845,
+      "learning_rate": 2.913727517648318e-07,
+      "loss": 0.0447,
+      "step": 9825
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.5773264693029911,
+      "learning_rate": 2.908772358894002e-07,
+      "loss": 0.0471,
+      "step": 9826
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.3602340679270009,
+      "learning_rate": 2.903821290926329e-07,
+      "loss": 0.0367,
+      "step": 9827
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.4474149091282598,
+      "learning_rate": 2.898874314175415e-07,
+      "loss": 0.0457,
+      "step": 9828
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.6105528413705201,
+      "learning_rate": 2.8939314290709784e-07,
+      "loss": 0.0465,
+      "step": 9829
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.2969374472911948,
+      "learning_rate": 2.888992636042437e-07,
+      "loss": 0.0345,
+      "step": 9830
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.6759926534803957,
+      "learning_rate": 2.8840579355187803e-07,
+      "loss": 0.0423,
+      "step": 9831
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.4730773547325942,
+      "learning_rate": 2.87912732792871e-07,
+      "loss": 0.0482,
+      "step": 9832
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.240197143859333,
+      "learning_rate": 2.874200813700534e-07,
+      "loss": 0.0345,
+      "step": 9833
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.5300998369846195,
+      "learning_rate": 2.869278393262226e-07,
+      "loss": 0.0487,
+      "step": 9834
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.6449191163018604,
+      "learning_rate": 2.8643600670413773e-07,
+      "loss": 0.047,
+      "step": 9835
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.365745756381367,
+      "learning_rate": 2.8594458354652687e-07,
+      "loss": 0.0391,
+      "step": 9836
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.4020974106830846,
+      "learning_rate": 2.8545356989607587e-07,
+      "loss": 0.0377,
+      "step": 9837
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.4993321914494795,
+      "learning_rate": 2.849629657954417e-07,
+      "loss": 0.0423,
+      "step": 9838
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.3176257290625997,
+      "learning_rate": 2.8447277128724136e-07,
+      "loss": 0.0396,
+      "step": 9839
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.5722986440128743,
+      "learning_rate": 2.839829864140586e-07,
+      "loss": 0.0505,
+      "step": 9840
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.3387003357080745,
+      "learning_rate": 2.8349361121844056e-07,
+      "loss": 0.0316,
+      "step": 9841
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.3888312025810354,
+      "learning_rate": 2.8300464574289866e-07,
+      "loss": 0.0473,
+      "step": 9842
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.7009623484664964,
+      "learning_rate": 2.8251609002990844e-07,
+      "loss": 0.05,
+      "step": 9843
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.309737893226172,
+      "learning_rate": 2.82027944121912e-07,
+      "loss": 0.0336,
+      "step": 9844
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.5967437592276221,
+      "learning_rate": 2.815402080613122e-07,
+      "loss": 0.0433,
+      "step": 9845
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.3585195529859975,
+      "learning_rate": 2.810528818904812e-07,
+      "loss": 0.036,
+      "step": 9846
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.4126556863302957,
+      "learning_rate": 2.8056596565175067e-07,
+      "loss": 0.0467,
+      "step": 9847
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.3086968936064312,
+      "learning_rate": 2.80079459387419e-07,
+      "loss": 0.0361,
+      "step": 9848
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.6818363077834921,
+      "learning_rate": 2.7959336313974847e-07,
+      "loss": 0.0504,
+      "step": 9849
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.5029960459553915,
+      "learning_rate": 2.7910767695096707e-07,
+      "loss": 0.0474,
+      "step": 9850
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.3907806419948412,
+      "learning_rate": 2.7862240086326486e-07,
+      "loss": 0.0435,
+      "step": 9851
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.452709734046454,
+      "learning_rate": 2.781375349187987e-07,
+      "loss": 0.0409,
+      "step": 9852
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.7429092616502795,
+      "learning_rate": 2.7765307915968763e-07,
+      "loss": 0.0567,
+      "step": 9853
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.705903219762938,
+      "learning_rate": 2.771690336280164e-07,
+      "loss": 0.0556,
+      "step": 9854
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.7142335408285376,
+      "learning_rate": 2.7668539836583295e-07,
+      "loss": 0.0466,
+      "step": 9855
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.637738099032581,
+      "learning_rate": 2.762021734151521e-07,
+      "loss": 0.0482,
+      "step": 9856
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.3529093302066764,
+      "learning_rate": 2.7571935881794963e-07,
+      "loss": 0.0359,
+      "step": 9857
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.5529970753734432,
+      "learning_rate": 2.7523695461616875e-07,
+      "loss": 0.0432,
+      "step": 9858
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.707764222059563,
+      "learning_rate": 2.747549608517147e-07,
+      "loss": 0.0516,
+      "step": 9859
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.4463722368483976,
+      "learning_rate": 2.74273377566458e-07,
+      "loss": 0.0423,
+      "step": 9860
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.349447549362689,
+      "learning_rate": 2.7379220480223345e-07,
+      "loss": 0.0442,
+      "step": 9861
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.5103472549282548,
+      "learning_rate": 2.7331144260084096e-07,
+      "loss": 0.0402,
+      "step": 9862
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.2552170688503217,
+      "learning_rate": 2.7283109100404323e-07,
+      "loss": 0.0355,
+      "step": 9863
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.6171469551718765,
+      "learning_rate": 2.7235115005356913e-07,
+      "loss": 0.0444,
+      "step": 9864
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.5979316493374738,
+      "learning_rate": 2.718716197911098e-07,
+      "loss": 0.0417,
+      "step": 9865
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.6841265472287308,
+      "learning_rate": 2.713925002583223e-07,
+      "loss": 0.0459,
+      "step": 9866
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.65455171703717,
+      "learning_rate": 2.7091379149682683e-07,
+      "loss": 0.0518,
+      "step": 9867
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.6998033465079585,
+      "learning_rate": 2.704354935482095e-07,
+      "loss": 0.0467,
+      "step": 9868
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.593785477475419,
+      "learning_rate": 2.699576064540188e-07,
+      "loss": 0.0432,
+      "step": 9869
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.4622081571625656,
+      "learning_rate": 2.6948013025576927e-07,
+      "loss": 0.0433,
+      "step": 9870
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.562749547860653,
+      "learning_rate": 2.6900306499493875e-07,
+      "loss": 0.0485,
+      "step": 9871
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.7302400372988471,
+      "learning_rate": 2.685264107129698e-07,
+      "loss": 0.0422,
+      "step": 9872
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.4753895557590084,
+      "learning_rate": 2.680501674512681e-07,
+      "loss": 0.0415,
+      "step": 9873
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5588022569303843,
+      "learning_rate": 2.675743352512061e-07,
+      "loss": 0.0484,
+      "step": 9874
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.0242744658091154,
+      "learning_rate": 2.6709891415411747e-07,
+      "loss": 0.048,
+      "step": 9875
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5001127430636434,
+      "learning_rate": 2.66623904201303e-07,
+      "loss": 0.0453,
+      "step": 9876
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.4175396211241404,
+      "learning_rate": 2.661493054340264e-07,
+      "loss": 0.0398,
+      "step": 9877
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5714798136902872,
+      "learning_rate": 2.656751178935146e-07,
+      "loss": 0.0461,
+      "step": 9878
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5532353574468123,
+      "learning_rate": 2.65201341620962e-07,
+      "loss": 0.0423,
+      "step": 9879
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5500560452724461,
+      "learning_rate": 2.647279766575228e-07,
+      "loss": 0.0442,
+      "step": 9880
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.2652048926676074,
+      "learning_rate": 2.6425502304432027e-07,
+      "loss": 0.037,
+      "step": 9881
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5625283143804822,
+      "learning_rate": 2.637824808224382e-07,
+      "loss": 0.0475,
+      "step": 9882
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.2854026331583217,
+      "learning_rate": 2.633103500329276e-07,
+      "loss": 0.0333,
+      "step": 9883
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.4775010185234378,
+      "learning_rate": 2.628386307167996e-07,
+      "loss": 0.0426,
+      "step": 9884
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.3513125354332844,
+      "learning_rate": 2.623673229150342e-07,
+      "loss": 0.0368,
+      "step": 9885
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.3130380737818335,
+      "learning_rate": 2.618964266685725e-07,
+      "loss": 0.0415,
+      "step": 9886
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5634218143317626,
+      "learning_rate": 2.6142594201832183e-07,
+      "loss": 0.0432,
+      "step": 9887
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5746765441141195,
+      "learning_rate": 2.6095586900515226e-07,
+      "loss": 0.0448,
+      "step": 9888
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.4473319624113612,
+      "learning_rate": 2.604862076699005e-07,
+      "loss": 0.0455,
+      "step": 9889
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.2966509031495084,
+      "learning_rate": 2.600169580533629e-07,
+      "loss": 0.0354,
+      "step": 9890
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.4992169509356985,
+      "learning_rate": 2.5954812019630515e-07,
+      "loss": 0.0473,
+      "step": 9891
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.3649809612611266,
+      "learning_rate": 2.59079694139453e-07,
+      "loss": 0.0443,
+      "step": 9892
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.388705188417682,
+      "learning_rate": 2.5861167992350055e-07,
+      "loss": 0.0425,
+      "step": 9893
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.3531723257502692,
+      "learning_rate": 2.5814407758910144e-07,
+      "loss": 0.0397,
+      "step": 9894
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.3001448137766514,
+      "learning_rate": 2.576768871768792e-07,
+      "loss": 0.0395,
+      "step": 9895
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.3223272335887373,
+      "learning_rate": 2.5721010872741536e-07,
+      "loss": 0.0414,
+      "step": 9896
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.4408639068099598,
+      "learning_rate": 2.567437422812602e-07,
+      "loss": 0.0493,
+      "step": 9897
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.72025937387094,
+      "learning_rate": 2.562777878789258e-07,
+      "loss": 0.0538,
+      "step": 9898
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.3346530631338769,
+      "learning_rate": 2.5581224556089024e-07,
+      "loss": 0.0368,
+      "step": 9899
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5658326787347971,
+      "learning_rate": 2.55347115367594e-07,
+      "loss": 0.046,
+      "step": 9900
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.4959811682537567,
+      "learning_rate": 2.548823973394449e-07,
+      "loss": 0.045,
+      "step": 9901
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5733053667046126,
+      "learning_rate": 2.544180915168093e-07,
+      "loss": 0.0464,
+      "step": 9902
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.654203123008192,
+      "learning_rate": 2.539541979400234e-07,
+      "loss": 0.0456,
+      "step": 9903
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.3246762189597465,
+      "learning_rate": 2.534907166493844e-07,
+      "loss": 0.0389,
+      "step": 9904
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5533663189290856,
+      "learning_rate": 2.530276476851562e-07,
+      "loss": 0.0416,
+      "step": 9905
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.2753222478415875,
+      "learning_rate": 2.525649910875627e-07,
+      "loss": 0.0397,
+      "step": 9906
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.483551393887903,
+      "learning_rate": 2.5210274689679793e-07,
+      "loss": 0.0381,
+      "step": 9907
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.3708261780449744,
+      "learning_rate": 2.5164091515301357e-07,
+      "loss": 0.0393,
+      "step": 9908
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.410171594673024,
+      "learning_rate": 2.511794958963309e-07,
+      "loss": 0.032,
+      "step": 9909
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.363809843280214,
+      "learning_rate": 2.507184891668313e-07,
+      "loss": 0.0394,
+      "step": 9910
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.5508238729801465,
+      "learning_rate": 2.502578950045642e-07,
+      "loss": 0.0484,
+      "step": 9911
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.2011950846821335,
+      "learning_rate": 2.4979771344953885e-07,
+      "loss": 0.0392,
+      "step": 9912
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.4010435126518643,
+      "learning_rate": 2.493379445417338e-07,
+      "loss": 0.0383,
+      "step": 9913
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.3276900789666153,
+      "learning_rate": 2.488785883210859e-07,
+      "loss": 0.0376,
+      "step": 9914
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.4916243828279114,
+      "learning_rate": 2.4841964482750114e-07,
+      "loss": 0.0465,
+      "step": 9915
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.357624878923833,
+      "learning_rate": 2.479611141008459e-07,
+      "loss": 0.0425,
+      "step": 9916
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.3952131153832825,
+      "learning_rate": 2.4750299618095496e-07,
+      "loss": 0.0465,
+      "step": 9917
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.6395275475593913,
+      "learning_rate": 2.470452911076227e-07,
+      "loss": 0.0427,
+      "step": 9918
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.6684740405017038,
+      "learning_rate": 2.46587998920611e-07,
+      "loss": 0.047,
+      "step": 9919
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.423408571455933,
+      "learning_rate": 2.461311196596433e-07,
+      "loss": 0.0419,
+      "step": 9920
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.3668621474877518,
+      "learning_rate": 2.4567465336440945e-07,
+      "loss": 0.0403,
+      "step": 9921
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.519298234099257,
+      "learning_rate": 2.4521860007456153e-07,
+      "loss": 0.045,
+      "step": 9922
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.434028095497877,
+      "learning_rate": 2.4476295982971744e-07,
+      "loss": 0.039,
+      "step": 9923
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.6306078409431466,
+      "learning_rate": 2.443077326694582e-07,
+      "loss": 0.0487,
+      "step": 9924
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.2544989117559147,
+      "learning_rate": 2.438529186333288e-07,
+      "loss": 0.0389,
+      "step": 9925
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.3657880574944892,
+      "learning_rate": 2.4339851776083833e-07,
+      "loss": 0.0394,
+      "step": 9926
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.4559968811636415,
+      "learning_rate": 2.4294453009146124e-07,
+      "loss": 0.0413,
+      "step": 9927
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.2071863618966017,
+      "learning_rate": 2.424909556646343e-07,
+      "loss": 0.0351,
+      "step": 9928
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.4972927043943733,
+      "learning_rate": 2.4203779451975996e-07,
+      "loss": 0.0453,
+      "step": 9929
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.687264033521166,
+      "learning_rate": 2.415850466962044e-07,
+      "loss": 0.0482,
+      "step": 9930
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.5271501050292777,
+      "learning_rate": 2.4113271223329625e-07,
+      "loss": 0.0469,
+      "step": 9931
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.6565599146316645,
+      "learning_rate": 2.4068079117033014e-07,
+      "loss": 0.0489,
+      "step": 9932
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.5123902046640314,
+      "learning_rate": 2.402292835465647e-07,
+      "loss": 0.0473,
+      "step": 9933
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.6855982618787908,
+      "learning_rate": 2.3977818940122076e-07,
+      "loss": 0.0479,
+      "step": 9934
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.6928475087006483,
+      "learning_rate": 2.393275087734864e-07,
+      "loss": 0.0539,
+      "step": 9935
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.3913679248040156,
+      "learning_rate": 2.3887724170251094e-07,
+      "loss": 0.0446,
+      "step": 9936
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.3818552292966644,
+      "learning_rate": 2.384273882274091e-07,
+      "loss": 0.0443,
+      "step": 9937
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.5723582684192356,
+      "learning_rate": 2.3797794838725853e-07,
+      "loss": 0.0481,
+      "step": 9938
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.435776775141003,
+      "learning_rate": 2.37528922221103e-07,
+      "loss": 0.0465,
+      "step": 9939
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.3581773442951404,
+      "learning_rate": 2.370803097679486e-07,
+      "loss": 0.0406,
+      "step": 9940
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.7031672105371836,
+      "learning_rate": 2.3663211106676632e-07,
+      "loss": 0.0524,
+      "step": 9941
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.2076553577210025,
+      "learning_rate": 2.3618432615649057e-07,
+      "loss": 0.0346,
+      "step": 9942
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.5094552801578587,
+      "learning_rate": 2.3573695507602024e-07,
+      "loss": 0.0435,
+      "step": 9943
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.3999864830432405,
+      "learning_rate": 2.3528999786421758e-07,
+      "loss": 0.0451,
+      "step": 9944
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.6972659242623396,
+      "learning_rate": 2.3484345455991042e-07,
+      "loss": 0.0469,
+      "step": 9945
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.3549589868860779,
+      "learning_rate": 2.343973252018894e-07,
+      "loss": 0.0372,
+      "step": 9946
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.2106799489140474,
+      "learning_rate": 2.3395160982890963e-07,
+      "loss": 0.0365,
+      "step": 9947
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.3687189059977076,
+      "learning_rate": 2.335063084796907e-07,
+      "loss": 0.0394,
+      "step": 9948
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.4694079445240011,
+      "learning_rate": 2.3306142119291442e-07,
+      "loss": 0.0395,
+      "step": 9949
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.5471581548787146,
+      "learning_rate": 2.3261694800722767e-07,
+      "loss": 0.0468,
+      "step": 9950
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.4013040981637193,
+      "learning_rate": 2.3217288896124347e-07,
+      "loss": 0.0427,
+      "step": 9951
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.3468518205794058,
+      "learning_rate": 2.317292440935348e-07,
+      "loss": 0.0407,
+      "step": 9952
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.3354133139502393,
+      "learning_rate": 2.3128601344264257e-07,
+      "loss": 0.0443,
+      "step": 9953
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.5329044954697,
+      "learning_rate": 2.3084319704706925e-07,
+      "loss": 0.0405,
+      "step": 9954
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.6971678156240182,
+      "learning_rate": 2.3040079494528244e-07,
+      "loss": 0.0489,
+      "step": 9955
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.3771926469195823,
+      "learning_rate": 2.2995880717571195e-07,
+      "loss": 0.0385,
+      "step": 9956
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.4183147383098516,
+      "learning_rate": 2.2951723377675484e-07,
+      "loss": 0.0397,
+      "step": 9957
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.5905894433028311,
+      "learning_rate": 2.2907607478676818e-07,
+      "loss": 0.0459,
+      "step": 9958
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.7465748885705494,
+      "learning_rate": 2.28635330244078e-07,
+      "loss": 0.045,
+      "step": 9959
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.4416913307087156,
+      "learning_rate": 2.2819500018696927e-07,
+      "loss": 0.042,
+      "step": 9960
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.878288101385799,
+      "learning_rate": 2.277550846536941e-07,
+      "loss": 0.0431,
+      "step": 9961
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.62103140011818,
+      "learning_rate": 2.2731558368246698e-07,
+      "loss": 0.0456,
+      "step": 9962
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.3686837715792661,
+      "learning_rate": 2.2687649731146844e-07,
+      "loss": 0.043,
+      "step": 9963
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.5648244152736621,
+      "learning_rate": 2.264378255788402e-07,
+      "loss": 0.0508,
+      "step": 9964
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.498949245338006,
+      "learning_rate": 2.2599956852269067e-07,
+      "loss": 0.0523,
+      "step": 9965
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.7187470608975022,
+      "learning_rate": 2.2556172618108996e-07,
+      "loss": 0.0406,
+      "step": 9966
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.3217549152220947,
+      "learning_rate": 2.2512429859207375e-07,
+      "loss": 0.0389,
+      "step": 9967
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.5951556901078363,
+      "learning_rate": 2.2468728579363997e-07,
+      "loss": 0.0426,
+      "step": 9968
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.2124102859218526,
+      "learning_rate": 2.242506878237538e-07,
+      "loss": 0.0292,
+      "step": 9969
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.7050345942832428,
+      "learning_rate": 2.2381450472033995e-07,
+      "loss": 0.0485,
+      "step": 9970
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.4818539513574767,
+      "learning_rate": 2.2337873652129084e-07,
+      "loss": 0.039,
+      "step": 9971
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.3533995033261557,
+      "learning_rate": 2.229433832644623e-07,
+      "loss": 0.0407,
+      "step": 9972
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.6947179935281724,
+      "learning_rate": 2.2250844498767077e-07,
+      "loss": 0.0438,
+      "step": 9973
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.5884973453111142,
+      "learning_rate": 2.2207392172870047e-07,
+      "loss": 0.0462,
+      "step": 9974
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.1457296533193817,
+      "learning_rate": 2.2163981352529728e-07,
+      "loss": 0.034,
+      "step": 9975
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.4100688335896259,
+      "learning_rate": 2.2120612041517387e-07,
+      "loss": 0.0426,
+      "step": 9976
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.8737929615612752,
+      "learning_rate": 2.2077284243600227e-07,
+      "loss": 0.0498,
+      "step": 9977
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.469506257465614,
+      "learning_rate": 2.203399796254241e-07,
+      "loss": 0.0522,
+      "step": 9978
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.421875687400047,
+      "learning_rate": 2.199075320210392e-07,
+      "loss": 0.0427,
+      "step": 9979
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.427953189732144,
+      "learning_rate": 2.1947549966041537e-07,
+      "loss": 0.0439,
+      "step": 9980
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.3106523566682518,
+      "learning_rate": 2.19043882581082e-07,
+      "loss": 0.0411,
+      "step": 9981
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.377868047063109,
+      "learning_rate": 2.1861268082053466e-07,
+      "loss": 0.0454,
+      "step": 9982
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.53894493699053,
+      "learning_rate": 2.1818189441623061e-07,
+      "loss": 0.0437,
+      "step": 9983
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.2742909729142209,
+      "learning_rate": 2.1775152340559325e-07,
+      "loss": 0.0412,
+      "step": 9984
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.3057018754904868,
+      "learning_rate": 2.173215678260071e-07,
+      "loss": 0.0391,
+      "step": 9985
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.6073207467725237,
+      "learning_rate": 2.1689202771482344e-07,
+      "loss": 0.0472,
+      "step": 9986
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.491532424567738,
+      "learning_rate": 2.164629031093546e-07,
+      "loss": 0.0415,
+      "step": 9987
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4890768238619554,
+      "learning_rate": 2.160341940468802e-07,
+      "loss": 0.0352,
+      "step": 9988
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.286049031602354,
+      "learning_rate": 2.156059005646405e-07,
+      "loss": 0.0357,
+      "step": 9989
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.5091075252335213,
+      "learning_rate": 2.151780226998429e-07,
+      "loss": 0.0506,
+      "step": 9990
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.3155509652951816,
+      "learning_rate": 2.1475056048965437e-07,
+      "loss": 0.0392,
+      "step": 9991
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4985696667725794,
+      "learning_rate": 2.1432351397121021e-07,
+      "loss": 0.0407,
+      "step": 9992
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.2810530446134696,
+      "learning_rate": 2.1389688318160683e-07,
+      "loss": 0.0432,
+      "step": 9993
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.3036727045533048,
+      "learning_rate": 2.1347066815790574e-07,
+      "loss": 0.0373,
+      "step": 9994
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.385305085361356,
+      "learning_rate": 2.1304486893713172e-07,
+      "loss": 0.0457,
+      "step": 9995
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4327354350419996,
+      "learning_rate": 2.1261948555627464e-07,
+      "loss": 0.0434,
+      "step": 9996
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.529576686493087,
+      "learning_rate": 2.1219451805228607e-07,
+      "loss": 0.0469,
+      "step": 9997
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.5254670769013934,
+      "learning_rate": 2.1176996646208313e-07,
+      "loss": 0.0394,
+      "step": 9998
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.577544313339664,
+      "learning_rate": 2.113458308225458e-07,
+      "loss": 0.053,
+      "step": 9999
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.3262685690873348,
+      "learning_rate": 2.109221111705201e-07,
+      "loss": 0.0423,
+      "step": 10000
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4070531625072271,
+      "learning_rate": 2.104988075428127e-07,
+      "loss": 0.0343,
+      "step": 10001
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.3782776834064139,
+      "learning_rate": 2.1007591997619703e-07,
+      "loss": 0.0419,
+      "step": 10002
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4366752438404735,
+      "learning_rate": 2.0965344850740698e-07,
+      "loss": 0.0357,
+      "step": 10003
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.5937770558453033,
+      "learning_rate": 2.092313931731449e-07,
+      "loss": 0.0446,
+      "step": 10004
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.8474194095435017,
+      "learning_rate": 2.0880975401007253e-07,
+      "loss": 0.0484,
+      "step": 10005
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.616401014842261,
+      "learning_rate": 2.0838853105481838e-07,
+      "loss": 0.0466,
+      "step": 10006
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.5626912153334243,
+      "learning_rate": 2.079677243439743e-07,
+      "loss": 0.0478,
+      "step": 10007
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4043662609913272,
+      "learning_rate": 2.0754733391409486e-07,
+      "loss": 0.0483,
+      "step": 10008
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4271267332321544,
+      "learning_rate": 2.0712735980169819e-07,
+      "loss": 0.0409,
+      "step": 10009
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.317669235773752,
+      "learning_rate": 2.067078020432689e-07,
+      "loss": 0.0398,
+      "step": 10010
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.2652711012815854,
+      "learning_rate": 2.0628866067525288e-07,
+      "loss": 0.0362,
+      "step": 10011
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.2568785258001451,
+      "learning_rate": 2.05869935734061e-07,
+      "loss": 0.0405,
+      "step": 10012
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4448691240660754,
+      "learning_rate": 2.0545162725606693e-07,
+      "loss": 0.039,
+      "step": 10013
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.6647068135034264,
+      "learning_rate": 2.0503373527760994e-07,
+      "loss": 0.0512,
+      "step": 10014
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4890586533894707,
+      "learning_rate": 2.04616259834991e-07,
+      "loss": 0.042,
+      "step": 10015
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.5012071702584902,
+      "learning_rate": 2.0419920096447666e-07,
+      "loss": 0.0415,
+      "step": 10016
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4980712082252385,
+      "learning_rate": 2.0378255870229625e-07,
+      "loss": 0.0531,
+      "step": 10017
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.7202242107131056,
+      "learning_rate": 2.033663330846436e-07,
+      "loss": 0.0493,
+      "step": 10018
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.7811727008363314,
+      "learning_rate": 2.0295052414767535e-07,
+      "loss": 0.0449,
+      "step": 10019
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.2868436309825144,
+      "learning_rate": 2.0253513192751374e-07,
+      "loss": 0.035,
+      "step": 10020
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.6101278480111838,
+      "learning_rate": 2.0212015646024152e-07,
+      "loss": 0.0493,
+      "step": 10021
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.595526022662632,
+      "learning_rate": 2.017055977819099e-07,
+      "loss": 0.0502,
+      "step": 10022
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.654964965433621,
+      "learning_rate": 2.0129145592852893e-07,
+      "loss": 0.0485,
+      "step": 10023
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.5614816234156104,
+      "learning_rate": 2.008777309360771e-07,
+      "loss": 0.0535,
+      "step": 10024
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.50195589348083,
+      "learning_rate": 2.0046442284049339e-07,
+      "loss": 0.0426,
+      "step": 10025
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.3773806432243265,
+      "learning_rate": 2.0005153167768133e-07,
+      "loss": 0.043,
+      "step": 10026
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.4055825170233591,
+      "learning_rate": 1.9963905748350888e-07,
+      "loss": 0.0455,
+      "step": 10027
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.2510483783281026,
+      "learning_rate": 1.9922700029380737e-07,
+      "loss": 0.0335,
+      "step": 10028
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.4439127659282065,
+      "learning_rate": 1.9881536014437153e-07,
+      "loss": 0.0479,
+      "step": 10029
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.3495244636676933,
+      "learning_rate": 1.9840413707096162e-07,
+      "loss": 0.0383,
+      "step": 10030
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.330462975670874,
+      "learning_rate": 1.9799333110929907e-07,
+      "loss": 0.0462,
+      "step": 10031
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.4918709077584789,
+      "learning_rate": 1.9758294229507092e-07,
+      "loss": 0.0522,
+      "step": 10032
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.340969991679822,
+      "learning_rate": 1.9717297066392638e-07,
+      "loss": 0.0416,
+      "step": 10033
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.3934030593697044,
+      "learning_rate": 1.9676341625148144e-07,
+      "loss": 0.0311,
+      "step": 10034
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.4321468062799512,
+      "learning_rate": 1.963542790933115e-07,
+      "loss": 0.0398,
+      "step": 10035
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.3765965114976613,
+      "learning_rate": 1.959455592249604e-07,
+      "loss": 0.0386,
+      "step": 10036
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.4293559334957437,
+      "learning_rate": 1.9553725668193192e-07,
+      "loss": 0.041,
+      "step": 10037
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.4517457291569,
+      "learning_rate": 1.9512937149969546e-07,
+      "loss": 0.0388,
+      "step": 10038
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.2496325312482004,
+      "learning_rate": 1.947219037136827e-07,
+      "loss": 0.0359,
+      "step": 10039
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.2256996888267322,
+      "learning_rate": 1.94314853359292e-07,
+      "loss": 0.0368,
+      "step": 10040
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.6438257390620437,
+      "learning_rate": 1.939082204718823e-07,
+      "loss": 0.0524,
+      "step": 10041
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.4835166261340051,
+      "learning_rate": 1.9350200508677863e-07,
+      "loss": 0.044,
+      "step": 10042
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.5909502228181926,
+      "learning_rate": 1.9309620723926725e-07,
+      "loss": 0.0513,
+      "step": 10043
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.3909438929297275,
+      "learning_rate": 1.9269082696460106e-07,
+      "loss": 0.0443,
+      "step": 10044
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.669266724077017,
+      "learning_rate": 1.9228586429799356e-07,
+      "loss": 0.0457,
+      "step": 10045
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.6168068068434518,
+      "learning_rate": 1.9188131927462493e-07,
+      "loss": 0.0548,
+      "step": 10046
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.5149264983368511,
+      "learning_rate": 1.9147719192963655e-07,
+      "loss": 0.0442,
+      "step": 10047
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.5974634272137342,
+      "learning_rate": 1.910734822981364e-07,
+      "loss": 0.0441,
+      "step": 10048
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.4004483993784096,
+      "learning_rate": 1.9067019041519363e-07,
+      "loss": 0.0402,
+      "step": 10049
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.1847403296842942,
+      "learning_rate": 1.9026731631584194e-07,
+      "loss": 0.0333,
+      "step": 10050
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.5837068412913442,
+      "learning_rate": 1.8986486003507776e-07,
+      "loss": 0.048,
+      "step": 10051
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.5652989833217095,
+      "learning_rate": 1.8946282160786421e-07,
+      "loss": 0.0497,
+      "step": 10052
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.281581513093705,
+      "learning_rate": 1.8906120106912452e-07,
+      "loss": 0.0379,
+      "step": 10053
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.410325844236411,
+      "learning_rate": 1.8865999845374794e-07,
+      "loss": 0.0445,
+      "step": 10054
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.3005625137015098,
+      "learning_rate": 1.8825921379658718e-07,
+      "loss": 0.0404,
+      "step": 10055
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.511000452274161,
+      "learning_rate": 1.8785884713245718e-07,
+      "loss": 0.0453,
+      "step": 10056
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.2323975640121196,
+      "learning_rate": 1.8745889849613786e-07,
+      "loss": 0.0351,
+      "step": 10057
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.517902785221159,
+      "learning_rate": 1.8705936792237255e-07,
+      "loss": 0.0508,
+      "step": 10058
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.4283797107413627,
+      "learning_rate": 1.8666025544586796e-07,
+      "loss": 0.0391,
+      "step": 10059
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.745277829113621,
+      "learning_rate": 1.862615611012958e-07,
+      "loss": 0.0517,
+      "step": 10060
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.289379322109334,
+      "learning_rate": 1.8586328492328942e-07,
+      "loss": 0.0417,
+      "step": 10061
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.3214675718049216,
+      "learning_rate": 1.854654269464473e-07,
+      "loss": 0.0329,
+      "step": 10062
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.440840097146051,
+      "learning_rate": 1.8506798720533014e-07,
+      "loss": 0.0402,
+      "step": 10063
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.5255387496552482,
+      "learning_rate": 1.8467096573446418e-07,
+      "loss": 0.0429,
+      "step": 10064
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.4651518674310213,
+      "learning_rate": 1.8427436256833853e-07,
+      "loss": 0.0393,
+      "step": 10065
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.2504394114205686,
+      "learning_rate": 1.838781777414056e-07,
+      "loss": 0.0384,
+      "step": 10066
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.3723061421069038,
+      "learning_rate": 1.8348241128808285e-07,
+      "loss": 0.0424,
+      "step": 10067
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.2943048882035808,
+      "learning_rate": 1.8308706324274783e-07,
+      "loss": 0.0381,
+      "step": 10068
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.269121210348574,
+      "learning_rate": 1.8269213363974637e-07,
+      "loss": 0.0337,
+      "step": 10069
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.645592678855142,
+      "learning_rate": 1.822976225133838e-07,
+      "loss": 0.0471,
+      "step": 10070
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.7757937018933823,
+      "learning_rate": 1.8190352989793325e-07,
+      "loss": 0.0601,
+      "step": 10071
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.494229272229063,
+      "learning_rate": 1.8150985582762792e-07,
+      "loss": 0.0424,
+      "step": 10072
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.2880897944663652,
+      "learning_rate": 1.8111660033666767e-07,
+      "loss": 0.039,
+      "step": 10073
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.515622071739428,
+      "learning_rate": 1.8072376345921127e-07,
+      "loss": 0.0432,
+      "step": 10074
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.4881606835057053,
+      "learning_rate": 1.8033134522938701e-07,
+      "loss": 0.0374,
+      "step": 10075
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.5407343913258063,
+      "learning_rate": 1.7993934568128256e-07,
+      "loss": 0.0453,
+      "step": 10076
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.4138531751618018,
+      "learning_rate": 1.7954776484895188e-07,
+      "loss": 0.0353,
+      "step": 10077
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.5924006685294518,
+      "learning_rate": 1.7915660276641045e-07,
+      "loss": 0.0452,
+      "step": 10078
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.5119785026649646,
+      "learning_rate": 1.7876585946763892e-07,
+      "loss": 0.045,
+      "step": 10079
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.5187261686703846,
+      "learning_rate": 1.7837553498657955e-07,
+      "loss": 0.0511,
+      "step": 10080
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.413272087148863,
+      "learning_rate": 1.7798562935714082e-07,
+      "loss": 0.0372,
+      "step": 10081
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.5096797310966612,
+      "learning_rate": 1.7759614261319337e-07,
+      "loss": 0.0461,
+      "step": 10082
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.5166389965927052,
+      "learning_rate": 1.772070747885718e-07,
+      "loss": 0.044,
+      "step": 10083
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.4355471934432087,
+      "learning_rate": 1.7681842591707465e-07,
+      "loss": 0.0378,
+      "step": 10084
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.5176718536389435,
+      "learning_rate": 1.764301960324627e-07,
+      "loss": 0.0414,
+      "step": 10085
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.5046547218333923,
+      "learning_rate": 1.7604238516846062e-07,
+      "loss": 0.0437,
+      "step": 10086
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.5010675303381087,
+      "learning_rate": 1.7565499335875924e-07,
+      "loss": 0.0392,
+      "step": 10087
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.5329800147829753,
+      "learning_rate": 1.7526802063700943e-07,
+      "loss": 0.0461,
+      "step": 10088
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.323494455746387,
+      "learning_rate": 1.748814670368282e-07,
+      "loss": 0.0348,
+      "step": 10089
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.4984739245776257,
+      "learning_rate": 1.744953325917953e-07,
+      "loss": 0.0426,
+      "step": 10090
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.4033968647536448,
+      "learning_rate": 1.741096173354534e-07,
+      "loss": 0.0396,
+      "step": 10091
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.3598023944094868,
+      "learning_rate": 1.7372432130130955e-07,
+      "loss": 0.0401,
+      "step": 10092
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.5191944219745634,
+      "learning_rate": 1.7333944452283425e-07,
+      "loss": 0.0458,
+      "step": 10093
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.6974850785826827,
+      "learning_rate": 1.729549870334607e-07,
+      "loss": 0.0514,
+      "step": 10094
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.082311200400368,
+      "learning_rate": 1.725709488665883e-07,
+      "loss": 0.0465,
+      "step": 10095
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.5926724684163749,
+      "learning_rate": 1.7218733005557707e-07,
+      "loss": 0.0494,
+      "step": 10096
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.432621228268689,
+      "learning_rate": 1.71804130633752e-07,
+      "loss": 0.0427,
+      "step": 10097
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.3469829644280789,
+      "learning_rate": 1.7142135063440034e-07,
+      "loss": 0.0433,
+      "step": 10098
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.5759139631588548,
+      "learning_rate": 1.7103899009077606e-07,
+      "loss": 0.0486,
+      "step": 10099
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.5959131723318278,
+      "learning_rate": 1.7065704903609259e-07,
+      "loss": 0.0461,
+      "step": 10100
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.313403489637816,
+      "learning_rate": 1.7027552750353005e-07,
+      "loss": 0.0355,
+      "step": 10101
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.4811476128264136,
+      "learning_rate": 1.6989442552623082e-07,
+      "loss": 0.0499,
+      "step": 10102
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.3768012253187047,
+      "learning_rate": 1.695137431373006e-07,
+      "loss": 0.039,
+      "step": 10103
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.4296769551684576,
+      "learning_rate": 1.6913348036980914e-07,
+      "loss": 0.0447,
+      "step": 10104
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.555174933504298,
+      "learning_rate": 1.6875363725679052e-07,
+      "loss": 0.0499,
+      "step": 10105
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.6434851184884691,
+      "learning_rate": 1.683742138312394e-07,
+      "loss": 0.0439,
+      "step": 10106
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.5512538620189686,
+      "learning_rate": 1.6799521012611843e-07,
+      "loss": 0.0452,
+      "step": 10107
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.4046692988104377,
+      "learning_rate": 1.676166261743506e-07,
+      "loss": 0.0425,
+      "step": 10108
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.6908210894622395,
+      "learning_rate": 1.67238462008823e-07,
+      "loss": 0.0542,
+      "step": 10109
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.6627767438976886,
+      "learning_rate": 1.66860717662386e-07,
+      "loss": 0.0431,
+      "step": 10110
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.7289206635384784,
+      "learning_rate": 1.6648339316785556e-07,
+      "loss": 0.0427,
+      "step": 10111
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.4980696733457581,
+      "learning_rate": 1.6610648855800772e-07,
+      "loss": 0.0415,
+      "step": 10112
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.5352621398930482,
+      "learning_rate": 1.657300038655857e-07,
+      "loss": 0.043,
+      "step": 10113
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.3654952467606503,
+      "learning_rate": 1.6535393912329388e-07,
+      "loss": 0.0388,
+      "step": 10114
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.2492204846769608,
+      "learning_rate": 1.6497829436380009e-07,
+      "loss": 0.0384,
+      "step": 10115
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.5246298453278262,
+      "learning_rate": 1.6460306961973705e-07,
+      "loss": 0.0444,
+      "step": 10116
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.5581795102414493,
+      "learning_rate": 1.6422826492370037e-07,
+      "loss": 0.05,
+      "step": 10117
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.3456830726366065,
+      "learning_rate": 1.6385388030824844e-07,
+      "loss": 0.0341,
+      "step": 10118
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.3718080844963219,
+      "learning_rate": 1.6347991580590472e-07,
+      "loss": 0.0508,
+      "step": 10119
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.6513661442086138,
+      "learning_rate": 1.6310637144915542e-07,
+      "loss": 0.0527,
+      "step": 10120
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.316832159190448,
+      "learning_rate": 1.6273324727044905e-07,
+      "loss": 0.0394,
+      "step": 10121
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.2826645293382917,
+      "learning_rate": 1.6236054330219853e-07,
+      "loss": 0.0348,
+      "step": 10122
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.3530485335628175,
+      "learning_rate": 1.619882595767819e-07,
+      "loss": 0.043,
+      "step": 10123
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.58148363927937,
+      "learning_rate": 1.6161639612653824e-07,
+      "loss": 0.0485,
+      "step": 10124
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.3947319197264723,
+      "learning_rate": 1.612449529837712e-07,
+      "loss": 0.0365,
+      "step": 10125
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.7408314190435925,
+      "learning_rate": 1.6087393018074825e-07,
+      "loss": 0.0492,
+      "step": 10126
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.5687440109595219,
+      "learning_rate": 1.605033277496998e-07,
+      "loss": 0.0471,
+      "step": 10127
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.352200757766226,
+      "learning_rate": 1.601331457228189e-07,
+      "loss": 0.0412,
+      "step": 10128
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4702556117474788,
+      "learning_rate": 1.597633841322638e-07,
+      "loss": 0.0409,
+      "step": 10129
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.697997663766319,
+      "learning_rate": 1.5939404301015537e-07,
+      "loss": 0.0403,
+      "step": 10130
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.5204800355367796,
+      "learning_rate": 1.590251223885786e-07,
+      "loss": 0.0486,
+      "step": 10131
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4234644478681038,
+      "learning_rate": 1.5865662229958112e-07,
+      "loss": 0.0428,
+      "step": 10132
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4673909107302767,
+      "learning_rate": 1.5828854277517404e-07,
+      "loss": 0.0397,
+      "step": 10133
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.6504606247051161,
+      "learning_rate": 1.5792088384733174e-07,
+      "loss": 0.0417,
+      "step": 10134
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4207983775597695,
+      "learning_rate": 1.5755364554799367e-07,
+      "loss": 0.0353,
+      "step": 10135
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4866392758209426,
+      "learning_rate": 1.5718682790906048e-07,
+      "loss": 0.0448,
+      "step": 10136
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.3952214728060652,
+      "learning_rate": 1.568204309623983e-07,
+      "loss": 0.0414,
+      "step": 10137
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4779424049223246,
+      "learning_rate": 1.5645445473983557e-07,
+      "loss": 0.0468,
+      "step": 10138
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4092837774680038,
+      "learning_rate": 1.5608889927316407e-07,
+      "loss": 0.04,
+      "step": 10139
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.6254130604515964,
+      "learning_rate": 1.5572376459413897e-07,
+      "loss": 0.045,
+      "step": 10140
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.3558297525643035,
+      "learning_rate": 1.55359050734481e-07,
+      "loss": 0.0402,
+      "step": 10141
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.384797794915866,
+      "learning_rate": 1.549947577258709e-07,
+      "loss": 0.0393,
+      "step": 10142
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.3470437893134084,
+      "learning_rate": 1.5463088559995564e-07,
+      "loss": 0.0356,
+      "step": 10143
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.3570426854956028,
+      "learning_rate": 1.5426743438834436e-07,
+      "loss": 0.0437,
+      "step": 10144
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.516786162655479,
+      "learning_rate": 1.5390440412260954e-07,
+      "loss": 0.0415,
+      "step": 10145
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.2551763850367763,
+      "learning_rate": 1.535417948342871e-07,
+      "loss": 0.0312,
+      "step": 10146
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.365626797717536,
+      "learning_rate": 1.531796065548774e-07,
+      "loss": 0.0444,
+      "step": 10147
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.3184970338463724,
+      "learning_rate": 1.5281783931584303e-07,
+      "loss": 0.0416,
+      "step": 10148
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.3837442956832846,
+      "learning_rate": 1.524564931486111e-07,
+      "loss": 0.0384,
+      "step": 10149
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.418513100036654,
+      "learning_rate": 1.5209556808457093e-07,
+      "loss": 0.0338,
+      "step": 10150
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.5049714897159012,
+      "learning_rate": 1.5173506415507632e-07,
+      "loss": 0.0465,
+      "step": 10151
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.413141116128674,
+      "learning_rate": 1.5137498139144336e-07,
+      "loss": 0.0406,
+      "step": 10152
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.6949414870341155,
+      "learning_rate": 1.510153198249531e-07,
+      "loss": 0.0443,
+      "step": 10153
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.356071886152123,
+      "learning_rate": 1.506560794868478e-07,
+      "loss": 0.0423,
+      "step": 10154
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.2555536510251804,
+      "learning_rate": 1.5029726040833638e-07,
+      "loss": 0.0403,
+      "step": 10155
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4800940028116727,
+      "learning_rate": 1.4993886262058833e-07,
+      "loss": 0.0476,
+      "step": 10156
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4765918271488911,
+      "learning_rate": 1.4958088615473598e-07,
+      "loss": 0.0463,
+      "step": 10157
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4769284088971608,
+      "learning_rate": 1.4922333104187892e-07,
+      "loss": 0.052,
+      "step": 10158
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4769669964561707,
+      "learning_rate": 1.4886619731307617e-07,
+      "loss": 0.0463,
+      "step": 10159
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.656253393306451,
+      "learning_rate": 1.485094849993529e-07,
+      "loss": 0.0507,
+      "step": 10160
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.4442834023060762,
+      "learning_rate": 1.481531941316955e-07,
+      "loss": 0.0403,
+      "step": 10161
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.3963816038453938,
+      "learning_rate": 1.4779732474105525e-07,
+      "loss": 0.0456,
+      "step": 10162
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.5250482284270535,
+      "learning_rate": 1.4744187685834576e-07,
+      "loss": 0.0418,
+      "step": 10163
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.2213055780938222,
+      "learning_rate": 1.4708685051444515e-07,
+      "loss": 0.033,
+      "step": 10164
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.4716080702731333,
+      "learning_rate": 1.4673224574019373e-07,
+      "loss": 0.0425,
+      "step": 10165
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.3351171757890703,
+      "learning_rate": 1.4637806256639685e-07,
+      "loss": 0.0334,
+      "step": 10166
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.753367743117955,
+      "learning_rate": 1.460243010238216e-07,
+      "loss": 0.0418,
+      "step": 10167
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.7239401447375389,
+      "learning_rate": 1.4567096114319833e-07,
+      "loss": 0.0544,
+      "step": 10168
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.359980583791488,
+      "learning_rate": 1.4531804295522256e-07,
+      "loss": 0.0385,
+      "step": 10169
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.760715249627928,
+      "learning_rate": 1.449655464905514e-07,
+      "loss": 0.0524,
+      "step": 10170
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.5336253529150983,
+      "learning_rate": 1.4461347177980644e-07,
+      "loss": 0.0441,
+      "step": 10171
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.2444795763847114,
+      "learning_rate": 1.4426181885357215e-07,
+      "loss": 0.0367,
+      "step": 10172
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.3773821883987891,
+      "learning_rate": 1.439105877423963e-07,
+      "loss": 0.0452,
+      "step": 10173
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.6067729145648395,
+      "learning_rate": 1.4355977847679004e-07,
+      "loss": 0.0405,
+      "step": 10174
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.557517681061879,
+      "learning_rate": 1.432093910872273e-07,
+      "loss": 0.0437,
+      "step": 10175
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.6638289089761185,
+      "learning_rate": 1.4285942560414768e-07,
+      "loss": 0.0412,
+      "step": 10176
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.7623261820886194,
+      "learning_rate": 1.4250988205795068e-07,
+      "loss": 0.0444,
+      "step": 10177
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.6768483942338641,
+      "learning_rate": 1.421607604790026e-07,
+      "loss": 0.0506,
+      "step": 10178
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.6491985570417915,
+      "learning_rate": 1.4181206089763033e-07,
+      "loss": 0.0487,
+      "step": 10179
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.3607398917969995,
+      "learning_rate": 1.414637833441257e-07,
+      "loss": 0.0414,
+      "step": 10180
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.2958749745842397,
+      "learning_rate": 1.4111592784874285e-07,
+      "loss": 0.0335,
+      "step": 10181
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.3815602599142944,
+      "learning_rate": 1.4076849444170036e-07,
+      "loss": 0.0403,
+      "step": 10182
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.1496738231043915,
+      "learning_rate": 1.4042148315317862e-07,
+      "loss": 0.0365,
+      "step": 10183
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.392121912593456,
+      "learning_rate": 1.400748940133234e-07,
+      "loss": 0.0442,
+      "step": 10184
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.3662152606491669,
+      "learning_rate": 1.3972872705224238e-07,
+      "loss": 0.0346,
+      "step": 10185
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.5519988603068833,
+      "learning_rate": 1.3938298230000646e-07,
+      "loss": 0.0486,
+      "step": 10186
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.4999385781894288,
+      "learning_rate": 1.3903765978665052e-07,
+      "loss": 0.0451,
+      "step": 10187
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.808520194515675,
+      "learning_rate": 1.3869275954217275e-07,
+      "loss": 0.042,
+      "step": 10188
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.6806842765022694,
+      "learning_rate": 1.3834828159653368e-07,
+      "loss": 0.0466,
+      "step": 10189
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.5624591276295847,
+      "learning_rate": 1.3800422597965935e-07,
+      "loss": 0.0474,
+      "step": 10190
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.4401612323497468,
+      "learning_rate": 1.376605927214364e-07,
+      "loss": 0.0387,
+      "step": 10191
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.3970698920584177,
+      "learning_rate": 1.373173818517165e-07,
+      "loss": 0.0448,
+      "step": 10192
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.6205673605008553,
+      "learning_rate": 1.369745934003136e-07,
+      "loss": 0.049,
+      "step": 10193
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.4896466474769081,
+      "learning_rate": 1.3663222739700665e-07,
+      "loss": 0.0448,
+      "step": 10194
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.5594991319235247,
+      "learning_rate": 1.362902838715352e-07,
+      "loss": 0.0448,
+      "step": 10195
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.4203446493838834,
+      "learning_rate": 1.3594876285360548e-07,
+      "loss": 0.0403,
+      "step": 10196
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.3779027645202082,
+      "learning_rate": 1.3560766437288432e-07,
+      "loss": 0.0391,
+      "step": 10197
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.594925848746612,
+      "learning_rate": 1.3526698845900244e-07,
+      "loss": 0.0513,
+      "step": 10198
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.384964692684855,
+      "learning_rate": 1.3492673514155452e-07,
+      "loss": 0.043,
+      "step": 10199
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.6021753219030994,
+      "learning_rate": 1.3458690445009804e-07,
+      "loss": 0.0444,
+      "step": 10200
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.3658660049668352,
+      "learning_rate": 1.342474964141538e-07,
+      "loss": 0.0414,
+      "step": 10201
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.5958701857333515,
+      "learning_rate": 1.3390851106320656e-07,
+      "loss": 0.0477,
+      "step": 10202
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.645845823444285,
+      "learning_rate": 1.3356994842670335e-07,
+      "loss": 0.053,
+      "step": 10203
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.5876968401139924,
+      "learning_rate": 1.3323180853405504e-07,
+      "loss": 0.0499,
+      "step": 10204
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.3767623368089479,
+      "learning_rate": 1.328940914146354e-07,
+      "loss": 0.0427,
+      "step": 10205
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.324832777356151,
+      "learning_rate": 1.3255679709778148e-07,
+      "loss": 0.034,
+      "step": 10206
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.6321385577055632,
+      "learning_rate": 1.322199256127943e-07,
+      "loss": 0.0483,
+      "step": 10207
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.4608467447509321,
+      "learning_rate": 1.3188347698893767e-07,
+      "loss": 0.0459,
+      "step": 10208
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.1990162029341689,
+      "learning_rate": 1.3154745125543877e-07,
+      "loss": 0.0386,
+      "step": 10209
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.6463833026468588,
+      "learning_rate": 1.312118484414876e-07,
+      "loss": 0.0451,
+      "step": 10210
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.4055817944023774,
+      "learning_rate": 1.308766685762375e-07,
+      "loss": 0.0387,
+      "step": 10211
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.3649988242859281,
+      "learning_rate": 1.3054191168880682e-07,
+      "loss": 0.0345,
+      "step": 10212
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.3899041587005065,
+      "learning_rate": 1.3020757780827343e-07,
+      "loss": 0.0434,
+      "step": 10213
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.2815803586900372,
+      "learning_rate": 1.2987366696368243e-07,
+      "loss": 0.0357,
+      "step": 10214
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.734277136420921,
+      "learning_rate": 1.2954017918404006e-07,
+      "loss": 0.0597,
+      "step": 10215
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.4887782261271665,
+      "learning_rate": 1.2920711449831646e-07,
+      "loss": 0.0493,
+      "step": 10216
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.472048747818957,
+      "learning_rate": 1.2887447293544353e-07,
+      "loss": 0.0415,
+      "step": 10217
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.6798398333796083,
+      "learning_rate": 1.2854225452431923e-07,
+      "loss": 0.0447,
+      "step": 10218
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.3257456437823179,
+      "learning_rate": 1.2821045929380162e-07,
+      "loss": 0.0443,
+      "step": 10219
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.8266360371144905,
+      "learning_rate": 1.2787908727271536e-07,
+      "loss": 0.045,
+      "step": 10220
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.4740762837088641,
+      "learning_rate": 1.2754813848984526e-07,
+      "loss": 0.05,
+      "step": 10221
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.6029878452905846,
+      "learning_rate": 1.2721761297394108e-07,
+      "loss": 0.0467,
+      "step": 10222
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.4101053076620393,
+      "learning_rate": 1.268875107537143e-07,
+      "loss": 0.043,
+      "step": 10223
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.3741626795655153,
+      "learning_rate": 1.2655783185784253e-07,
+      "loss": 0.0376,
+      "step": 10224
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.3306873939405592,
+      "learning_rate": 1.2622857631496344e-07,
+      "loss": 0.0465,
+      "step": 10225
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.4822898465031749,
+      "learning_rate": 1.2589974415367968e-07,
+      "loss": 0.0466,
+      "step": 10226
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.5327377675320248,
+      "learning_rate": 1.2557133540255728e-07,
+      "loss": 0.0452,
+      "step": 10227
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.379059637986405,
+      "learning_rate": 1.25243350090124e-07,
+      "loss": 0.0428,
+      "step": 10228
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.6152608978483405,
+      "learning_rate": 1.2491578824487204e-07,
+      "loss": 0.0459,
+      "step": 10229
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.502272124704391,
+      "learning_rate": 1.24588649895257e-07,
+      "loss": 0.0405,
+      "step": 10230
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.6826585532978697,
+      "learning_rate": 1.2426193506969607e-07,
+      "loss": 0.0471,
+      "step": 10231
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.3696286621718514,
+      "learning_rate": 1.2393564379657163e-07,
+      "loss": 0.0366,
+      "step": 10232
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.601824766206709,
+      "learning_rate": 1.2360977610422874e-07,
+      "loss": 0.0425,
+      "step": 10233
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.1938427245129186,
+      "learning_rate": 1.2328433202097422e-07,
+      "loss": 0.0322,
+      "step": 10234
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.6093412248909347,
+      "learning_rate": 1.229593115750799e-07,
+      "loss": 0.0485,
+      "step": 10235
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.5378740047703965,
+      "learning_rate": 1.2263471479477984e-07,
+      "loss": 0.0426,
+      "step": 10236
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.5700360144658767,
+      "learning_rate": 1.2231054170827205e-07,
+      "loss": 0.0444,
+      "step": 10237
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.352477567417536,
+      "learning_rate": 1.219867923437168e-07,
+      "loss": 0.0376,
+      "step": 10238
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.261433457523689,
+      "learning_rate": 1.2166346672923824e-07,
+      "loss": 0.0407,
+      "step": 10239
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.5882594161659969,
+      "learning_rate": 1.2134056489292335e-07,
+      "loss": 0.051,
+      "step": 10240
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.0624647775817864,
+      "learning_rate": 1.210180868628219e-07,
+      "loss": 0.0448,
+      "step": 10241
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.4300234273668058,
+      "learning_rate": 1.2069603266694873e-07,
+      "loss": 0.0417,
+      "step": 10242
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.524139481692172,
+      "learning_rate": 1.2037440233327868e-07,
+      "loss": 0.0386,
+      "step": 10243
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.6800072034563645,
+      "learning_rate": 1.2005319588975328e-07,
+      "loss": 0.0466,
+      "step": 10244
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.572036328377728,
+      "learning_rate": 1.1973241336427522e-07,
+      "loss": 0.0422,
+      "step": 10245
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.039479993750843,
+      "learning_rate": 1.1941205478470995e-07,
+      "loss": 0.0488,
+      "step": 10246
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.3253889628535431,
+      "learning_rate": 1.1909212017888639e-07,
+      "loss": 0.0388,
+      "step": 10247
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.370221469506656,
+      "learning_rate": 1.1877260957459835e-07,
+      "loss": 0.042,
+      "step": 10248
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.5202802116543876,
+      "learning_rate": 1.1845352299960089e-07,
+      "loss": 0.0495,
+      "step": 10249
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.4972730505640341,
+      "learning_rate": 1.1813486048161348e-07,
+      "loss": 0.0452,
+      "step": 10250
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.8616685997059415,
+      "learning_rate": 1.1781662204831735e-07,
+      "loss": 0.0509,
+      "step": 10251
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.6096214226019423,
+      "learning_rate": 1.1749880772735811e-07,
+      "loss": 0.0405,
+      "step": 10252
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.5009625917660852,
+      "learning_rate": 1.1718141754634371e-07,
+      "loss": 0.0449,
+      "step": 10253
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.4716946698278233,
+      "learning_rate": 1.1686445153284598e-07,
+      "loss": 0.0412,
+      "step": 10254
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.4623878498518141,
+      "learning_rate": 1.1654790971439956e-07,
+      "loss": 0.0475,
+      "step": 10255
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.5477689084291164,
+      "learning_rate": 1.1623179211850244e-07,
+      "loss": 0.049,
+      "step": 10256
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.4660251036140532,
+      "learning_rate": 1.1591609877261545e-07,
+      "loss": 0.0439,
+      "step": 10257
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.2390179348252883,
+      "learning_rate": 1.1560082970416164e-07,
+      "loss": 0.038,
+      "step": 10258
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.6294912710546787,
+      "learning_rate": 1.1528598494052967e-07,
+      "loss": 0.0395,
+      "step": 10259
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.8053669469188616,
+      "learning_rate": 1.1497156450906933e-07,
+      "loss": 0.0507,
+      "step": 10260
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.4246623450905862,
+      "learning_rate": 1.1465756843709431e-07,
+      "loss": 0.0406,
+      "step": 10261
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.4200724216566583,
+      "learning_rate": 1.1434399675188112e-07,
+      "loss": 0.0483,
+      "step": 10262
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.4818500816572715,
+      "learning_rate": 1.1403084948067023e-07,
+      "loss": 0.046,
+      "step": 10263
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.6169302805486885,
+      "learning_rate": 1.1371812665066262e-07,
+      "loss": 0.0407,
+      "step": 10264
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.3403597403887053,
+      "learning_rate": 1.1340582828902658e-07,
+      "loss": 0.0436,
+      "step": 10265
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.4519373707005698,
+      "learning_rate": 1.1309395442288928e-07,
+      "loss": 0.0421,
+      "step": 10266
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.3980574560668704,
+      "learning_rate": 1.1278250507934518e-07,
+      "loss": 0.0442,
+      "step": 10267
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.4310280768496546,
+      "learning_rate": 1.1247148028544819e-07,
+      "loss": 0.041,
+      "step": 10268
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.60816964126377,
+      "learning_rate": 1.1216088006821724e-07,
+      "loss": 0.0447,
+      "step": 10269
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.9797915469700673,
+      "learning_rate": 1.1185070445463352e-07,
+      "loss": 0.0551,
+      "step": 10270
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.5612072625321713,
+      "learning_rate": 1.1154095347164274e-07,
+      "loss": 0.0521,
+      "step": 10271
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.5114783108727903,
+      "learning_rate": 1.1123162714615221e-07,
+      "loss": 0.04,
+      "step": 10272
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.3942306456436464,
+      "learning_rate": 1.1092272550503269e-07,
+      "loss": 0.0404,
+      "step": 10273
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.5997001004650417,
+      "learning_rate": 1.1061424857511937e-07,
+      "loss": 0.0527,
+      "step": 10274
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.6502340585635868,
+      "learning_rate": 1.1030619638320805e-07,
+      "loss": 0.0446,
+      "step": 10275
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.4094479406994191,
+      "learning_rate": 1.0999856895605953e-07,
+      "loss": 0.045,
+      "step": 10276
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.5145313916333254,
+      "learning_rate": 1.0969136632039746e-07,
+      "loss": 0.0449,
+      "step": 10277
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.4800230600559876,
+      "learning_rate": 1.0938458850290823e-07,
+      "loss": 0.0461,
+      "step": 10278
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.509031164988642,
+      "learning_rate": 1.0907823553024166e-07,
+      "loss": 0.0462,
+      "step": 10279
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.8100307906192197,
+      "learning_rate": 1.0877230742901035e-07,
+      "loss": 0.0478,
+      "step": 10280
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.6272359398029264,
+      "learning_rate": 1.084668042257897e-07,
+      "loss": 0.0391,
+      "step": 10281
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.6515151883720558,
+      "learning_rate": 1.0816172594711904e-07,
+      "loss": 0.0553,
+      "step": 10282
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.5686126414868269,
+      "learning_rate": 1.0785707261949996e-07,
+      "loss": 0.0418,
+      "step": 10283
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.327101290021838,
+      "learning_rate": 1.0755284426939794e-07,
+      "loss": 0.0375,
+      "step": 10284
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.3216015845474436,
+      "learning_rate": 1.0724904092324074e-07,
+      "loss": 0.0427,
+      "step": 10285
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.4432631845120643,
+      "learning_rate": 1.0694566260742001e-07,
+      "loss": 0.0379,
+      "step": 10286
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.5885397278330393,
+      "learning_rate": 1.0664270934828969e-07,
+      "loss": 0.0476,
+      "step": 10287
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.5102568216053067,
+      "learning_rate": 1.0634018117216705e-07,
+      "loss": 0.0482,
+      "step": 10288
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.5269141424697574,
+      "learning_rate": 1.0603807810533273e-07,
+      "loss": 0.0421,
+      "step": 10289
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.6185465992239596,
+      "learning_rate": 1.0573640017402964e-07,
+      "loss": 0.0481,
+      "step": 10290
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.5601938207928556,
+      "learning_rate": 1.054351474044657e-07,
+      "loss": 0.0448,
+      "step": 10291
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.4916135845255432,
+      "learning_rate": 1.0513431982280997e-07,
+      "loss": 0.0476,
+      "step": 10292
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.5692871454727004,
+      "learning_rate": 1.0483391745519488e-07,
+      "loss": 0.0545,
+      "step": 10293
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.3951246053325876,
+      "learning_rate": 1.0453394032771569e-07,
+      "loss": 0.0455,
+      "step": 10294
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.321654252823857,
+      "learning_rate": 1.0423438846643264e-07,
+      "loss": 0.0382,
+      "step": 10295
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.285175199321555,
+      "learning_rate": 1.0393526189736602e-07,
+      "loss": 0.0378,
+      "step": 10296
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.4703895036001293,
+      "learning_rate": 1.0363656064650174e-07,
+      "loss": 0.0333,
+      "step": 10297
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.319733881728103,
+      "learning_rate": 1.0333828473978846e-07,
+      "loss": 0.0356,
+      "step": 10298
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.9693758319323303,
+      "learning_rate": 1.0304043420313602e-07,
+      "loss": 0.0282,
+      "step": 10299
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.3891653156750152,
+      "learning_rate": 1.0274300906241819e-07,
+      "loss": 0.0445,
+      "step": 10300
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.6528772992668859,
+      "learning_rate": 1.0244600934347371e-07,
+      "loss": 0.0494,
+      "step": 10301
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.4941120781891513,
+      "learning_rate": 1.0214943507210085e-07,
+      "loss": 0.0458,
+      "step": 10302
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.3169203646237786,
+      "learning_rate": 1.018532862740651e-07,
+      "loss": 0.0435,
+      "step": 10303
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.412247627194948,
+      "learning_rate": 1.015575629750909e-07,
+      "loss": 0.0401,
+      "step": 10304
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.4512937381357316,
+      "learning_rate": 1.0126226520086823e-07,
+      "loss": 0.0429,
+      "step": 10305
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.3243465959762477,
+      "learning_rate": 1.0096739297704938e-07,
+      "loss": 0.0405,
+      "step": 10306
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.7011885514836016,
+      "learning_rate": 1.006729463292494e-07,
+      "loss": 0.0516,
+      "step": 10307
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.5913459298847301,
+      "learning_rate": 1.0037892528304726e-07,
+      "loss": 0.048,
+      "step": 10308
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.4834376671130916,
+      "learning_rate": 1.0008532986398422e-07,
+      "loss": 0.0497,
+      "step": 10309
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.6195733931655385,
+      "learning_rate": 9.979216009756488e-08,
+      "loss": 0.0411,
+      "step": 10310
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.474512514504948,
+      "learning_rate": 9.949941600925606e-08,
+      "loss": 0.0438,
+      "step": 10311
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.3162705760507007,
+      "learning_rate": 9.920709762448854e-08,
+      "loss": 0.038,
+      "step": 10312
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.303891864528856,
+      "learning_rate": 9.891520496865647e-08,
+      "loss": 0.0394,
+      "step": 10313
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.6789717579052865,
+      "learning_rate": 9.862373806711567e-08,
+      "loss": 0.0444,
+      "step": 10314
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.5782821863584564,
+      "learning_rate": 9.833269694518587e-08,
+      "loss": 0.0478,
+      "step": 10315
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.4184351865591265,
+      "learning_rate": 9.804208162815021e-08,
+      "loss": 0.0439,
+      "step": 10316
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.514993088182654,
+      "learning_rate": 9.775189214125347e-08,
+      "loss": 0.0432,
+      "step": 10317
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.4028780395722584,
+      "learning_rate": 9.746212850970383e-08,
+      "loss": 0.042,
+      "step": 10318
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.3638358319960358,
+      "learning_rate": 9.717279075867448e-08,
+      "loss": 0.0388,
+      "step": 10319
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.255442066749372,
+      "learning_rate": 9.688387891329864e-08,
+      "loss": 0.038,
+      "step": 10320
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.7533958667775111,
+      "learning_rate": 9.65953929986746e-08,
+      "loss": 0.0514,
+      "step": 10321
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.8962955796887333,
+      "learning_rate": 9.630733303986283e-08,
+      "loss": 0.0527,
+      "step": 10322
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.3924174917053918,
+      "learning_rate": 9.601969906188723e-08,
+      "loss": 0.0437,
+      "step": 10323
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.575135355359502,
+      "learning_rate": 9.573249108973281e-08,
+      "loss": 0.0526,
+      "step": 10324
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.427146965064548,
+      "learning_rate": 9.544570914835128e-08,
+      "loss": 0.0397,
+      "step": 10325
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.5062534836147499,
+      "learning_rate": 9.51593532626538e-08,
+      "loss": 0.0507,
+      "step": 10326
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.5246331564463163,
+      "learning_rate": 9.48734234575166e-08,
+      "loss": 0.0436,
+      "step": 10327
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.6828932792811744,
+      "learning_rate": 9.45879197577787e-08,
+      "loss": 0.0506,
+      "step": 10328
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.4481151660985787,
+      "learning_rate": 9.430284218824026e-08,
+      "loss": 0.0436,
+      "step": 10329
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.442306868238276,
+      "learning_rate": 9.401819077366648e-08,
+      "loss": 0.0432,
+      "step": 10330
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.284158726469798,
+      "learning_rate": 9.373396553878533e-08,
+      "loss": 0.0345,
+      "step": 10331
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.57870026596695,
+      "learning_rate": 9.345016650828598e-08,
+      "loss": 0.0444,
+      "step": 10332
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.4688552879807097,
+      "learning_rate": 9.316679370682368e-08,
+      "loss": 0.0451,
+      "step": 10333
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.5659604489405252,
+      "learning_rate": 9.288384715901377e-08,
+      "loss": 0.0513,
+      "step": 10334
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.5777080898318339,
+      "learning_rate": 9.260132688943546e-08,
+      "loss": 0.0441,
+      "step": 10335
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.2774441832928285,
+      "learning_rate": 9.231923292263134e-08,
+      "loss": 0.0378,
+      "step": 10336
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.5296833588321208,
+      "learning_rate": 9.203756528310737e-08,
+      "loss": 0.0437,
+      "step": 10337
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.889128590443198,
+      "learning_rate": 9.175632399533118e-08,
+      "loss": 0.0516,
+      "step": 10338
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.6249260949307074,
+      "learning_rate": 9.147550908373381e-08,
+      "loss": 0.0434,
+      "step": 10339
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.5555338711340916,
+      "learning_rate": 9.119512057271074e-08,
+      "loss": 0.0496,
+      "step": 10340
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.1433943411402412,
+      "learning_rate": 9.091515848661747e-08,
+      "loss": 0.0315,
+      "step": 10341
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.5240935042432777,
+      "learning_rate": 9.063562284977512e-08,
+      "loss": 0.0441,
+      "step": 10342
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.614214158248411,
+      "learning_rate": 9.035651368646647e-08,
+      "loss": 0.0442,
+      "step": 10343
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.4971010822561845,
+      "learning_rate": 9.00778310209377e-08,
+      "loss": 0.0385,
+      "step": 10344
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.4149919038604564,
+      "learning_rate": 8.979957487739832e-08,
+      "loss": 0.0363,
+      "step": 10345
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.6562788039679477,
+      "learning_rate": 8.952174528001955e-08,
+      "loss": 0.0426,
+      "step": 10346
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.3299042671092614,
+      "learning_rate": 8.924434225293654e-08,
+      "loss": 0.0396,
+      "step": 10347
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.3912319361516368,
+      "learning_rate": 8.896736582024667e-08,
+      "loss": 0.0391,
+      "step": 10348
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.3235075232631488,
+      "learning_rate": 8.869081600601126e-08,
+      "loss": 0.0399,
+      "step": 10349
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.5196248668925225,
+      "learning_rate": 8.84146928342544e-08,
+      "loss": 0.0444,
+      "step": 10350
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.3314955621203368,
+      "learning_rate": 8.813899632896194e-08,
+      "loss": 0.0382,
+      "step": 10351
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.5136119482377008,
+      "learning_rate": 8.786372651408359e-08,
+      "loss": 0.0391,
+      "step": 10352
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.6113514667425146,
+      "learning_rate": 8.758888341353189e-08,
+      "loss": 0.0436,
+      "step": 10353
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.4628482667214961,
+      "learning_rate": 8.731446705118274e-08,
+      "loss": 0.0392,
+      "step": 10354
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.4541406677547066,
+      "learning_rate": 8.704047745087429e-08,
+      "loss": 0.0401,
+      "step": 10355
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.47844119824978,
+      "learning_rate": 8.676691463640752e-08,
+      "loss": 0.0483,
+      "step": 10356
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.2880468902319846,
+      "learning_rate": 8.649377863154728e-08,
+      "loss": 0.0375,
+      "step": 10357
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.7535988972256478,
+      "learning_rate": 8.622106946002074e-08,
+      "loss": 0.0514,
+      "step": 10358
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.2373807504773728,
+      "learning_rate": 8.594878714551669e-08,
+      "loss": 0.0333,
+      "step": 10359
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.3325768562989997,
+      "learning_rate": 8.567693171168956e-08,
+      "loss": 0.0412,
+      "step": 10360
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.9237394540285586,
+      "learning_rate": 8.540550318215434e-08,
+      "loss": 0.0618,
+      "step": 10361
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.6409395837992216,
+      "learning_rate": 8.513450158049109e-08,
+      "loss": 0.0519,
+      "step": 10362
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.6383801130934434,
+      "learning_rate": 8.486392693024038e-08,
+      "loss": 0.0493,
+      "step": 10363
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.512018204354824,
+      "learning_rate": 8.459377925490786e-08,
+      "loss": 0.05,
+      "step": 10364
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.4367333890512126,
+      "learning_rate": 8.432405857795978e-08,
+      "loss": 0.0432,
+      "step": 10365
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.4262197437763022,
+      "learning_rate": 8.405476492282739e-08,
+      "loss": 0.0418,
+      "step": 10366
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.3326549464357016,
+      "learning_rate": 8.378589831290363e-08,
+      "loss": 0.035,
+      "step": 10367
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.7137729101554846,
+      "learning_rate": 8.351745877154594e-08,
+      "loss": 0.0459,
+      "step": 10368
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.3902164569559086,
+      "learning_rate": 8.324944632207288e-08,
+      "loss": 0.0381,
+      "step": 10369
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.448264673910247,
+      "learning_rate": 8.298186098776583e-08,
+      "loss": 0.04,
+      "step": 10370
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.670068270432286,
+      "learning_rate": 8.27147027918701e-08,
+      "loss": 0.0461,
+      "step": 10371
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.3463922980741976,
+      "learning_rate": 8.244797175759434e-08,
+      "loss": 0.0342,
+      "step": 10372
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.386949657687647,
+      "learning_rate": 8.218166790810833e-08,
+      "loss": 0.0392,
+      "step": 10373
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.4742312064066943,
+      "learning_rate": 8.191579126654637e-08,
+      "loss": 0.045,
+      "step": 10374
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.4694893700869942,
+      "learning_rate": 8.165034185600496e-08,
+      "loss": 0.043,
+      "step": 10375
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.5876849825880603,
+      "learning_rate": 8.138531969954289e-08,
+      "loss": 0.0435,
+      "step": 10376
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.7141144862744118,
+      "learning_rate": 8.11207248201834e-08,
+      "loss": 0.0527,
+      "step": 10377
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.6802114553234544,
+      "learning_rate": 8.08565572409109e-08,
+      "loss": 0.0471,
+      "step": 10378
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.7638751949800415,
+      "learning_rate": 8.059281698467369e-08,
+      "loss": 0.0515,
+      "step": 10379
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.4159406269711217,
+      "learning_rate": 8.032950407438289e-08,
+      "loss": 0.0418,
+      "step": 10380
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.4966568094806383,
+      "learning_rate": 8.006661853291298e-08,
+      "loss": 0.0407,
+      "step": 10381
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.381330016620923,
+      "learning_rate": 7.980416038309902e-08,
+      "loss": 0.0408,
+      "step": 10382
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.31171745013247,
+      "learning_rate": 7.954212964774166e-08,
+      "loss": 0.0362,
+      "step": 10383
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.2703144126665844,
+      "learning_rate": 7.928052634960382e-08,
+      "loss": 0.0431,
+      "step": 10384
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.4227961100887532,
+      "learning_rate": 7.901935051140952e-08,
+      "loss": 0.0478,
+      "step": 10385
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.7906918875077478,
+      "learning_rate": 7.875860215584841e-08,
+      "loss": 0.0431,
+      "step": 10386
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.617959422979846,
+      "learning_rate": 7.849828130557013e-08,
+      "loss": 0.039,
+      "step": 10387
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.3952306771581602,
+      "learning_rate": 7.823838798318995e-08,
+      "loss": 0.0413,
+      "step": 10388
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.3842298734069025,
+      "learning_rate": 7.797892221128311e-08,
+      "loss": 0.0379,
+      "step": 10389
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.4615336826799024,
+      "learning_rate": 7.77198840123905e-08,
+      "loss": 0.0464,
+      "step": 10390
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.5553569264543106,
+      "learning_rate": 7.746127340901411e-08,
+      "loss": 0.0487,
+      "step": 10391
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.3813967362153239,
+      "learning_rate": 7.720309042361984e-08,
+      "loss": 0.0381,
+      "step": 10392
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.567256096422835,
+      "learning_rate": 7.694533507863477e-08,
+      "loss": 0.0458,
+      "step": 10393
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.241073047242047,
+      "learning_rate": 7.6688007396451e-08,
+      "loss": 0.0426,
+      "step": 10394
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.4485734996317654,
+      "learning_rate": 7.643110739942172e-08,
+      "loss": 0.0453,
+      "step": 10395
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.5640042434759203,
+      "learning_rate": 7.617463510986466e-08,
+      "loss": 0.0495,
+      "step": 10396
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.554320891888565,
+      "learning_rate": 7.591859055005813e-08,
+      "loss": 0.0455,
+      "step": 10397
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.3364329243032316,
+      "learning_rate": 7.56629737422454e-08,
+      "loss": 0.0388,
+      "step": 10398
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.536956510138542,
+      "learning_rate": 7.540778470863153e-08,
+      "loss": 0.0465,
+      "step": 10399
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.4399696994715812,
+      "learning_rate": 7.515302347138486e-08,
+      "loss": 0.0466,
+      "step": 10400
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.4785707142507865,
+      "learning_rate": 7.48986900526355e-08,
+      "loss": 0.046,
+      "step": 10401
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.2844883705836427,
+      "learning_rate": 7.464478447447854e-08,
+      "loss": 0.0356,
+      "step": 10402
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.6988691429939067,
+      "learning_rate": 7.439130675896966e-08,
+      "loss": 0.0443,
+      "step": 10403
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.3681141683345532,
+      "learning_rate": 7.413825692812848e-08,
+      "loss": 0.0405,
+      "step": 10404
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.5333713080712394,
+      "learning_rate": 7.388563500393742e-08,
+      "loss": 0.0444,
+      "step": 10405
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.4071943098207074,
+      "learning_rate": 7.363344100834225e-08,
+      "loss": 0.0416,
+      "step": 10406
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.6049639714903283,
+      "learning_rate": 7.338167496324933e-08,
+      "loss": 0.0451,
+      "step": 10407
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.4602801919560122,
+      "learning_rate": 7.313033689053061e-08,
+      "loss": 0.0422,
+      "step": 10408
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.4583022540475874,
+      "learning_rate": 7.287942681201921e-08,
+      "loss": 0.043,
+      "step": 10409
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.6096720733180037,
+      "learning_rate": 7.26289447495121e-08,
+      "loss": 0.05,
+      "step": 10410
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.3400683299466933,
+      "learning_rate": 7.237889072476856e-08,
+      "loss": 0.0436,
+      "step": 10411
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.5331925092574545,
+      "learning_rate": 7.212926475950954e-08,
+      "loss": 0.0475,
+      "step": 10412
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.3190404757539123,
+      "learning_rate": 7.188006687542048e-08,
+      "loss": 0.0403,
+      "step": 10413
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.464292288372917,
+      "learning_rate": 7.163129709414962e-08,
+      "loss": 0.0444,
+      "step": 10414
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.361384891605114,
+      "learning_rate": 7.138295543730634e-08,
+      "loss": 0.0376,
+      "step": 10415
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.4097894002658897,
+      "learning_rate": 7.113504192646503e-08,
+      "loss": 0.0471,
+      "step": 10416
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.410768994742681,
+      "learning_rate": 7.088755658316127e-08,
+      "loss": 0.0427,
+      "step": 10417
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.4381431534130324,
+      "learning_rate": 7.064049942889395e-08,
+      "loss": 0.0445,
+      "step": 10418
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.5946808364596332,
+      "learning_rate": 7.03938704851248e-08,
+      "loss": 0.0441,
+      "step": 10419
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.3293688540349755,
+      "learning_rate": 7.014766977327836e-08,
+      "loss": 0.0392,
+      "step": 10420
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.3937211122060589,
+      "learning_rate": 6.990189731474195e-08,
+      "loss": 0.04,
+      "step": 10421
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.7143530019462123,
+      "learning_rate": 6.965655313086572e-08,
+      "loss": 0.0521,
+      "step": 10422
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.4004485772801822,
+      "learning_rate": 6.941163724296263e-08,
+      "loss": 0.0387,
+      "step": 10423
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.5712393627263477,
+      "learning_rate": 6.916714967230786e-08,
+      "loss": 0.0439,
+      "step": 10424
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.3161319352546244,
+      "learning_rate": 6.892309044014056e-08,
+      "loss": 0.0354,
+      "step": 10425
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.5744980343029722,
+      "learning_rate": 6.867945956766154e-08,
+      "loss": 0.0444,
+      "step": 10426
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.3715513200862968,
+      "learning_rate": 6.843625707603496e-08,
+      "loss": 0.0408,
+      "step": 10427
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.4724034197542435,
+      "learning_rate": 6.819348298638839e-08,
+      "loss": 0.045,
+      "step": 10428
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.6140175253905769,
+      "learning_rate": 6.795113731981052e-08,
+      "loss": 0.0482,
+      "step": 10429
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.355919530318479,
+      "learning_rate": 6.770922009735392e-08,
+      "loss": 0.0403,
+      "step": 10430
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.2826238635166742,
+      "learning_rate": 6.746773134003404e-08,
+      "loss": 0.0357,
+      "step": 10431
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.4886024664071593,
+      "learning_rate": 6.722667106882907e-08,
+      "loss": 0.0442,
+      "step": 10432
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.7302810325645976,
+      "learning_rate": 6.698603930467951e-08,
+      "loss": 0.0454,
+      "step": 10433
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.7181990315340228,
+      "learning_rate": 6.674583606848862e-08,
+      "loss": 0.0574,
+      "step": 10434
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.489644584243883,
+      "learning_rate": 6.650606138112358e-08,
+      "loss": 0.0395,
+      "step": 10435
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.317091627112999,
+      "learning_rate": 6.626671526341222e-08,
+      "loss": 0.0381,
+      "step": 10436
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.49765090451203,
+      "learning_rate": 6.602779773614731e-08,
+      "loss": 0.0397,
+      "step": 10437
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.4742928506958795,
+      "learning_rate": 6.578930882008283e-08,
+      "loss": 0.0378,
+      "step": 10438
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.6911730204639146,
+      "learning_rate": 6.555124853593719e-08,
+      "loss": 0.0522,
+      "step": 10439
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.3221733205049593,
+      "learning_rate": 6.531361690438942e-08,
+      "loss": 0.0366,
+      "step": 10440
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.6437206403539906,
+      "learning_rate": 6.507641394608355e-08,
+      "loss": 0.0514,
+      "step": 10441
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.41723331166098,
+      "learning_rate": 6.483963968162421e-08,
+      "loss": 0.0411,
+      "step": 10442
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.7832281710388262,
+      "learning_rate": 6.460329413157996e-08,
+      "loss": 0.0519,
+      "step": 10443
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.5145694967997823,
+      "learning_rate": 6.436737731648268e-08,
+      "loss": 0.0449,
+      "step": 10444
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.514990966274188,
+      "learning_rate": 6.4131889256826e-08,
+      "loss": 0.0442,
+      "step": 10445
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.5178262509714926,
+      "learning_rate": 6.389682997306689e-08,
+      "loss": 0.0367,
+      "step": 10446
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.4268360447473367,
+      "learning_rate": 6.366219948562402e-08,
+      "loss": 0.0406,
+      "step": 10447
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.6302697018110215,
+      "learning_rate": 6.342799781487997e-08,
+      "loss": 0.0501,
+      "step": 10448
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.7688867703098903,
+      "learning_rate": 6.31942249811801e-08,
+      "loss": 0.0491,
+      "step": 10449
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.5442890796211093,
+      "learning_rate": 6.296088100483155e-08,
+      "loss": 0.0745,
+      "step": 10450
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.5245711706203575,
+      "learning_rate": 6.27279659061053e-08,
+      "loss": 0.0401,
+      "step": 10451
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.4076630845364102,
+      "learning_rate": 6.249547970523407e-08,
+      "loss": 0.046,
+      "step": 10452
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.4982870536977921,
+      "learning_rate": 6.22634224224139e-08,
+      "loss": 0.0511,
+      "step": 10453
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.5676202407555635,
+      "learning_rate": 6.203179407780368e-08,
+      "loss": 0.042,
+      "step": 10454
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.3329055305244886,
+      "learning_rate": 6.18005946915251e-08,
+      "loss": 0.0373,
+      "step": 10455
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.2716843715746093,
+      "learning_rate": 6.156982428366154e-08,
+      "loss": 0.0386,
+      "step": 10456
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.4250545852685859,
+      "learning_rate": 6.133948287426028e-08,
+      "loss": 0.0399,
+      "step": 10457
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.3659099017277354,
+      "learning_rate": 6.110957048333088e-08,
+      "loss": 0.043,
+      "step": 10458
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.5264272316243634,
+      "learning_rate": 6.088008713084626e-08,
+      "loss": 0.0469,
+      "step": 10459
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.586533580149427,
+      "learning_rate": 6.065103283674045e-08,
+      "loss": 0.04,
+      "step": 10460
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.3908812825700378,
+      "learning_rate": 6.0422407620912e-08,
+      "loss": 0.0398,
+      "step": 10461
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.4789001012786605,
+      "learning_rate": 6.019421150322114e-08,
+      "loss": 0.0457,
+      "step": 10462
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.3846931614588163,
+      "learning_rate": 5.996644450349142e-08,
+      "loss": 0.0346,
+      "step": 10463
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.7817040833377382,
+      "learning_rate": 5.973910664150818e-08,
+      "loss": 0.0502,
+      "step": 10464
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.3354937827456324,
+      "learning_rate": 5.951219793702112e-08,
+      "loss": 0.036,
+      "step": 10465
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.6107870625726324,
+      "learning_rate": 5.928571840974062e-08,
+      "loss": 0.0445,
+      "step": 10466
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.7054474609660477,
+      "learning_rate": 5.905966807934205e-08,
+      "loss": 0.0436,
+      "step": 10467
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.5140400889771781,
+      "learning_rate": 5.88340469654608e-08,
+      "loss": 0.0443,
+      "step": 10468
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.6091141052733897,
+      "learning_rate": 5.8608855087697314e-08,
+      "loss": 0.0482,
+      "step": 10469
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.4973352005753136,
+      "learning_rate": 5.8384092465614274e-08,
+      "loss": 0.0465,
+      "step": 10470
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.4628731345895611,
+      "learning_rate": 5.815975911873606e-08,
+      "loss": 0.0435,
+      "step": 10471
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.6299086228498854,
+      "learning_rate": 5.7935855066549863e-08,
+      "loss": 0.0569,
+      "step": 10472
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.390736585950677,
+      "learning_rate": 5.771238032850679e-08,
+      "loss": 0.0394,
+      "step": 10473
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.4745681489958136,
+      "learning_rate": 5.74893349240202e-08,
+      "loss": 0.0421,
+      "step": 10474
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.4167210859699482,
+      "learning_rate": 5.726671887246515e-08,
+      "loss": 0.0459,
+      "step": 10475
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.60975529460272,
+      "learning_rate": 5.704453219318118e-08,
+      "loss": 0.0497,
+      "step": 10476
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.7771808830047289,
+      "learning_rate": 5.682277490546839e-08,
+      "loss": 0.0515,
+      "step": 10477
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.6642573367482645,
+      "learning_rate": 5.6601447028591384e-08,
+      "loss": 0.0503,
+      "step": 10478
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.4304031850761514,
+      "learning_rate": 5.638054858177644e-08,
+      "loss": 0.0421,
+      "step": 10479
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.6052916801962822,
+      "learning_rate": 5.616007958421321e-08,
+      "loss": 0.0536,
+      "step": 10480
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.3406679097447967,
+      "learning_rate": 5.5940040055053604e-08,
+      "loss": 0.041,
+      "step": 10481
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.7884976405084136,
+      "learning_rate": 5.572043001341232e-08,
+      "loss": 0.0508,
+      "step": 10482
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.5050543149581124,
+      "learning_rate": 5.550124947836688e-08,
+      "loss": 0.0405,
+      "step": 10483
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.299849223166704,
+      "learning_rate": 5.5282498468957056e-08,
+      "loss": 0.0352,
+      "step": 10484
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.7670370218793927,
+      "learning_rate": 5.5064177004185424e-08,
+      "loss": 0.0558,
+      "step": 10485
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.368751530837964,
+      "learning_rate": 5.484628510301793e-08,
+      "loss": 0.0417,
+      "step": 10486
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.5111358559904722,
+      "learning_rate": 5.462882278438275e-08,
+      "loss": 0.0466,
+      "step": 10487
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.5182220647113223,
+      "learning_rate": 5.4411790067170345e-08,
+      "loss": 0.0423,
+      "step": 10488
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.5593558009746167,
+      "learning_rate": 5.4195186970234514e-08,
+      "loss": 0.0516,
+      "step": 10489
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.5236994174620118,
+      "learning_rate": 5.397901351239077e-08,
+      "loss": 0.0414,
+      "step": 10490
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.4920153706616395,
+      "learning_rate": 5.3763269712419076e-08,
+      "loss": 0.0419,
+      "step": 10491
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.6839827935738112,
+      "learning_rate": 5.354795558906001e-08,
+      "loss": 0.0431,
+      "step": 10492
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.4922375139022215,
+      "learning_rate": 5.333307116101804e-08,
+      "loss": 0.0507,
+      "step": 10493
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.2345045699868673,
+      "learning_rate": 5.3118616446960484e-08,
+      "loss": 0.0378,
+      "step": 10494
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.387778502893405,
+      "learning_rate": 5.2904591465516855e-08,
+      "loss": 0.0407,
+      "step": 10495
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.4615672710949077,
+      "learning_rate": 5.26909962352784e-08,
+      "loss": 0.0412,
+      "step": 10496
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.5345797616087047,
+      "learning_rate": 5.24778307748014e-08,
+      "loss": 0.0437,
+      "step": 10497
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.392118549458176,
+      "learning_rate": 5.226509510260214e-08,
+      "loss": 0.0435,
+      "step": 10498
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.56025630476497,
+      "learning_rate": 5.2052789237161395e-08,
+      "loss": 0.0437,
+      "step": 10499
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.3696839715655833,
+      "learning_rate": 5.184091319692219e-08,
+      "loss": 0.0408,
+      "step": 10500
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.436605168578194,
+      "learning_rate": 5.1629467000290365e-08,
+      "loss": 0.0402,
+      "step": 10501
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.3792406807047526,
+      "learning_rate": 5.1418450665633445e-08,
+      "loss": 0.0411,
+      "step": 10502
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.39238645674009,
+      "learning_rate": 5.1207864211282324e-08,
+      "loss": 0.0364,
+      "step": 10503
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.613105318669477,
+      "learning_rate": 5.099770765553069e-08,
+      "loss": 0.0453,
+      "step": 10504
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.3801880881184851,
+      "learning_rate": 5.078798101663507e-08,
+      "loss": 0.0418,
+      "step": 10505
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.1254802287805858,
+      "learning_rate": 5.057868431281421e-08,
+      "loss": 0.0309,
+      "step": 10506
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.4453504735243323,
+      "learning_rate": 5.0369817562249126e-08,
+      "loss": 0.0407,
+      "step": 10507
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.625518649218529,
+      "learning_rate": 5.016138078308364e-08,
+      "loss": 0.0534,
+      "step": 10508
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.4075947882287927,
+      "learning_rate": 4.9953373993426036e-08,
+      "loss": 0.04,
+      "step": 10509
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.4844752833913288,
+      "learning_rate": 4.9745797211344096e-08,
+      "loss": 0.0387,
+      "step": 10510
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.3116317884502549,
+      "learning_rate": 4.953865045487061e-08,
+      "loss": 0.0394,
+      "step": 10511
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.4743315774940526,
+      "learning_rate": 4.9331933742000627e-08,
+      "loss": 0.044,
+      "step": 10512
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.390183659734006,
+      "learning_rate": 4.91256470906909e-08,
+      "loss": 0.035,
+      "step": 10513
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.6374208737038816,
+      "learning_rate": 4.891979051886153e-08,
+      "loss": 0.0517,
+      "step": 10514
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.3329173715421005,
+      "learning_rate": 4.8714364044396e-08,
+      "loss": 0.0397,
+      "step": 10515
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.4074261514127087,
+      "learning_rate": 4.850936768513781e-08,
+      "loss": 0.036,
+      "step": 10516
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.4525680209089034,
+      "learning_rate": 4.83048014588966e-08,
+      "loss": 0.0413,
+      "step": 10517
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.5016630148150008,
+      "learning_rate": 4.81006653834426e-08,
+      "loss": 0.0385,
+      "step": 10518
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.3765627716657398,
+      "learning_rate": 4.7896959476508296e-08,
+      "loss": 0.0375,
+      "step": 10519
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.9114280290934447,
+      "learning_rate": 4.7693683755788975e-08,
+      "loss": 0.0427,
+      "step": 10520
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.595345851984939,
+      "learning_rate": 4.7490838238944957e-08,
+      "loss": 0.0465,
+      "step": 10521
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.6092765283419799,
+      "learning_rate": 4.7288422943596035e-08,
+      "loss": 0.045,
+      "step": 10522
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.36869387181434,
+      "learning_rate": 4.708643788732592e-08,
+      "loss": 0.0432,
+      "step": 10523
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.5157135052333635,
+      "learning_rate": 4.6884883087681686e-08,
+      "loss": 0.0371,
+      "step": 10524
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.3869862708440757,
+      "learning_rate": 4.668375856217156e-08,
+      "loss": 0.0398,
+      "step": 10525
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.7007158361387813,
+      "learning_rate": 4.64830643282671e-08,
+      "loss": 0.0326,
+      "step": 10526
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.885647406751286,
+      "learning_rate": 4.628280040340272e-08,
+      "loss": 0.0516,
+      "step": 10527
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.6061252363205898,
+      "learning_rate": 4.608296680497559e-08,
+      "loss": 0.0452,
+      "step": 10528
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.4256128240117376,
+      "learning_rate": 4.588356355034462e-08,
+      "loss": 0.0499,
+      "step": 10529
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.5792852714537684,
+      "learning_rate": 4.568459065683206e-08,
+      "loss": 0.0449,
+      "step": 10530
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.500084858103287,
+      "learning_rate": 4.5486048141721863e-08,
+      "loss": 0.0508,
+      "step": 10531
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.3541734735358653,
+      "learning_rate": 4.528793602226245e-08,
+      "loss": 0.0411,
+      "step": 10532
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4951113871859114,
+      "learning_rate": 4.509025431566283e-08,
+      "loss": 0.0419,
+      "step": 10533
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.7373534476859782,
+      "learning_rate": 4.4893003039096494e-08,
+      "loss": 0.0474,
+      "step": 10534
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.483227027649943,
+      "learning_rate": 4.4696182209697515e-08,
+      "loss": 0.0431,
+      "step": 10535
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4612562224454717,
+      "learning_rate": 4.449979184456388e-08,
+      "loss": 0.043,
+      "step": 10536
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4390181830254363,
+      "learning_rate": 4.4303831960756385e-08,
+      "loss": 0.0397,
+      "step": 10537
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.3125925501348044,
+      "learning_rate": 4.410830257529752e-08,
+      "loss": 0.0399,
+      "step": 10538
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.6130170533053534,
+      "learning_rate": 4.391320370517205e-08,
+      "loss": 0.0491,
+      "step": 10539
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.535329865789914,
+      "learning_rate": 4.371853536732973e-08,
+      "loss": 0.0454,
+      "step": 10540
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4742641489493817,
+      "learning_rate": 4.3524297578680375e-08,
+      "loss": 0.0435,
+      "step": 10541
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.6370064759300502,
+      "learning_rate": 4.333049035609715e-08,
+      "loss": 0.0383,
+      "step": 10542
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4106480480866852,
+      "learning_rate": 4.3137113716416044e-08,
+      "loss": 0.0386,
+      "step": 10543
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.433474046013542,
+      "learning_rate": 4.294416767643639e-08,
+      "loss": 0.0378,
+      "step": 10544
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.217668720277976,
+      "learning_rate": 4.275165225291755e-08,
+      "loss": 0.0386,
+      "step": 10545
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.229271611677372,
+      "learning_rate": 4.255956746258505e-08,
+      "loss": 0.0376,
+      "step": 10546
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4680991228486628,
+      "learning_rate": 4.236791332212498e-08,
+      "loss": 0.0441,
+      "step": 10547
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4890777102554775,
+      "learning_rate": 4.217668984818513e-08,
+      "loss": 0.0382,
+      "step": 10548
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.307591590139095,
+      "learning_rate": 4.198589705737721e-08,
+      "loss": 0.0425,
+      "step": 10549
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4639753207858603,
+      "learning_rate": 4.1795534966275754e-08,
+      "loss": 0.0457,
+      "step": 10550
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.2184259958529033,
+      "learning_rate": 4.1605603591416964e-08,
+      "loss": 0.0392,
+      "step": 10551
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4344470733187191,
+      "learning_rate": 4.141610294930043e-08,
+      "loss": 0.0429,
+      "step": 10552
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.3010250129450645,
+      "learning_rate": 4.1227033056388535e-08,
+      "loss": 0.0403,
+      "step": 10553
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.600697120343962,
+      "learning_rate": 4.103839392910425e-08,
+      "loss": 0.0449,
+      "step": 10554
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4712729959667414,
+      "learning_rate": 4.085018558383558e-08,
+      "loss": 0.0384,
+      "step": 10555
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4646342061481719,
+      "learning_rate": 4.0662408036931664e-08,
+      "loss": 0.0419,
+      "step": 10556
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.468054644516981,
+      "learning_rate": 4.0475061304704465e-08,
+      "loss": 0.0389,
+      "step": 10557
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.6829702234847743,
+      "learning_rate": 4.028814540342985e-08,
+      "loss": 0.0473,
+      "step": 10558
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.595935391284207,
+      "learning_rate": 4.0101660349343706e-08,
+      "loss": 0.0509,
+      "step": 10559
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.3820917856356962,
+      "learning_rate": 3.991560615864587e-08,
+      "loss": 0.0296,
+      "step": 10560
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4516269972202702,
+      "learning_rate": 3.972998284749952e-08,
+      "loss": 0.0449,
+      "step": 10561
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4199581421938772,
+      "learning_rate": 3.9544790432029526e-08,
+      "loss": 0.0395,
+      "step": 10562
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4408793517747758,
+      "learning_rate": 3.936002892832302e-08,
+      "loss": 0.0431,
+      "step": 10563
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.4950793659775383,
+      "learning_rate": 3.917569835243107e-08,
+      "loss": 0.0456,
+      "step": 10564
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.3451696835761318,
+      "learning_rate": 3.8991798720365296e-08,
+      "loss": 0.0391,
+      "step": 10565
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.6242160414864777,
+      "learning_rate": 3.880833004810125e-08,
+      "loss": 0.0498,
+      "step": 10566
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.5004353431183932,
+      "learning_rate": 3.862529235157675e-08,
+      "loss": 0.0452,
+      "step": 10567
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.4781423286236863,
+      "learning_rate": 3.84426856466924e-08,
+      "loss": 0.0407,
+      "step": 10568
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.500867271292968,
+      "learning_rate": 3.826050994931052e-08,
+      "loss": 0.0429,
+      "step": 10569
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.4592679869171348,
+      "learning_rate": 3.807876527525789e-08,
+      "loss": 0.0387,
+      "step": 10570
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.6942215312920208,
+      "learning_rate": 3.7897451640321326e-08,
+      "loss": 0.0488,
+      "step": 10571
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.5777491719668757,
+      "learning_rate": 3.771656906025212e-08,
+      "loss": 0.0516,
+      "step": 10572
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.3124044562305939,
+      "learning_rate": 3.753611755076269e-08,
+      "loss": 0.0405,
+      "step": 10573
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.4259480081810934,
+      "learning_rate": 3.7356097127529414e-08,
+      "loss": 0.0369,
+      "step": 10574
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.351997642312581,
+      "learning_rate": 3.717650780619031e-08,
+      "loss": 0.0466,
+      "step": 10575
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.6707054177671112,
+      "learning_rate": 3.6997349602346244e-08,
+      "loss": 0.0433,
+      "step": 10576
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.4839038870689747,
+      "learning_rate": 3.681862253156088e-08,
+      "loss": 0.0471,
+      "step": 10577
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.421182300195976,
+      "learning_rate": 3.6640326609359566e-08,
+      "loss": 0.0448,
+      "step": 10578
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.3548519537328862,
+      "learning_rate": 3.646246185123103e-08,
+      "loss": 0.0381,
+      "step": 10579
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.6092962307539156,
+      "learning_rate": 3.6285028272626255e-08,
+      "loss": 0.0461,
+      "step": 10580
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.372974395435489,
+      "learning_rate": 3.610802588895845e-08,
+      "loss": 0.0408,
+      "step": 10581
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.5017275799961416,
+      "learning_rate": 3.593145471560477e-08,
+      "loss": 0.0454,
+      "step": 10582
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.4930879491599516,
+      "learning_rate": 3.575531476790295e-08,
+      "loss": 0.042,
+      "step": 10583
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.408209697329299,
+      "learning_rate": 3.5579606061154626e-08,
+      "loss": 0.0402,
+      "step": 10584
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.6603268217118023,
+      "learning_rate": 3.5404328610622593e-08,
+      "loss": 0.0437,
+      "step": 10585
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.725128325057117,
+      "learning_rate": 3.522948243153412e-08,
+      "loss": 0.0478,
+      "step": 10586
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.4743678820373143,
+      "learning_rate": 3.505506753907761e-08,
+      "loss": 0.0414,
+      "step": 10587
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.4854398040055226,
+      "learning_rate": 3.488108394840428e-08,
+      "loss": 0.0366,
+      "step": 10588
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.2474605478670773,
+      "learning_rate": 3.470753167462815e-08,
+      "loss": 0.0354,
+      "step": 10589
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.5363711041663948,
+      "learning_rate": 3.4534410732825485e-08,
+      "loss": 0.0489,
+      "step": 10590
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.3012499187284288,
+      "learning_rate": 3.4361721138035375e-08,
+      "loss": 0.0404,
+      "step": 10591
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.627693139689471,
+      "learning_rate": 3.4189462905259154e-08,
+      "loss": 0.0458,
+      "step": 10592
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.3848549171136908,
+      "learning_rate": 3.40176360494604e-08,
+      "loss": 0.0405,
+      "step": 10593
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.4540320628321397,
+      "learning_rate": 3.3846240585566074e-08,
+      "loss": 0.0369,
+      "step": 10594
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.3811105194800495,
+      "learning_rate": 3.367527652846536e-08,
+      "loss": 0.0405,
+      "step": 10595
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.47851606732552,
+      "learning_rate": 3.3504743893009726e-08,
+      "loss": 0.0406,
+      "step": 10596
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.307038979217443,
+      "learning_rate": 3.333464269401232e-08,
+      "loss": 0.0339,
+      "step": 10597
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.482110737868351,
+      "learning_rate": 3.316497294625132e-08,
+      "loss": 0.0402,
+      "step": 10598
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.5565306127406477,
+      "learning_rate": 3.2995734664464373e-08,
+      "loss": 0.0467,
+      "step": 10599
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.4052942799305057,
+      "learning_rate": 3.2826927863354174e-08,
+      "loss": 0.0348,
+      "step": 10600
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.4495009220449029,
+      "learning_rate": 3.2658552557583986e-08,
+      "loss": 0.0414,
+      "step": 10601
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.2496892731146683,
+      "learning_rate": 3.249060876178156e-08,
+      "loss": 0.032,
+      "step": 10602
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.3961382752405715,
+      "learning_rate": 3.232309649053467e-08,
+      "loss": 0.0407,
+      "step": 10603
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.6402632661595828,
+      "learning_rate": 3.2156015758396106e-08,
+      "loss": 0.0459,
+      "step": 10604
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.337580064108498,
+      "learning_rate": 3.198936657987928e-08,
+      "loss": 0.0357,
+      "step": 10605
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.3446132601068361,
+      "learning_rate": 3.182314896946204e-08,
+      "loss": 0.0387,
+      "step": 10606
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.43150128901666,
+      "learning_rate": 3.165736294158228e-08,
+      "loss": 0.0418,
+      "step": 10607
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.4174040032684947,
+      "learning_rate": 3.1492008510642935e-08,
+      "loss": 0.0351,
+      "step": 10608
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.405948665860412,
+      "learning_rate": 3.1327085691006954e-08,
+      "loss": 0.0447,
+      "step": 10609
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.1744522397206334,
+      "learning_rate": 3.116259449700232e-08,
+      "loss": 0.0374,
+      "step": 10610
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.4198323465996756,
+      "learning_rate": 3.09985349429176e-08,
+      "loss": 0.0401,
+      "step": 10611
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.3802792528284662,
+      "learning_rate": 3.083490704300529e-08,
+      "loss": 0.0446,
+      "step": 10612
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.259705200914064,
+      "learning_rate": 3.067171081147846e-08,
+      "loss": 0.035,
+      "step": 10613
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.8744874348942742,
+      "learning_rate": 3.050894626251466e-08,
+      "loss": 0.0509,
+      "step": 10614
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.7165307269300245,
+      "learning_rate": 3.034661341025258e-08,
+      "loss": 0.0545,
+      "step": 10615
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.6557424568776744,
+      "learning_rate": 3.0184712268794824e-08,
+      "loss": 0.0435,
+      "step": 10616
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.3624022671095553,
+      "learning_rate": 3.002324285220515e-08,
+      "loss": 0.0396,
+      "step": 10617
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.3157527987512172,
+      "learning_rate": 2.9862205174510104e-08,
+      "loss": 0.0337,
+      "step": 10618
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.1927533065961073,
+      "learning_rate": 2.970159924969962e-08,
+      "loss": 0.0341,
+      "step": 10619
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.4321051428770024,
+      "learning_rate": 2.9541425091724195e-08,
+      "loss": 0.043,
+      "step": 10620
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.2734553073270665,
+      "learning_rate": 2.9381682714499372e-08,
+      "loss": 0.0373,
+      "step": 10621
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.2211377567700816,
+      "learning_rate": 2.9222372131901266e-08,
+      "loss": 0.0318,
+      "step": 10622
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.2595193009598002,
+      "learning_rate": 2.9063493357769368e-08,
+      "loss": 0.0406,
+      "step": 10623
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.669762379640789,
+      "learning_rate": 2.8905046405905412e-08,
+      "loss": 0.0523,
+      "step": 10624
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.5509933566261285,
+      "learning_rate": 2.8747031290072834e-08,
+      "loss": 0.0462,
+      "step": 10625
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.6493909013553099,
+      "learning_rate": 2.858944802399899e-08,
+      "loss": 0.0444,
+      "step": 10626
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.6550358463538808,
+      "learning_rate": 2.8432296621373478e-08,
+      "loss": 0.0588,
+      "step": 10627
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.4231544392548225,
+      "learning_rate": 2.8275577095846495e-08,
+      "loss": 0.0437,
+      "step": 10628
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.2458509630893728,
+      "learning_rate": 2.8119289461033817e-08,
+      "loss": 0.0399,
+      "step": 10629
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.7910770037532278,
+      "learning_rate": 2.796343373051069e-08,
+      "loss": 0.0489,
+      "step": 10630
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.6393441870289993,
+      "learning_rate": 2.7808009917817402e-08,
+      "loss": 0.0455,
+      "step": 10631
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.5643553821934986,
+      "learning_rate": 2.765301803645426e-08,
+      "loss": 0.0444,
+      "step": 10632
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.4419643872100028,
+      "learning_rate": 2.7498458099886605e-08,
+      "loss": 0.0367,
+      "step": 10633
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.5201198650682082,
+      "learning_rate": 2.7344330121539807e-08,
+      "loss": 0.0487,
+      "step": 10634
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.4085618178721064,
+      "learning_rate": 2.7190634114803717e-08,
+      "loss": 0.0446,
+      "step": 10635
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.6583413173150827,
+      "learning_rate": 2.7037370093029868e-08,
+      "loss": 0.0553,
+      "step": 10636
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.5811792046777655,
+      "learning_rate": 2.6884538069531506e-08,
+      "loss": 0.0457,
+      "step": 10637
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.5155842228125431,
+      "learning_rate": 2.6732138057585232e-08,
+      "loss": 0.0424,
+      "step": 10638
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.5936853011347158,
+      "learning_rate": 2.6580170070430457e-08,
+      "loss": 0.0449,
+      "step": 10639
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.1087122341446085,
+      "learning_rate": 2.6428634121267726e-08,
+      "loss": 0.0339,
+      "step": 10640
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.2050498543844859,
+      "learning_rate": 2.627753022326207e-08,
+      "loss": 0.0379,
+      "step": 10641
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.5615352432417156,
+      "learning_rate": 2.61268583895391e-08,
+      "loss": 0.0464,
+      "step": 10642
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.727585114779161,
+      "learning_rate": 2.5976618633187233e-08,
+      "loss": 0.0468,
+      "step": 10643
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4229766279107658,
+      "learning_rate": 2.58268109672577e-08,
+      "loss": 0.0411,
+      "step": 10644
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.468224347708062,
+      "learning_rate": 2.5677435404765082e-08,
+      "loss": 0.0425,
+      "step": 10645
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4893922463100215,
+      "learning_rate": 2.5528491958684565e-08,
+      "loss": 0.044,
+      "step": 10646
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.509416337041036,
+      "learning_rate": 2.5379980641955792e-08,
+      "loss": 0.0446,
+      "step": 10647
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.3348902660693356,
+      "learning_rate": 2.5231901467479004e-08,
+      "loss": 0.0413,
+      "step": 10648
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4358113102302523,
+      "learning_rate": 2.5084254448117794e-08,
+      "loss": 0.0425,
+      "step": 10649
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4168325170032015,
+      "learning_rate": 2.4937039596698576e-08,
+      "loss": 0.0411,
+      "step": 10650
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4003557393614532,
+      "learning_rate": 2.4790256926010003e-08,
+      "loss": 0.04,
+      "step": 10651
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.9397362452968625,
+      "learning_rate": 2.4643906448801878e-08,
+      "loss": 0.0465,
+      "step": 10652
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.3819049866061375,
+      "learning_rate": 2.4497988177789034e-08,
+      "loss": 0.041,
+      "step": 10653
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.5329520554347635,
+      "learning_rate": 2.4352502125646882e-08,
+      "loss": 0.0388,
+      "step": 10654
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.7876571659896303,
+      "learning_rate": 2.4207448305012538e-08,
+      "loss": 0.0534,
+      "step": 10655
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.768251136088922,
+      "learning_rate": 2.406282672848814e-08,
+      "loss": 0.054,
+      "step": 10656
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.7662025099317198,
+      "learning_rate": 2.3918637408636425e-08,
+      "loss": 0.0535,
+      "step": 10657
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.7250325793466337,
+      "learning_rate": 2.3774880357982922e-08,
+      "loss": 0.0407,
+      "step": 10658
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.597163184804839,
+      "learning_rate": 2.363155558901542e-08,
+      "loss": 0.0446,
+      "step": 10659
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4798246531559354,
+      "learning_rate": 2.3488663114185628e-08,
+      "loss": 0.0497,
+      "step": 10660
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.5893666045051908,
+      "learning_rate": 2.3346202945905284e-08,
+      "loss": 0.0423,
+      "step": 10661
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4695499344695997,
+      "learning_rate": 2.320417509655004e-08,
+      "loss": 0.044,
+      "step": 10662
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4960748319277428,
+      "learning_rate": 2.3062579578458365e-08,
+      "loss": 0.0451,
+      "step": 10663
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.523977657211241,
+      "learning_rate": 2.2921416403929863e-08,
+      "loss": 0.0404,
+      "step": 10664
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.7230267713806449,
+      "learning_rate": 2.2780685585227504e-08,
+      "loss": 0.0577,
+      "step": 10665
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.617890761392945,
+      "learning_rate": 2.264038713457706e-08,
+      "loss": 0.0437,
+      "step": 10666
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.3440721676419964,
+      "learning_rate": 2.250052106416545e-08,
+      "loss": 0.0393,
+      "step": 10667
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.6007104281582487,
+      "learning_rate": 2.2361087386142954e-08,
+      "loss": 0.0455,
+      "step": 10668
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4334373708446035,
+      "learning_rate": 2.2222086112622665e-08,
+      "loss": 0.0467,
+      "step": 10669
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.474125347782493,
+      "learning_rate": 2.208351725567881e-08,
+      "loss": 0.0395,
+      "step": 10670
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4044074695879036,
+      "learning_rate": 2.1945380827348985e-08,
+      "loss": 0.0396,
+      "step": 10671
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.755600460382403,
+      "learning_rate": 2.180767683963303e-08,
+      "loss": 0.0428,
+      "step": 10672
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.5821853299495003,
+      "learning_rate": 2.1670405304493047e-08,
+      "loss": 0.0442,
+      "step": 10673
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4700043696136404,
+      "learning_rate": 2.1533566233853942e-08,
+      "loss": 0.0417,
+      "step": 10674
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.7028030461164436,
+      "learning_rate": 2.139715963960287e-08,
+      "loss": 0.0435,
+      "step": 10675
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.7007027839298752,
+      "learning_rate": 2.1261185533589246e-08,
+      "loss": 0.0393,
+      "step": 10676
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.3895439241406904,
+      "learning_rate": 2.1125643927625838e-08,
+      "loss": 0.0407,
+      "step": 10677
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.3811205442440373,
+      "learning_rate": 2.0990534833485455e-08,
+      "loss": 0.0412,
+      "step": 10678
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.448575969141517,
+      "learning_rate": 2.085585826290648e-08,
+      "loss": 0.0383,
+      "step": 10679
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.4932019895325057,
+      "learning_rate": 2.0721614227587338e-08,
+      "loss": 0.0428,
+      "step": 10680
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.3679061540017559,
+      "learning_rate": 2.058780273918981e-08,
+      "loss": 0.041,
+      "step": 10681
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.5050632579938947,
+      "learning_rate": 2.0454423809338487e-08,
+      "loss": 0.0431,
+      "step": 10682
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.5588729395985919,
+      "learning_rate": 2.0321477449619098e-08,
+      "loss": 0.0385,
+      "step": 10683
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.3685898383044695,
+      "learning_rate": 2.0188963671581852e-08,
+      "loss": 0.0401,
+      "step": 10684
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.7847647381851344,
+      "learning_rate": 2.0056882486736982e-08,
+      "loss": 0.0506,
+      "step": 10685
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.294559231599737,
+      "learning_rate": 1.992523390655865e-08,
+      "loss": 0.0362,
+      "step": 10686
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.1469227324379572,
+      "learning_rate": 1.9794017942483258e-08,
+      "loss": 0.0308,
+      "step": 10687
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.3411783014450693,
+      "learning_rate": 1.9663234605909465e-08,
+      "loss": 0.0419,
+      "step": 10688
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.63278633193149,
+      "learning_rate": 1.9532883908198185e-08,
+      "loss": 0.0426,
+      "step": 10689
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.5424335388261954,
+      "learning_rate": 1.9402965860672584e-08,
+      "loss": 0.0417,
+      "step": 10690
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.5409700082670148,
+      "learning_rate": 1.927348047461919e-08,
+      "loss": 0.0537,
+      "step": 10691
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.4289587527834118,
+      "learning_rate": 1.914442776128622e-08,
+      "loss": 0.0409,
+      "step": 10692
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.2158352555373946,
+      "learning_rate": 1.9015807731884163e-08,
+      "loss": 0.0304,
+      "step": 10693
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.348840203398288,
+      "learning_rate": 1.8887620397586292e-08,
+      "loss": 0.0408,
+      "step": 10694
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.6089578505695998,
+      "learning_rate": 1.8759865769528153e-08,
+      "loss": 0.0522,
+      "step": 10695
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.4076272186639576,
+      "learning_rate": 1.8632543858807528e-08,
+      "loss": 0.0411,
+      "step": 10696
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.3349742004464702,
+      "learning_rate": 1.850565467648502e-08,
+      "loss": 0.0431,
+      "step": 10697
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.4985578050175992,
+      "learning_rate": 1.8379198233583472e-08,
+      "loss": 0.0474,
+      "step": 10698
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.5666280567917392,
+      "learning_rate": 1.8253174541087437e-08,
+      "loss": 0.0491,
+      "step": 10699
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.4397053228698713,
+      "learning_rate": 1.8127583609945376e-08,
+      "loss": 0.0394,
+      "step": 10700
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.8288756968783817,
+      "learning_rate": 1.8002425451067452e-08,
+      "loss": 0.0426,
+      "step": 10701
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.3427518951434703,
+      "learning_rate": 1.7877700075324966e-08,
+      "loss": 0.0351,
+      "step": 10702
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.5922547653910373,
+      "learning_rate": 1.7753407493553698e-08,
+      "loss": 0.0526,
+      "step": 10703
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.5562393843258577,
+      "learning_rate": 1.762954771655001e-08,
+      "loss": 0.0421,
+      "step": 10704
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.365655872556554,
+      "learning_rate": 1.750612075507474e-08,
+      "loss": 0.0393,
+      "step": 10705
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.46759862072386,
+      "learning_rate": 1.7383126619848756e-08,
+      "loss": 0.0453,
+      "step": 10706
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.345111268280631,
+      "learning_rate": 1.7260565321556843e-08,
+      "loss": 0.0413,
+      "step": 10707
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.4855784706720545,
+      "learning_rate": 1.7138436870846598e-08,
+      "loss": 0.0429,
+      "step": 10708
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.506809403891813,
+      "learning_rate": 1.70167412783262e-08,
+      "loss": 0.0445,
+      "step": 10709
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.315210769585116,
+      "learning_rate": 1.689547855456719e-08,
+      "loss": 0.0389,
+      "step": 10710
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.6047562802574138,
+      "learning_rate": 1.677464871010448e-08,
+      "loss": 0.0457,
+      "step": 10711
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.353384694078905,
+      "learning_rate": 1.6654251755434115e-08,
+      "loss": 0.0372,
+      "step": 10712
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.5232199605381727,
+      "learning_rate": 1.653428770101495e-08,
+      "loss": 0.0411,
+      "step": 10713
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.2670046797244183,
+      "learning_rate": 1.6414756557267542e-08,
+      "loss": 0.0337,
+      "step": 10714
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.5716873497739476,
+      "learning_rate": 1.629565833457636e-08,
+      "loss": 0.0495,
+      "step": 10715
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.5697445399666754,
+      "learning_rate": 1.617699304328757e-08,
+      "loss": 0.0465,
+      "step": 10716
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.3800726614807521,
+      "learning_rate": 1.6058760693708487e-08,
+      "loss": 0.0409,
+      "step": 10717
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.34750262465537,
+      "learning_rate": 1.5940961296110335e-08,
+      "loss": 0.0448,
+      "step": 10718
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4988958278377509,
+      "learning_rate": 1.5823594860726598e-08,
+      "loss": 0.0471,
+      "step": 10719
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.3945111778973382,
+      "learning_rate": 1.5706661397753008e-08,
+      "loss": 0.0412,
+      "step": 10720
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.5667550513175472,
+      "learning_rate": 1.5590160917346443e-08,
+      "loss": 0.0487,
+      "step": 10721
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.7088042943121222,
+      "learning_rate": 1.5474093429628246e-08,
+      "loss": 0.0462,
+      "step": 10722
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.530968566214379,
+      "learning_rate": 1.5358458944680356e-08,
+      "loss": 0.0444,
+      "step": 10723
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.620501369165027,
+      "learning_rate": 1.5243257472549178e-08,
+      "loss": 0.0554,
+      "step": 10724
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.3434823417206685,
+      "learning_rate": 1.51284890232406e-08,
+      "loss": 0.0392,
+      "step": 10725
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.6184885690484216,
+      "learning_rate": 1.5014153606725535e-08,
+      "loss": 0.0531,
+      "step": 10726
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.6383499239853183,
+      "learning_rate": 1.4900251232935482e-08,
+      "loss": 0.0468,
+      "step": 10727
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.420321950898719,
+      "learning_rate": 1.4786781911765857e-08,
+      "loss": 0.0417,
+      "step": 10728
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4099910906282243,
+      "learning_rate": 1.4673745653073223e-08,
+      "loss": 0.0403,
+      "step": 10729
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.5804191938807965,
+      "learning_rate": 1.4561142466677502e-08,
+      "loss": 0.0487,
+      "step": 10730
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.2233289338773374,
+      "learning_rate": 1.4448972362359759e-08,
+      "loss": 0.0344,
+      "step": 10731
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4893824474879294,
+      "learning_rate": 1.433723534986442e-08,
+      "loss": 0.0418,
+      "step": 10732
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.300561263564656,
+      "learning_rate": 1.4225931438897612e-08,
+      "loss": 0.0348,
+      "step": 10733
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.3828893883781943,
+      "learning_rate": 1.411506063912882e-08,
+      "loss": 0.0426,
+      "step": 10734
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.7011372756451106,
+      "learning_rate": 1.4004622960189229e-08,
+      "loss": 0.0489,
+      "step": 10735
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4554272950162535,
+      "learning_rate": 1.3894618411672278e-08,
+      "loss": 0.0427,
+      "step": 10736
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.419167376558054,
+      "learning_rate": 1.3785047003134211e-08,
+      "loss": 0.0455,
+      "step": 10737
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4955770762044773,
+      "learning_rate": 1.3675908744093524e-08,
+      "loss": 0.0418,
+      "step": 10738
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4166534640261086,
+      "learning_rate": 1.3567203644030414e-08,
+      "loss": 0.0371,
+      "step": 10739
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.8107884742851748,
+      "learning_rate": 1.3458931712388434e-08,
+      "loss": 0.0516,
+      "step": 10740
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.2615096438842115,
+      "learning_rate": 1.3351092958573397e-08,
+      "loss": 0.0343,
+      "step": 10741
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.3032461606273722,
+      "learning_rate": 1.3243687391952809e-08,
+      "loss": 0.0427,
+      "step": 10742
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.303561856417339,
+      "learning_rate": 1.3136715021856983e-08,
+      "loss": 0.0374,
+      "step": 10743
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.6545778761700545,
+      "learning_rate": 1.3030175857578487e-08,
+      "loss": 0.0437,
+      "step": 10744
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4699715611802815,
+      "learning_rate": 1.292406990837214e-08,
+      "loss": 0.0408,
+      "step": 10745
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.5282628525093926,
+      "learning_rate": 1.2818397183456122e-08,
+      "loss": 0.045,
+      "step": 10746
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.387864496928228,
+      "learning_rate": 1.2713157692008648e-08,
+      "loss": 0.0349,
+      "step": 10747
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4244961667150988,
+      "learning_rate": 1.2608351443173516e-08,
+      "loss": 0.0405,
+      "step": 10748
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4531064337285766,
+      "learning_rate": 1.2503978446054555e-08,
+      "loss": 0.0427,
+      "step": 10749
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.385745563514159,
+      "learning_rate": 1.2400038709717843e-08,
+      "loss": 0.045,
+      "step": 10750
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.8587841268297296,
+      "learning_rate": 1.2296532243193382e-08,
+      "loss": 0.0558,
+      "step": 10751
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.334795964776019,
+      "learning_rate": 1.2193459055472867e-08,
+      "loss": 0.0359,
+      "step": 10752
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.6644759635299555,
+      "learning_rate": 1.2090819155509137e-08,
+      "loss": 0.0409,
+      "step": 10753
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.5957644796857635,
+      "learning_rate": 1.1988612552219503e-08,
+      "loss": 0.041,
+      "step": 10754
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.4370771175671393,
+      "learning_rate": 1.1886839254482419e-08,
+      "loss": 0.0434,
+      "step": 10755
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.7611440443201483,
+      "learning_rate": 1.178549927113859e-08,
+      "loss": 0.0438,
+      "step": 10756
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.5167430800846882,
+      "learning_rate": 1.168459261099153e-08,
+      "loss": 0.0465,
+      "step": 10757
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.544796971352805,
+      "learning_rate": 1.158411928280645e-08,
+      "loss": 0.0533,
+      "step": 10758
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.274658353266541,
+      "learning_rate": 1.1484079295311923e-08,
+      "loss": 0.0417,
+      "step": 10759
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.4947542967215883,
+      "learning_rate": 1.1384472657198775e-08,
+      "loss": 0.0389,
+      "step": 10760
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.373873197759611,
+      "learning_rate": 1.1285299377118974e-08,
+      "loss": 0.0431,
+      "step": 10761
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.6425062783641782,
+      "learning_rate": 1.1186559463687851e-08,
+      "loss": 0.0457,
+      "step": 10762
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.4095498211526,
+      "learning_rate": 1.1088252925482989e-08,
+      "loss": 0.0412,
+      "step": 10763
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.349540029139377,
+      "learning_rate": 1.0990379771044223e-08,
+      "loss": 0.0434,
+      "step": 10764
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.627067391370231,
+      "learning_rate": 1.0892940008873642e-08,
+      "loss": 0.0494,
+      "step": 10765
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.3155204605664943,
+      "learning_rate": 1.0795933647436141e-08,
+      "loss": 0.0356,
+      "step": 10766
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.3276536092192552,
+      "learning_rate": 1.0699360695158311e-08,
+      "loss": 0.0416,
+      "step": 10767
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.6432370559531282,
+      "learning_rate": 1.0603221160429e-08,
+      "loss": 0.0465,
+      "step": 10768
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.629532341534546,
+      "learning_rate": 1.0507515051600415e-08,
+      "loss": 0.0387,
+      "step": 10769
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.6497503371717654,
+      "learning_rate": 1.0412242376985903e-08,
+      "loss": 0.0538,
+      "step": 10770
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.3868137418797266,
+      "learning_rate": 1.0317403144862182e-08,
+      "loss": 0.0436,
+      "step": 10771
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.2919589733922905,
+      "learning_rate": 1.0222997363468213e-08,
+      "loss": 0.0369,
+      "step": 10772
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.4929805937932075,
+      "learning_rate": 1.0129025041004659e-08,
+      "loss": 0.0405,
+      "step": 10773
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.3469095328420277,
+      "learning_rate": 1.003548618563388e-08,
+      "loss": 0.0428,
+      "step": 10774
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.4184959989244899,
+      "learning_rate": 9.942380805483266e-09,
+      "loss": 0.0428,
+      "step": 10775
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.5614177386903487,
+      "learning_rate": 9.849708908639677e-09,
+      "loss": 0.0479,
+      "step": 10776
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.5611753328101294,
+      "learning_rate": 9.757470503153344e-09,
+      "loss": 0.045,
+      "step": 10777
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.4504061320816564,
+      "learning_rate": 9.6656655970373e-09,
+      "loss": 0.0493,
+      "step": 10778
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.3698716107193216,
+      "learning_rate": 9.574294198267387e-09,
+      "loss": 0.0387,
+      "step": 10779
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.4573844024879703,
+      "learning_rate": 9.48335631477948e-09,
+      "loss": 0.0395,
+      "step": 10780
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.3734655329160572,
+      "learning_rate": 9.39285195447448e-09,
+      "loss": 0.0363,
+      "step": 10781
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.3251746429178852,
+      "learning_rate": 9.302781125213878e-09,
+      "loss": 0.0383,
+      "step": 10782
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.5693584948225776,
+      "learning_rate": 9.213143834822524e-09,
+      "loss": 0.0546,
+      "step": 10783
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.6551295003690372,
+      "learning_rate": 9.123940091086414e-09,
+      "loss": 0.047,
+      "step": 10784
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.5290974854365542,
+      "learning_rate": 9.035169901754902e-09,
+      "loss": 0.0353,
+      "step": 10785
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.62696547558024,
+      "learning_rate": 8.946833274540157e-09,
+      "loss": 0.0466,
+      "step": 10786
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.3240798031847774,
+      "learning_rate": 8.858930217114925e-09,
+      "loss": 0.0354,
+      "step": 10787
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.267955964357376,
+      "learning_rate": 8.771460737115878e-09,
+      "loss": 0.0373,
+      "step": 10788
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.7453964815444112,
+      "learning_rate": 8.684424842140825e-09,
+      "loss": 0.0465,
+      "step": 10789
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.307256082800548,
+      "learning_rate": 8.5978225397515e-09,
+      "loss": 0.0359,
+      "step": 10790
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.4961167991739515,
+      "learning_rate": 8.511653837470212e-09,
+      "loss": 0.0486,
+      "step": 10791
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.6846583699787798,
+      "learning_rate": 8.425918742782646e-09,
+      "loss": 0.051,
+      "step": 10792
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.414505837094659,
+      "learning_rate": 8.340617263136175e-09,
+      "loss": 0.0405,
+      "step": 10793
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.6030134358880264,
+      "learning_rate": 8.255749405941538e-09,
+      "loss": 0.0434,
+      "step": 10794
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.602428884836434,
+      "learning_rate": 8.171315178570616e-09,
+      "loss": 0.0419,
+      "step": 10795
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.842711205773103,
+      "learning_rate": 8.087314588358653e-09,
+      "loss": 0.0483,
+      "step": 10796
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.4852357560728653,
+      "learning_rate": 8.003747642602588e-09,
+      "loss": 0.0444,
+      "step": 10797
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.4311937899110352,
+      "learning_rate": 7.920614348561618e-09,
+      "loss": 0.0429,
+      "step": 10798
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.441337972264241,
+      "learning_rate": 7.837914713457184e-09,
+      "loss": 0.0464,
+      "step": 10799
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.4898772018645408,
+      "learning_rate": 7.755648744474097e-09,
+      "loss": 0.043,
+      "step": 10800
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.4585072850147554,
+      "learning_rate": 7.67381644875831e-09,
+      "loss": 0.0488,
+      "step": 10801
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.6279724664834248,
+      "learning_rate": 7.59241783341913e-09,
+      "loss": 0.0396,
+      "step": 10802
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.498728209324664,
+      "learning_rate": 7.511452905526462e-09,
+      "loss": 0.0464,
+      "step": 10803
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.3981991639232685,
+      "learning_rate": 7.430921672114677e-09,
+      "loss": 0.0381,
+      "step": 10804
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.464355203458042,
+      "learning_rate": 7.350824140178736e-09,
+      "loss": 0.0435,
+      "step": 10805
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.592620134381601,
+      "learning_rate": 7.271160316677517e-09,
+      "loss": 0.0501,
+      "step": 10806
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.3458280643468141,
+      "learning_rate": 7.191930208530485e-09,
+      "loss": 0.0408,
+      "step": 10807
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.6062566831242322,
+      "learning_rate": 7.113133822621021e-09,
+      "loss": 0.0504,
+      "step": 10808
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.4190359223821862,
+      "learning_rate": 7.034771165794208e-09,
+      "loss": 0.0394,
+      "step": 10809
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.5126871254060108,
+      "learning_rate": 6.956842244856266e-09,
+      "loss": 0.0447,
+      "step": 10810
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.322002025450512,
+      "learning_rate": 6.87934706657789e-09,
+      "loss": 0.0392,
+      "step": 10811
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.6501446334555008,
+      "learning_rate": 6.802285637690364e-09,
+      "loss": 0.0524,
+      "step": 10812
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.573952576165649,
+      "learning_rate": 6.725657964888887e-09,
+      "loss": 0.0546,
+      "step": 10813
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.5134178138175283,
+      "learning_rate": 6.649464054829246e-09,
+      "loss": 0.0306,
+      "step": 10814
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.243906246997328,
+      "learning_rate": 6.573703914130591e-09,
+      "loss": 0.0346,
+      "step": 10815
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.7778733964028812,
+      "learning_rate": 6.498377549374324e-09,
+      "loss": 0.0459,
+      "step": 10816
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.5185647835941254,
+      "learning_rate": 6.423484967103544e-09,
+      "loss": 0.0458,
+      "step": 10817
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.4506850808515908,
+      "learning_rate": 6.349026173824713e-09,
+      "loss": 0.0417,
+      "step": 10818
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.316557314909194,
+      "learning_rate": 6.2750011760054355e-09,
+      "loss": 0.0377,
+      "step": 10819
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.613417561484535,
+      "learning_rate": 6.201409980076678e-09,
+      "loss": 0.0508,
+      "step": 10820
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.2832144289214522,
+      "learning_rate": 6.128252592431105e-09,
+      "loss": 0.035,
+      "step": 10821
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.7144320753603994,
+      "learning_rate": 6.055529019423634e-09,
+      "loss": 0.0478,
+      "step": 10822
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.6079462766544612,
+      "learning_rate": 5.983239267371987e-09,
+      "loss": 0.0448,
+      "step": 10823
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.4592229649283994,
+      "learning_rate": 5.911383342556143e-09,
+      "loss": 0.0403,
+      "step": 10824
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.5591005189827973,
+      "learning_rate": 5.839961251217774e-09,
+      "loss": 0.0422,
+      "step": 10825
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.4264799517020308,
+      "learning_rate": 5.768972999561362e-09,
+      "loss": 0.0385,
+      "step": 10826
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.7476446332307087,
+      "learning_rate": 5.698418593754196e-09,
+      "loss": 0.0473,
+      "step": 10827
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.2939486277703305,
+      "learning_rate": 5.628298039924152e-09,
+      "loss": 0.0426,
+      "step": 10828
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.5195819848685794,
+      "learning_rate": 5.558611344163023e-09,
+      "loss": 0.0519,
+      "step": 10829
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.4159970819293244,
+      "learning_rate": 5.489358512524856e-09,
+      "loss": 0.0337,
+      "step": 10830
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.2766911446969194,
+      "learning_rate": 5.4205395510253944e-09,
+      "loss": 0.0362,
+      "step": 10831
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.3779434282763896,
+      "learning_rate": 5.352154465643189e-09,
+      "loss": 0.0377,
+      "step": 10832
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.296236610449945,
+      "learning_rate": 5.284203262318488e-09,
+      "loss": 0.0362,
+      "step": 10833
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.3996965138864614,
+      "learning_rate": 5.216685946953237e-09,
+      "loss": 0.0393,
+      "step": 10834
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.4030681677837804,
+      "learning_rate": 5.14960252541441e-09,
+      "loss": 0.0344,
+      "step": 10835
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.5777290966746682,
+      "learning_rate": 5.082953003528457e-09,
+      "loss": 0.0379,
+      "step": 10836
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.7139913975929897,
+      "learning_rate": 5.016737387085191e-09,
+      "loss": 0.0564,
+      "step": 10837
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.4141745498864458,
+      "learning_rate": 4.950955681837233e-09,
+      "loss": 0.0455,
+      "step": 10838
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.4399323882780701,
+      "learning_rate": 4.885607893498345e-09,
+      "loss": 0.0422,
+      "step": 10839
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.4369208235777924,
+      "learning_rate": 4.8206940277456534e-09,
+      "loss": 0.0387,
+      "step": 10840
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.3570717895802908,
+      "learning_rate": 4.7562140902185364e-09,
+      "loss": 0.0401,
+      "step": 10841
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.7270394626888796,
+      "learning_rate": 4.6921680865169574e-09,
+      "loss": 0.0441,
+      "step": 10842
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.5126939251320637,
+      "learning_rate": 4.6285560222064655e-09,
+      "loss": 0.0413,
+      "step": 10843
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.6160771446346398,
+      "learning_rate": 4.565377902811529e-09,
+      "loss": 0.0427,
+      "step": 10844
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.5580995191338787,
+      "learning_rate": 4.502633733821093e-09,
+      "loss": 0.0414,
+      "step": 10845
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.5602462455976787,
+      "learning_rate": 4.440323520685241e-09,
+      "loss": 0.0524,
+      "step": 10846
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.8680690286145887,
+      "learning_rate": 4.378447268817421e-09,
+      "loss": 0.0518,
+      "step": 10847
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.2513158544551304,
+      "learning_rate": 4.317004983592221e-09,
+      "loss": 0.035,
+      "step": 10848
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.2011321173739833,
+      "learning_rate": 4.25599667034704e-09,
+      "loss": 0.0351,
+      "step": 10849
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.5202535962272616,
+      "learning_rate": 4.195422334382638e-09,
+      "loss": 0.0399,
+      "step": 10850
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.5785499276976593,
+      "learning_rate": 4.1352819809598045e-09,
+      "loss": 0.0471,
+      "step": 10851
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.4711722796643292,
+      "learning_rate": 4.075575615303807e-09,
+      "loss": 0.0446,
+      "step": 10852
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.3915687685053666,
+      "learning_rate": 4.016303242600495e-09,
+      "loss": 0.0421,
+      "step": 10853
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.5622287834229704,
+      "learning_rate": 3.957464867999638e-09,
+      "loss": 0.0438,
+      "step": 10854
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.5061642361001593,
+      "learning_rate": 3.8990604966121504e-09,
+      "loss": 0.0461,
+      "step": 10855
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.843311463282505,
+      "learning_rate": 3.841090133511749e-09,
+      "loss": 0.0555,
+      "step": 10856
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.4738266521322858,
+      "learning_rate": 3.783553783733851e-09,
+      "loss": 0.0458,
+      "step": 10857
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.654708273593146,
+      "learning_rate": 3.72645145227668e-09,
+      "loss": 0.0481,
+      "step": 10858
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.2932316918101079,
+      "learning_rate": 3.6697831441007136e-09,
+      "loss": 0.0349,
+      "step": 10859
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.243118157193055,
+      "learning_rate": 3.613548864129235e-09,
+      "loss": 0.0388,
+      "step": 10860
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.262885487662697,
+      "learning_rate": 3.5577486172466703e-09,
+      "loss": 0.0328,
+      "step": 10861
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.6348210138976664,
+      "learning_rate": 3.5023824083008083e-09,
+      "loss": 0.0399,
+      "step": 10862
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.394156585320726,
+      "learning_rate": 3.4474502421005805e-09,
+      "loss": 0.0428,
+      "step": 10863
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.6725504248972054,
+      "learning_rate": 3.3929521234188358e-09,
+      "loss": 0.0474,
+      "step": 10864
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.5900750531295955,
+      "learning_rate": 3.338888056989009e-09,
+      "loss": 0.0493,
+      "step": 10865
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.53002408772992,
+      "learning_rate": 3.2852580475078997e-09,
+      "loss": 0.0458,
+      "step": 10866
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.8165536663958906,
+      "learning_rate": 3.2320620996345586e-09,
+      "loss": 0.0443,
+      "step": 10867
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.4582502274869606,
+      "learning_rate": 3.1793002179897337e-09,
+      "loss": 0.0421,
+      "step": 10868
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.6026475678728451,
+      "learning_rate": 3.126972407156981e-09,
+      "loss": 0.0518,
+      "step": 10869
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.3202264582545662,
+      "learning_rate": 3.075078671682108e-09,
+      "loss": 0.0331,
+      "step": 10870
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.504591880791957,
+      "learning_rate": 3.023619016072621e-09,
+      "loss": 0.043,
+      "step": 10871
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.6565918996954974,
+      "learning_rate": 2.9725934447993875e-09,
+      "loss": 0.0497,
+      "step": 10872
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.4107936064471835,
+      "learning_rate": 2.9220019622944184e-09,
+      "loss": 0.037,
+      "step": 10873
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.606013548084908,
+      "learning_rate": 2.8718445729530862e-09,
+      "loss": 0.044,
+      "step": 10874
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.2578276030307272,
+      "learning_rate": 2.8221212811324616e-09,
+      "loss": 0.039,
+      "step": 10875
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.49840643551066,
+      "learning_rate": 2.772832091151312e-09,
+      "loss": 0.0453,
+      "step": 10876
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.3789989912753413,
+      "learning_rate": 2.7239770072923223e-09,
+      "loss": 0.0411,
+      "step": 10877
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.5225099461071243,
+      "learning_rate": 2.675556033798765e-09,
+      "loss": 0.0494,
+      "step": 10878
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.2759255895957398,
+      "learning_rate": 2.6275691748767207e-09,
+      "loss": 0.0405,
+      "step": 10879
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.457528019426865,
+      "learning_rate": 2.5800164346961864e-09,
+      "loss": 0.0442,
+      "step": 10880
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.5310093899628099,
+      "learning_rate": 2.532897817386637e-09,
+      "loss": 0.0456,
+      "step": 10881
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.5675491949601992,
+      "learning_rate": 2.4862133270414644e-09,
+      "loss": 0.0407,
+      "step": 10882
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.5227049475229266,
+      "learning_rate": 2.439962967716869e-09,
+      "loss": 0.0499,
+      "step": 10883
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.3894211815770774,
+      "learning_rate": 2.3941467434296372e-09,
+      "loss": 0.0443,
+      "step": 10884
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.3684873567453495,
+      "learning_rate": 2.348764658160474e-09,
+      "loss": 0.0398,
+      "step": 10885
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.5040855349948603,
+      "learning_rate": 2.303816715851781e-09,
+      "loss": 0.0423,
+      "step": 10886
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.8301923480515854,
+      "learning_rate": 2.2593029204076578e-09,
+      "loss": 0.0541,
+      "step": 10887
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.479666917008766,
+      "learning_rate": 2.215223275695011e-09,
+      "loss": 0.0478,
+      "step": 10888
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.553745389209031,
+      "learning_rate": 2.171577785543e-09,
+      "loss": 0.0466,
+      "step": 10889
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.4864910519163095,
+      "learning_rate": 2.128366453743591e-09,
+      "loss": 0.0395,
+      "step": 10890
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.3934181250881392,
+      "learning_rate": 2.085589284050449e-09,
+      "loss": 0.0427,
+      "step": 10891
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.2006737840195014,
+      "learning_rate": 2.0432462801789344e-09,
+      "loss": 0.0365,
+      "step": 10892
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.497205629063281,
+      "learning_rate": 2.0013374458077718e-09,
+      "loss": 0.0418,
+      "step": 10893
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.34771897200011,
+      "learning_rate": 1.959862784577937e-09,
+      "loss": 0.0376,
+      "step": 10894
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.7462236523832473,
+      "learning_rate": 1.9188223000915496e-09,
+      "loss": 0.0551,
+      "step": 10895
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.5678687103542155,
+      "learning_rate": 1.8782159959140898e-09,
+      "loss": 0.0457,
+      "step": 10896
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.5797821269564256,
+      "learning_rate": 1.8380438755738472e-09,
+      "loss": 0.0452,
+      "step": 10897
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.4492515090472524,
+      "learning_rate": 1.798305942559142e-09,
+      "loss": 0.041,
+      "step": 10898
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.5478121971483612,
+      "learning_rate": 1.759002200322768e-09,
+      "loss": 0.0433,
+      "step": 10899
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.5204610573668493,
+      "learning_rate": 1.720132652278661e-09,
+      "loss": 0.0477,
+      "step": 10900
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.4364415337562024,
+      "learning_rate": 1.6816973018035642e-09,
+      "loss": 0.0409,
+      "step": 10901
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.7175872077160956,
+      "learning_rate": 1.6436961522364737e-09,
+      "loss": 0.052,
+      "step": 10902
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.8682631936416245,
+      "learning_rate": 1.6061292068786372e-09,
+      "loss": 0.0597,
+      "step": 10903
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.3354585640062273,
+      "learning_rate": 1.5689964689935555e-09,
+      "loss": 0.04,
+      "step": 10904
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.891281490714632,
+      "learning_rate": 1.5322979418058714e-09,
+      "loss": 0.0561,
+      "step": 10905
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.2928408089002037,
+      "learning_rate": 1.4960336285047005e-09,
+      "loss": 0.0349,
+      "step": 10906
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.420691371532058,
+      "learning_rate": 1.4602035322397456e-09,
+      "loss": 0.0422,
+      "step": 10907
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.477545062235664,
+      "learning_rate": 1.424807656124072e-09,
+      "loss": 0.0397,
+      "step": 10908
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.5958664640141942,
+      "learning_rate": 1.3898460032318872e-09,
+      "loss": 0.0537,
+      "step": 10909
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.3871411859594693,
+      "learning_rate": 1.3553185766007616e-09,
+      "loss": 0.0407,
+      "step": 10910
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.3130826320213982,
+      "learning_rate": 1.3212253792299624e-09,
+      "loss": 0.0387,
+      "step": 10911
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.667983640787204,
+      "learning_rate": 1.2875664140804545e-09,
+      "loss": 0.0528,
+      "step": 10912
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.2482181672550523,
+      "learning_rate": 1.2543416840771206e-09,
+      "loss": 0.0434,
+      "step": 10913
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.6023723463161241,
+      "learning_rate": 1.2215511921059852e-09,
+      "loss": 0.0471,
+      "step": 10914
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.3836231954049276,
+      "learning_rate": 1.189194941015326e-09,
+      "loss": 0.039,
+      "step": 10915
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.4885924304484868,
+      "learning_rate": 1.157272933615672e-09,
+      "loss": 0.0522,
+      "step": 10916
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.4915465173248015,
+      "learning_rate": 1.1257851726809154e-09,
+      "loss": 0.0463,
+      "step": 10917
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.351826167693264,
+      "learning_rate": 1.094731660945536e-09,
+      "loss": 0.0334,
+      "step": 10918
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.5812667154922224,
+      "learning_rate": 1.0641124011068205e-09,
+      "loss": 0.0527,
+      "step": 10919
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.594638045291966,
+      "learning_rate": 1.0339273958259732e-09,
+      "loss": 0.0448,
+      "step": 10920
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.3517937994813007,
+      "learning_rate": 1.004176647724231e-09,
+      "loss": 0.0364,
+      "step": 10921
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.5341344236005836,
+      "learning_rate": 9.748601593861929e-10,
+      "loss": 0.0411,
+      "step": 10922
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.4796411696306626,
+      "learning_rate": 9.459779333587104e-10,
+      "loss": 0.0476,
+      "step": 10923
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.5099239128241653,
+      "learning_rate": 9.175299721503328e-10,
+      "loss": 0.0491,
+      "step": 10924
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.5823395724123108,
+      "learning_rate": 8.895162782324163e-10,
+      "loss": 0.0459,
+      "step": 10925
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.5086060090207962,
+      "learning_rate": 8.619368540391248e-10,
+      "loss": 0.038,
+      "step": 10926
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.5210457148475258,
+      "learning_rate": 8.347917019657647e-10,
+      "loss": 0.0468,
+      "step": 10927
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.4443675287545779,
+      "learning_rate": 8.080808243704496e-10,
+      "loss": 0.0444,
+      "step": 10928
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.3457093965889122,
+      "learning_rate": 7.818042235735457e-10,
+      "loss": 0.0398,
+      "step": 10929
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.27019573383568,
+      "learning_rate": 7.559619018576714e-10,
+      "loss": 0.035,
+      "step": 10930
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.7464677059014857,
+      "learning_rate": 7.305538614682528e-10,
+      "loss": 0.0532,
+      "step": 10931
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.5521159457023526,
+      "learning_rate": 7.055801046113031e-10,
+      "loss": 0.0508,
+      "step": 10932
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.7303688372664419,
+      "learning_rate": 6.810406334573084e-10,
+      "loss": 0.0534,
+      "step": 10933
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.355671809820054,
+      "learning_rate": 6.569354501378966e-10,
+      "loss": 0.0383,
+      "step": 10934
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.4943001310646458,
+      "learning_rate": 6.332645567463935e-10,
+      "loss": 0.0392,
+      "step": 10935
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.5269695855631886,
+      "learning_rate": 6.100279553400424e-10,
+      "loss": 0.0448,
+      "step": 10936
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.6549043342421343,
+      "learning_rate": 5.872256479361182e-10,
+      "loss": 0.0401,
+      "step": 10937
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.5735513500680245,
+      "learning_rate": 5.648576365169245e-10,
+      "loss": 0.0384,
+      "step": 10938
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.4356358875042432,
+      "learning_rate": 5.429239230242411e-10,
+      "loss": 0.0465,
+      "step": 10939
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.4486137075123762,
+      "learning_rate": 5.214245093643211e-10,
+      "loss": 0.0406,
+      "step": 10940
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.4488762550065764,
+      "learning_rate": 5.003593974045596e-10,
+      "loss": 0.0406,
+      "step": 10941
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.389571561467369,
+      "learning_rate": 4.797285889746039e-10,
+      "loss": 0.0368,
+      "step": 10942
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.596618911010025,
+      "learning_rate": 4.5953208586690947e-10,
+      "loss": 0.0504,
+      "step": 10943
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.7320840699695563,
+      "learning_rate": 4.397698898361835e-10,
+      "loss": 0.0525,
+      "step": 10944
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.5823792318049978,
+      "learning_rate": 4.2044200259883095e-10,
+      "loss": 0.0496,
+      "step": 10945
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.397425095126363,
+      "learning_rate": 4.0154842583350895e-10,
+      "loss": 0.0474,
+      "step": 10946
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.6356798080497905,
+      "learning_rate": 3.830891611822374e-10,
+      "loss": 0.0496,
+      "step": 10947
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.6431303351633053,
+      "learning_rate": 3.6506421024762315e-10,
+      "loss": 0.0487,
+      "step": 10948
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.4735679237357644,
+      "learning_rate": 3.4747357459674614e-10,
+      "loss": 0.0452,
+      "step": 10949
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.2572602536817783,
+      "learning_rate": 3.303172557561629e-10,
+      "loss": 0.0399,
+      "step": 10950
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.3881766214196973,
+      "learning_rate": 3.1359525521801326e-10,
+      "loss": 0.0346,
+      "step": 10951
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.6869686987053554,
+      "learning_rate": 2.9730757443335867e-10,
+      "loss": 0.0472,
+      "step": 10952
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.7139200856000782,
+      "learning_rate": 2.814542148177335e-10,
+      "loss": 0.0458,
+      "step": 10953
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.534488477017455,
+      "learning_rate": 2.660351777483694e-10,
+      "loss": 0.0435,
+      "step": 10954
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.624282335540079,
+      "learning_rate": 2.5105046456475047e-10,
+      "loss": 0.0499,
+      "step": 10955
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.3362921800439473,
+      "learning_rate": 2.3650007656805804e-10,
+      "loss": 0.0421,
+      "step": 10956
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.4230969674878908,
+      "learning_rate": 2.2238401502339136e-10,
+      "loss": 0.0468,
+      "step": 10957
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.844619776161194,
+      "learning_rate": 2.0870228115588142e-10,
+      "loss": 0.0503,
+      "step": 10958
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.430522189451049,
+      "learning_rate": 1.9545487615402204e-10,
+      "loss": 0.0407,
+      "step": 10959
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.3522080482763055,
+      "learning_rate": 1.8264180116966956e-10,
+      "loss": 0.0386,
+      "step": 10960
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.5747266886203286,
+      "learning_rate": 1.702630573152675e-10,
+      "loss": 0.0372,
+      "step": 10961
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.3986460141251411,
+      "learning_rate": 1.583186456660668e-10,
+      "loss": 0.0464,
+      "step": 10962
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.435492933928032,
+      "learning_rate": 1.4680856725957094e-10,
+      "loss": 0.0384,
+      "step": 10963
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.2943468033829362,
+      "learning_rate": 1.3573282309609082e-10,
+      "loss": 0.0382,
+      "step": 10964
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.4257102002799198,
+      "learning_rate": 1.250914141370796e-10,
+      "loss": 0.0409,
+      "step": 10965
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.5938799952554827,
+      "learning_rate": 1.1488434130790815e-10,
+      "loss": 0.0462,
+      "step": 10966
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.5987602633605487,
+      "learning_rate": 1.0511160549453448e-10,
+      "loss": 0.0489,
+      "step": 10967
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.786183730671755,
+      "learning_rate": 9.577320754627917e-11,
+      "loss": 0.0528,
+      "step": 10968
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.46937398014273,
+      "learning_rate": 8.686914827416015e-11,
+      "loss": 0.0483,
+      "step": 10969
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.1937786460409998,
+      "learning_rate": 7.839942845144777e-11,
+      "loss": 0.0341,
+      "step": 10970
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.44079617756281,
+      "learning_rate": 7.036404881421988e-11,
+      "loss": 0.0434,
+      "step": 10971
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.7941135845578364,
+      "learning_rate": 6.276301006080676e-11,
+      "loss": 0.0495,
+      "step": 10972
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.5234947691103542,
+      "learning_rate": 5.55963128506809e-11,
+      "loss": 0.0427,
+      "step": 10973
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.445840999807596,
+      "learning_rate": 4.886395780723252e-11,
+      "loss": 0.0334,
+      "step": 10974
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.550023526099395,
+      "learning_rate": 4.256594551499405e-11,
+      "loss": 0.0411,
+      "step": 10975
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.3735763245528048,
+      "learning_rate": 3.6702276520750316e-11,
+      "loss": 0.0434,
+      "step": 10976
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.6566842345750339,
+      "learning_rate": 3.127295133409369e-11,
+      "loss": 0.041,
+      "step": 10977
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.3697366871544565,
+      "learning_rate": 2.6277970426868972e-11,
+      "loss": 0.0471,
+      "step": 10978
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.3988212399486728,
+      "learning_rate": 2.1717334232618235e-11,
+      "loss": 0.0416,
+      "step": 10979
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.6106847688772814,
+      "learning_rate": 1.7591043148246222e-11,
+      "loss": 0.0551,
+      "step": 10980
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.5175493867414471,
+      "learning_rate": 1.3899097531244744e-11,
+      "loss": 0.0467,
+      "step": 10981
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.883069086105246,
+      "learning_rate": 1.0641497703023363e-11,
+      "loss": 0.0578,
+      "step": 10982
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.5729843921762448,
+      "learning_rate": 7.818243946133841e-12,
+      "loss": 0.0428,
+      "step": 10983
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.5778767449570963,
+      "learning_rate": 5.4293365059354676e-12,
+      "loss": 0.0438,
+      "step": 10984
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.6096525037457174,
+      "learning_rate": 3.474775590039947e-12,
+      "loss": 0.0407,
+      "step": 10985
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.584474226349137,
+      "learning_rate": 1.9545613683114027e-12,
+      "loss": 0.0505,
+      "step": 10986
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.4404310764741348,
+      "learning_rate": 8.686939728663746e-13,
+      "loss": 0.0386,
+      "step": 10987
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.524871018292121,
+      "learning_rate": 2.1717349807381937e-13,
+      "loss": 0.0404,
+      "step": 10988
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.3380365262766343,
+      "learning_rate": 0.0,
+      "loss": 0.0335,
+      "step": 10989
+    },
+    {
+      "epoch": 3.0,
+      "step": 10989,
+      "total_flos": 1316175770025984.0,
+      "train_loss": 0.13379205959944065,
+      "train_runtime": 24670.7836,
+      "train_samples_per_second": 28.505,
+      "train_steps_per_second": 0.445
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 10989,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 24000,
+  "total_flos": 1316175770025984.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}