diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,42021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8952551477170994,
+  "eval_steps": 500,
+  "global_step": 6000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.61997274421151,
+      "learning_rate": 9.900990099009901e-08,
+      "loss": 0.3005,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.96129971490389,
+      "learning_rate": 1.9801980198019803e-07,
+      "loss": 1.903,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.059955618049987,
+      "learning_rate": 2.9702970297029703e-07,
+      "loss": 1.9233,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.103831500070694,
+      "learning_rate": 3.9603960396039606e-07,
+      "loss": 1.991,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.392721025912955,
+      "learning_rate": 4.950495049504951e-07,
+      "loss": 2.1935,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.140573034359793,
+      "learning_rate": 5.940594059405941e-07,
+      "loss": 1.9898,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.046640372445655,
+      "learning_rate": 6.930693069306931e-07,
+      "loss": 2.067,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.629696916819253,
+      "learning_rate": 7.920792079207921e-07,
+      "loss": 1.976,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.205985254680062,
+      "learning_rate": 8.910891089108911e-07,
+      "loss": 2.1618,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.46713316887076,
+      "learning_rate": 9.900990099009902e-07,
+      "loss": 1.9472,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.58775167810242,
+      "learning_rate": 1.0891089108910893e-06,
+      "loss": 1.8545,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8016952473974713,
+      "learning_rate": 1.1881188118811881e-06,
+      "loss": 0.2673,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.19759391544971,
+      "learning_rate": 1.2871287128712872e-06,
+      "loss": 1.8575,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.692865737683729,
+      "learning_rate": 1.3861386138613863e-06,
+      "loss": 1.6309,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.933875275893488,
+      "learning_rate": 1.4851485148514852e-06,
+      "loss": 1.7404,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.478750370188157,
+      "learning_rate": 1.5841584158415842e-06,
+      "loss": 1.625,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.872117125031288,
+      "learning_rate": 1.6831683168316833e-06,
+      "loss": 1.8698,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.316127487534764,
+      "learning_rate": 1.7821782178217822e-06,
+      "loss": 1.721,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.345460825151939,
+      "learning_rate": 1.8811881188118813e-06,
+      "loss": 1.7876,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.154912232618392,
+      "learning_rate": 1.9801980198019803e-06,
+      "loss": 1.7299,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.423809728704567,
+      "learning_rate": 2.0792079207920794e-06,
+      "loss": 1.5514,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.057358450273484,
+      "learning_rate": 2.1782178217821785e-06,
+      "loss": 1.4899,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.609038984044864,
+      "learning_rate": 2.2772277227722776e-06,
+      "loss": 1.521,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.523777715195209,
+      "learning_rate": 2.3762376237623762e-06,
+      "loss": 1.4854,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.917318190494613,
+      "learning_rate": 2.4752475247524753e-06,
+      "loss": 1.4383,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6832101082075663,
+      "learning_rate": 2.5742574257425744e-06,
+      "loss": 1.3966,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.11950097422884,
+      "learning_rate": 2.6732673267326735e-06,
+      "loss": 1.3713,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.870341350061341,
+      "learning_rate": 2.7722772277227726e-06,
+      "loss": 1.3955,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.538406561042297,
+      "learning_rate": 2.8712871287128712e-06,
+      "loss": 1.4885,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.897265745516444,
+      "learning_rate": 2.9702970297029703e-06,
+      "loss": 1.3748,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.110082061925306,
+      "learning_rate": 3.0693069306930694e-06,
+      "loss": 1.3526,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1958344879587637,
+      "learning_rate": 3.1683168316831685e-06,
+      "loss": 1.3512,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4997592400440376,
+      "learning_rate": 3.2673267326732676e-06,
+      "loss": 1.343,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5120153248850117,
+      "learning_rate": 3.3663366336633666e-06,
+      "loss": 1.3795,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.966950103870316,
+      "learning_rate": 3.4653465346534653e-06,
+      "loss": 0.2754,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.984780628539569,
+      "learning_rate": 3.5643564356435644e-06,
+      "loss": 1.3494,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9492507200213534,
+      "learning_rate": 3.6633663366336635e-06,
+      "loss": 1.3834,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9985573150804936,
+      "learning_rate": 3.7623762376237625e-06,
+      "loss": 1.3432,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.85356604200625,
+      "learning_rate": 3.861386138613862e-06,
+      "loss": 1.3271,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.693328464433035,
+      "learning_rate": 3.960396039603961e-06,
+      "loss": 1.3504,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6422555223277864,
+      "learning_rate": 4.05940594059406e-06,
+      "loss": 1.3333,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0890843341925964,
+      "learning_rate": 4.158415841584159e-06,
+      "loss": 1.3359,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6806443316179953,
+      "learning_rate": 4.2574257425742575e-06,
+      "loss": 1.3349,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.40103168013117,
+      "learning_rate": 4.356435643564357e-06,
+      "loss": 1.2479,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6170663606307594,
+      "learning_rate": 4.455445544554456e-06,
+      "loss": 1.2362,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9458924900429565,
+      "learning_rate": 4.554455445544555e-06,
+      "loss": 0.3158,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.018752566429295,
+      "learning_rate": 4.653465346534654e-06,
+      "loss": 1.2055,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9931288795038689,
+      "learning_rate": 4.7524752475247525e-06,
+      "loss": 1.2299,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1795775560871347,
+      "learning_rate": 4.851485148514852e-06,
+      "loss": 1.3277,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0673924508492887,
+      "learning_rate": 4.950495049504951e-06,
+      "loss": 0.2725,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2237872077237144,
+      "learning_rate": 5.04950495049505e-06,
+      "loss": 1.2136,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0788688977429324,
+      "learning_rate": 5.148514851485149e-06,
+      "loss": 1.22,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0966509718753015,
+      "learning_rate": 5.247524752475248e-06,
+      "loss": 1.2273,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.734413949144073,
+      "learning_rate": 5.346534653465347e-06,
+      "loss": 1.149,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.121929724048015,
+      "learning_rate": 5.4455445544554465e-06,
+      "loss": 1.3127,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7037759311356657,
+      "learning_rate": 5.544554455445545e-06,
+      "loss": 1.2493,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6789600438668426,
+      "learning_rate": 5.643564356435644e-06,
+      "loss": 1.2193,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4444259630024483,
+      "learning_rate": 5.7425742574257425e-06,
+      "loss": 1.2431,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.556882297525969,
+      "learning_rate": 5.841584158415842e-06,
+      "loss": 1.2741,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.238520138493385,
+      "learning_rate": 5.940594059405941e-06,
+      "loss": 1.157,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.415764043872665,
+      "learning_rate": 6.03960396039604e-06,
+      "loss": 1.1484,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3190309713539,
+      "learning_rate": 6.138613861386139e-06,
+      "loss": 1.2549,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1736754940496277,
+      "learning_rate": 6.237623762376238e-06,
+      "loss": 1.1193,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.095907947880233,
+      "learning_rate": 6.336633663366337e-06,
+      "loss": 1.1703,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.730595662137738,
+      "learning_rate": 6.4356435643564364e-06,
+      "loss": 1.1338,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3472798025027144,
+      "learning_rate": 6.534653465346535e-06,
+      "loss": 1.0909,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.437408914759312,
+      "learning_rate": 6.633663366336635e-06,
+      "loss": 1.2125,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8735235958026046,
+      "learning_rate": 6.732673267326733e-06,
+      "loss": 1.2269,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1360912316858367,
+      "learning_rate": 6.831683168316833e-06,
+      "loss": 1.161,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.134595139964946,
+      "learning_rate": 6.930693069306931e-06,
+      "loss": 1.1717,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.288370730967829,
+      "learning_rate": 7.02970297029703e-06,
+      "loss": 1.1607,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9900389604402784,
+      "learning_rate": 7.128712871287129e-06,
+      "loss": 1.0875,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8977512995416834,
+      "learning_rate": 7.227722772277228e-06,
+      "loss": 1.037,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.38588961482632,
+      "learning_rate": 7.326732673267327e-06,
+      "loss": 1.0846,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9794592682896341,
+      "learning_rate": 7.425742574257426e-06,
+      "loss": 1.2375,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9059594403889988,
+      "learning_rate": 7.524752475247525e-06,
+      "loss": 1.1174,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0406430966361695,
+      "learning_rate": 7.6237623762376246e-06,
+      "loss": 0.3022,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3048006434065953,
+      "learning_rate": 7.722772277227724e-06,
+      "loss": 1.065,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.286878240228043,
+      "learning_rate": 7.821782178217822e-06,
+      "loss": 1.2009,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.950766978671876,
+      "learning_rate": 7.920792079207921e-06,
+      "loss": 1.0968,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2562402312894143,
+      "learning_rate": 8.019801980198021e-06,
+      "loss": 1.0641,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4273820131869353,
+      "learning_rate": 8.11881188118812e-06,
+      "loss": 1.093,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.748782091404448,
+      "learning_rate": 8.217821782178218e-06,
+      "loss": 1.0757,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1066063495283025,
+      "learning_rate": 8.316831683168318e-06,
+      "loss": 1.0849,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.226804248334672,
+      "learning_rate": 8.415841584158416e-06,
+      "loss": 1.2182,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.465113508694537,
+      "learning_rate": 8.514851485148515e-06,
+      "loss": 1.1054,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0314927432927967,
+      "learning_rate": 8.613861386138615e-06,
+      "loss": 1.2359,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9051492201816909,
+      "learning_rate": 8.712871287128714e-06,
+      "loss": 1.1209,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5454329213950686,
+      "learning_rate": 8.811881188118812e-06,
+      "loss": 1.2238,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.277898763351694,
+      "learning_rate": 8.910891089108911e-06,
+      "loss": 1.1183,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1921439263709614,
+      "learning_rate": 9.009900990099011e-06,
+      "loss": 1.2191,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1118277735762523,
+      "learning_rate": 9.10891089108911e-06,
+      "loss": 1.1306,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.094034560815147,
+      "learning_rate": 9.20792079207921e-06,
+      "loss": 1.0911,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1956751649437387,
+      "learning_rate": 9.306930693069308e-06,
+      "loss": 1.1,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.568304117862986,
+      "learning_rate": 9.405940594059405e-06,
+      "loss": 1.1316,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0568845808895038,
+      "learning_rate": 9.504950495049505e-06,
+      "loss": 1.1461,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9863236557601718,
+      "learning_rate": 9.603960396039604e-06,
+      "loss": 1.0696,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9849082068957913,
+      "learning_rate": 9.702970297029704e-06,
+      "loss": 1.1677,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.161881947466254,
+      "learning_rate": 9.801980198019802e-06,
+      "loss": 1.2309,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5290291810230108,
+      "learning_rate": 9.900990099009901e-06,
+      "loss": 1.0963,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8701775563508618,
+      "learning_rate": 1e-05,
+      "loss": 1.1526,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9515200801588286,
+      "learning_rate": 1.00990099009901e-05,
+      "loss": 1.0315,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9314667803813104,
+      "learning_rate": 1.01980198019802e-05,
+      "loss": 1.1808,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.030072975262262,
+      "learning_rate": 1.0297029702970298e-05,
+      "loss": 1.1477,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5820036091674017,
+      "learning_rate": 1.0396039603960397e-05,
+      "loss": 1.1251,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.281488203650231,
+      "learning_rate": 1.0495049504950497e-05,
+      "loss": 1.1243,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1578590256842682,
+      "learning_rate": 1.0594059405940596e-05,
+      "loss": 1.1695,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1702469274118794,
+      "learning_rate": 1.0693069306930694e-05,
+      "loss": 1.214,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.958633320084963,
+      "learning_rate": 1.0792079207920793e-05,
+      "loss": 1.0675,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7990099787040412,
+      "learning_rate": 1.0891089108910893e-05,
+      "loss": 1.0354,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3100199739513276,
+      "learning_rate": 1.0990099009900992e-05,
+      "loss": 1.1947,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9146437443620536,
+      "learning_rate": 1.108910891089109e-05,
+      "loss": 1.1562,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0933911830979057,
+      "learning_rate": 1.118811881188119e-05,
+      "loss": 1.1122,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.147201848587997,
+      "learning_rate": 1.1287128712871288e-05,
+      "loss": 1.1182,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.224570092314074,
+      "learning_rate": 1.1386138613861385e-05,
+      "loss": 1.1405,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.353149108522584,
+      "learning_rate": 1.1485148514851485e-05,
+      "loss": 1.1476,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2498791946987926,
+      "learning_rate": 1.1584158415841584e-05,
+      "loss": 1.051,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3488301438133083,
+      "learning_rate": 1.1683168316831684e-05,
+      "loss": 1.1357,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.536534949887731,
+      "learning_rate": 1.1782178217821782e-05,
+      "loss": 1.1184,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4491744166716476,
+      "learning_rate": 1.1881188118811881e-05,
+      "loss": 1.1291,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.254394405663255,
+      "learning_rate": 1.198019801980198e-05,
+      "loss": 1.0794,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.302067290368141,
+      "learning_rate": 1.207920792079208e-05,
+      "loss": 1.0474,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.133676800950005,
+      "learning_rate": 1.217821782178218e-05,
+      "loss": 1.1153,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0990622122422082,
+      "learning_rate": 1.2277227722772278e-05,
+      "loss": 1.1387,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3939641900262445,
+      "learning_rate": 1.2376237623762377e-05,
+      "loss": 1.1263,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.469167927484086,
+      "learning_rate": 1.2475247524752477e-05,
+      "loss": 1.1487,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.11531848687401,
+      "learning_rate": 1.2574257425742576e-05,
+      "loss": 1.1051,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9339220307775706,
+      "learning_rate": 1.2673267326732674e-05,
+      "loss": 1.0821,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.363460636424139,
+      "learning_rate": 1.2772277227722773e-05,
+      "loss": 1.1555,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2165294099352177,
+      "learning_rate": 1.2871287128712873e-05,
+      "loss": 1.0388,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.136400737843224,
+      "learning_rate": 1.2970297029702972e-05,
+      "loss": 1.0771,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9733133562414973,
+      "learning_rate": 1.306930693069307e-05,
+      "loss": 1.0488,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.831830200762256,
+      "learning_rate": 1.316831683168317e-05,
+      "loss": 1.0305,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4462152054957795,
+      "learning_rate": 1.326732673267327e-05,
+      "loss": 1.0456,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9833745142641062,
+      "learning_rate": 1.3366336633663369e-05,
+      "loss": 1.1242,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.245339693291035,
+      "learning_rate": 1.3465346534653467e-05,
+      "loss": 1.0501,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.141274120592458,
+      "learning_rate": 1.3564356435643566e-05,
+      "loss": 1.0582,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.374277998011458,
+      "learning_rate": 1.3663366336633666e-05,
+      "loss": 1.0625,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9395677478874662,
+      "learning_rate": 1.3762376237623762e-05,
+      "loss": 1.0507,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0312424081880573,
+      "learning_rate": 1.3861386138613861e-05,
+      "loss": 1.1991,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3908910540964516,
+      "learning_rate": 1.396039603960396e-05,
+      "loss": 1.1068,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8709075890872888,
+      "learning_rate": 1.405940594059406e-05,
+      "loss": 1.0952,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.000991169628112,
+      "learning_rate": 1.4158415841584158e-05,
+      "loss": 1.0709,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0796126377623683,
+      "learning_rate": 1.4257425742574257e-05,
+      "loss": 1.0188,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.781014562824781,
+      "learning_rate": 1.4356435643564357e-05,
+      "loss": 1.1252,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3517136145035034,
+      "learning_rate": 1.4455445544554456e-05,
+      "loss": 1.0104,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8762707854357925,
+      "learning_rate": 1.4554455445544556e-05,
+      "loss": 1.0909,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.380889665506398,
+      "learning_rate": 1.4653465346534654e-05,
+      "loss": 1.0343,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3265227494389067,
+      "learning_rate": 1.4752475247524753e-05,
+      "loss": 1.0973,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8184422639009856,
+      "learning_rate": 1.4851485148514853e-05,
+      "loss": 1.1392,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3622326550140116,
+      "learning_rate": 1.4950495049504952e-05,
+      "loss": 1.0573,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3223206984371676,
+      "learning_rate": 1.504950495049505e-05,
+      "loss": 1.084,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9331405224070533,
+      "learning_rate": 1.514851485148515e-05,
+      "loss": 1.0064,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.824550576019055,
+      "learning_rate": 1.5247524752475249e-05,
+      "loss": 1.1104,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8952197937491162,
+      "learning_rate": 1.534653465346535e-05,
+      "loss": 1.0337,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2666424543945978,
+      "learning_rate": 1.5445544554455448e-05,
+      "loss": 0.3005,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8195748164659915,
+      "learning_rate": 1.5544554455445548e-05,
+      "loss": 1.0188,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9899061164398222,
+      "learning_rate": 1.5643564356435644e-05,
+      "loss": 1.0611,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8363408315251817,
+      "learning_rate": 1.5742574257425743e-05,
+      "loss": 1.0527,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8801020982340528,
+      "learning_rate": 1.5841584158415843e-05,
+      "loss": 1.1418,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1893456795945476,
+      "learning_rate": 1.5940594059405942e-05,
+      "loss": 1.1231,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1579716933776347,
+      "learning_rate": 1.6039603960396042e-05,
+      "loss": 1.065,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1190706654616207,
+      "learning_rate": 1.613861386138614e-05,
+      "loss": 1.0472,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9383266093102474,
+      "learning_rate": 1.623762376237624e-05,
+      "loss": 1.0465,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.882368165924385,
+      "learning_rate": 1.6336633663366337e-05,
+      "loss": 1.0295,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.063286986987607,
+      "learning_rate": 1.6435643564356436e-05,
+      "loss": 1.0524,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1721790601730753,
+      "learning_rate": 1.6534653465346536e-05,
+      "loss": 1.0208,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7842244798772222,
+      "learning_rate": 1.6633663366336635e-05,
+      "loss": 1.0414,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8412095101139438,
+      "learning_rate": 1.6732673267326735e-05,
+      "loss": 1.0799,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0135741890005785,
+      "learning_rate": 1.683168316831683e-05,
+      "loss": 1.1115,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9897759857740758,
+      "learning_rate": 1.693069306930693e-05,
+      "loss": 1.0938,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9745107771408377,
+      "learning_rate": 1.702970297029703e-05,
+      "loss": 1.1316,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8723737443904078,
+      "learning_rate": 1.712871287128713e-05,
+      "loss": 1.1484,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9620152153533466,
+      "learning_rate": 1.722772277227723e-05,
+      "loss": 1.0567,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6571345160418192,
+      "learning_rate": 1.732673267326733e-05,
+      "loss": 0.9909,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3155871128220933,
+      "learning_rate": 1.7425742574257428e-05,
+      "loss": 1.0136,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8975985315668868,
+      "learning_rate": 1.7524752475247528e-05,
+      "loss": 1.0782,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0331616612731165,
+      "learning_rate": 1.7623762376237624e-05,
+      "loss": 1.1173,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.292007435673564,
+      "learning_rate": 1.7722772277227723e-05,
+      "loss": 1.0251,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2625782524323026,
+      "learning_rate": 1.7821782178217823e-05,
+      "loss": 1.0902,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.44840691692822,
+      "learning_rate": 1.7920792079207922e-05,
+      "loss": 1.012,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1918713608144023,
+      "learning_rate": 1.8019801980198022e-05,
+      "loss": 1.0646,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6412360098068943,
+      "learning_rate": 1.811881188118812e-05,
+      "loss": 1.0053,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1715811901191255,
+      "learning_rate": 1.821782178217822e-05,
+      "loss": 1.0353,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.721914291625859,
+      "learning_rate": 1.831683168316832e-05,
+      "loss": 1.0385,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8854165035497128,
+      "learning_rate": 1.841584158415842e-05,
+      "loss": 1.0852,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.094190246998848,
+      "learning_rate": 1.8514851485148516e-05,
+      "loss": 0.9552,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7715648330031164,
+      "learning_rate": 1.8613861386138615e-05,
+      "loss": 1.0648,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9993014093448769,
+      "learning_rate": 1.8712871287128715e-05,
+      "loss": 1.0375,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1229396887764107,
+      "learning_rate": 1.881188118811881e-05,
+      "loss": 1.1171,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.675628067421174,
+      "learning_rate": 1.891089108910891e-05,
+      "loss": 1.1088,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.967944255637049,
+      "learning_rate": 1.900990099009901e-05,
+      "loss": 1.0119,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.140395665588024,
+      "learning_rate": 1.910891089108911e-05,
+      "loss": 1.1057,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.041794683901522,
+      "learning_rate": 1.920792079207921e-05,
+      "loss": 1.047,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1757371151486833,
+      "learning_rate": 1.930693069306931e-05,
+      "loss": 1.05,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0122452148872787,
+      "learning_rate": 1.9405940594059408e-05,
+      "loss": 1.0408,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.453726450269774,
+      "learning_rate": 1.9504950495049508e-05,
+      "loss": 1.05,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9361060861916906,
+      "learning_rate": 1.9603960396039604e-05,
+      "loss": 0.9558,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0977788104405524,
+      "learning_rate": 1.9702970297029703e-05,
+      "loss": 1.0304,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7380497408748563,
+      "learning_rate": 1.9801980198019803e-05,
+      "loss": 1.0248,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.061102223700244,
+      "learning_rate": 1.9900990099009902e-05,
+      "loss": 1.066,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8769054208304206,
+      "learning_rate": 2e-05,
+      "loss": 1.0021,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9700349198522595,
+      "learning_rate": 1.9999998831999504e-05,
+      "loss": 1.0194,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1649539812066867,
+      "learning_rate": 1.9999995327998284e-05,
+      "loss": 1.0191,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.244717044191121,
+      "learning_rate": 1.9999989487997156e-05,
+      "loss": 1.0838,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.165347941470964,
+      "learning_rate": 1.9999981311997488e-05,
+      "loss": 0.9735,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8396468310253444,
+      "learning_rate": 1.999997080000119e-05,
+      "loss": 1.0186,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.084120466227254,
+      "learning_rate": 1.999995795201072e-05,
+      "loss": 1.0204,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.012804798367043,
+      "learning_rate": 1.9999942768029073e-05,
+      "loss": 0.9858,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8934152718067327,
+      "learning_rate": 1.99999252480598e-05,
+      "loss": 1.051,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8814071954750615,
+      "learning_rate": 1.9999905392106993e-05,
+      "loss": 1.0256,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.015441666257054,
+      "learning_rate": 1.9999883200175286e-05,
+      "loss": 0.9485,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.884512155912271,
+      "learning_rate": 1.9999858672269874e-05,
+      "loss": 1.0199,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9580818739678238,
+      "learning_rate": 1.9999831808396477e-05,
+      "loss": 0.3055,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9852759613789064,
+      "learning_rate": 1.999980260856137e-05,
+      "loss": 1.1423,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.966574727366982,
+      "learning_rate": 1.9999771072771384e-05,
+      "loss": 1.1323,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7852930104230016,
+      "learning_rate": 1.9999737201033877e-05,
+      "loss": 1.0521,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7433767151616615,
+      "learning_rate": 1.999970099335676e-05,
+      "loss": 1.082,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.237094436350509,
+      "learning_rate": 1.99996624497485e-05,
+      "loss": 1.1611,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.724866004461281,
+      "learning_rate": 1.9999621570218092e-05,
+      "loss": 1.0491,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9428030588375604,
+      "learning_rate": 1.999957835477509e-05,
+      "loss": 1.0146,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0752598382230425,
+      "learning_rate": 1.999953280342959e-05,
+      "loss": 1.115,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.759065001688073,
+      "learning_rate": 1.9999484916192225e-05,
+      "loss": 0.9559,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.037851146449414,
+      "learning_rate": 1.9999434693074192e-05,
+      "loss": 1.0458,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8568434491521704,
+      "learning_rate": 1.999938213408722e-05,
+      "loss": 1.1218,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.082233029119249,
+      "learning_rate": 1.9999327239243586e-05,
+      "loss": 1.0666,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.073497027934702,
+      "learning_rate": 1.9999270008556108e-05,
+      "loss": 1.0257,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8903854115847758,
+      "learning_rate": 1.9999210442038164e-05,
+      "loss": 1.0818,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5869163536214277,
+      "learning_rate": 1.9999148539703662e-05,
+      "loss": 1.0493,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.266002510945593,
+      "learning_rate": 1.9999084301567066e-05,
+      "loss": 1.1063,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9129603073518342,
+      "learning_rate": 1.9999017727643378e-05,
+      "loss": 1.0675,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8782768186563432,
+      "learning_rate": 1.9998948817948157e-05,
+      "loss": 1.0875,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2753283414379606,
+      "learning_rate": 1.9998877572497493e-05,
+      "loss": 0.9763,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0831375564536687,
+      "learning_rate": 1.9998803991308036e-05,
+      "loss": 1.1391,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.766648181745103,
+      "learning_rate": 1.999872807439697e-05,
+      "loss": 0.9554,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.726816959099485,
+      "learning_rate": 1.9998649821782027e-05,
+      "loss": 1.0681,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.955048102606591,
+      "learning_rate": 1.999856923348149e-05,
+      "loss": 1.0445,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8429625110431678,
+      "learning_rate": 1.9998486309514184e-05,
+      "loss": 1.0983,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.985602888795277,
+      "learning_rate": 1.999840104989948e-05,
+      "loss": 1.1041,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.080965162810293,
+      "learning_rate": 1.9998313454657295e-05,
+      "loss": 1.1206,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9744983192855934,
+      "learning_rate": 1.9998223523808092e-05,
+      "loss": 0.9096,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7847360350104817,
+      "learning_rate": 1.9998131257372878e-05,
+      "loss": 1.0918,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9233913080632643,
+      "learning_rate": 1.9998036655373206e-05,
+      "loss": 1.0834,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9163129655517637,
+      "learning_rate": 1.9997939717831173e-05,
+      "loss": 0.9963,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0982153571844986,
+      "learning_rate": 1.9997840444769428e-05,
+      "loss": 1.0588,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9450012708748048,
+      "learning_rate": 1.9997738836211157e-05,
+      "loss": 1.023,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.908639429438158,
+      "learning_rate": 1.99976348921801e-05,
+      "loss": 0.9814,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8994625310916364,
+      "learning_rate": 1.9997528612700536e-05,
+      "loss": 1.0355,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9360007341939889,
+      "learning_rate": 1.999741999779729e-05,
+      "loss": 1.0732,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8855786798393648,
+      "learning_rate": 1.999730904749574e-05,
+      "loss": 1.0306,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8928087643137284,
+      "learning_rate": 1.9997195761821797e-05,
+      "loss": 1.0616,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5177678894190167,
+      "learning_rate": 1.9997080140801932e-05,
+      "loss": 0.9914,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9171654850753623,
+      "learning_rate": 1.9996962184463147e-05,
+      "loss": 1.0288,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.696907802479002,
+      "learning_rate": 1.9996841892833e-05,
+      "loss": 1.0123,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.159671670460747,
+      "learning_rate": 1.9996719265939594e-05,
+      "loss": 1.0899,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.026016334585346,
+      "learning_rate": 1.999659430381157e-05,
+      "loss": 0.9954,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.046475194876582,
+      "learning_rate": 1.999646700647812e-05,
+      "loss": 1.006,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7231000352423613,
+      "learning_rate": 1.999633737396898e-05,
+      "loss": 1.0283,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8544222339937602,
+      "learning_rate": 1.9996205406314434e-05,
+      "loss": 1.0276,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6873838916765134,
+      "learning_rate": 1.9996071103545313e-05,
+      "loss": 1.0372,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8972473662410652,
+      "learning_rate": 1.9995934465692984e-05,
+      "loss": 1.0258,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.82234148294883,
+      "learning_rate": 1.9995795492789368e-05,
+      "loss": 1.0849,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.862813177036648,
+      "learning_rate": 1.999565418486693e-05,
+      "loss": 1.0573,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.090188627218271,
+      "learning_rate": 1.999551054195868e-05,
+      "loss": 0.9734,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7376482121628392,
+      "learning_rate": 1.9995364564098166e-05,
+      "loss": 0.9733,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2116963201468174,
+      "learning_rate": 1.99952162513195e-05,
+      "loss": 1.0725,
+      "step": 266
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.250349703077166,
+      "learning_rate": 1.9995065603657317e-05,
+      "loss": 0.9851,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0398888877351142,
+      "learning_rate": 1.9994912621146814e-05,
+      "loss": 1.0597,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9515067854595212,
+      "learning_rate": 1.999475730382373e-05,
+      "loss": 1.0157,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.018554443890871,
+      "learning_rate": 1.9994599651724345e-05,
+      "loss": 1.1006,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7905254733302385,
+      "learning_rate": 1.9994439664885484e-05,
+      "loss": 1.0788,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9693304466436599,
+      "learning_rate": 1.999427734334452e-05,
+      "loss": 0.9886,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7373140794779989,
+      "learning_rate": 1.9994112687139373e-05,
+      "loss": 0.9662,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9108285900577464,
+      "learning_rate": 1.9993945696308508e-05,
+      "loss": 1.0296,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7065143719243163,
+      "learning_rate": 1.9993776370890932e-05,
+      "loss": 1.0147,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2489597462270825,
+      "learning_rate": 1.9993604710926203e-05,
+      "loss": 0.9634,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7126272748330709,
+      "learning_rate": 1.9993430716454415e-05,
+      "loss": 0.8735,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.121608926231665,
+      "learning_rate": 1.9993254387516216e-05,
+      "loss": 0.9658,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.95073604202899,
+      "learning_rate": 1.99930757241528e-05,
+      "loss": 1.033,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7314585208120084,
+      "learning_rate": 1.9992894726405894e-05,
+      "loss": 0.9172,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.828127203842204,
+      "learning_rate": 1.9992711394317787e-05,
+      "loss": 0.9255,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.807205312990042,
+      "learning_rate": 1.9992525727931303e-05,
+      "loss": 1.0163,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9628912323148937,
+      "learning_rate": 1.9992337727289813e-05,
+      "loss": 0.9669,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0469560934703117,
+      "learning_rate": 1.9992147392437235e-05,
+      "loss": 1.0973,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3830503454477934,
+      "learning_rate": 1.999195472341803e-05,
+      "loss": 0.9828,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0586659085839196,
+      "learning_rate": 1.999175972027721e-05,
+      "loss": 1.057,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9353614779926795,
+      "learning_rate": 1.9991562383060316e-05,
+      "loss": 1.0074,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8419681886062762,
+      "learning_rate": 1.999136271181346e-05,
+      "loss": 1.0666,
+      "step": 288
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7879527348103896,
+      "learning_rate": 1.999116070658328e-05,
+      "loss": 1.0565,
+      "step": 289
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9199725384008703,
+      "learning_rate": 1.999095636741696e-05,
+      "loss": 1.0567,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9017182933614696,
+      "learning_rate": 1.999074969436224e-05,
+      "loss": 1.0614,
+      "step": 291
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8817085773345408,
+      "learning_rate": 1.9990540687467394e-05,
+      "loss": 0.9546,
+      "step": 292
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7456010052817308,
+      "learning_rate": 1.999032934678125e-05,
+      "loss": 1.09,
+      "step": 293
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8772906893987726,
+      "learning_rate": 1.9990115672353176e-05,
+      "loss": 1.0998,
+      "step": 294
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7140211617623706,
+      "learning_rate": 1.998989966423308e-05,
+      "loss": 0.9885,
+      "step": 295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.020093630177855,
+      "learning_rate": 1.9989681322471434e-05,
+      "loss": 0.9418,
+      "step": 296
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7061585698765522,
+      "learning_rate": 1.9989460647119232e-05,
+      "loss": 1.0078,
+      "step": 297
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.677045096579522,
+      "learning_rate": 1.998923763822803e-05,
+      "loss": 1.0491,
+      "step": 298
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.736300799709409,
+      "learning_rate": 1.9989012295849917e-05,
+      "loss": 1.0532,
+      "step": 299
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.672784489958878,
+      "learning_rate": 1.998878462003754e-05,
+      "loss": 1.0298,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6542057723864698,
+      "learning_rate": 1.998855461084408e-05,
+      "loss": 0.9925,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.024508696247414,
+      "learning_rate": 1.998832226832327e-05,
+      "loss": 1.096,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8182449765794237,
+      "learning_rate": 1.998808759252938e-05,
+      "loss": 1.0475,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.992777576197344,
+      "learning_rate": 1.9987850583517232e-05,
+      "loss": 1.0334,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.784599012720452,
+      "learning_rate": 1.9987611241342196e-05,
+      "loss": 1.0242,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7102947181663875,
+      "learning_rate": 1.998736956606018e-05,
+      "loss": 0.9414,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.82319059334704,
+      "learning_rate": 1.9987125557727633e-05,
+      "loss": 1.0131,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8249288275636182,
+      "learning_rate": 1.9986879216401562e-05,
+      "loss": 1.029,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1554929681087978,
+      "learning_rate": 1.9986630542139513e-05,
+      "loss": 0.3005,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.283899568144886,
+      "learning_rate": 1.9986379534999577e-05,
+      "loss": 1.0279,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8502366709886288,
+      "learning_rate": 1.9986126195040384e-05,
+      "loss": 0.3008,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9804678198381456,
+      "learning_rate": 1.9985870522321118e-05,
+      "loss": 1.027,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8523264206389816,
+      "learning_rate": 1.9985612516901504e-05,
+      "loss": 0.9465,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7362400017550708,
+      "learning_rate": 1.9985352178841807e-05,
+      "loss": 0.9668,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8851263402754863,
+      "learning_rate": 1.998508950820285e-05,
+      "loss": 1.0016,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.880179349622182,
+      "learning_rate": 1.9984824505045988e-05,
+      "loss": 1.0792,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6622596229035727,
+      "learning_rate": 1.9984557169433126e-05,
+      "loss": 0.9558,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7383853473997237,
+      "learning_rate": 1.998428750142672e-05,
+      "loss": 1.082,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7835925618124706,
+      "learning_rate": 1.998401550108975e-05,
+      "loss": 0.8812,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6013586634606467,
+      "learning_rate": 1.9983741168485772e-05,
+      "loss": 1.0246,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8094094972336086,
+      "learning_rate": 1.998346450367886e-05,
+      "loss": 0.9858,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8180493040233767,
+      "learning_rate": 1.9983185506733643e-05,
+      "loss": 1.0272,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2021418975259297,
+      "learning_rate": 1.9982904177715297e-05,
+      "loss": 1.05,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0115654312892914,
+      "learning_rate": 1.9982620516689544e-05,
+      "loss": 1.0401,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0398671984922796,
+      "learning_rate": 1.9982334523722643e-05,
+      "loss": 0.9548,
+      "step": 325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1646276400547193,
+      "learning_rate": 1.9982046198881403e-05,
+      "loss": 1.0689,
+      "step": 326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1778496983117752,
+      "learning_rate": 1.9981755542233175e-05,
+      "loss": 0.3349,
+      "step": 327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9221729245359622,
+      "learning_rate": 1.998146255384586e-05,
+      "loss": 1.0653,
+      "step": 328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.958836273193452,
+      "learning_rate": 1.9981167233787898e-05,
+      "loss": 1.0531,
+      "step": 329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.855521367983362,
+      "learning_rate": 1.9980869582128274e-05,
+      "loss": 0.9553,
+      "step": 330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2028664105751212,
+      "learning_rate": 1.9980569598936524e-05,
+      "loss": 0.9979,
+      "step": 331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7059139757042052,
+      "learning_rate": 1.9980267284282718e-05,
+      "loss": 1.0646,
+      "step": 332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.864030614349971,
+      "learning_rate": 1.997996263823748e-05,
+      "loss": 1.012,
+      "step": 333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0264614662391174,
+      "learning_rate": 1.997965566087198e-05,
+      "loss": 0.9927,
+      "step": 334
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7346730357420757,
+      "learning_rate": 1.997934635225792e-05,
+      "loss": 1.0399,
+      "step": 335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8750773135775893,
+      "learning_rate": 1.9979034712467556e-05,
+      "loss": 0.9405,
+      "step": 336
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0466712475331534,
+      "learning_rate": 1.9978720741573693e-05,
+      "loss": 1.0562,
+      "step": 337
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0050362975564058,
+      "learning_rate": 1.997840443964967e-05,
+      "loss": 1.0619,
+      "step": 338
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7409079251133657,
+      "learning_rate": 1.9978085806769375e-05,
+      "loss": 1.0322,
+      "step": 339
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9528789498793981,
+      "learning_rate": 1.9977764843007242e-05,
+      "loss": 0.9718,
+      "step": 340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6951967762418776,
+      "learning_rate": 1.9977441548438246e-05,
+      "loss": 0.993,
+      "step": 341
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.801089539690508,
+      "learning_rate": 1.9977115923137912e-05,
+      "loss": 1.0848,
+      "step": 342
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.924276458472716,
+      "learning_rate": 1.9976787967182303e-05,
+      "loss": 1.1116,
+      "step": 343
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.911339529842714,
+      "learning_rate": 1.9976457680648033e-05,
+      "loss": 1.0106,
+      "step": 344
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8795885286729412,
+      "learning_rate": 1.9976125063612254e-05,
+      "loss": 1.0892,
+      "step": 345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8747873776798474,
+      "learning_rate": 1.9975790116152668e-05,
+      "loss": 1.0552,
+      "step": 346
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6759695411928928,
+      "learning_rate": 1.9975452838347513e-05,
+      "loss": 0.9651,
+      "step": 347
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.892483736446108,
+      "learning_rate": 1.9975113230275583e-05,
+      "loss": 0.9574,
+      "step": 348
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8709728862932533,
+      "learning_rate": 1.997477129201621e-05,
+      "loss": 1.0756,
+      "step": 349
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9454781450273466,
+      "learning_rate": 1.997442702364927e-05,
+      "loss": 1.0243,
+      "step": 350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2211417729946594,
+      "learning_rate": 1.997408042525518e-05,
+      "loss": 0.2988,
+      "step": 351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0725325602143942,
+      "learning_rate": 1.9973731496914914e-05,
+      "loss": 1.0043,
+      "step": 352
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.869161955493485,
+      "learning_rate": 1.9973380238709974e-05,
+      "loss": 0.9723,
+      "step": 353
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8280401841890608,
+      "learning_rate": 1.9973026650722417e-05,
+      "loss": 0.973,
+      "step": 354
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9158082367614362,
+      "learning_rate": 1.997267073303484e-05,
+      "loss": 1.0206,
+      "step": 355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9184700180705447,
+      "learning_rate": 1.9972312485730384e-05,
+      "loss": 0.9818,
+      "step": 356
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8360498617161698,
+      "learning_rate": 1.9971951908892743e-05,
+      "loss": 0.9858,
+      "step": 357
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.879228401720701,
+      "learning_rate": 1.997158900260614e-05,
+      "loss": 0.9905,
+      "step": 358
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9168440780518097,
+      "learning_rate": 1.9971223766955353e-05,
+      "loss": 1.0587,
+      "step": 359
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6694300100059905,
+      "learning_rate": 1.99708562020257e-05,
+      "loss": 0.9729,
+      "step": 360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0261356800612624,
+      "learning_rate": 1.997048630790305e-05,
+      "loss": 1.022,
+      "step": 361
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6654939561470525,
+      "learning_rate": 1.9970114084673796e-05,
+      "loss": 1.0084,
+      "step": 362
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.653267932249111,
+      "learning_rate": 1.9969739532424907e-05,
+      "loss": 1.081,
+      "step": 363
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6893220464759946,
+      "learning_rate": 1.9969362651243865e-05,
+      "loss": 0.9857,
+      "step": 364
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.642384659467501,
+      "learning_rate": 1.9968983441218715e-05,
+      "loss": 1.0841,
+      "step": 365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9159525667443391,
+      "learning_rate": 1.9968601902438043e-05,
+      "loss": 0.996,
+      "step": 366
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7973358848616572,
+      "learning_rate": 1.996821803499097e-05,
+      "loss": 1.0523,
+      "step": 367
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6687826721336216,
+      "learning_rate": 1.9967831838967175e-05,
+      "loss": 1.0016,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7746941099314142,
+      "learning_rate": 1.9967443314456867e-05,
+      "loss": 1.0465,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2004750933586537,
+      "learning_rate": 1.996705246155081e-05,
+      "loss": 1.0773,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7655020719540753,
+      "learning_rate": 1.99666592803403e-05,
+      "loss": 1.0387,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.365683017709636,
+      "learning_rate": 1.9966263770917192e-05,
+      "loss": 1.0124,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0366524707880345,
+      "learning_rate": 1.9965865933373874e-05,
+      "loss": 0.874,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.671223843411921,
+      "learning_rate": 1.9965465767803283e-05,
+      "loss": 0.9557,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9471098475001651,
+      "learning_rate": 1.9965063274298893e-05,
+      "loss": 0.9588,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8134046364665488,
+      "learning_rate": 1.996465845295473e-05,
+      "loss": 1.1017,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9016930565882155,
+      "learning_rate": 1.9964251303865362e-05,
+      "loss": 0.9739,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.271598534245101,
+      "learning_rate": 1.9963841827125897e-05,
+      "loss": 0.9472,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.896049194221097,
+      "learning_rate": 1.9963430022831988e-05,
+      "loss": 0.9424,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.780642059590237,
+      "learning_rate": 1.996301589107983e-05,
+      "loss": 0.9207,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.270826279650387,
+      "learning_rate": 1.9962599431966168e-05,
+      "loss": 1.0037,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8704117316418303,
+      "learning_rate": 1.996218064558829e-05,
+      "loss": 0.9496,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8661718269613556,
+      "learning_rate": 1.9961759532044017e-05,
+      "loss": 1.0117,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8417651698598574,
+      "learning_rate": 1.9961336091431728e-05,
+      "loss": 0.9955,
+      "step": 384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6470843282071816,
+      "learning_rate": 1.9960910323850333e-05,
+      "loss": 0.9237,
+      "step": 385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.084013593695897,
+      "learning_rate": 1.9960482229399294e-05,
+      "loss": 0.9585,
+      "step": 386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7429179544113134,
+      "learning_rate": 1.9960051808178616e-05,
+      "loss": 0.979,
+      "step": 387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7214252112920558,
+      "learning_rate": 1.995961906028884e-05,
+      "loss": 0.9104,
+      "step": 388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.710460520148069,
+      "learning_rate": 1.9959183985831063e-05,
+      "loss": 1.025,
+      "step": 389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7839407392827988,
+      "learning_rate": 1.9958746584906914e-05,
+      "loss": 0.9947,
+      "step": 390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.816552667971807,
+      "learning_rate": 1.995830685761857e-05,
+      "loss": 0.929,
+      "step": 391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9080235563689414,
+      "learning_rate": 1.9957864804068752e-05,
+      "loss": 1.0407,
+      "step": 392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.609038295754576,
+      "learning_rate": 1.9957420424360726e-05,
+      "loss": 1.0248,
+      "step": 393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6180642787300732,
+      "learning_rate": 1.9956973718598292e-05,
+      "loss": 1.0125,
+      "step": 394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.837362902823947,
+      "learning_rate": 1.9956524686885807e-05,
+      "loss": 0.9714,
+      "step": 395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.692729518219842,
+      "learning_rate": 1.9956073329328168e-05,
+      "loss": 0.9608,
+      "step": 396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.262973964163377,
+      "learning_rate": 1.99556196460308e-05,
+      "loss": 1.0805,
+      "step": 397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.486665981438739,
+      "learning_rate": 1.9955163637099698e-05,
+      "loss": 0.973,
+      "step": 398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6279921082061088,
+      "learning_rate": 1.9954705302641373e-05,
+      "loss": 1.0719,
+      "step": 399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8690799137248861,
+      "learning_rate": 1.99542446427629e-05,
+      "loss": 0.9738,
+      "step": 400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8688923352953541,
+      "learning_rate": 1.9953781657571887e-05,
+      "loss": 0.8855,
+      "step": 401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9138325981333588,
+      "learning_rate": 1.995331634717649e-05,
+      "loss": 1.008,
+      "step": 402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9804330084088011,
+      "learning_rate": 1.9952848711685398e-05,
+      "loss": 0.3712,
+      "step": 403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7239681728136025,
+      "learning_rate": 1.9952378751207858e-05,
+      "loss": 0.9922,
+      "step": 404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.272422800043243,
+      "learning_rate": 1.995190646585365e-05,
+      "loss": 1.0908,
+      "step": 405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9734144396729554,
+      "learning_rate": 1.99514318557331e-05,
+      "loss": 1.0069,
+      "step": 406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6705814464427644,
+      "learning_rate": 1.9950954920957074e-05,
+      "loss": 0.973,
+      "step": 407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9164283455695768,
+      "learning_rate": 1.995047566163699e-05,
+      "loss": 0.967,
+      "step": 408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0825054548353537,
+      "learning_rate": 1.99499940778848e-05,
+      "loss": 1.0111,
+      "step": 409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6946507603916696,
+      "learning_rate": 1.9949510169813006e-05,
+      "loss": 1.0424,
+      "step": 410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6963439332985022,
+      "learning_rate": 1.994902393753464e-05,
+      "loss": 0.925,
+      "step": 411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7813881184752924,
+      "learning_rate": 1.994853538116329e-05,
+      "loss": 0.98,
+      "step": 412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9476875471598322,
+      "learning_rate": 1.9948044500813085e-05,
+      "loss": 0.2952,
+      "step": 413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.631977815926575,
+      "learning_rate": 1.9947551296598698e-05,
+      "loss": 1.04,
+      "step": 414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8654933240402845,
+      "learning_rate": 1.9947055768635333e-05,
+      "loss": 1.0291,
+      "step": 415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.724831197476821,
+      "learning_rate": 1.9946557917038752e-05,
+      "loss": 1.0285,
+      "step": 416
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7055810949008126,
+      "learning_rate": 1.994605774192525e-05,
+      "loss": 0.9754,
+      "step": 417
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.902655632725055,
+      "learning_rate": 1.9945555243411666e-05,
+      "loss": 0.9884,
+      "step": 418
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7521010796747778,
+      "learning_rate": 1.994505042161539e-05,
+      "loss": 1.0392,
+      "step": 419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9706370538272724,
+      "learning_rate": 1.9944543276654343e-05,
+      "loss": 1.0162,
+      "step": 420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5913916826868926,
+      "learning_rate": 1.9944033808646997e-05,
+      "loss": 0.9555,
+      "step": 421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.947416664127582,
+      "learning_rate": 1.994352201771236e-05,
+      "loss": 1.0022,
+      "step": 422
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6636265488106525,
+      "learning_rate": 1.994300790396999e-05,
+      "loss": 0.9972,
+      "step": 423
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8301763145490835,
+      "learning_rate": 1.9942491467539984e-05,
+      "loss": 0.2936,
+      "step": 424
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7860710337754988,
+      "learning_rate": 1.994197270854298e-05,
+      "loss": 1.1105,
+      "step": 425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7246742095943215,
+      "learning_rate": 1.9941451627100163e-05,
+      "loss": 1.0,
+      "step": 426
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7204170411400848,
+      "learning_rate": 1.9940928223333254e-05,
+      "loss": 1.0085,
+      "step": 427
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7230008336438516,
+      "learning_rate": 1.994040249736452e-05,
+      "loss": 1.0553,
+      "step": 428
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9500428810203743,
+      "learning_rate": 1.9939874449316776e-05,
+      "loss": 0.9951,
+      "step": 429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9835723244770853,
+      "learning_rate": 1.9939344079313368e-05,
+      "loss": 0.9606,
+      "step": 430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7612908926450928,
+      "learning_rate": 1.9938811387478193e-05,
+      "loss": 1.068,
+      "step": 431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.804672956335246,
+      "learning_rate": 1.9938276373935688e-05,
+      "loss": 0.2874,
+      "step": 432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1189658835656644,
+      "learning_rate": 1.9937739038810833e-05,
+      "loss": 0.9592,
+      "step": 433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.907383084649791,
+      "learning_rate": 1.9937199382229147e-05,
+      "loss": 0.9922,
+      "step": 434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8673974960543036,
+      "learning_rate": 1.9936657404316694e-05,
+      "loss": 1.0396,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7609749617395538,
+      "learning_rate": 1.9936113105200085e-05,
+      "loss": 1.016,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.857177064408329,
+      "learning_rate": 1.9935566485006464e-05,
+      "loss": 1.0517,
+      "step": 437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7881466257826963,
+      "learning_rate": 1.9935017543863522e-05,
+      "loss": 1.0058,
+      "step": 438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7818711326981459,
+      "learning_rate": 1.993446628189949e-05,
+      "loss": 0.2837,
+      "step": 439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.732664123594324,
+      "learning_rate": 1.993391269924315e-05,
+      "loss": 1.0294,
+      "step": 440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9324930738137451,
+      "learning_rate": 1.9933356796023808e-05,
+      "loss": 1.0167,
+      "step": 441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8240254481494775,
+      "learning_rate": 1.993279857237133e-05,
+      "loss": 1.0563,
+      "step": 442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6751234246176034,
+      "learning_rate": 1.9932238028416118e-05,
+      "loss": 0.969,
+      "step": 443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8773278134733776,
+      "learning_rate": 1.9931675164289114e-05,
+      "loss": 0.3523,
+      "step": 444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8829661421393336,
+      "learning_rate": 1.9931109980121797e-05,
+      "loss": 1.0271,
+      "step": 445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.670229941389474,
+      "learning_rate": 1.9930542476046204e-05,
+      "loss": 0.9869,
+      "step": 446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9529735258492515,
+      "learning_rate": 1.99299726521949e-05,
+      "loss": 1.0093,
+      "step": 447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8337061013793297,
+      "learning_rate": 1.9929400508700994e-05,
+      "loss": 1.0088,
+      "step": 448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8868987054874113,
+      "learning_rate": 1.9928826045698138e-05,
+      "loss": 0.9401,
+      "step": 449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7736822119106102,
+      "learning_rate": 1.992824926332053e-05,
+      "loss": 0.8959,
+      "step": 450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8443091482171898,
+      "learning_rate": 1.9927670161702906e-05,
+      "loss": 0.9301,
+      "step": 451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8673399290410109,
+      "learning_rate": 1.992708874098054e-05,
+      "loss": 1.0448,
+      "step": 452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7480975521989213,
+      "learning_rate": 1.992650500128926e-05,
+      "loss": 1.0478,
+      "step": 453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7284452650329931,
+      "learning_rate": 1.992591894276542e-05,
+      "loss": 0.9936,
+      "step": 454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.158507005520441,
+      "learning_rate": 1.9925330565545927e-05,
+      "loss": 1.0103,
+      "step": 455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8754450935331193,
+      "learning_rate": 1.9924739869768222e-05,
+      "loss": 0.9605,
+      "step": 456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.597953439972833,
+      "learning_rate": 1.9924146855570298e-05,
+      "loss": 0.9656,
+      "step": 457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7630046266649517,
+      "learning_rate": 1.992355152309068e-05,
+      "loss": 0.9888,
+      "step": 458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.673044052785185,
+      "learning_rate": 1.9922953872468436e-05,
+      "loss": 1.0132,
+      "step": 459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.93742425159078,
+      "learning_rate": 1.992235390384318e-05,
+      "loss": 0.9908,
+      "step": 460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7712795240065684,
+      "learning_rate": 1.9921751617355063e-05,
+      "loss": 0.9565,
+      "step": 461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.691828531846784,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 0.9299,
+      "step": 462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6994929004809474,
+      "learning_rate": 1.9920540091353567e-05,
+      "loss": 0.9843,
+      "step": 463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6426793406324183,
+      "learning_rate": 1.99199308521232e-05,
+      "loss": 1.0106,
+      "step": 464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9115990999206212,
+      "learning_rate": 1.9919319295596e-05,
+      "loss": 1.0126,
+      "step": 465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8678655610247155,
+      "learning_rate": 1.9918705421914816e-05,
+      "loss": 0.9274,
+      "step": 466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6057461570853822,
+      "learning_rate": 1.9918089231223066e-05,
+      "loss": 1.0117,
+      "step": 467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6413082675476167,
+      "learning_rate": 1.991747072366468e-05,
+      "loss": 1.0145,
+      "step": 468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7508791020380656,
+      "learning_rate": 1.9916849899384147e-05,
+      "loss": 0.9372,
+      "step": 469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6100942765885784,
+      "learning_rate": 1.9916226758526487e-05,
+      "loss": 0.8925,
+      "step": 470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7453075149314676,
+      "learning_rate": 1.9915601301237268e-05,
+      "loss": 1.0288,
+      "step": 471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7705674758427683,
+      "learning_rate": 1.99149735276626e-05,
+      "loss": 0.9981,
+      "step": 472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.772355931431959,
+      "learning_rate": 1.9914343437949127e-05,
+      "loss": 1.0421,
+      "step": 473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.046502924909874,
+      "learning_rate": 1.991371103224404e-05,
+      "loss": 0.9943,
+      "step": 474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4070735930320404,
+      "learning_rate": 1.9913076310695068e-05,
+      "loss": 1.0179,
+      "step": 475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.595258698737812,
+      "learning_rate": 1.991243927345048e-05,
+      "loss": 0.9025,
+      "step": 476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7908365270325948,
+      "learning_rate": 1.9911799920659093e-05,
+      "loss": 0.9881,
+      "step": 477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5227477921046528,
+      "learning_rate": 1.9911158252470257e-05,
+      "loss": 1.063,
+      "step": 478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.791504372040175,
+      "learning_rate": 1.9910514269033866e-05,
+      "loss": 0.985,
+      "step": 479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.850521202263624,
+      "learning_rate": 1.9909867970500353e-05,
+      "loss": 0.9251,
+      "step": 480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8951528050679687,
+      "learning_rate": 1.9909219357020695e-05,
+      "loss": 1.0234,
+      "step": 481
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8356998614462292,
+      "learning_rate": 1.9908568428746408e-05,
+      "loss": 0.9865,
+      "step": 482
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6852686719838303,
+      "learning_rate": 1.990791518582955e-05,
+      "loss": 0.965,
+      "step": 483
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7180072631786487,
+      "learning_rate": 1.9907259628422718e-05,
+      "loss": 1.0102,
+      "step": 484
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7700524194227036,
+      "learning_rate": 1.9906601756679048e-05,
+      "loss": 0.9603,
+      "step": 485
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6828417658585442,
+      "learning_rate": 1.990594157075222e-05,
+      "loss": 0.9792,
+      "step": 486
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7557589626348973,
+      "learning_rate": 1.9905279070796454e-05,
+      "loss": 0.8588,
+      "step": 487
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7317388665444549,
+      "learning_rate": 1.9904614256966514e-05,
+      "loss": 0.9986,
+      "step": 488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6931663180939196,
+      "learning_rate": 1.9903947129417696e-05,
+      "loss": 1.0184,
+      "step": 489
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6999141717029091,
+      "learning_rate": 1.990327768830584e-05,
+      "loss": 1.0062,
+      "step": 490
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.990725713993706,
+      "learning_rate": 1.990260593378733e-05,
+      "loss": 1.0831,
+      "step": 491
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6705278936628876,
+      "learning_rate": 1.9901931866019087e-05,
+      "loss": 0.9517,
+      "step": 492
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9700353571405922,
+      "learning_rate": 1.9901255485158574e-05,
+      "loss": 1.024,
+      "step": 493
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0101515433726713,
+      "learning_rate": 1.9900576791363795e-05,
+      "loss": 0.9651,
+      "step": 494
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9079881135031231,
+      "learning_rate": 1.989989578479329e-05,
+      "loss": 1.0301,
+      "step": 495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1563214017269066,
+      "learning_rate": 1.989921246560614e-05,
+      "loss": 0.9411,
+      "step": 496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9433340483122155,
+      "learning_rate": 1.989852683396198e-05,
+      "loss": 1.0592,
+      "step": 497
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6650423245663724,
+      "learning_rate": 1.989783889002096e-05,
+      "loss": 0.9298,
+      "step": 498
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9234651813792842,
+      "learning_rate": 1.989714863394379e-05,
+      "loss": 0.9825,
+      "step": 499
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0082280975700524,
+      "learning_rate": 1.989645606589171e-05,
+      "loss": 1.0092,
+      "step": 500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9815040662391932,
+      "learning_rate": 1.989576118602651e-05,
+      "loss": 1.0367,
+      "step": 501
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7069392081266068,
+      "learning_rate": 1.9895063994510512e-05,
+      "loss": 0.9731,
+      "step": 502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.085711779508058,
+      "learning_rate": 1.989436449150658e-05,
+      "loss": 1.0642,
+      "step": 503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6401934567058296,
+      "learning_rate": 1.9893662677178116e-05,
+      "loss": 1.0061,
+      "step": 504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8772927472174918,
+      "learning_rate": 1.9892958551689065e-05,
+      "loss": 1.0555,
+      "step": 505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6825233296028157,
+      "learning_rate": 1.989225211520391e-05,
+      "loss": 0.9436,
+      "step": 506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5274675019123192,
+      "learning_rate": 1.9891543367887675e-05,
+      "loss": 0.9617,
+      "step": 507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5803756492693821,
+      "learning_rate": 1.9890832309905927e-05,
+      "loss": 0.963,
+      "step": 508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6939482132596444,
+      "learning_rate": 1.989011894142476e-05,
+      "loss": 0.9643,
+      "step": 509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0225263659357182,
+      "learning_rate": 1.9889403262610827e-05,
+      "loss": 0.3143,
+      "step": 510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.589048564009571,
+      "learning_rate": 1.9888685273631307e-05,
+      "loss": 0.9786,
+      "step": 511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6228493525542471,
+      "learning_rate": 1.988796497465392e-05,
+      "loss": 0.9555,
+      "step": 512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8677355778429002,
+      "learning_rate": 1.9887242365846933e-05,
+      "loss": 0.3073,
+      "step": 513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.947119784848312,
+      "learning_rate": 1.988651744737914e-05,
+      "loss": 0.907,
+      "step": 514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.786175594370177,
+      "learning_rate": 1.9885790219419888e-05,
+      "loss": 0.9491,
+      "step": 515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2095671154084395,
+      "learning_rate": 1.988506068213906e-05,
+      "loss": 0.9752,
+      "step": 516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1107332157479002,
+      "learning_rate": 1.988432883570707e-05,
+      "loss": 0.3077,
+      "step": 517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0106255199643956,
+      "learning_rate": 1.9883594680294878e-05,
+      "loss": 0.9505,
+      "step": 518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9913934845663013,
+      "learning_rate": 1.9882858216073982e-05,
+      "loss": 1.0423,
+      "step": 519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5683550787091347,
+      "learning_rate": 1.9882119443216426e-05,
+      "loss": 1.0429,
+      "step": 520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6720639321493218,
+      "learning_rate": 1.988137836189478e-05,
+      "loss": 0.9568,
+      "step": 521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6874564869445212,
+      "learning_rate": 1.9880634972282168e-05,
+      "loss": 0.9808,
+      "step": 522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9911132917647867,
+      "learning_rate": 1.987988927455224e-05,
+      "loss": 1.035,
+      "step": 523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8655671766479962,
+      "learning_rate": 1.9879141268879194e-05,
+      "loss": 0.9435,
+      "step": 524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7108612549789455,
+      "learning_rate": 1.9878390955437764e-05,
+      "loss": 1.0306,
+      "step": 525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.903099821347723,
+      "learning_rate": 1.987763833440322e-05,
+      "loss": 0.2912,
+      "step": 526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7776897303273496,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 1.0163,
+      "step": 527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0838132509155445,
+      "learning_rate": 1.9876126170258588e-05,
+      "loss": 1.0462,
+      "step": 528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1040942707606436,
+      "learning_rate": 1.9875366627501743e-05,
+      "loss": 0.9781,
+      "step": 529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7766394928123865,
+      "learning_rate": 1.9874604777858272e-05,
+      "loss": 0.9783,
+      "step": 530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7587165044446451,
+      "learning_rate": 1.9873840621506138e-05,
+      "loss": 1.0506,
+      "step": 531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8520954719057352,
+      "learning_rate": 1.987307415862385e-05,
+      "loss": 1.0355,
+      "step": 532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.738546160539183,
+      "learning_rate": 1.9872305389390454e-05,
+      "loss": 1.0503,
+      "step": 533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7963801967360402,
+      "learning_rate": 1.987153431398554e-05,
+      "loss": 1.0747,
+      "step": 534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.051071892987532,
+      "learning_rate": 1.9870760932589225e-05,
+      "loss": 0.9607,
+      "step": 535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6462638295482712,
+      "learning_rate": 1.9869985245382172e-05,
+      "loss": 1.0054,
+      "step": 536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9253676833173057,
+      "learning_rate": 1.9869207252545582e-05,
+      "loss": 0.9846,
+      "step": 537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.864572095881507,
+      "learning_rate": 1.9868426954261198e-05,
+      "loss": 1.0285,
+      "step": 538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8687462565056039,
+      "learning_rate": 1.986764435071129e-05,
+      "loss": 0.9482,
+      "step": 539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.117948055904973,
+      "learning_rate": 1.986685944207868e-05,
+      "loss": 1.0006,
+      "step": 540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8121410787300227,
+      "learning_rate": 1.9866072228546724e-05,
+      "loss": 0.9498,
+      "step": 541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5886336473246425,
+      "learning_rate": 1.986528271029931e-05,
+      "loss": 0.889,
+      "step": 542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9319387230966267,
+      "learning_rate": 1.9864490887520877e-05,
+      "loss": 0.8756,
+      "step": 543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5417456523426636,
+      "learning_rate": 1.986369676039638e-05,
+      "loss": 0.8807,
+      "step": 544
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1763364590287626,
+      "learning_rate": 1.9862900329111344e-05,
+      "loss": 0.9774,
+      "step": 545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9791657103778806,
+      "learning_rate": 1.986210159385181e-05,
+      "loss": 0.9612,
+      "step": 546
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6572905484841525,
+      "learning_rate": 1.9861300554804357e-05,
+      "loss": 0.9916,
+      "step": 547
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6777328584462108,
+      "learning_rate": 1.9860497212156114e-05,
+      "loss": 1.0715,
+      "step": 548
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7792046059260411,
+      "learning_rate": 1.985969156609474e-05,
+      "loss": 1.0082,
+      "step": 549
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5246067255277136,
+      "learning_rate": 1.9858883616808434e-05,
+      "loss": 1.0161,
+      "step": 550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7274715039470692,
+      "learning_rate": 1.9858073364485933e-05,
+      "loss": 1.0161,
+      "step": 551
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6498690937955873,
+      "learning_rate": 1.985726080931651e-05,
+      "loss": 0.9486,
+      "step": 552
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8807515037160574,
+      "learning_rate": 1.9856445951489984e-05,
+      "loss": 0.9449,
+      "step": 553
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8208768935597974,
+      "learning_rate": 1.9855628791196698e-05,
+      "loss": 0.9693,
+      "step": 554
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6833041720935167,
+      "learning_rate": 1.9854809328627546e-05,
+      "loss": 1.0078,
+      "step": 555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6543281933906169,
+      "learning_rate": 1.985398756397395e-05,
+      "loss": 0.9228,
+      "step": 556
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.583195277532729,
+      "learning_rate": 1.9853163497427885e-05,
+      "loss": 0.9595,
+      "step": 557
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.867090988857162,
+      "learning_rate": 1.985233712918184e-05,
+      "loss": 0.9916,
+      "step": 558
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5232151688440378,
+      "learning_rate": 1.9851508459428858e-05,
+      "loss": 1.0172,
+      "step": 559
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7777277343045925,
+      "learning_rate": 1.985067748836252e-05,
+      "loss": 0.9679,
+      "step": 560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.812436360647807,
+      "learning_rate": 1.9849844216176945e-05,
+      "loss": 0.9489,
+      "step": 561
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7085347777656354,
+      "learning_rate": 1.9849008643066774e-05,
+      "loss": 0.9593,
+      "step": 562
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.73334444203699,
+      "learning_rate": 1.9848170769227203e-05,
+      "loss": 0.9853,
+      "step": 563
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.728386131651639,
+      "learning_rate": 1.984733059485396e-05,
+      "loss": 0.9522,
+      "step": 564
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9050766614994752,
+      "learning_rate": 1.984648812014331e-05,
+      "loss": 1.0545,
+      "step": 565
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6863897587600656,
+      "learning_rate": 1.9845643345292055e-05,
+      "loss": 0.889,
+      "step": 566
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6819698208684524,
+      "learning_rate": 1.984479627049753e-05,
+      "loss": 0.9936,
+      "step": 567
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8885363255340635,
+      "learning_rate": 1.984394689595762e-05,
+      "loss": 0.3346,
+      "step": 568
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8871134864353314,
+      "learning_rate": 1.9843095221870736e-05,
+      "loss": 0.9714,
+      "step": 569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4777184199148228,
+      "learning_rate": 1.984224124843582e-05,
+      "loss": 0.9833,
+      "step": 570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.829762953791349,
+      "learning_rate": 1.9841384975852373e-05,
+      "loss": 1.0405,
+      "step": 571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7528872110219917,
+      "learning_rate": 1.9840526404320415e-05,
+      "loss": 1.0123,
+      "step": 572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8760102619994743,
+      "learning_rate": 1.9839665534040507e-05,
+      "loss": 0.9951,
+      "step": 573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5714402952846405,
+      "learning_rate": 1.9838802365213752e-05,
+      "loss": 0.945,
+      "step": 574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2546242292651044,
+      "learning_rate": 1.9837936898041783e-05,
+      "loss": 0.8821,
+      "step": 575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7115642594977378,
+      "learning_rate": 1.9837069132726775e-05,
+      "loss": 0.8517,
+      "step": 576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6632874064960919,
+      "learning_rate": 1.983619906947144e-05,
+      "loss": 1.0078,
+      "step": 577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6181599335511319,
+      "learning_rate": 1.9835326708479015e-05,
+      "loss": 1.009,
+      "step": 578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7747866544314281,
+      "learning_rate": 1.98344520499533e-05,
+      "loss": 1.0198,
+      "step": 579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8244597406979507,
+      "learning_rate": 1.9833575094098602e-05,
+      "loss": 0.3072,
+      "step": 580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.978657081005872,
+      "learning_rate": 1.9832695841119784e-05,
+      "loss": 1.0135,
+      "step": 581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7488868677099108,
+      "learning_rate": 1.9831814291222233e-05,
+      "loss": 0.9888,
+      "step": 582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6212122804901654,
+      "learning_rate": 1.983093044461189e-05,
+      "loss": 0.963,
+      "step": 583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5699787639287242,
+      "learning_rate": 1.9830044301495213e-05,
+      "loss": 0.9359,
+      "step": 584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0507971672167007,
+      "learning_rate": 1.9829155862079207e-05,
+      "loss": 0.9246,
+      "step": 585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4874152469933142,
+      "learning_rate": 1.9828265126571414e-05,
+      "loss": 0.9327,
+      "step": 586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.718251280605203,
+      "learning_rate": 1.982737209517991e-05,
+      "loss": 0.9475,
+      "step": 587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7479648720031697,
+      "learning_rate": 1.9826476768113304e-05,
+      "loss": 0.9925,
+      "step": 588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7225239112520367,
+      "learning_rate": 1.9825579145580747e-05,
+      "loss": 0.9955,
+      "step": 589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8573163420724126,
+      "learning_rate": 1.982467922779192e-05,
+      "loss": 0.96,
+      "step": 590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9061968201895618,
+      "learning_rate": 1.982377701495705e-05,
+      "loss": 0.9714,
+      "step": 591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5896774926760178,
+      "learning_rate": 1.982287250728689e-05,
+      "loss": 0.9445,
+      "step": 592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8734917685883975,
+      "learning_rate": 1.982196570499273e-05,
+      "loss": 0.3112,
+      "step": 593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.674042343599907,
+      "learning_rate": 1.9821056608286406e-05,
+      "loss": 0.9634,
+      "step": 594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6146271670195458,
+      "learning_rate": 1.982014521738028e-05,
+      "loss": 0.9387,
+      "step": 595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6757936015929897,
+      "learning_rate": 1.9819231532487252e-05,
+      "loss": 0.979,
+      "step": 596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9101585434975927,
+      "learning_rate": 1.981831555382076e-05,
+      "loss": 0.9978,
+      "step": 597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.638509774175569,
+      "learning_rate": 1.9817397281594778e-05,
+      "loss": 0.9985,
+      "step": 598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6597048566534132,
+      "learning_rate": 1.981647671602381e-05,
+      "loss": 1.0344,
+      "step": 599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5224908736480791,
+      "learning_rate": 1.9815553857322905e-05,
+      "loss": 0.9918,
+      "step": 600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9985248529060036,
+      "learning_rate": 1.9814628705707643e-05,
+      "loss": 1.0289,
+      "step": 601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5327615466480906,
+      "learning_rate": 1.9813701261394136e-05,
+      "loss": 0.9655,
+      "step": 602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6229098866994003,
+      "learning_rate": 1.9812771524599036e-05,
+      "loss": 0.9565,
+      "step": 603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5051253765063912,
+      "learning_rate": 1.981183949553953e-05,
+      "loss": 0.9854,
+      "step": 604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.706387313271035,
+      "learning_rate": 1.981090517443334e-05,
+      "loss": 0.8863,
+      "step": 605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6174941774190525,
+      "learning_rate": 1.9809968561498728e-05,
+      "loss": 0.9844,
+      "step": 606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7553072826263136,
+      "learning_rate": 1.980902965695448e-05,
+      "loss": 0.9222,
+      "step": 607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8186270151144839,
+      "learning_rate": 1.9808088461019926e-05,
+      "loss": 0.96,
+      "step": 608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.891269831009788,
+      "learning_rate": 1.980714497391493e-05,
+      "loss": 0.9335,
+      "step": 609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6874262635801682,
+      "learning_rate": 1.9806199195859893e-05,
+      "loss": 0.9475,
+      "step": 610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7359133537168558,
+      "learning_rate": 1.9805251127075746e-05,
+      "loss": 1.0473,
+      "step": 611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.621453926565137,
+      "learning_rate": 1.9804300767783958e-05,
+      "loss": 0.9227,
+      "step": 612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6230725930169774,
+      "learning_rate": 1.9803348118206536e-05,
+      "loss": 0.9708,
+      "step": 613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.983605922888492,
+      "learning_rate": 1.9802393178566017e-05,
+      "loss": 1.0104,
+      "step": 614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9554202974102726,
+      "learning_rate": 1.9801435949085475e-05,
+      "loss": 0.9732,
+      "step": 615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.849294609509034,
+      "learning_rate": 1.9800476429988516e-05,
+      "loss": 0.9606,
+      "step": 616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.563357603037707,
+      "learning_rate": 1.979951462149929e-05,
+      "loss": 0.9642,
+      "step": 617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.15545925295218,
+      "learning_rate": 1.979855052384247e-05,
+      "loss": 0.9849,
+      "step": 618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.621135612314959,
+      "learning_rate": 1.9797584137243272e-05,
+      "loss": 1.0143,
+      "step": 619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5509560147836148,
+      "learning_rate": 1.9796615461927443e-05,
+      "loss": 0.9174,
+      "step": 620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7656446578189005,
+      "learning_rate": 1.9795644498121266e-05,
+      "loss": 0.9895,
+      "step": 621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.611425933181333,
+      "learning_rate": 1.979467124605156e-05,
+      "loss": 1.0008,
+      "step": 622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8890890708045418,
+      "learning_rate": 1.9793695705945674e-05,
+      "loss": 1.017,
+      "step": 623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0334166640022753,
+      "learning_rate": 1.9792717878031498e-05,
+      "loss": 0.8839,
+      "step": 624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7456661635297466,
+      "learning_rate": 1.979173776253745e-05,
+      "loss": 0.9794,
+      "step": 625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5294370940175432,
+      "learning_rate": 1.979075535969248e-05,
+      "loss": 0.9557,
+      "step": 626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0296542777234383,
+      "learning_rate": 1.9789770669726088e-05,
+      "loss": 0.9594,
+      "step": 627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5689454503809042,
+      "learning_rate": 1.9788783692868288e-05,
+      "loss": 1.0005,
+      "step": 628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6518207403192,
+      "learning_rate": 1.9787794429349645e-05,
+      "loss": 0.994,
+      "step": 629
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6208285390544763,
+      "learning_rate": 1.9786802879401248e-05,
+      "loss": 0.9192,
+      "step": 630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6425782996323863,
+      "learning_rate": 1.978580904325472e-05,
+      "loss": 0.97,
+      "step": 631
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6699630286070308,
+      "learning_rate": 1.9784812921142232e-05,
+      "loss": 0.9691,
+      "step": 632
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.716663219773858,
+      "learning_rate": 1.9783814513296464e-05,
+      "loss": 0.8943,
+      "step": 633
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.83085374874018,
+      "learning_rate": 1.9782813819950656e-05,
+      "loss": 0.849,
+      "step": 634
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.576647997372439,
+      "learning_rate": 1.9781810841338563e-05,
+      "loss": 0.9435,
+      "step": 635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5873338689852323,
+      "learning_rate": 1.978080557769448e-05,
+      "loss": 0.9405,
+      "step": 636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8983319227664757,
+      "learning_rate": 1.977979802925324e-05,
+      "loss": 0.9337,
+      "step": 637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.001087829787801,
+      "learning_rate": 1.977878819625021e-05,
+      "loss": 0.9547,
+      "step": 638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5462695415537155,
+      "learning_rate": 1.9777776078921282e-05,
+      "loss": 0.988,
+      "step": 639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8648083574798346,
+      "learning_rate": 1.9776761677502888e-05,
+      "loss": 0.9448,
+      "step": 640
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7401880360865456,
+      "learning_rate": 1.977574499223199e-05,
+      "loss": 0.8825,
+      "step": 641
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.588603384090599,
+      "learning_rate": 1.977472602334609e-05,
+      "loss": 0.9502,
+      "step": 642
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.013363154080898,
+      "learning_rate": 1.9773704771083217e-05,
+      "loss": 0.8083,
+      "step": 643
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.789818531110137,
+      "learning_rate": 1.9772681235681936e-05,
+      "loss": 0.9824,
+      "step": 644
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.604910024704947,
+      "learning_rate": 1.9771655417381342e-05,
+      "loss": 0.9318,
+      "step": 645
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5895748848355196,
+      "learning_rate": 1.9770627316421074e-05,
+      "loss": 1.0216,
+      "step": 646
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7181986683114863,
+      "learning_rate": 1.976959693304129e-05,
+      "loss": 0.971,
+      "step": 647
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.090795138197271,
+      "learning_rate": 1.9768564267482685e-05,
+      "loss": 1.0202,
+      "step": 648
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7739944108727936,
+      "learning_rate": 1.97675293199865e-05,
+      "loss": 0.9814,
+      "step": 649
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8472601771501216,
+      "learning_rate": 1.9766492090794488e-05,
+      "loss": 1.0092,
+      "step": 650
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7088780593500077,
+      "learning_rate": 1.9765452580148954e-05,
+      "loss": 0.9903,
+      "step": 651
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6144693677940285,
+      "learning_rate": 1.9764410788292724e-05,
+      "loss": 0.919,
+      "step": 652
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5807718951238905,
+      "learning_rate": 1.976336671546916e-05,
+      "loss": 0.9575,
+      "step": 653
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6875454987025957,
+      "learning_rate": 1.9762320361922156e-05,
+      "loss": 1.046,
+      "step": 654
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4835502843129302,
+      "learning_rate": 1.9761271727896148e-05,
+      "loss": 0.9545,
+      "step": 655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.552094217209983,
+      "learning_rate": 1.9760220813636087e-05,
+      "loss": 0.884,
+      "step": 656
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6201406423148235,
+      "learning_rate": 1.9759167619387474e-05,
+      "loss": 0.907,
+      "step": 657
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5753068772105716,
+      "learning_rate": 1.9758112145396335e-05,
+      "loss": 0.8854,
+      "step": 658
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8855973676547777,
+      "learning_rate": 1.9757054391909224e-05,
+      "loss": 1.0453,
+      "step": 659
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.15880228589204,
+      "learning_rate": 1.9755994359173238e-05,
+      "loss": 0.9453,
+      "step": 660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5953451254625532,
+      "learning_rate": 1.9754932047435994e-05,
+      "loss": 0.924,
+      "step": 661
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.914323680924929,
+      "learning_rate": 1.9753867456945653e-05,
+      "loss": 0.9879,
+      "step": 662
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4891585281224864,
+      "learning_rate": 1.9752800587950903e-05,
+      "loss": 0.9308,
+      "step": 663
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9770560967582835,
+      "learning_rate": 1.9751731440700964e-05,
+      "loss": 0.9479,
+      "step": 664
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9414212346368325,
+      "learning_rate": 1.975066001544559e-05,
+      "loss": 0.3177,
+      "step": 665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4875219115077953,
+      "learning_rate": 1.9749586312435065e-05,
+      "loss": 1.0379,
+      "step": 666
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.696156265428028,
+      "learning_rate": 1.9748510331920204e-05,
+      "loss": 0.9922,
+      "step": 667
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.833725443792948,
+      "learning_rate": 1.974743207415236e-05,
+      "loss": 0.9119,
+      "step": 668
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5354239446380398,
+      "learning_rate": 1.9746351539383412e-05,
+      "loss": 1.0048,
+      "step": 669
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5072036363757007,
+      "learning_rate": 1.9745268727865774e-05,
+      "loss": 0.9452,
+      "step": 670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5520227210932784,
+      "learning_rate": 1.974418363985239e-05,
+      "loss": 1.019,
+      "step": 671
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7194076439498294,
+      "learning_rate": 1.9743096275596735e-05,
+      "loss": 0.9544,
+      "step": 672
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9746300444384288,
+      "learning_rate": 1.9742006635352822e-05,
+      "loss": 1.0593,
+      "step": 673
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4515460624215812,
+      "learning_rate": 1.9740914719375186e-05,
+      "loss": 0.9031,
+      "step": 674
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7203249098913107,
+      "learning_rate": 1.9739820527918904e-05,
+      "loss": 1.0139,
+      "step": 675
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6701071452379914,
+      "learning_rate": 1.9738724061239574e-05,
+      "loss": 0.9522,
+      "step": 676
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9822445033734273,
+      "learning_rate": 1.9737625319593338e-05,
+      "loss": 1.023,
+      "step": 677
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8078491756589703,
+      "learning_rate": 1.9736524303236852e-05,
+      "loss": 0.9974,
+      "step": 678
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.620720465896709,
+      "learning_rate": 1.973542101242732e-05,
+      "loss": 0.9633,
+      "step": 679
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8149811065822787,
+      "learning_rate": 1.973431544742247e-05,
+      "loss": 0.9773,
+      "step": 680
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0294651699681596,
+      "learning_rate": 1.9733207608480563e-05,
+      "loss": 0.9759,
+      "step": 681
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8415138223323133,
+      "learning_rate": 1.9732097495860388e-05,
+      "loss": 0.892,
+      "step": 682
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.184193274643253,
+      "learning_rate": 1.9730985109821268e-05,
+      "loss": 0.9952,
+      "step": 683
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.829076071264722,
+      "learning_rate": 1.9729870450623056e-05,
+      "loss": 0.9834,
+      "step": 684
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9336315247670188,
+      "learning_rate": 1.972875351852614e-05,
+      "loss": 0.9684,
+      "step": 685
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.843499861737653,
+      "learning_rate": 1.972763431379143e-05,
+      "loss": 1.0203,
+      "step": 686
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.280939490502182,
+      "learning_rate": 1.972651283668038e-05,
+      "loss": 0.2995,
+      "step": 687
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7276681600565358,
+      "learning_rate": 1.9725389087454955e-05,
+      "loss": 1.0055,
+      "step": 688
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7860889549356194,
+      "learning_rate": 1.9724263066377678e-05,
+      "loss": 0.9907,
+      "step": 689
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.580169419290622,
+      "learning_rate": 1.9723134773711577e-05,
+      "loss": 0.9858,
+      "step": 690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7798074038021647,
+      "learning_rate": 1.972200420972022e-05,
+      "loss": 0.9992,
+      "step": 691
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6603925458254127,
+      "learning_rate": 1.9720871374667714e-05,
+      "loss": 0.9673,
+      "step": 692
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.553690559546088,
+      "learning_rate": 1.971973626881869e-05,
+      "loss": 0.9464,
+      "step": 693
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5521020597463204,
+      "learning_rate": 1.97185988924383e-05,
+      "loss": 1.0286,
+      "step": 694
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1051505101313315,
+      "learning_rate": 1.9717459245792244e-05,
+      "loss": 0.9472,
+      "step": 695
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4890965884130531,
+      "learning_rate": 1.971631732914674e-05,
+      "loss": 1.0221,
+      "step": 696
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6719007066729015,
+      "learning_rate": 1.971517314276854e-05,
+      "loss": 0.9156,
+      "step": 697
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7548339892151512,
+      "learning_rate": 1.9714026686924925e-05,
+      "loss": 1.0017,
+      "step": 698
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6078271673560387,
+      "learning_rate": 1.971287796188371e-05,
+      "loss": 0.9495,
+      "step": 699
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9569815517061826,
+      "learning_rate": 1.971172696791323e-05,
+      "loss": 0.9591,
+      "step": 700
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5622121712700883,
+      "learning_rate": 1.971057370528237e-05,
+      "loss": 0.9649,
+      "step": 701
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5798150776059081,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.9079,
+      "step": 702
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.964148706586664,
+      "learning_rate": 1.970826037511762e-05,
+      "loss": 0.9367,
+      "step": 703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5867877522598344,
+      "learning_rate": 1.9707100308124128e-05,
+      "loss": 0.9014,
+      "step": 704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5013019479606504,
+      "learning_rate": 1.9705937973551038e-05,
+      "loss": 0.8933,
+      "step": 705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.795385433687663,
+      "learning_rate": 1.9704773371669872e-05,
+      "loss": 0.8796,
+      "step": 706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6536167372499206,
+      "learning_rate": 1.9703606502752674e-05,
+      "loss": 0.9731,
+      "step": 707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6204598508801429,
+      "learning_rate": 1.9702437367072035e-05,
+      "loss": 1.0082,
+      "step": 708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7234877711751868,
+      "learning_rate": 1.970126596490106e-05,
+      "loss": 0.9431,
+      "step": 709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7680801758734348,
+      "learning_rate": 1.9700092296513386e-05,
+      "loss": 1.0432,
+      "step": 710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6312546439488396,
+      "learning_rate": 1.969891636218319e-05,
+      "loss": 0.9005,
+      "step": 711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6630434018495426,
+      "learning_rate": 1.9697738162185163e-05,
+      "loss": 0.9447,
+      "step": 712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5979663983867929,
+      "learning_rate": 1.9696557696794537e-05,
+      "loss": 0.9786,
+      "step": 713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.010511880721655,
+      "learning_rate": 1.9695374966287065e-05,
+      "loss": 0.9573,
+      "step": 714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.451426944418813,
+      "learning_rate": 1.9694189970939033e-05,
+      "loss": 1.0436,
+      "step": 715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5830455766111613,
+      "learning_rate": 1.9693002711027264e-05,
+      "loss": 0.957,
+      "step": 716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.850718215200661,
+      "learning_rate": 1.969181318682909e-05,
+      "loss": 0.9834,
+      "step": 717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6516973570348639,
+      "learning_rate": 1.9690621398622394e-05,
+      "loss": 0.9334,
+      "step": 718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7337003718934916,
+      "learning_rate": 1.968942734668557e-05,
+      "loss": 0.9681,
+      "step": 719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.620868496009823,
+      "learning_rate": 1.9688231031297556e-05,
+      "loss": 1.0357,
+      "step": 720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5329059221659178,
+      "learning_rate": 1.9687032452737806e-05,
+      "loss": 0.9337,
+      "step": 721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7384823136362557,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 0.9651,
+      "step": 722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6136505602497995,
+      "learning_rate": 1.9684628507223588e-05,
+      "loss": 0.9546,
+      "step": 723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7034927122136125,
+      "learning_rate": 1.968342314083068e-05,
+      "loss": 0.9271,
+      "step": 724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.607731683848297,
+      "learning_rate": 1.9682215512389163e-05,
+      "loss": 1.026,
+      "step": 725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5943989963344436,
+      "learning_rate": 1.9681005622181137e-05,
+      "loss": 0.9623,
+      "step": 726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7812770707599777,
+      "learning_rate": 1.967979347048923e-05,
+      "loss": 0.9548,
+      "step": 727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8161840733654948,
+      "learning_rate": 1.9678579057596608e-05,
+      "loss": 1.063,
+      "step": 728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.508235574944762,
+      "learning_rate": 1.967736238378695e-05,
+      "loss": 0.9271,
+      "step": 729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6764547210723733,
+      "learning_rate": 1.9676143449344477e-05,
+      "loss": 0.8779,
+      "step": 730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.620570709692484,
+      "learning_rate": 1.9674922254553933e-05,
+      "loss": 0.9964,
+      "step": 731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.580417804019471,
+      "learning_rate": 1.9673698799700582e-05,
+      "loss": 0.9744,
+      "step": 732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5791244835449232,
+      "learning_rate": 1.967247308507023e-05,
+      "loss": 0.9742,
+      "step": 733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.792692921600134,
+      "learning_rate": 1.9671245110949202e-05,
+      "loss": 0.9572,
+      "step": 734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5785025119118827,
+      "learning_rate": 1.9670014877624353e-05,
+      "loss": 0.9293,
+      "step": 735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0725073120663478,
+      "learning_rate": 1.9668782385383065e-05,
+      "loss": 1.046,
+      "step": 736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7035287151083034,
+      "learning_rate": 1.9667547634513248e-05,
+      "loss": 0.967,
+      "step": 737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7512804545280596,
+      "learning_rate": 1.966631062530334e-05,
+      "loss": 0.9975,
+      "step": 738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6133321229889348,
+      "learning_rate": 1.9665071358042307e-05,
+      "loss": 1.0493,
+      "step": 739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4723658218407412,
+      "learning_rate": 1.9663829833019643e-05,
+      "loss": 0.8976,
+      "step": 740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8201845036066666,
+      "learning_rate": 1.9662586050525365e-05,
+      "loss": 0.998,
+      "step": 741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7125991303578467,
+      "learning_rate": 1.9661340010850025e-05,
+      "loss": 0.8724,
+      "step": 742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5276290567322122,
+      "learning_rate": 1.9660091714284694e-05,
+      "loss": 1.0208,
+      "step": 743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7215956625771214,
+      "learning_rate": 1.9658841161120982e-05,
+      "loss": 0.888,
+      "step": 744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5381710598112661,
+      "learning_rate": 1.9657588351651007e-05,
+      "loss": 1.0088,
+      "step": 745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8559954819162459,
+      "learning_rate": 1.9656333286167432e-05,
+      "loss": 0.996,
+      "step": 746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6472589044431567,
+      "learning_rate": 1.9655075964963443e-05,
+      "loss": 0.9381,
+      "step": 747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7694855146202892,
+      "learning_rate": 1.965381638833274e-05,
+      "loss": 0.9218,
+      "step": 748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7817060959538993,
+      "learning_rate": 1.965255455656957e-05,
+      "loss": 0.8943,
+      "step": 749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.264459651465797,
+      "learning_rate": 1.9651290469968694e-05,
+      "loss": 1.0297,
+      "step": 750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3211454847753379,
+      "learning_rate": 1.9650024128825406e-05,
+      "loss": 0.3338,
+      "step": 751
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.787921990712669,
+      "learning_rate": 1.9648755533435517e-05,
+      "loss": 0.9523,
+      "step": 752
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4703297869429213,
+      "learning_rate": 1.9647484684095373e-05,
+      "loss": 1.0012,
+      "step": 753
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7385139352458288,
+      "learning_rate": 1.964621158110185e-05,
+      "loss": 1.0068,
+      "step": 754
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.907685612810227,
+      "learning_rate": 1.9644936224752336e-05,
+      "loss": 0.9179,
+      "step": 755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8933415826630855,
+      "learning_rate": 1.9643658615344762e-05,
+      "loss": 0.9486,
+      "step": 756
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.992017226328773,
+      "learning_rate": 1.9642378753177573e-05,
+      "loss": 0.9536,
+      "step": 757
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.023337729951108,
+      "learning_rate": 1.964109663854975e-05,
+      "loss": 1.0767,
+      "step": 758
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4346591890770255,
+      "learning_rate": 1.9639812271760784e-05,
+      "loss": 0.9116,
+      "step": 759
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.518443313881122,
+      "learning_rate": 1.963852565311072e-05,
+      "loss": 0.9837,
+      "step": 760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8199084318474694,
+      "learning_rate": 1.96372367829001e-05,
+      "loss": 0.9258,
+      "step": 761
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6245527643511086,
+      "learning_rate": 1.9635945661430006e-05,
+      "loss": 0.9377,
+      "step": 762
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5378969603202852,
+      "learning_rate": 1.9634652289002047e-05,
+      "loss": 0.9842,
+      "step": 763
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.977930319428025,
+      "learning_rate": 1.9633356665918354e-05,
+      "loss": 0.9865,
+      "step": 764
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5555691411716785,
+      "learning_rate": 1.9632058792481582e-05,
+      "loss": 0.8887,
+      "step": 765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9052615270432982,
+      "learning_rate": 1.963075866899492e-05,
+      "loss": 0.9831,
+      "step": 766
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.515443301247266,
+      "learning_rate": 1.9629456295762067e-05,
+      "loss": 0.983,
+      "step": 767
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5899357968239571,
+      "learning_rate": 1.962815167308727e-05,
+      "loss": 0.9628,
+      "step": 768
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6180424261794053,
+      "learning_rate": 1.962684480127528e-05,
+      "loss": 0.9495,
+      "step": 769
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.12702526476482,
+      "learning_rate": 1.9625535680631386e-05,
+      "loss": 0.3225,
+      "step": 770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6043543503912514,
+      "learning_rate": 1.96242243114614e-05,
+      "loss": 0.9655,
+      "step": 771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.832430901685091,
+      "learning_rate": 1.9622910694071654e-05,
+      "loss": 0.9929,
+      "step": 772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.59752167598961,
+      "learning_rate": 1.962159482876901e-05,
+      "loss": 0.9495,
+      "step": 773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7301326515733286,
+      "learning_rate": 1.962027671586086e-05,
+      "loss": 0.9566,
+      "step": 774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4314960664362162,
+      "learning_rate": 1.9618956355655107e-05,
+      "loss": 0.9669,
+      "step": 775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6231886139640732,
+      "learning_rate": 1.9617633748460193e-05,
+      "loss": 0.9891,
+      "step": 776
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9323902546254417,
+      "learning_rate": 1.9616308894585078e-05,
+      "loss": 0.9722,
+      "step": 777
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.878864606294589,
+      "learning_rate": 1.9614981794339244e-05,
+      "loss": 0.3285,
+      "step": 778
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7802191332111326,
+      "learning_rate": 1.961365244803271e-05,
+      "loss": 1.0153,
+      "step": 779
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7960769954528806,
+      "learning_rate": 1.9612320855976002e-05,
+      "loss": 0.9685,
+      "step": 780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6217763647045618,
+      "learning_rate": 1.9610987018480186e-05,
+      "loss": 1.0031,
+      "step": 781
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.533360438567821,
+      "learning_rate": 1.9609650935856847e-05,
+      "loss": 0.8413,
+      "step": 782
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8564985910763983,
+      "learning_rate": 1.9608312608418087e-05,
+      "loss": 0.9401,
+      "step": 783
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7005211804337508,
+      "learning_rate": 1.960697203647655e-05,
+      "loss": 1.0817,
+      "step": 784
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7265343383287515,
+      "learning_rate": 1.9605629220345382e-05,
+      "loss": 1.0034,
+      "step": 785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6573616078460243,
+      "learning_rate": 1.9604284160338276e-05,
+      "loss": 0.9444,
+      "step": 786
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5693599542419174,
+      "learning_rate": 1.9602936856769432e-05,
+      "loss": 0.9385,
+      "step": 787
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5807249079892538,
+      "learning_rate": 1.9601587309953584e-05,
+      "loss": 0.9423,
+      "step": 788
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.595139212401982,
+      "learning_rate": 1.960023552020598e-05,
+      "loss": 0.9456,
+      "step": 789
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6573991073134278,
+      "learning_rate": 1.9598881487842406e-05,
+      "loss": 0.9606,
+      "step": 790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5280299828053654,
+      "learning_rate": 1.9597525213179157e-05,
+      "loss": 1.0081,
+      "step": 791
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4739930349034827,
+      "learning_rate": 1.9596166696533062e-05,
+      "loss": 0.965,
+      "step": 792
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.605344258189759,
+      "learning_rate": 1.9594805938221473e-05,
+      "loss": 0.9415,
+      "step": 793
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.890047419613228,
+      "learning_rate": 1.959344293856226e-05,
+      "loss": 0.984,
+      "step": 794
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9133741187045141,
+      "learning_rate": 1.959207769787382e-05,
+      "loss": 0.9391,
+      "step": 795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6111552983857254,
+      "learning_rate": 1.959071021647507e-05,
+      "loss": 0.9868,
+      "step": 796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5936306249565788,
+      "learning_rate": 1.9589340494685464e-05,
+      "loss": 0.9812,
+      "step": 797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5988902267485232,
+      "learning_rate": 1.9587968532824963e-05,
+      "loss": 1.0153,
+      "step": 798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.468728497784049,
+      "learning_rate": 1.958659433121406e-05,
+      "loss": 0.915,
+      "step": 799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.659695382942566,
+      "learning_rate": 1.958521789017376e-05,
+      "loss": 0.9228,
+      "step": 800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5599141479387162,
+      "learning_rate": 1.958383921002561e-05,
+      "loss": 0.9287,
+      "step": 801
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7991215852729405,
+      "learning_rate": 1.9582458291091664e-05,
+      "loss": 1.0624,
+      "step": 802
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6836577013884875,
+      "learning_rate": 1.9581075133694508e-05,
+      "loss": 0.9117,
+      "step": 803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1281370925839833,
+      "learning_rate": 1.9579689738157245e-05,
+      "loss": 0.9825,
+      "step": 804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4650976673035854,
+      "learning_rate": 1.9578302104803506e-05,
+      "loss": 0.9993,
+      "step": 805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8172287385584375,
+      "learning_rate": 1.957691223395744e-05,
+      "loss": 0.3352,
+      "step": 806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6686968375695956,
+      "learning_rate": 1.957552012594372e-05,
+      "loss": 0.9538,
+      "step": 807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.597347173663805,
+      "learning_rate": 1.957412578108755e-05,
+      "loss": 0.9667,
+      "step": 808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4868611847346722,
+      "learning_rate": 1.957272919971464e-05,
+      "loss": 1.0481,
+      "step": 809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8735716920337183,
+      "learning_rate": 1.9571330382151236e-05,
+      "loss": 0.3028,
+      "step": 810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4596733830732393,
+      "learning_rate": 1.95699293287241e-05,
+      "loss": 0.9571,
+      "step": 811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6699330837029742,
+      "learning_rate": 1.956852603976052e-05,
+      "loss": 0.958,
+      "step": 812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7239402323941375,
+      "learning_rate": 1.9567120515588307e-05,
+      "loss": 0.9794,
+      "step": 813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6432445752609868,
+      "learning_rate": 1.9565712756535785e-05,
+      "loss": 0.987,
+      "step": 814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6719034696097996,
+      "learning_rate": 1.9564302762931812e-05,
+      "loss": 0.8879,
+      "step": 815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9281853724039748,
+      "learning_rate": 1.956289053510576e-05,
+      "loss": 0.35,
+      "step": 816
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7122314178552518,
+      "learning_rate": 1.9561476073387527e-05,
+      "loss": 0.9109,
+      "step": 817
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.605043130921607,
+      "learning_rate": 1.956005937810753e-05,
+      "loss": 0.8973,
+      "step": 818
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.767181525865368,
+      "learning_rate": 1.955864044959671e-05,
+      "loss": 0.9653,
+      "step": 819
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.071947557845971,
+      "learning_rate": 1.955721928818653e-05,
+      "loss": 0.9143,
+      "step": 820
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0343881098299943,
+      "learning_rate": 1.955579589420897e-05,
+      "loss": 0.9355,
+      "step": 821
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7729730228894678,
+      "learning_rate": 1.9554370267996537e-05,
+      "loss": 0.9691,
+      "step": 822
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8539727953985209,
+      "learning_rate": 1.9552942409882257e-05,
+      "loss": 0.3385,
+      "step": 823
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6703362614428694,
+      "learning_rate": 1.9551512320199684e-05,
+      "loss": 0.9498,
+      "step": 824
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.874488032500137,
+      "learning_rate": 1.9550079999282874e-05,
+      "loss": 0.9428,
+      "step": 825
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5722811011032523,
+      "learning_rate": 1.9548645447466433e-05,
+      "loss": 0.9365,
+      "step": 826
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.611663630842481,
+      "learning_rate": 1.954720866508546e-05,
+      "loss": 0.9423,
+      "step": 827
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7398318468448093,
+      "learning_rate": 1.954576965247559e-05,
+      "loss": 0.9502,
+      "step": 828
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.476052281238449,
+      "learning_rate": 1.9544328409972978e-05,
+      "loss": 0.9546,
+      "step": 829
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8316432385069856,
+      "learning_rate": 1.9542884937914302e-05,
+      "loss": 0.3392,
+      "step": 830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6086582432602319,
+      "learning_rate": 1.9541439236636752e-05,
+      "loss": 0.9685,
+      "step": 831
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.627499851444343,
+      "learning_rate": 1.9539991306478046e-05,
+      "loss": 0.9678,
+      "step": 832
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4601392681712042,
+      "learning_rate": 1.953854114777642e-05,
+      "loss": 0.8833,
+      "step": 833
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6841304801946464,
+      "learning_rate": 1.9537088760870632e-05,
+      "loss": 0.9526,
+      "step": 834
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6870119704868776,
+      "learning_rate": 1.953563414609996e-05,
+      "loss": 0.9043,
+      "step": 835
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6331167376822397,
+      "learning_rate": 1.95341773038042e-05,
+      "loss": 1.1093,
+      "step": 836
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3930486162495799,
+      "learning_rate": 1.953271823432367e-05,
+      "loss": 0.9266,
+      "step": 837
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.923470220822032,
+      "learning_rate": 1.9531256937999217e-05,
+      "loss": 0.94,
+      "step": 838
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6023026640119524,
+      "learning_rate": 1.952979341517219e-05,
+      "loss": 1.0197,
+      "step": 839
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8427734011117904,
+      "learning_rate": 1.9528327666184472e-05,
+      "loss": 0.9667,
+      "step": 840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7607792127358646,
+      "learning_rate": 1.9526859691378465e-05,
+      "loss": 0.87,
+      "step": 841
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.738292970510681,
+      "learning_rate": 1.952538949109708e-05,
+      "loss": 0.9786,
+      "step": 842
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7149022077185407,
+      "learning_rate": 1.9523917065683764e-05,
+      "loss": 0.9481,
+      "step": 843
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6248001906539882,
+      "learning_rate": 1.952244241548247e-05,
+      "loss": 1.0003,
+      "step": 844
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5887165132952656,
+      "learning_rate": 1.952096554083768e-05,
+      "loss": 0.9199,
+      "step": 845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4189858020589843,
+      "learning_rate": 1.9519486442094397e-05,
+      "loss": 0.9986,
+      "step": 846
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6247350070913582,
+      "learning_rate": 1.9518005119598124e-05,
+      "loss": 0.989,
+      "step": 847
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5457802659279052,
+      "learning_rate": 1.951652157369491e-05,
+      "loss": 0.9394,
+      "step": 848
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9914760328946208,
+      "learning_rate": 1.9515035804731313e-05,
+      "loss": 0.8794,
+      "step": 849
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5605248412676282,
+      "learning_rate": 1.9513547813054397e-05,
+      "loss": 0.9008,
+      "step": 850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6121720699064361,
+      "learning_rate": 1.951205759901177e-05,
+      "loss": 1.0623,
+      "step": 851
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6745778433427454,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.8868,
+      "step": 852
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7653379671174583,
+      "learning_rate": 1.9509070505222336e-05,
+      "loss": 0.9075,
+      "step": 853
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7888150269683845,
+      "learning_rate": 1.9507573626173317e-05,
+      "loss": 0.3222,
+      "step": 854
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7344204965414827,
+      "learning_rate": 1.9506074526154155e-05,
+      "loss": 0.9156,
+      "step": 855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8533577253818907,
+      "learning_rate": 1.950457320551503e-05,
+      "loss": 0.9556,
+      "step": 856
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6143951594129295,
+      "learning_rate": 1.9503069664606663e-05,
+      "loss": 0.9598,
+      "step": 857
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.818599168474104,
+      "learning_rate": 1.950156390378027e-05,
+      "loss": 0.3028,
+      "step": 858
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7817867682255593,
+      "learning_rate": 1.9500055923387608e-05,
+      "loss": 0.9002,
+      "step": 859
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5324585947406468,
+      "learning_rate": 1.9498545723780932e-05,
+      "loss": 0.912,
+      "step": 860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6974837115858863,
+      "learning_rate": 1.9497033305313033e-05,
+      "loss": 0.9645,
+      "step": 861
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5706612451183646,
+      "learning_rate": 1.9495518668337204e-05,
+      "loss": 0.7744,
+      "step": 862
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.638377908930441,
+      "learning_rate": 1.949400181320727e-05,
+      "loss": 0.9316,
+      "step": 863
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.627236425754178,
+      "learning_rate": 1.9492482740277564e-05,
+      "loss": 0.9285,
+      "step": 864
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5474317221945242,
+      "learning_rate": 1.9490961449902946e-05,
+      "loss": 1.0384,
+      "step": 865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8115569633404542,
+      "learning_rate": 1.948943794243879e-05,
+      "loss": 0.9834,
+      "step": 866
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5825512490629434,
+      "learning_rate": 1.9487912218240983e-05,
+      "loss": 0.9289,
+      "step": 867
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7189175199020104,
+      "learning_rate": 1.9486384277665938e-05,
+      "loss": 0.9621,
+      "step": 868
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8574553595973367,
+      "learning_rate": 1.9484854121070578e-05,
+      "loss": 1.0111,
+      "step": 869
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6584234855089184,
+      "learning_rate": 1.9483321748812353e-05,
+      "loss": 0.9371,
+      "step": 870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4512580612383406,
+      "learning_rate": 1.948178716124922e-05,
+      "loss": 0.9797,
+      "step": 871
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6227528687952613,
+      "learning_rate": 1.9480250358739667e-05,
+      "loss": 1.0155,
+      "step": 872
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3941354403582422,
+      "learning_rate": 1.947871134164268e-05,
+      "loss": 0.9815,
+      "step": 873
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.505416291604214,
+      "learning_rate": 1.9477170110317783e-05,
+      "loss": 0.9639,
+      "step": 874
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8965056361215689,
+      "learning_rate": 1.9475626665125e-05,
+      "loss": 0.9266,
+      "step": 875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.718278169891963,
+      "learning_rate": 1.947408100642489e-05,
+      "loss": 1.0076,
+      "step": 876
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5414496706199177,
+      "learning_rate": 1.947253313457851e-05,
+      "loss": 0.9159,
+      "step": 877
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.523081742831939,
+      "learning_rate": 1.9470983049947446e-05,
+      "loss": 0.8902,
+      "step": 878
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7716021093844976,
+      "learning_rate": 1.9469430752893796e-05,
+      "loss": 0.976,
+      "step": 879
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5944098305813688,
+      "learning_rate": 1.946787624378018e-05,
+      "loss": 0.8876,
+      "step": 880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.582416853005702,
+      "learning_rate": 1.946631952296973e-05,
+      "loss": 0.9828,
+      "step": 881
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7543406872405687,
+      "learning_rate": 1.94647605908261e-05,
+      "loss": 0.9286,
+      "step": 882
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7509437192016315,
+      "learning_rate": 1.946319944771345e-05,
+      "loss": 0.9437,
+      "step": 883
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5976541584439565,
+      "learning_rate": 1.9461636093996468e-05,
+      "loss": 1.0232,
+      "step": 884
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3474434836678684,
+      "learning_rate": 1.9460070530040348e-05,
+      "loss": 0.8281,
+      "step": 885
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4156152856463997,
+      "learning_rate": 1.9458502756210814e-05,
+      "loss": 0.9457,
+      "step": 886
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8484432165552453,
+      "learning_rate": 1.9456932772874092e-05,
+      "loss": 0.8671,
+      "step": 887
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2946776434619025,
+      "learning_rate": 1.9455360580396934e-05,
+      "loss": 0.947,
+      "step": 888
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7082894116566696,
+      "learning_rate": 1.94537861791466e-05,
+      "loss": 0.9695,
+      "step": 889
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4976028657233036,
+      "learning_rate": 1.9452209569490874e-05,
+      "loss": 0.981,
+      "step": 890
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8254217524765315,
+      "learning_rate": 1.945063075179805e-05,
+      "loss": 0.9312,
+      "step": 891
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.476431507148291,
+      "learning_rate": 1.944904972643694e-05,
+      "loss": 0.979,
+      "step": 892
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8768589768232298,
+      "learning_rate": 1.9447466493776877e-05,
+      "loss": 0.9469,
+      "step": 893
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5379945213156887,
+      "learning_rate": 1.9445881054187694e-05,
+      "loss": 0.9659,
+      "step": 894
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5909786619757784,
+      "learning_rate": 1.944429340803976e-05,
+      "loss": 0.9913,
+      "step": 895
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5529905341837291,
+      "learning_rate": 1.9442703555703945e-05,
+      "loss": 0.9874,
+      "step": 896
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4110654742267894,
+      "learning_rate": 1.944111149755164e-05,
+      "loss": 0.9531,
+      "step": 897
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9139645541321378,
+      "learning_rate": 1.9439517233954744e-05,
+      "loss": 0.3246,
+      "step": 898
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7918989713367353,
+      "learning_rate": 1.9437920765285683e-05,
+      "loss": 1.0115,
+      "step": 899
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6180711626310682,
+      "learning_rate": 1.9436322091917392e-05,
+      "loss": 1.0617,
+      "step": 900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7613615976563735,
+      "learning_rate": 1.943472121422332e-05,
+      "loss": 0.8652,
+      "step": 901
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.838596921762269,
+      "learning_rate": 1.9433118132577432e-05,
+      "loss": 0.9465,
+      "step": 902
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7593066826882449,
+      "learning_rate": 1.943151284735421e-05,
+      "loss": 0.9267,
+      "step": 903
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7286081547944228,
+      "learning_rate": 1.9429905358928648e-05,
+      "loss": 0.9482,
+      "step": 904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6013748989484846,
+      "learning_rate": 1.9428295667676253e-05,
+      "loss": 0.9392,
+      "step": 905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.477979721648062,
+      "learning_rate": 1.942668377397305e-05,
+      "loss": 0.9403,
+      "step": 906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8116506629543034,
+      "learning_rate": 1.9425069678195577e-05,
+      "loss": 0.9181,
+      "step": 907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.463518801071589,
+      "learning_rate": 1.9423453380720892e-05,
+      "loss": 0.9259,
+      "step": 908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7889208815673736,
+      "learning_rate": 1.9421834881926558e-05,
+      "loss": 0.9983,
+      "step": 909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6250521486424425,
+      "learning_rate": 1.9420214182190657e-05,
+      "loss": 0.9634,
+      "step": 910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5477731149331013,
+      "learning_rate": 1.941859128189178e-05,
+      "loss": 0.961,
+      "step": 911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9274218565859795,
+      "learning_rate": 1.9416966181409047e-05,
+      "loss": 0.973,
+      "step": 912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7070551266080638,
+      "learning_rate": 1.9415338881122074e-05,
+      "loss": 0.9625,
+      "step": 913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.570203737031849,
+      "learning_rate": 1.9413709381411003e-05,
+      "loss": 0.8757,
+      "step": 914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5458424859877031,
+      "learning_rate": 1.9412077682656477e-05,
+      "loss": 0.9537,
+      "step": 915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4804993993661473,
+      "learning_rate": 1.941044378523967e-05,
+      "loss": 0.8397,
+      "step": 916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7101371762548216,
+      "learning_rate": 1.9408807689542257e-05,
+      "loss": 0.924,
+      "step": 917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7586849018709336,
+      "learning_rate": 1.9407169395946427e-05,
+      "loss": 0.9729,
+      "step": 918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6309936159814857,
+      "learning_rate": 1.9405528904834895e-05,
+      "loss": 0.9666,
+      "step": 919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6603488103603472,
+      "learning_rate": 1.940388621659087e-05,
+      "loss": 0.9298,
+      "step": 920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5612286879365345,
+      "learning_rate": 1.9402241331598092e-05,
+      "loss": 0.9558,
+      "step": 921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7696040056441986,
+      "learning_rate": 1.94005942502408e-05,
+      "loss": 0.9099,
+      "step": 922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8667631018292109,
+      "learning_rate": 1.939894497290375e-05,
+      "loss": 0.3391,
+      "step": 923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8132504592887957,
+      "learning_rate": 1.9397293499972224e-05,
+      "loss": 0.9872,
+      "step": 924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8360100975331577,
+      "learning_rate": 1.9395639831831997e-05,
+      "loss": 0.3111,
+      "step": 925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6230927918752514,
+      "learning_rate": 1.939398396886937e-05,
+      "loss": 0.9175,
+      "step": 926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8194081757435474,
+      "learning_rate": 1.9392325911471154e-05,
+      "loss": 0.9112,
+      "step": 927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6459187817926038,
+      "learning_rate": 1.939066566002467e-05,
+      "loss": 0.9796,
+      "step": 928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4737549725477763,
+      "learning_rate": 1.938900321491775e-05,
+      "loss": 0.9073,
+      "step": 929
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9749829234217935,
+      "learning_rate": 1.9387338576538743e-05,
+      "loss": 0.9278,
+      "step": 930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7038633436916437,
+      "learning_rate": 1.938567174527651e-05,
+      "loss": 0.9326,
+      "step": 931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6000795965773242,
+      "learning_rate": 1.9384002721520423e-05,
+      "loss": 0.9664,
+      "step": 932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.528057646802909,
+      "learning_rate": 1.9382331505660364e-05,
+      "loss": 0.9458,
+      "step": 933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3866788253946987,
+      "learning_rate": 1.938065809808673e-05,
+      "loss": 0.903,
+      "step": 934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0752714504340406,
+      "learning_rate": 1.9378982499190434e-05,
+      "loss": 0.3537,
+      "step": 935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6387625806708923,
+      "learning_rate": 1.937730470936289e-05,
+      "loss": 0.9443,
+      "step": 936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6095798903658123,
+      "learning_rate": 1.937562472899603e-05,
+      "loss": 0.9536,
+      "step": 937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8058085489516018,
+      "learning_rate": 1.9373942558482303e-05,
+      "loss": 1.0234,
+      "step": 938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4679773865385315,
+      "learning_rate": 1.9372258198214654e-05,
+      "loss": 0.8131,
+      "step": 939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.504810337076543,
+      "learning_rate": 1.937057164858656e-05,
+      "loss": 0.9705,
+      "step": 940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5347373957621673,
+      "learning_rate": 1.9368882909991996e-05,
+      "loss": 0.9555,
+      "step": 941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8589917753550486,
+      "learning_rate": 1.936719198282545e-05,
+      "loss": 0.2939,
+      "step": 942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8675059590818638,
+      "learning_rate": 1.9365498867481926e-05,
+      "loss": 0.3401,
+      "step": 943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6451040502696443,
+      "learning_rate": 1.9363803564356932e-05,
+      "loss": 0.9992,
+      "step": 944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.741057516475174,
+      "learning_rate": 1.936210607384649e-05,
+      "loss": 0.9762,
+      "step": 945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5599441319175684,
+      "learning_rate": 1.936040639634714e-05,
+      "loss": 0.8611,
+      "step": 946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9412636173986958,
+      "learning_rate": 1.935870453225592e-05,
+      "loss": 0.353,
+      "step": 947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.142613107261814,
+      "learning_rate": 1.935700048197039e-05,
+      "loss": 0.9236,
+      "step": 948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5365310075242953,
+      "learning_rate": 1.9355294245888617e-05,
+      "loss": 1.0025,
+      "step": 949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.589850715087705,
+      "learning_rate": 1.9353585824409178e-05,
+      "loss": 1.0179,
+      "step": 950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6281875161324748,
+      "learning_rate": 1.9351875217931154e-05,
+      "loss": 0.9697,
+      "step": 951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.58210676742971,
+      "learning_rate": 1.9350162426854152e-05,
+      "loss": 0.9674,
+      "step": 952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.684299553382832,
+      "learning_rate": 1.9348447451578273e-05,
+      "loss": 1.0601,
+      "step": 953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8102414769658999,
+      "learning_rate": 1.9346730292504134e-05,
+      "loss": 1.0156,
+      "step": 954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7662935932176484,
+      "learning_rate": 1.934501095003287e-05,
+      "loss": 0.8664,
+      "step": 955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.456918291463753,
+      "learning_rate": 1.9343289424566122e-05,
+      "loss": 0.9181,
+      "step": 956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7153072415109856,
+      "learning_rate": 1.934156571650603e-05,
+      "loss": 0.8278,
+      "step": 957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.693980805834375,
+      "learning_rate": 1.9339839826255257e-05,
+      "loss": 0.9765,
+      "step": 958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6779858641600216,
+      "learning_rate": 1.9338111754216968e-05,
+      "loss": 0.9532,
+      "step": 959
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.584152107553659,
+      "learning_rate": 1.9336381500794845e-05,
+      "loss": 0.8465,
+      "step": 960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5607130262991993,
+      "learning_rate": 1.9334649066393072e-05,
+      "loss": 0.9158,
+      "step": 961
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4843447600597182,
+      "learning_rate": 1.933291445141635e-05,
+      "loss": 0.8929,
+      "step": 962
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6243375668224997,
+      "learning_rate": 1.9331177656269878e-05,
+      "loss": 0.9765,
+      "step": 963
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5112402630613377,
+      "learning_rate": 1.932943868135938e-05,
+      "loss": 0.976,
+      "step": 964
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5868170324152127,
+      "learning_rate": 1.9327697527091076e-05,
+      "loss": 0.8916,
+      "step": 965
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.600139313552218,
+      "learning_rate": 1.9325954193871698e-05,
+      "loss": 0.9441,
+      "step": 966
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.423233759805251,
+      "learning_rate": 1.9324208682108493e-05,
+      "loss": 0.8994,
+      "step": 967
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8389678335153199,
+      "learning_rate": 1.932246099220921e-05,
+      "loss": 0.3334,
+      "step": 968
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6511189112258235,
+      "learning_rate": 1.932071112458211e-05,
+      "loss": 0.8791,
+      "step": 969
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5945456870031507,
+      "learning_rate": 1.9318959079635965e-05,
+      "loss": 0.9591,
+      "step": 970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.573552942160686,
+      "learning_rate": 1.931720485778005e-05,
+      "loss": 1.0872,
+      "step": 971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4453527792911074,
+      "learning_rate": 1.931544845942415e-05,
+      "loss": 0.9623,
+      "step": 972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8379703185945078,
+      "learning_rate": 1.9313689884978567e-05,
+      "loss": 0.3613,
+      "step": 973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6831574449960423,
+      "learning_rate": 1.9311929134854093e-05,
+      "loss": 0.9332,
+      "step": 974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6223101659621977,
+      "learning_rate": 1.931016620946205e-05,
+      "loss": 0.9178,
+      "step": 975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.855097133121107,
+      "learning_rate": 1.930840110921425e-05,
+      "loss": 0.2847,
+      "step": 976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5152435523234884,
+      "learning_rate": 1.9306633834523022e-05,
+      "loss": 0.8288,
+      "step": 977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8569361260096856,
+      "learning_rate": 1.9304864385801204e-05,
+      "loss": 0.3056,
+      "step": 978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8565481660094034,
+      "learning_rate": 1.9303092763462142e-05,
+      "loss": 0.8711,
+      "step": 979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9333937936525332,
+      "learning_rate": 1.9301318967919684e-05,
+      "loss": 0.9718,
+      "step": 980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9502049743382441,
+      "learning_rate": 1.9299542999588184e-05,
+      "loss": 1.0259,
+      "step": 981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7067858532691014,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 0.9195,
+      "step": 982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.705227669431673,
+      "learning_rate": 1.929598454621805e-05,
+      "loss": 0.858,
+      "step": 983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5549062019144357,
+      "learning_rate": 1.9294202062010667e-05,
+      "loss": 0.9408,
+      "step": 984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8191871479844153,
+      "learning_rate": 1.929241740667676e-05,
+      "loss": 0.9941,
+      "step": 985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.564111482985053,
+      "learning_rate": 1.9290630580633215e-05,
+      "loss": 0.9335,
+      "step": 986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9801970291709687,
+      "learning_rate": 1.9288841584297445e-05,
+      "loss": 0.8279,
+      "step": 987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.567372245990633,
+      "learning_rate": 1.9287050418087355e-05,
+      "loss": 0.9691,
+      "step": 988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5953372252161375,
+      "learning_rate": 1.9285257082421363e-05,
+      "loss": 0.9838,
+      "step": 989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7250035247041102,
+      "learning_rate": 1.9283461577718387e-05,
+      "loss": 1.0088,
+      "step": 990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6945413668318634,
+      "learning_rate": 1.9281663904397868e-05,
+      "loss": 1.0042,
+      "step": 991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7227946602108561,
+      "learning_rate": 1.927986406287973e-05,
+      "loss": 0.96,
+      "step": 992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7345490411262316,
+      "learning_rate": 1.9278062053584426e-05,
+      "loss": 0.879,
+      "step": 993
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6858241979163227,
+      "learning_rate": 1.92762578769329e-05,
+      "loss": 0.9862,
+      "step": 994
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4507724525460193,
+      "learning_rate": 1.9274451533346617e-05,
+      "loss": 0.8984,
+      "step": 995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6060884328583993,
+      "learning_rate": 1.9272643023247527e-05,
+      "loss": 0.8906,
+      "step": 996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5068881235338356,
+      "learning_rate": 1.92708323470581e-05,
+      "loss": 0.8976,
+      "step": 997
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.458801622575446,
+      "learning_rate": 1.9269019505201316e-05,
+      "loss": 0.9613,
+      "step": 998
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7842201725250648,
+      "learning_rate": 1.926720449810065e-05,
+      "loss": 0.9744,
+      "step": 999
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7434290655106102,
+      "learning_rate": 1.9265387326180094e-05,
+      "loss": 0.9213,
+      "step": 1000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5809507488983856,
+      "learning_rate": 1.9263567989864135e-05,
+      "loss": 0.94,
+      "step": 1001
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0103325391784943,
+      "learning_rate": 1.9261746489577767e-05,
+      "loss": 0.9846,
+      "step": 1002
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4733187161883579,
+      "learning_rate": 1.92599228257465e-05,
+      "loss": 0.9268,
+      "step": 1003
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.73848092346793,
+      "learning_rate": 1.9258096998796335e-05,
+      "loss": 0.9287,
+      "step": 1004
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5786852190432847,
+      "learning_rate": 1.9256269009153793e-05,
+      "loss": 1.0006,
+      "step": 1005
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.047706222288261,
+      "learning_rate": 1.9254438857245885e-05,
+      "loss": 1.0241,
+      "step": 1006
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.564794242473782,
+      "learning_rate": 1.925260654350014e-05,
+      "loss": 0.8822,
+      "step": 1007
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7244582623023423,
+      "learning_rate": 1.925077206834458e-05,
+      "loss": 0.9899,
+      "step": 1008
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5106414287874854,
+      "learning_rate": 1.924893543220775e-05,
+      "loss": 0.9203,
+      "step": 1009
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6940817584575845,
+      "learning_rate": 1.924709663551868e-05,
+      "loss": 0.9331,
+      "step": 1010
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4736767061444838,
+      "learning_rate": 1.924525567870691e-05,
+      "loss": 0.9139,
+      "step": 1011
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5892533394013502,
+      "learning_rate": 1.92434125622025e-05,
+      "loss": 0.9577,
+      "step": 1012
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5069840037690432,
+      "learning_rate": 1.924156728643599e-05,
+      "loss": 0.9924,
+      "step": 1013
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.827463580569672,
+      "learning_rate": 1.923971985183844e-05,
+      "loss": 1.0114,
+      "step": 1014
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8072351096196546,
+      "learning_rate": 1.9237870258841412e-05,
+      "loss": 0.9514,
+      "step": 1015
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6544125718032203,
+      "learning_rate": 1.9236018507876973e-05,
+      "loss": 0.8919,
+      "step": 1016
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6331111598313437,
+      "learning_rate": 1.9234164599377692e-05,
+      "loss": 0.874,
+      "step": 1017
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6305193472351633,
+      "learning_rate": 1.923230853377664e-05,
+      "loss": 0.8717,
+      "step": 1018
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8152351603497061,
+      "learning_rate": 1.9230450311507393e-05,
+      "loss": 1.0009,
+      "step": 1019
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6114396940966587,
+      "learning_rate": 1.9228589933004038e-05,
+      "loss": 0.975,
+      "step": 1020
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6012989408836813,
+      "learning_rate": 1.922672739870115e-05,
+      "loss": 0.9333,
+      "step": 1021
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7855007219011148,
+      "learning_rate": 1.9224862709033823e-05,
+      "loss": 0.2865,
+      "step": 1022
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.483599431306749,
+      "learning_rate": 1.922299586443765e-05,
+      "loss": 0.8616,
+      "step": 1023
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6727535577627035,
+      "learning_rate": 1.9221126865348726e-05,
+      "loss": 0.9666,
+      "step": 1024
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6338024218663487,
+      "learning_rate": 1.9219255712203643e-05,
+      "loss": 0.9909,
+      "step": 1025
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5637048760714245,
+      "learning_rate": 1.921738240543951e-05,
+      "loss": 0.8844,
+      "step": 1026
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5686945741102496,
+      "learning_rate": 1.9215506945493933e-05,
+      "loss": 0.9415,
+      "step": 1027
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8983981362302563,
+      "learning_rate": 1.921362933280501e-05,
+      "loss": 0.3279,
+      "step": 1028
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6451691310090992,
+      "learning_rate": 1.9211749567811357e-05,
+      "loss": 1.0003,
+      "step": 1029
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4457971868499322,
+      "learning_rate": 1.9209867650952088e-05,
+      "loss": 0.8923,
+      "step": 1030
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5679541975371474,
+      "learning_rate": 1.920798358266682e-05,
+      "loss": 0.8575,
+      "step": 1031
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3624673288260447,
+      "learning_rate": 1.9206097363395668e-05,
+      "loss": 0.9169,
+      "step": 1032
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4873129881378602,
+      "learning_rate": 1.9204208993579256e-05,
+      "loss": 0.9479,
+      "step": 1033
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5513706432478374,
+      "learning_rate": 1.9202318473658707e-05,
+      "loss": 0.9757,
+      "step": 1034
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.557317703777887,
+      "learning_rate": 1.9200425804075643e-05,
+      "loss": 0.8763,
+      "step": 1035
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5147214892815681,
+      "learning_rate": 1.91985309852722e-05,
+      "loss": 0.907,
+      "step": 1036
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8336115611527413,
+      "learning_rate": 1.9196634017690993e-05,
+      "loss": 0.8366,
+      "step": 1037
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5462916898783394,
+      "learning_rate": 1.919473490177517e-05,
+      "loss": 0.9545,
+      "step": 1038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.553231495625522,
+      "learning_rate": 1.9192833637968357e-05,
+      "loss": 0.9268,
+      "step": 1039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7533040956484711,
+      "learning_rate": 1.919093022671469e-05,
+      "loss": 0.8703,
+      "step": 1040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.603529183148066,
+      "learning_rate": 1.9189024668458803e-05,
+      "loss": 0.9518,
+      "step": 1041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6093992907348331,
+      "learning_rate": 1.9187116963645845e-05,
+      "loss": 0.8954,
+      "step": 1042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5781791454648255,
+      "learning_rate": 1.9185207112721443e-05,
+      "loss": 0.9686,
+      "step": 1043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4886508899381776,
+      "learning_rate": 1.9183295116131747e-05,
+      "loss": 0.9872,
+      "step": 1044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7310737376825676,
+      "learning_rate": 1.91813809743234e-05,
+      "loss": 0.981,
+      "step": 1045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6699307837443829,
+      "learning_rate": 1.917946468774354e-05,
+      "loss": 0.9904,
+      "step": 1046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5569391143917253,
+      "learning_rate": 1.9177546256839814e-05,
+      "loss": 0.9039,
+      "step": 1047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5562505197332657,
+      "learning_rate": 1.9175625682060367e-05,
+      "loss": 1.0236,
+      "step": 1048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6843176124223667,
+      "learning_rate": 1.917370296385385e-05,
+      "loss": 0.9393,
+      "step": 1049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8842551839750727,
+      "learning_rate": 1.9171778102669404e-05,
+      "loss": 0.9301,
+      "step": 1050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5283218590969267,
+      "learning_rate": 1.916985109895668e-05,
+      "loss": 0.9344,
+      "step": 1051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7811991972956647,
+      "learning_rate": 1.9167921953165827e-05,
+      "loss": 0.9073,
+      "step": 1052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.009274141812004,
+      "learning_rate": 1.916599066574749e-05,
+      "loss": 1.0302,
+      "step": 1053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5560398545916263,
+      "learning_rate": 1.916405723715282e-05,
+      "loss": 0.8723,
+      "step": 1054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6982062556988202,
+      "learning_rate": 1.9162121667833473e-05,
+      "loss": 0.827,
+      "step": 1055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4390011295161496,
+      "learning_rate": 1.9160183958241584e-05,
+      "loss": 1.012,
+      "step": 1056
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5675652703775997,
+      "learning_rate": 1.9158244108829815e-05,
+      "loss": 0.9237,
+      "step": 1057
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7182232678190625,
+      "learning_rate": 1.9156302120051308e-05,
+      "loss": 0.8548,
+      "step": 1058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.540934584726451,
+      "learning_rate": 1.915435799235971e-05,
+      "loss": 0.9531,
+      "step": 1059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5662737954471622,
+      "learning_rate": 1.9152411726209176e-05,
+      "loss": 0.9333,
+      "step": 1060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5147867502384917,
+      "learning_rate": 1.9150463322054352e-05,
+      "loss": 0.8942,
+      "step": 1061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.455687265455594,
+      "learning_rate": 1.9148512780350384e-05,
+      "loss": 0.9758,
+      "step": 1062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.542549480743184,
+      "learning_rate": 1.914656010155292e-05,
+      "loss": 1.0144,
+      "step": 1063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6262299732539058,
+      "learning_rate": 1.9144605286118104e-05,
+      "loss": 1.0009,
+      "step": 1064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6034253997602659,
+      "learning_rate": 1.914264833450258e-05,
+      "loss": 1.0239,
+      "step": 1065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6515694758828192,
+      "learning_rate": 1.9140689247163497e-05,
+      "loss": 0.996,
+      "step": 1066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5371943126802028,
+      "learning_rate": 1.9138728024558494e-05,
+      "loss": 0.9552,
+      "step": 1067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4234567087135699,
+      "learning_rate": 1.9136764667145715e-05,
+      "loss": 0.9426,
+      "step": 1068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.584905264976068,
+      "learning_rate": 1.91347991753838e-05,
+      "loss": 0.9271,
+      "step": 1069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5283414309081558,
+      "learning_rate": 1.9132831549731886e-05,
+      "loss": 0.9339,
+      "step": 1070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6230574262071782,
+      "learning_rate": 1.9130861790649613e-05,
+      "loss": 0.9209,
+      "step": 1071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.782268243589682,
+      "learning_rate": 1.9128889898597117e-05,
+      "loss": 0.8811,
+      "step": 1072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.642954884070177,
+      "learning_rate": 1.912691587403503e-05,
+      "loss": 0.9269,
+      "step": 1073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.611167321706687,
+      "learning_rate": 1.9124939717424486e-05,
+      "loss": 0.9532,
+      "step": 1074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.039632935772377,
+      "learning_rate": 1.9122961429227115e-05,
+      "loss": 0.9914,
+      "step": 1075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.507891765917394,
+      "learning_rate": 1.9120981009905044e-05,
+      "loss": 0.9597,
+      "step": 1076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7353146859857143,
+      "learning_rate": 1.91189984599209e-05,
+      "loss": 0.9525,
+      "step": 1077
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.743926213446444,
+      "learning_rate": 1.911701377973781e-05,
+      "loss": 0.9592,
+      "step": 1078
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4958115191969878,
+      "learning_rate": 1.9115026969819396e-05,
+      "loss": 0.9158,
+      "step": 1079
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5499496067703662,
+      "learning_rate": 1.9113038030629767e-05,
+      "loss": 0.9213,
+      "step": 1080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.632258271537784,
+      "learning_rate": 1.9111046962633547e-05,
+      "loss": 0.9798,
+      "step": 1081
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6584206293562322,
+      "learning_rate": 1.910905376629585e-05,
+      "loss": 0.912,
+      "step": 1082
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8075650041200193,
+      "learning_rate": 1.9107058442082288e-05,
+      "loss": 0.9683,
+      "step": 1083
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5794052075073737,
+      "learning_rate": 1.9105060990458964e-05,
+      "loss": 0.9788,
+      "step": 1084
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4310453616359426,
+      "learning_rate": 1.9103061411892488e-05,
+      "loss": 0.932,
+      "step": 1085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4454411410741406,
+      "learning_rate": 1.9101059706849957e-05,
+      "loss": 0.9622,
+      "step": 1086
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6760446379265568,
+      "learning_rate": 1.9099055875798974e-05,
+      "loss": 0.9857,
+      "step": 1087
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6486076508974432,
+      "learning_rate": 1.909704991920763e-05,
+      "loss": 0.9774,
+      "step": 1088
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.612687760287242,
+      "learning_rate": 1.909504183754452e-05,
+      "loss": 0.9426,
+      "step": 1089
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.751900434499064,
+      "learning_rate": 1.909303163127873e-05,
+      "loss": 0.9245,
+      "step": 1090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4777626648740374,
+      "learning_rate": 1.9091019300879848e-05,
+      "loss": 0.9625,
+      "step": 1091
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.553647698794413,
+      "learning_rate": 1.9089004846817947e-05,
+      "loss": 0.9355,
+      "step": 1092
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3314385018182198,
+      "learning_rate": 1.908698826956361e-05,
+      "loss": 0.9244,
+      "step": 1093
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.546789986990871,
+      "learning_rate": 1.9084969569587908e-05,
+      "loss": 0.9455,
+      "step": 1094
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6272980769666485,
+      "learning_rate": 1.9082948747362412e-05,
+      "loss": 0.8533,
+      "step": 1095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.646582490796409,
+      "learning_rate": 1.908092580335918e-05,
+      "loss": 1.0128,
+      "step": 1096
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6106610836389794,
+      "learning_rate": 1.9078900738050776e-05,
+      "loss": 0.9394,
+      "step": 1097
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.850712706921773,
+      "learning_rate": 1.9076873551910256e-05,
+      "loss": 0.9306,
+      "step": 1098
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.629630802435573,
+      "learning_rate": 1.907484424541117e-05,
+      "loss": 0.8473,
+      "step": 1099
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5917835047809785,
+      "learning_rate": 1.9072812819027566e-05,
+      "loss": 0.8673,
+      "step": 1100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6063218553405059,
+      "learning_rate": 1.907077927323398e-05,
+      "loss": 0.9599,
+      "step": 1101
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0558726669264535,
+      "learning_rate": 1.9068743608505454e-05,
+      "loss": 1.0059,
+      "step": 1102
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8377801434250658,
+      "learning_rate": 1.9066705825317518e-05,
+      "loss": 0.9009,
+      "step": 1103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6201705153542958,
+      "learning_rate": 1.90646659241462e-05,
+      "loss": 0.8804,
+      "step": 1104
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5761230040469598,
+      "learning_rate": 1.9062623905468015e-05,
+      "loss": 0.8529,
+      "step": 1105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5993850233949765,
+      "learning_rate": 1.9060579769759986e-05,
+      "loss": 0.9292,
+      "step": 1106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5057654530863425,
+      "learning_rate": 1.905853351749962e-05,
+      "loss": 0.862,
+      "step": 1107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7542682308151063,
+      "learning_rate": 1.905648514916492e-05,
+      "loss": 0.8259,
+      "step": 1108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8702799615906952,
+      "learning_rate": 1.905443466523439e-05,
+      "loss": 0.3283,
+      "step": 1109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4645305788331042,
+      "learning_rate": 1.9052382066187017e-05,
+      "loss": 0.9252,
+      "step": 1110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5128347426173903,
+      "learning_rate": 1.9050327352502292e-05,
+      "loss": 0.9566,
+      "step": 1111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5390629762319612,
+      "learning_rate": 1.9048270524660197e-05,
+      "loss": 0.9492,
+      "step": 1112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4759078605606883,
+      "learning_rate": 1.9046211583141206e-05,
+      "loss": 0.9014,
+      "step": 1113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.647019446415463,
+      "learning_rate": 1.9044150528426288e-05,
+      "loss": 0.8662,
+      "step": 1114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8630314364105205,
+      "learning_rate": 1.9042087360996904e-05,
+      "loss": 0.3188,
+      "step": 1115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5735208914477208,
+      "learning_rate": 1.904002208133501e-05,
+      "loss": 0.9512,
+      "step": 1116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5933487202315462,
+      "learning_rate": 1.903795468992306e-05,
+      "loss": 1.0019,
+      "step": 1117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6957806433073173,
+      "learning_rate": 1.9035885187243996e-05,
+      "loss": 0.9519,
+      "step": 1118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.751716539320122,
+      "learning_rate": 1.903381357378125e-05,
+      "loss": 0.9308,
+      "step": 1119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5846124102495645,
+      "learning_rate": 1.9031739850018755e-05,
+      "loss": 0.8757,
+      "step": 1120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6299300138382484,
+      "learning_rate": 1.902966401644093e-05,
+      "loss": 0.9869,
+      "step": 1121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.75999031791679,
+      "learning_rate": 1.902758607353269e-05,
+      "loss": 0.8777,
+      "step": 1122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5477360968137088,
+      "learning_rate": 1.9025506021779446e-05,
+      "loss": 0.9461,
+      "step": 1123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5465128298842241,
+      "learning_rate": 1.9023423861667093e-05,
+      "loss": 0.9638,
+      "step": 1124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.458325709737424,
+      "learning_rate": 1.902133959368203e-05,
+      "loss": 0.8923,
+      "step": 1125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6242994588879356,
+      "learning_rate": 1.901925321831114e-05,
+      "loss": 0.9359,
+      "step": 1126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5322664977753728,
+      "learning_rate": 1.9017164736041795e-05,
+      "loss": 0.9671,
+      "step": 1127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4499265652951752,
+      "learning_rate": 1.9015074147361875e-05,
+      "loss": 0.9617,
+      "step": 1128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.629563184268512,
+      "learning_rate": 1.901298145275973e-05,
+      "loss": 0.869,
+      "step": 1129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6874952305178883,
+      "learning_rate": 1.9010886652724226e-05,
+      "loss": 0.9653,
+      "step": 1130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6908336566611315,
+      "learning_rate": 1.9008789747744697e-05,
+      "loss": 0.9849,
+      "step": 1131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4852033800777031,
+      "learning_rate": 1.9006690738310988e-05,
+      "loss": 0.9168,
+      "step": 1132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.319541720274812,
+      "learning_rate": 1.9004589624913428e-05,
+      "loss": 0.9015,
+      "step": 1133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6475923414574336,
+      "learning_rate": 1.900248640804283e-05,
+      "loss": 0.8672,
+      "step": 1134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5780109993436227,
+      "learning_rate": 1.9000381088190512e-05,
+      "loss": 0.9191,
+      "step": 1135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9846952079335375,
+      "learning_rate": 1.8998273665848273e-05,
+      "loss": 0.3237,
+      "step": 1136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3882339815916798,
+      "learning_rate": 1.8996164141508412e-05,
+      "loss": 0.9764,
+      "step": 1137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.534498214435999,
+      "learning_rate": 1.899405251566371e-05,
+      "loss": 0.9173,
+      "step": 1138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4067702006750886,
+      "learning_rate": 1.8991938788807446e-05,
+      "loss": 0.8883,
+      "step": 1139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4152944938468568,
+      "learning_rate": 1.8989822961433386e-05,
+      "loss": 0.9221,
+      "step": 1140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.604223557388208,
+      "learning_rate": 1.8987705034035784e-05,
+      "loss": 0.9298,
+      "step": 1141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8280467427137816,
+      "learning_rate": 1.898558500710939e-05,
+      "loss": 0.308,
+      "step": 1142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.424759975225943,
+      "learning_rate": 1.8983462881149447e-05,
+      "loss": 0.9473,
+      "step": 1143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8005634391281962,
+      "learning_rate": 1.898133865665168e-05,
+      "loss": 0.978,
+      "step": 1144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8356695202795086,
+      "learning_rate": 1.89792123341123e-05,
+      "loss": 0.9625,
+      "step": 1145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4477431954667923,
+      "learning_rate": 1.8977083914028034e-05,
+      "loss": 0.9643,
+      "step": 1146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4744031225049494,
+      "learning_rate": 1.8974953396896066e-05,
+      "loss": 0.8575,
+      "step": 1147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9251500327455193,
+      "learning_rate": 1.897282078321409e-05,
+      "loss": 0.9276,
+      "step": 1148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0091543750979954,
+      "learning_rate": 1.8970686073480287e-05,
+      "loss": 0.9273,
+      "step": 1149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.711553824038662,
+      "learning_rate": 1.8968549268193325e-05,
+      "loss": 0.9475,
+      "step": 1150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.593556430926505,
+      "learning_rate": 1.896641036785236e-05,
+      "loss": 0.9742,
+      "step": 1151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6311827847181575,
+      "learning_rate": 1.896426937295704e-05,
+      "loss": 0.8693,
+      "step": 1152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5218338485437497,
+      "learning_rate": 1.89621262840075e-05,
+      "loss": 0.979,
+      "step": 1153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7871557978104736,
+      "learning_rate": 1.895998110150437e-05,
+      "loss": 0.3205,
+      "step": 1154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5994081568588547,
+      "learning_rate": 1.8957833825948756e-05,
+      "loss": 0.9642,
+      "step": 1155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5861992995747178,
+      "learning_rate": 1.8955684457842275e-05,
+      "loss": 0.9777,
+      "step": 1156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3731965923484906,
+      "learning_rate": 1.8953532997687008e-05,
+      "loss": 0.9178,
+      "step": 1157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.573120856031462,
+      "learning_rate": 1.895137944598554e-05,
+      "loss": 0.8593,
+      "step": 1158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7175529012027326,
+      "learning_rate": 1.8949223803240945e-05,
+      "loss": 0.9878,
+      "step": 1159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4005092480162458,
+      "learning_rate": 1.8947066069956777e-05,
+      "loss": 0.9378,
+      "step": 1160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5569368233785106,
+      "learning_rate": 1.8944906246637084e-05,
+      "loss": 0.9185,
+      "step": 1161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6432224375309423,
+      "learning_rate": 1.89427443337864e-05,
+      "loss": 0.9815,
+      "step": 1162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7008021551835373,
+      "learning_rate": 1.8940580331909747e-05,
+      "loss": 0.9366,
+      "step": 1163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5144996188733308,
+      "learning_rate": 1.893841424151264e-05,
+      "loss": 0.8872,
+      "step": 1164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4102849274200462,
+      "learning_rate": 1.8936246063101077e-05,
+      "loss": 0.9337,
+      "step": 1165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4799118869337513,
+      "learning_rate": 1.893407579718154e-05,
+      "loss": 0.9644,
+      "step": 1166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4522204519140673,
+      "learning_rate": 1.8931903444261007e-05,
+      "loss": 0.9241,
+      "step": 1167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.853809030195236,
+      "learning_rate": 1.892972900484694e-05,
+      "loss": 0.3487,
+      "step": 1168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.747228098226903,
+      "learning_rate": 1.892755247944729e-05,
+      "loss": 0.9473,
+      "step": 1169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8855540718253962,
+      "learning_rate": 1.892537386857049e-05,
+      "loss": 0.2987,
+      "step": 1170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5213810875862679,
+      "learning_rate": 1.8923193172725467e-05,
+      "loss": 0.9122,
+      "step": 1171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5748516388964353,
+      "learning_rate": 1.8921010392421628e-05,
+      "loss": 0.9491,
+      "step": 1172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6928939630742308,
+      "learning_rate": 1.8918825528168872e-05,
+      "loss": 0.9216,
+      "step": 1173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5499795107198675,
+      "learning_rate": 1.8916638580477586e-05,
+      "loss": 0.937,
+      "step": 1174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6389989492389447,
+      "learning_rate": 1.891444954985864e-05,
+      "loss": 1.0057,
+      "step": 1175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5387430525153407,
+      "learning_rate": 1.891225843682339e-05,
+      "loss": 0.9499,
+      "step": 1176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6905031491277491,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 0.9389,
+      "step": 1177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3827014638750614,
+      "learning_rate": 1.890786996555184e-05,
+      "loss": 0.8842,
+      "step": 1178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9882403410484893,
+      "learning_rate": 1.8905672608340693e-05,
+      "loss": 0.3373,
+      "step": 1179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.936884862689874,
+      "learning_rate": 1.890347317076354e-05,
+      "loss": 0.8439,
+      "step": 1180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5826834988298497,
+      "learning_rate": 1.890127165333416e-05,
+      "loss": 0.9413,
+      "step": 1181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.716550978669316,
+      "learning_rate": 1.889906805656684e-05,
+      "loss": 0.9028,
+      "step": 1182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.65347920176481,
+      "learning_rate": 1.8896862380976332e-05,
+      "loss": 0.9535,
+      "step": 1183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3065422708988446,
+      "learning_rate": 1.889465462707789e-05,
+      "loss": 0.9364,
+      "step": 1184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5315127743317618,
+      "learning_rate": 1.8892444795387237e-05,
+      "loss": 1.0257,
+      "step": 1185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6037148550513256,
+      "learning_rate": 1.8890232886420598e-05,
+      "loss": 0.9283,
+      "step": 1186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8328336039368817,
+      "learning_rate": 1.888801890069467e-05,
+      "loss": 0.3108,
+      "step": 1187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.551529222916193,
+      "learning_rate": 1.8885802838726634e-05,
+      "loss": 0.957,
+      "step": 1188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6219782082568708,
+      "learning_rate": 1.8883584701034177e-05,
+      "loss": 0.9161,
+      "step": 1189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4742246808113189,
+      "learning_rate": 1.8881364488135448e-05,
+      "loss": 0.9947,
+      "step": 1190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5864959917297368,
+      "learning_rate": 1.887914220054909e-05,
+      "loss": 0.8545,
+      "step": 1191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6162266422211726,
+      "learning_rate": 1.8876917838794226e-05,
+      "loss": 0.8937,
+      "step": 1192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5502100079784191,
+      "learning_rate": 1.8874691403390476e-05,
+      "loss": 0.9681,
+      "step": 1193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6388995327313163,
+      "learning_rate": 1.8872462894857927e-05,
+      "loss": 0.8787,
+      "step": 1194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6493475252940657,
+      "learning_rate": 1.887023231371716e-05,
+      "loss": 0.9474,
+      "step": 1195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5760642833929563,
+      "learning_rate": 1.886799966048924e-05,
+      "loss": 0.9447,
+      "step": 1196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.807192015145044,
+      "learning_rate": 1.886576493569572e-05,
+      "loss": 0.9359,
+      "step": 1197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4519760309115668,
+      "learning_rate": 1.8863528139858628e-05,
+      "loss": 0.9556,
+      "step": 1198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4077583230853012,
+      "learning_rate": 1.886128927350048e-05,
+      "loss": 0.953,
+      "step": 1199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.52294725819737,
+      "learning_rate": 1.885904833714427e-05,
+      "loss": 0.8942,
+      "step": 1200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4446690602737893,
+      "learning_rate": 1.8856805331313487e-05,
+      "loss": 0.8931,
+      "step": 1201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4177219935426382,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.8838,
+      "step": 1202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.544077419393447,
+      "learning_rate": 1.8852313113324553e-05,
+      "loss": 0.9048,
+      "step": 1203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5659039029571216,
+      "learning_rate": 1.885006390221578e-05,
+      "loss": 0.909,
+      "step": 1204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4326474232723647,
+      "learning_rate": 1.8847812623731202e-05,
+      "loss": 0.9173,
+      "step": 1205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.505348227046058,
+      "learning_rate": 1.8845559278396707e-05,
+      "loss": 0.8552,
+      "step": 1206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8805963484830883,
+      "learning_rate": 1.884330386673869e-05,
+      "loss": 0.8872,
+      "step": 1207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6435418743284602,
+      "learning_rate": 1.8841046389284004e-05,
+      "loss": 0.9942,
+      "step": 1208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5910597996206866,
+      "learning_rate": 1.8838786846560003e-05,
+      "loss": 0.9474,
+      "step": 1209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6840176723500377,
+      "learning_rate": 1.883652523909451e-05,
+      "loss": 0.9131,
+      "step": 1210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8808549938368748,
+      "learning_rate": 1.883426156741585e-05,
+      "loss": 0.96,
+      "step": 1211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5601920463302617,
+      "learning_rate": 1.8831995832052802e-05,
+      "loss": 0.942,
+      "step": 1212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.673192219474004,
+      "learning_rate": 1.8829728033534644e-05,
+      "loss": 0.941,
+      "step": 1213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6446156400833298,
+      "learning_rate": 1.882745817239114e-05,
+      "loss": 0.9321,
+      "step": 1214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6079547527209246,
+      "learning_rate": 1.882518624915253e-05,
+      "loss": 0.9555,
+      "step": 1215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5859759724230382,
+      "learning_rate": 1.8822912264349535e-05,
+      "loss": 0.9227,
+      "step": 1216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6065715125698727,
+      "learning_rate": 1.8820636218513354e-05,
+      "loss": 0.9624,
+      "step": 1217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5036676341415365,
+      "learning_rate": 1.881835811217567e-05,
+      "loss": 0.8987,
+      "step": 1218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.723636847104978,
+      "learning_rate": 1.8816077945868656e-05,
+      "loss": 0.9751,
+      "step": 1219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4359919454953791,
+      "learning_rate": 1.8813795720124958e-05,
+      "loss": 1.0122,
+      "step": 1220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4045787191543881,
+      "learning_rate": 1.8811511435477695e-05,
+      "loss": 0.8864,
+      "step": 1221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5753337937586187,
+      "learning_rate": 1.8809225092460488e-05,
+      "loss": 0.9649,
+      "step": 1222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4494026059745404,
+      "learning_rate": 1.880693669160742e-05,
+      "loss": 0.9843,
+      "step": 1223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7448198121680767,
+      "learning_rate": 1.8804646233453067e-05,
+      "loss": 0.917,
+      "step": 1224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5492769850059651,
+      "learning_rate": 1.8802353718532472e-05,
+      "loss": 0.902,
+      "step": 1225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7016026852506534,
+      "learning_rate": 1.8800059147381172e-05,
+      "loss": 0.9395,
+      "step": 1226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6215535350617172,
+      "learning_rate": 1.8797762520535178e-05,
+      "loss": 0.9146,
+      "step": 1227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6084653427022122,
+      "learning_rate": 1.8795463838530988e-05,
+      "loss": 0.8814,
+      "step": 1228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6905071988948548,
+      "learning_rate": 1.8793163101905562e-05,
+      "loss": 0.8176,
+      "step": 1229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6941768912386646,
+      "learning_rate": 1.8790860311196365e-05,
+      "loss": 0.9573,
+      "step": 1230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8326464679439716,
+      "learning_rate": 1.878855546694132e-05,
+      "loss": 0.933,
+      "step": 1231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6187182912151123,
+      "learning_rate": 1.8786248569678847e-05,
+      "loss": 0.9439,
+      "step": 1232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8292645668436696,
+      "learning_rate": 1.8783939619947827e-05,
+      "loss": 0.8819,
+      "step": 1233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4424933382792418,
+      "learning_rate": 1.878162861828764e-05,
+      "loss": 0.9702,
+      "step": 1234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.629760378746899,
+      "learning_rate": 1.8779315565238133e-05,
+      "loss": 0.9306,
+      "step": 1235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7883182293496125,
+      "learning_rate": 1.8777000461339635e-05,
+      "loss": 0.9406,
+      "step": 1236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8801254801532837,
+      "learning_rate": 1.8774683307132956e-05,
+      "loss": 0.3198,
+      "step": 1237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5185368527656098,
+      "learning_rate": 1.877236410315938e-05,
+      "loss": 0.948,
+      "step": 1238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4700319581192989,
+      "learning_rate": 1.8770042849960676e-05,
+      "loss": 0.9333,
+      "step": 1239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6535375410700721,
+      "learning_rate": 1.8767719548079088e-05,
+      "loss": 0.9977,
+      "step": 1240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4919605419559654,
+      "learning_rate": 1.8765394198057342e-05,
+      "loss": 0.8905,
+      "step": 1241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5638896984825594,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 1.023,
+      "step": 1242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.643108876942023,
+      "learning_rate": 1.8760737355766654e-05,
+      "loss": 0.9504,
+      "step": 1243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5382238193047761,
+      "learning_rate": 1.8758405864585554e-05,
+      "loss": 0.9877,
+      "step": 1244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6923843660033813,
+      "learning_rate": 1.875607232743997e-05,
+      "loss": 0.9433,
+      "step": 1245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4801961313132213,
+      "learning_rate": 1.875373674487502e-05,
+      "loss": 0.8968,
+      "step": 1246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3989568237535377,
+      "learning_rate": 1.8751399117436292e-05,
+      "loss": 0.9798,
+      "step": 1247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6318968276986034,
+      "learning_rate": 1.8749059445669856e-05,
+      "loss": 0.8227,
+      "step": 1248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5987330878890313,
+      "learning_rate": 1.8746717730122266e-05,
+      "loss": 0.8551,
+      "step": 1249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4963770665364888,
+      "learning_rate": 1.874437397134054e-05,
+      "loss": 0.9298,
+      "step": 1250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6008725261776349,
+      "learning_rate": 1.8742028169872188e-05,
+      "loss": 0.8638,
+      "step": 1251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9130527674558184,
+      "learning_rate": 1.873968032626518e-05,
+      "loss": 0.9344,
+      "step": 1252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7082103393248815,
+      "learning_rate": 1.873733044106798e-05,
+      "loss": 0.8369,
+      "step": 1253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.505975577772877,
+      "learning_rate": 1.8734978514829518e-05,
+      "loss": 0.9899,
+      "step": 1254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3955283226871127,
+      "learning_rate": 1.8732624548099204e-05,
+      "loss": 0.9152,
+      "step": 1255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4906288739719873,
+      "learning_rate": 1.8730268541426924e-05,
+      "loss": 0.8999,
+      "step": 1256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8773049411359727,
+      "learning_rate": 1.8727910495363043e-05,
+      "loss": 0.3199,
+      "step": 1257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6588613464064355,
+      "learning_rate": 1.8725550410458403e-05,
+      "loss": 0.9285,
+      "step": 1258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3533282989792614,
+      "learning_rate": 1.872318828726432e-05,
+      "loss": 0.9319,
+      "step": 1259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.84850224972434,
+      "learning_rate": 1.8720824126332583e-05,
+      "loss": 0.9614,
+      "step": 1260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6700730043477208,
+      "learning_rate": 1.8718457928215458e-05,
+      "loss": 0.9567,
+      "step": 1261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.642011708088758,
+      "learning_rate": 1.8716089693465696e-05,
+      "loss": 0.9584,
+      "step": 1262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.708630877310769,
+      "learning_rate": 1.871371942263651e-05,
+      "loss": 0.8787,
+      "step": 1263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.491084515022238,
+      "learning_rate": 1.8711347116281598e-05,
+      "loss": 0.9189,
+      "step": 1264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.678290646274311,
+      "learning_rate": 1.8708972774955136e-05,
+      "loss": 0.9222,
+      "step": 1265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4672324007677358,
+      "learning_rate": 1.8706596399211765e-05,
+      "loss": 0.863,
+      "step": 1266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5234800650690186,
+      "learning_rate": 1.8704217989606606e-05,
+      "loss": 0.9324,
+      "step": 1267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5853127502177646,
+      "learning_rate": 1.870183754669526e-05,
+      "loss": 0.9476,
+      "step": 1268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.504131171511284,
+      "learning_rate": 1.8699455071033795e-05,
+      "loss": 0.9881,
+      "step": 1269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.203410522410484,
+      "learning_rate": 1.8697070563178758e-05,
+      "loss": 0.9066,
+      "step": 1270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5035306343812098,
+      "learning_rate": 1.869468402368717e-05,
+      "loss": 0.95,
+      "step": 1271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5599186836032735,
+      "learning_rate": 1.869229545311653e-05,
+      "loss": 1.0106,
+      "step": 1272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.835861925842248,
+      "learning_rate": 1.8689904852024803e-05,
+      "loss": 1.0088,
+      "step": 1273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4560103873428691,
+      "learning_rate": 1.868751222097044e-05,
+      "loss": 0.9692,
+      "step": 1274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3895527095621782,
+      "learning_rate": 1.868511756051236e-05,
+      "loss": 0.8148,
+      "step": 1275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8882937550870484,
+      "learning_rate": 1.868272087120995e-05,
+      "loss": 0.8738,
+      "step": 1276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3671512469526015,
+      "learning_rate": 1.8680322153623077e-05,
+      "loss": 0.8906,
+      "step": 1277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5260697060885258,
+      "learning_rate": 1.8677921408312083e-05,
+      "loss": 0.8557,
+      "step": 1278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7329632279907663,
+      "learning_rate": 1.8675518635837788e-05,
+      "loss": 0.9841,
+      "step": 1279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5317587579021834,
+      "learning_rate": 1.8673113836761475e-05,
+      "loss": 0.9284,
+      "step": 1280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5342315444098116,
+      "learning_rate": 1.86707070116449e-05,
+      "loss": 0.9132,
+      "step": 1281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6717295817387856,
+      "learning_rate": 1.8668298161050308e-05,
+      "loss": 0.8795,
+      "step": 1282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5656831730049623,
+      "learning_rate": 1.8665887285540405e-05,
+      "loss": 1.0018,
+      "step": 1283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6604674743194432,
+      "learning_rate": 1.8663474385678364e-05,
+      "loss": 0.8799,
+      "step": 1284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.646276807088851,
+      "learning_rate": 1.8661059462027844e-05,
+      "loss": 0.8531,
+      "step": 1285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4665391825162069,
+      "learning_rate": 1.8658642515152973e-05,
+      "loss": 0.9083,
+      "step": 1286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4433325441109526,
+      "learning_rate": 1.8656223545618345e-05,
+      "loss": 0.9448,
+      "step": 1287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5484993310241129,
+      "learning_rate": 1.8653802553989037e-05,
+      "loss": 0.9323,
+      "step": 1288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3402972566022855,
+      "learning_rate": 1.865137954083059e-05,
+      "loss": 0.8927,
+      "step": 1289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7226551730613198,
+      "learning_rate": 1.864895450670902e-05,
+      "loss": 0.9696,
+      "step": 1290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.428140063383765,
+      "learning_rate": 1.8646527452190815e-05,
+      "loss": 0.8815,
+      "step": 1291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6296343514031562,
+      "learning_rate": 1.8644098377842934e-05,
+      "loss": 0.9921,
+      "step": 1292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5240842326767499,
+      "learning_rate": 1.8641667284232813e-05,
+      "loss": 0.9861,
+      "step": 1293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5892538011792583,
+      "learning_rate": 1.8639234171928355e-05,
+      "loss": 0.9365,
+      "step": 1294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6202418582308802,
+      "learning_rate": 1.863679904149793e-05,
+      "loss": 0.9119,
+      "step": 1295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6208106241825433,
+      "learning_rate": 1.8634361893510393e-05,
+      "loss": 0.8552,
+      "step": 1296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6827653785313705,
+      "learning_rate": 1.8631922728535054e-05,
+      "loss": 0.9716,
+      "step": 1297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5840536530440106,
+      "learning_rate": 1.8629481547141708e-05,
+      "loss": 0.8995,
+      "step": 1298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6552926791037665,
+      "learning_rate": 1.862703834990061e-05,
+      "loss": 0.9875,
+      "step": 1299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7296687832433923,
+      "learning_rate": 1.8624593137382495e-05,
+      "loss": 0.9829,
+      "step": 1300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9887302056486602,
+      "learning_rate": 1.8622145910158568e-05,
+      "loss": 0.3243,
+      "step": 1301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5895416783536622,
+      "learning_rate": 1.8619696668800494e-05,
+      "loss": 0.9121,
+      "step": 1302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.727160176659264,
+      "learning_rate": 1.861724541388042e-05,
+      "loss": 0.9216,
+      "step": 1303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3950993785765546,
+      "learning_rate": 1.8614792145970958e-05,
+      "loss": 0.8511,
+      "step": 1304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.833936936452527,
+      "learning_rate": 1.861233686564519e-05,
+      "loss": 0.9771,
+      "step": 1305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.464209391884905,
+      "learning_rate": 1.860987957347668e-05,
+      "loss": 0.9382,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5161525371094917,
+      "learning_rate": 1.860742027003944e-05,
+      "loss": 0.9019,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.603007359166619,
+      "learning_rate": 1.8604958955907964e-05,
+      "loss": 0.9364,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.272688820219678,
+      "learning_rate": 1.8602495631657222e-05,
+      "loss": 0.8541,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5642350446073157,
+      "learning_rate": 1.8600030297862643e-05,
+      "loss": 0.987,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5720007495605721,
+      "learning_rate": 1.859756295510013e-05,
+      "loss": 1.0131,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3758413622015309,
+      "learning_rate": 1.8595093603946053e-05,
+      "loss": 0.9656,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5162712819548991,
+      "learning_rate": 1.8592622244977255e-05,
+      "loss": 0.912,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5771425028591872,
+      "learning_rate": 1.8590148878771043e-05,
+      "loss": 0.9851,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5543978543493784,
+      "learning_rate": 1.8587673505905198e-05,
+      "loss": 0.976,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.477787802433991,
+      "learning_rate": 1.858519612695797e-05,
+      "loss": 0.8555,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7477959223642614,
+      "learning_rate": 1.8582716742508066e-05,
+      "loss": 0.9811,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6315112141858157,
+      "learning_rate": 1.8580235353134678e-05,
+      "loss": 0.9515,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.499480271700626,
+      "learning_rate": 1.8577751959417458e-05,
+      "loss": 0.9578,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5070908875280902,
+      "learning_rate": 1.8575266561936526e-05,
+      "loss": 0.9274,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.652018021206146,
+      "learning_rate": 1.857277916127247e-05,
+      "loss": 0.9642,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6129189943556552,
+      "learning_rate": 1.8570289758006346e-05,
+      "loss": 0.9857,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.531419336555744,
+      "learning_rate": 1.856779835271968e-05,
+      "loss": 0.9487,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.543964428092662,
+      "learning_rate": 1.856530494599447e-05,
+      "loss": 0.9843,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5111762121811114,
+      "learning_rate": 1.856280953841317e-05,
+      "loss": 0.9859,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4074756380707643,
+      "learning_rate": 1.8560312130558706e-05,
+      "loss": 0.9449,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.520628428223058,
+      "learning_rate": 1.8557812723014476e-05,
+      "loss": 0.889,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4257094365154306,
+      "learning_rate": 1.8555311316364344e-05,
+      "loss": 0.892,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6365941433827504,
+      "learning_rate": 1.8552807911192636e-05,
+      "loss": 0.8999,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5991657175142358,
+      "learning_rate": 1.8550302508084145e-05,
+      "loss": 0.8945,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3884756966401848,
+      "learning_rate": 1.8547795107624137e-05,
+      "loss": 0.9064,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3479605524683398,
+      "learning_rate": 1.8545285710398343e-05,
+      "loss": 0.8934,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4691756727343244,
+      "learning_rate": 1.8542774316992953e-05,
+      "loss": 0.937,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4131267847512172,
+      "learning_rate": 1.8540260927994633e-05,
+      "loss": 0.8723,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4669830614267512,
+      "learning_rate": 1.853774554399051e-05,
+      "loss": 0.9234,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4830741377153693,
+      "learning_rate": 1.8535228165568177e-05,
+      "loss": 0.9726,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6042513861104033,
+      "learning_rate": 1.853270879331569e-05,
+      "loss": 0.9202,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5331887627190128,
+      "learning_rate": 1.8530187427821585e-05,
+      "loss": 0.9193,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0493620964174841,
+      "learning_rate": 1.8527664069674844e-05,
+      "loss": 0.3261,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5734458237459092,
+      "learning_rate": 1.8525138719464927e-05,
+      "loss": 0.8838,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5167431946381404,
+      "learning_rate": 1.8522611377781758e-05,
+      "loss": 0.8644,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6394667431036494,
+      "learning_rate": 1.852008204521572e-05,
+      "loss": 0.9634,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6506746623286161,
+      "learning_rate": 1.8517550722357667e-05,
+      "loss": 0.9378,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5567760887218947,
+      "learning_rate": 1.8515017409798915e-05,
+      "loss": 0.9221,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4543054935979685,
+      "learning_rate": 1.8512482108131254e-05,
+      "loss": 0.9022,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1048299823681966,
+      "learning_rate": 1.850994481794692e-05,
+      "loss": 0.9204,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5253089487284166,
+      "learning_rate": 1.850740553983863e-05,
+      "loss": 0.8562,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6580663092147898,
+      "learning_rate": 1.8504864274399557e-05,
+      "loss": 0.8681,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6978845861442777,
+      "learning_rate": 1.8502321022223344e-05,
+      "loss": 0.9746,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3833270476083062,
+      "learning_rate": 1.8499775783904094e-05,
+      "loss": 0.9414,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5016279513457007,
+      "learning_rate": 1.849722856003637e-05,
+      "loss": 0.9382,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6086146088387632,
+      "learning_rate": 1.8494679351215212e-05,
+      "loss": 0.9381,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3333134192572778,
+      "learning_rate": 1.8492128158036113e-05,
+      "loss": 0.9227,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6278614849135744,
+      "learning_rate": 1.8489574981095026e-05,
+      "loss": 1.0032,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4688545935501023,
+      "learning_rate": 1.8487019820988378e-05,
+      "loss": 0.8512,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5330555705119522,
+      "learning_rate": 1.8484462678313053e-05,
+      "loss": 0.9969,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.348580242682029,
+      "learning_rate": 1.8481903553666405e-05,
+      "loss": 0.9258,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4913611220880396,
+      "learning_rate": 1.847934244764624e-05,
+      "loss": 0.9432,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5106434669616648,
+      "learning_rate": 1.8476779360850833e-05,
+      "loss": 0.9214,
+      "step": 1359
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.676985565785228,
+      "learning_rate": 1.8474214293878926e-05,
+      "loss": 0.9706,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5161578382250267,
+      "learning_rate": 1.8471647247329714e-05,
+      "loss": 0.9006,
+      "step": 1361
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8451989408657201,
+      "learning_rate": 1.846907822180286e-05,
+      "loss": 0.8809,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2645698139980488,
+      "learning_rate": 1.8466507217898493e-05,
+      "loss": 0.9171,
+      "step": 1363
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5400524767749244,
+      "learning_rate": 1.8463934236217195e-05,
+      "loss": 0.9166,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6959936733583316,
+      "learning_rate": 1.8461359277360014e-05,
+      "loss": 0.955,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5706814669330444,
+      "learning_rate": 1.8458782341928465e-05,
+      "loss": 0.9506,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3933633781270796,
+      "learning_rate": 1.845620343052452e-05,
+      "loss": 0.8393,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5434039962717498,
+      "learning_rate": 1.8453622543750608e-05,
+      "loss": 1.0014,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4432493853182402,
+      "learning_rate": 1.8451039682209626e-05,
+      "loss": 0.9955,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7283538149943536,
+      "learning_rate": 1.8448454846504936e-05,
+      "loss": 0.9312,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6093431598096921,
+      "learning_rate": 1.844586803724035e-05,
+      "loss": 0.823,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.420925494064239,
+      "learning_rate": 1.8443279255020153e-05,
+      "loss": 0.9405,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4719271100644025,
+      "learning_rate": 1.8440688500449076e-05,
+      "loss": 1.0121,
+      "step": 1373
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5046113355419708,
+      "learning_rate": 1.8438095774132327e-05,
+      "loss": 0.8513,
+      "step": 1374
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.58328749970449,
+      "learning_rate": 1.8435501076675566e-05,
+      "loss": 0.9453,
+      "step": 1375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5273273984224014,
+      "learning_rate": 1.8432904408684912e-05,
+      "loss": 0.9004,
+      "step": 1376
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6405097693893897,
+      "learning_rate": 1.8430305770766947e-05,
+      "loss": 0.8508,
+      "step": 1377
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.431987793216744,
+      "learning_rate": 1.8427705163528716e-05,
+      "loss": 0.9321,
+      "step": 1378
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4307210572382882,
+      "learning_rate": 1.8425102587577716e-05,
+      "loss": 0.9552,
+      "step": 1379
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4030733256180892,
+      "learning_rate": 1.8422498043521915e-05,
+      "loss": 0.8972,
+      "step": 1380
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5824955859280896,
+      "learning_rate": 1.841989153196973e-05,
+      "loss": 0.8686,
+      "step": 1381
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5346678546610544,
+      "learning_rate": 1.8417283053530047e-05,
+      "loss": 0.9393,
+      "step": 1382
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6932440897803103,
+      "learning_rate": 1.84146726088122e-05,
+      "loss": 1.0061,
+      "step": 1383
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.730670733105885,
+      "learning_rate": 1.8412060198426e-05,
+      "loss": 0.9637,
+      "step": 1384
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5837054389199716,
+      "learning_rate": 1.8409445822981694e-05,
+      "loss": 0.9567,
+      "step": 1385
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4366452640659975,
+      "learning_rate": 1.8406829483090006e-05,
+      "loss": 0.9026,
+      "step": 1386
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.445942618934189,
+      "learning_rate": 1.8404211179362116e-05,
+      "loss": 0.8645,
+      "step": 1387
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.470152848227919,
+      "learning_rate": 1.8401590912409653e-05,
+      "loss": 0.8691,
+      "step": 1388
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8476267080384925,
+      "learning_rate": 1.839896868284472e-05,
+      "loss": 0.7976,
+      "step": 1389
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3481521481882888,
+      "learning_rate": 1.8396344491279864e-05,
+      "loss": 0.9588,
+      "step": 1390
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6982467282779086,
+      "learning_rate": 1.8393718338328102e-05,
+      "loss": 0.9483,
+      "step": 1391
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.375222365594476,
+      "learning_rate": 1.8391090224602895e-05,
+      "loss": 0.8921,
+      "step": 1392
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4177302969215262,
+      "learning_rate": 1.838846015071818e-05,
+      "loss": 0.9851,
+      "step": 1393
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1595223674721813,
+      "learning_rate": 1.8385828117288333e-05,
+      "loss": 0.3101,
+      "step": 1394
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4966780701730207,
+      "learning_rate": 1.838319412492821e-05,
+      "loss": 0.8755,
+      "step": 1395
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4258036193660553,
+      "learning_rate": 1.8380558174253095e-05,
+      "loss": 0.9083,
+      "step": 1396
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5377659722200108,
+      "learning_rate": 1.837792026587876e-05,
+      "loss": 0.8627,
+      "step": 1397
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.697825250670764,
+      "learning_rate": 1.837528040042142e-05,
+      "loss": 0.9405,
+      "step": 1398
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6569503302197386,
+      "learning_rate": 1.8372638578497738e-05,
+      "loss": 0.8927,
+      "step": 1399
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4769447528205317,
+      "learning_rate": 1.8369994800724855e-05,
+      "loss": 0.8755,
+      "step": 1400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5028198194593256,
+      "learning_rate": 1.836734906772035e-05,
+      "loss": 0.9232,
+      "step": 1401
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7897250123186603,
+      "learning_rate": 1.8364701380102267e-05,
+      "loss": 0.8772,
+      "step": 1402
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7545370866299816,
+      "learning_rate": 1.836205173848911e-05,
+      "loss": 0.797,
+      "step": 1403
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3680504500535342,
+      "learning_rate": 1.835940014349984e-05,
+      "loss": 0.9338,
+      "step": 1404
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.627172584661305,
+      "learning_rate": 1.8356746595753856e-05,
+      "loss": 0.9072,
+      "step": 1405
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6118572545415228,
+      "learning_rate": 1.8354091095871038e-05,
+      "loss": 0.9417,
+      "step": 1406
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5496130790598868,
+      "learning_rate": 1.8351433644471708e-05,
+      "loss": 0.8422,
+      "step": 1407
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6138938151875928,
+      "learning_rate": 1.8348774242176642e-05,
+      "loss": 0.8675,
+      "step": 1408
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0340059348763608,
+      "learning_rate": 1.8346112889607086e-05,
+      "loss": 0.3633,
+      "step": 1409
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3929871898635133,
+      "learning_rate": 1.8343449587384727e-05,
+      "loss": 0.8767,
+      "step": 1410
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8613657702374764,
+      "learning_rate": 1.8340784336131715e-05,
+      "loss": 0.9883,
+      "step": 1411
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4416846972977517,
+      "learning_rate": 1.8338117136470648e-05,
+      "loss": 0.9644,
+      "step": 1412
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4801650062828753,
+      "learning_rate": 1.833544798902459e-05,
+      "loss": 0.9389,
+      "step": 1413
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4552923172875576,
+      "learning_rate": 1.8332776894417047e-05,
+      "loss": 0.8727,
+      "step": 1414
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5582812457274473,
+      "learning_rate": 1.8330103853271993e-05,
+      "loss": 0.9203,
+      "step": 1415
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.602036635066962,
+      "learning_rate": 1.832742886621385e-05,
+      "loss": 0.9371,
+      "step": 1416
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5894637268745953,
+      "learning_rate": 1.8324751933867496e-05,
+      "loss": 0.9851,
+      "step": 1417
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4416083359620568,
+      "learning_rate": 1.8322073056858257e-05,
+      "loss": 0.9316,
+      "step": 1418
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6005260021073704,
+      "learning_rate": 1.8319392235811927e-05,
+      "loss": 0.8721,
+      "step": 1419
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5302405606397658,
+      "learning_rate": 1.831670947135474e-05,
+      "loss": 0.8851,
+      "step": 1420
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7482184246000811,
+      "learning_rate": 1.8314024764113392e-05,
+      "loss": 0.9732,
+      "step": 1421
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4686236650583988,
+      "learning_rate": 1.831133811471503e-05,
+      "loss": 0.9718,
+      "step": 1422
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.866156273345178,
+      "learning_rate": 1.830864952378726e-05,
+      "loss": 0.9626,
+      "step": 1423
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3176770062046157,
+      "learning_rate": 1.830595899195813e-05,
+      "loss": 0.8406,
+      "step": 1424
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.320715112076083,
+      "learning_rate": 1.830326651985615e-05,
+      "loss": 0.8682,
+      "step": 1425
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9674957023138261,
+      "learning_rate": 1.8300572108110287e-05,
+      "loss": 0.3209,
+      "step": 1426
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7141530219710719,
+      "learning_rate": 1.829787575734995e-05,
+      "loss": 0.9555,
+      "step": 1427
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.46450866299843,
+      "learning_rate": 1.8295177468205015e-05,
+      "loss": 0.8932,
+      "step": 1428
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8058981880342446,
+      "learning_rate": 1.8292477241305794e-05,
+      "loss": 0.9201,
+      "step": 1429
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.384683821808224,
+      "learning_rate": 1.8289775077283063e-05,
+      "loss": 1.0247,
+      "step": 1430
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2925014335623914,
+      "learning_rate": 1.8287070976768045e-05,
+      "loss": 0.9926,
+      "step": 1431
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3516160590524255,
+      "learning_rate": 1.8284364940392426e-05,
+      "loss": 0.8744,
+      "step": 1432
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4822768315059653,
+      "learning_rate": 1.828165696878833e-05,
+      "loss": 0.8601,
+      "step": 1433
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6360883130209174,
+      "learning_rate": 1.8278947062588343e-05,
+      "loss": 0.8558,
+      "step": 1434
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5637041161020004,
+      "learning_rate": 1.8276235222425494e-05,
+      "loss": 0.8817,
+      "step": 1435
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7258729237288886,
+      "learning_rate": 1.8273521448933277e-05,
+      "loss": 0.9295,
+      "step": 1436
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6557778931673972,
+      "learning_rate": 1.827080574274562e-05,
+      "loss": 0.903,
+      "step": 1437
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4290332417607585,
+      "learning_rate": 1.826808810449692e-05,
+      "loss": 0.9376,
+      "step": 1438
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3708815417218,
+      "learning_rate": 1.8265368534822016e-05,
+      "loss": 0.904,
+      "step": 1439
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5470844016284202,
+      "learning_rate": 1.8262647034356197e-05,
+      "loss": 0.9463,
+      "step": 1440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5947478381137556,
+      "learning_rate": 1.825992360373521e-05,
+      "loss": 0.8913,
+      "step": 1441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.446940972521378,
+      "learning_rate": 1.825719824359524e-05,
+      "loss": 0.8848,
+      "step": 1442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.455984981257585,
+      "learning_rate": 1.8254470954572946e-05,
+      "loss": 0.9032,
+      "step": 1443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5308337765028341,
+      "learning_rate": 1.825174173730541e-05,
+      "loss": 0.9168,
+      "step": 1444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4858221441011372,
+      "learning_rate": 1.8249010592430186e-05,
+      "loss": 0.9031,
+      "step": 1445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2818499561887657,
+      "learning_rate": 1.8246277520585262e-05,
+      "loss": 0.9196,
+      "step": 1446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.503553425905122,
+      "learning_rate": 1.824354252240909e-05,
+      "loss": 0.8541,
+      "step": 1447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7404696372944537,
+      "learning_rate": 1.8240805598540562e-05,
+      "loss": 0.9065,
+      "step": 1448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4865176497344108,
+      "learning_rate": 1.8238066749619026e-05,
+      "loss": 0.9083,
+      "step": 1449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5013071116892325,
+      "learning_rate": 1.8235325976284276e-05,
+      "loss": 0.9602,
+      "step": 1450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9342158491412124,
+      "learning_rate": 1.823258327917656e-05,
+      "loss": 0.3407,
+      "step": 1451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7922600558941737,
+      "learning_rate": 1.8229838658936566e-05,
+      "loss": 0.8983,
+      "step": 1452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.471219040705767,
+      "learning_rate": 1.822709211620544e-05,
+      "loss": 0.9212,
+      "step": 1453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.79779651791726,
+      "learning_rate": 1.822434365162478e-05,
+      "loss": 0.8914,
+      "step": 1454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7298716910087544,
+      "learning_rate": 1.8221593265836624e-05,
+      "loss": 0.9112,
+      "step": 1455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3886159390640143,
+      "learning_rate": 1.821884095948346e-05,
+      "loss": 0.9563,
+      "step": 1456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.529580158241215,
+      "learning_rate": 1.821608673320823e-05,
+      "loss": 0.9526,
+      "step": 1457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6158602711557444,
+      "learning_rate": 1.8213330587654324e-05,
+      "loss": 0.9298,
+      "step": 1458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.458628492345118,
+      "learning_rate": 1.8210572523465567e-05,
+      "loss": 0.9128,
+      "step": 1459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6006041008487177,
+      "learning_rate": 1.8207812541286256e-05,
+      "loss": 0.8498,
+      "step": 1460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4658261355564617,
+      "learning_rate": 1.820505064176112e-05,
+      "loss": 0.8809,
+      "step": 1461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.505954233538878,
+      "learning_rate": 1.820228682553533e-05,
+      "loss": 0.8566,
+      "step": 1462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6084062301641522,
+      "learning_rate": 1.8199521093254524e-05,
+      "loss": 0.8057,
+      "step": 1463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6311049225116951,
+      "learning_rate": 1.8196753445564775e-05,
+      "loss": 0.9693,
+      "step": 1464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3278004768731315,
+      "learning_rate": 1.8193983883112603e-05,
+      "loss": 0.8407,
+      "step": 1465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.426047660811788,
+      "learning_rate": 1.819121240654498e-05,
+      "loss": 0.945,
+      "step": 1466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5451472349527837,
+      "learning_rate": 1.818843901650932e-05,
+      "loss": 0.839,
+      "step": 1467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6821455734600956,
+      "learning_rate": 1.8185663713653495e-05,
+      "loss": 0.9481,
+      "step": 1468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4708004598843258,
+      "learning_rate": 1.8182886498625806e-05,
+      "loss": 0.851,
+      "step": 1469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3762300466315378,
+      "learning_rate": 1.818010737207502e-05,
+      "loss": 0.9502,
+      "step": 1470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4969819510026572,
+      "learning_rate": 1.8177326334650335e-05,
+      "loss": 0.9317,
+      "step": 1471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7290544898393814,
+      "learning_rate": 1.8174543387001403e-05,
+      "loss": 0.8434,
+      "step": 1472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.415050729001895,
+      "learning_rate": 1.817175852977832e-05,
+      "loss": 0.9118,
+      "step": 1473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6752822495386923,
+      "learning_rate": 1.8168971763631636e-05,
+      "loss": 0.9153,
+      "step": 1474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.370087380605017,
+      "learning_rate": 1.816618308921233e-05,
+      "loss": 0.9445,
+      "step": 1475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5811721293966912,
+      "learning_rate": 1.816339250717184e-05,
+      "loss": 1.0104,
+      "step": 1476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6250835223007674,
+      "learning_rate": 1.816060001816205e-05,
+      "loss": 0.9256,
+      "step": 1477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.502584890655715,
+      "learning_rate": 1.815780562283528e-05,
+      "loss": 0.8978,
+      "step": 1478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5892699633215046,
+      "learning_rate": 1.8155009321844306e-05,
+      "loss": 0.8817,
+      "step": 1479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5803467565813663,
+      "learning_rate": 1.8152211115842342e-05,
+      "loss": 0.9955,
+      "step": 1480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.651876969836039,
+      "learning_rate": 1.8149411005483047e-05,
+      "loss": 0.9283,
+      "step": 1481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.743679780298233,
+      "learning_rate": 1.8146608991420533e-05,
+      "loss": 0.9177,
+      "step": 1482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5462032940470876,
+      "learning_rate": 1.8143805074309344e-05,
+      "loss": 0.9605,
+      "step": 1483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6143460454441354,
+      "learning_rate": 1.814099925480448e-05,
+      "loss": 0.8757,
+      "step": 1484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4543950045459417,
+      "learning_rate": 1.813819153356138e-05,
+      "loss": 0.9228,
+      "step": 1485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3778814108628574,
+      "learning_rate": 1.8135381911235924e-05,
+      "loss": 0.937,
+      "step": 1486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.426541612969419,
+      "learning_rate": 1.8132570388484442e-05,
+      "loss": 0.7878,
+      "step": 1487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.556011293504534,
+      "learning_rate": 1.8129756965963712e-05,
+      "loss": 0.899,
+      "step": 1488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6689503961511356,
+      "learning_rate": 1.812694164433094e-05,
+      "loss": 0.9025,
+      "step": 1489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.599667296566135,
+      "learning_rate": 1.8124124424243794e-05,
+      "loss": 0.9317,
+      "step": 1490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6482370715757464,
+      "learning_rate": 1.8121305306360374e-05,
+      "loss": 0.9498,
+      "step": 1491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6130410170244356,
+      "learning_rate": 1.811848429133922e-05,
+      "loss": 0.8094,
+      "step": 1492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5177496551723935,
+      "learning_rate": 1.8115661379839328e-05,
+      "loss": 0.818,
+      "step": 1493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3777150791970385,
+      "learning_rate": 1.8112836572520132e-05,
+      "loss": 0.942,
+      "step": 1494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.06139906827602,
+      "learning_rate": 1.8110009870041505e-05,
+      "loss": 0.331,
+      "step": 1495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5670786161336097,
+      "learning_rate": 1.8107181273063764e-05,
+      "loss": 0.8789,
+      "step": 1496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.397744676453817,
+      "learning_rate": 1.810435078224767e-05,
+      "loss": 0.9252,
+      "step": 1497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5655471855230654,
+      "learning_rate": 1.8101518398254423e-05,
+      "loss": 0.9022,
+      "step": 1498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4664594056337388,
+      "learning_rate": 1.8098684121745675e-05,
+      "loss": 0.9256,
+      "step": 1499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.492371839850668,
+      "learning_rate": 1.8095847953383506e-05,
+      "loss": 0.9655,
+      "step": 1500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3361464728772487,
+      "learning_rate": 1.809300989383045e-05,
+      "loss": 0.8374,
+      "step": 1501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7483009846294821,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.9746,
+      "step": 1502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.554544434482407,
+      "learning_rate": 1.8087328103803998e-05,
+      "loss": 0.9189,
+      "step": 1503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5815414912754846,
+      "learning_rate": 1.808448437465787e-05,
+      "loss": 0.8423,
+      "step": 1504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5686762241478212,
+      "learning_rate": 1.8081638756975385e-05,
+      "loss": 0.9415,
+      "step": 1505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7599626212831148,
+      "learning_rate": 1.8078791251421282e-05,
+      "loss": 0.9791,
+      "step": 1506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4598071792420808,
+      "learning_rate": 1.8075941858660737e-05,
+      "loss": 0.9578,
+      "step": 1507
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6101342396386267,
+      "learning_rate": 1.8073090579359373e-05,
+      "loss": 0.9315,
+      "step": 1508
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4658079826654757,
+      "learning_rate": 1.807023741418324e-05,
+      "loss": 0.9002,
+      "step": 1509
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.428926679758196,
+      "learning_rate": 1.8067382363798845e-05,
+      "loss": 0.9682,
+      "step": 1510
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7658554455291877,
+      "learning_rate": 1.8064525428873128e-05,
+      "loss": 0.9315,
+      "step": 1511
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5955447068129873,
+      "learning_rate": 1.8061666610073465e-05,
+      "loss": 0.9593,
+      "step": 1512
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.473041064226398,
+      "learning_rate": 1.805880590806768e-05,
+      "loss": 0.9342,
+      "step": 1513
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7178878290796751,
+      "learning_rate": 1.8055943323524033e-05,
+      "loss": 0.3207,
+      "step": 1514
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0193566545685027,
+      "learning_rate": 1.8053078857111218e-05,
+      "loss": 0.9605,
+      "step": 1515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.629310915558816,
+      "learning_rate": 1.805021250949839e-05,
+      "loss": 0.9409,
+      "step": 1516
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5853255658087846,
+      "learning_rate": 1.8047344281355112e-05,
+      "loss": 0.8924,
+      "step": 1517
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5117666257872096,
+      "learning_rate": 1.8044474173351403e-05,
+      "loss": 0.8893,
+      "step": 1518
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4301520769232796,
+      "learning_rate": 1.8041602186157732e-05,
+      "loss": 0.9124,
+      "step": 1519
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4755765686141669,
+      "learning_rate": 1.803872832044499e-05,
+      "loss": 0.8578,
+      "step": 1520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.573526723741205,
+      "learning_rate": 1.8035852576884508e-05,
+      "loss": 0.9123,
+      "step": 1521
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.582495436445243,
+      "learning_rate": 1.8032974956148064e-05,
+      "loss": 0.9056,
+      "step": 1522
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.433860248616746,
+      "learning_rate": 1.803009545890787e-05,
+      "loss": 0.935,
+      "step": 1523
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4019894756760587,
+      "learning_rate": 1.8027214085836578e-05,
+      "loss": 0.942,
+      "step": 1524
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7583527407692845,
+      "learning_rate": 1.8024330837607275e-05,
+      "loss": 0.8886,
+      "step": 1525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4990834528561294,
+      "learning_rate": 1.802144571489349e-05,
+      "loss": 0.9415,
+      "step": 1526
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4385328780349365,
+      "learning_rate": 1.8018558718369187e-05,
+      "loss": 0.9743,
+      "step": 1527
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4935455084531502,
+      "learning_rate": 1.8015669848708768e-05,
+      "loss": 0.9042,
+      "step": 1528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2882483382153245,
+      "learning_rate": 1.8012779106587073e-05,
+      "loss": 0.8525,
+      "step": 1529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.529216023691471,
+      "learning_rate": 1.800988649267938e-05,
+      "loss": 0.9304,
+      "step": 1530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4827389836552516,
+      "learning_rate": 1.8006992007661407e-05,
+      "loss": 0.9116,
+      "step": 1531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.703636851520783,
+      "learning_rate": 1.8004095652209304e-05,
+      "loss": 1.0178,
+      "step": 1532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6171629582491727,
+      "learning_rate": 1.8001197426999658e-05,
+      "loss": 0.9361,
+      "step": 1533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.350755787731342,
+      "learning_rate": 1.7998297332709492e-05,
+      "loss": 0.8575,
+      "step": 1534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4670535763301058,
+      "learning_rate": 1.7995395370016274e-05,
+      "loss": 1.0027,
+      "step": 1535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5501017036498101,
+      "learning_rate": 1.7992491539597904e-05,
+      "loss": 0.9615,
+      "step": 1536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4328751731537759,
+      "learning_rate": 1.7989585842132713e-05,
+      "loss": 0.9616,
+      "step": 1537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.70064371803186,
+      "learning_rate": 1.7986678278299473e-05,
+      "loss": 0.8806,
+      "step": 1538
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4205590953877154,
+      "learning_rate": 1.798376884877739e-05,
+      "loss": 0.9058,
+      "step": 1539
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6090932293860116,
+      "learning_rate": 1.798085755424611e-05,
+      "loss": 0.9362,
+      "step": 1540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6654992270354068,
+      "learning_rate": 1.7977944395385713e-05,
+      "loss": 0.9807,
+      "step": 1541
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.419143038794105,
+      "learning_rate": 1.7975029372876706e-05,
+      "loss": 0.8858,
+      "step": 1542
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6965653057851617,
+      "learning_rate": 1.7972112487400047e-05,
+      "loss": 0.8889,
+      "step": 1543
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4143447455987608,
+      "learning_rate": 1.7969193739637113e-05,
+      "loss": 0.8719,
+      "step": 1544
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4752442060782491,
+      "learning_rate": 1.7966273130269727e-05,
+      "loss": 0.8251,
+      "step": 1545
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4645123412033758,
+      "learning_rate": 1.796335065998015e-05,
+      "loss": 0.9551,
+      "step": 1546
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4545903687669843,
+      "learning_rate": 1.7960426329451062e-05,
+      "loss": 0.8886,
+      "step": 1547
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6639257574088695,
+      "learning_rate": 1.795750013936559e-05,
+      "loss": 0.8962,
+      "step": 1548
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.378138793621645,
+      "learning_rate": 1.7954572090407292e-05,
+      "loss": 0.9416,
+      "step": 1549
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5078069755966894,
+      "learning_rate": 1.7951642183260163e-05,
+      "loss": 0.9405,
+      "step": 1550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4714863757146972,
+      "learning_rate": 1.7948710418608626e-05,
+      "loss": 0.8232,
+      "step": 1551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5203309743382125,
+      "learning_rate": 1.7945776797137544e-05,
+      "loss": 0.9125,
+      "step": 1552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4394416976847793,
+      "learning_rate": 1.794284131953221e-05,
+      "loss": 0.8864,
+      "step": 1553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6192554529792336,
+      "learning_rate": 1.7939903986478354e-05,
+      "loss": 0.9913,
+      "step": 1554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6718502906283552,
+      "learning_rate": 1.7936964798662134e-05,
+      "loss": 0.8929,
+      "step": 1555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.40202113587381,
+      "learning_rate": 1.793402375677015e-05,
+      "loss": 0.7959,
+      "step": 1556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4785852844175729,
+      "learning_rate": 1.7931080861489425e-05,
+      "loss": 0.8941,
+      "step": 1557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6389009817960503,
+      "learning_rate": 1.7928136113507416e-05,
+      "loss": 0.8848,
+      "step": 1558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6442349522622586,
+      "learning_rate": 1.7925189513512026e-05,
+      "loss": 0.943,
+      "step": 1559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6843259031486602,
+      "learning_rate": 1.7922241062191576e-05,
+      "loss": 0.8991,
+      "step": 1560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4130152069717505,
+      "learning_rate": 1.791929076023482e-05,
+      "loss": 0.8297,
+      "step": 1561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.521629480217419,
+      "learning_rate": 1.791633860833096e-05,
+      "loss": 0.9784,
+      "step": 1562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5847138266858074,
+      "learning_rate": 1.7913384607169608e-05,
+      "loss": 0.912,
+      "step": 1563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.589646845065804,
+      "learning_rate": 1.7910428757440826e-05,
+      "loss": 0.9449,
+      "step": 1564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5248021958024043,
+      "learning_rate": 1.7907471059835097e-05,
+      "loss": 0.9189,
+      "step": 1565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5770284034236917,
+      "learning_rate": 1.7904511515043338e-05,
+      "loss": 0.8986,
+      "step": 1566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6235388826026316,
+      "learning_rate": 1.7901550123756906e-05,
+      "loss": 0.9351,
+      "step": 1567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6085182688439044,
+      "learning_rate": 1.7898586886667574e-05,
+      "loss": 0.9453,
+      "step": 1568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4297041475129335,
+      "learning_rate": 1.7895621804467562e-05,
+      "loss": 0.9748,
+      "step": 1569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5316494607998903,
+      "learning_rate": 1.789265487784951e-05,
+      "loss": 0.9373,
+      "step": 1570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.471618927636049,
+      "learning_rate": 1.788968610750649e-05,
+      "loss": 0.9069,
+      "step": 1571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5583034862952156,
+      "learning_rate": 1.7886715494132008e-05,
+      "loss": 0.9885,
+      "step": 1572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.556130305248774,
+      "learning_rate": 1.7883743038420002e-05,
+      "loss": 0.9727,
+      "step": 1573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4781080752587312,
+      "learning_rate": 1.788076874106484e-05,
+      "loss": 0.8378,
+      "step": 1574
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5675583559886217,
+      "learning_rate": 1.7877792602761312e-05,
+      "loss": 0.9611,
+      "step": 1575
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5171448654937751,
+      "learning_rate": 1.787481462420465e-05,
+      "loss": 0.845,
+      "step": 1576
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3741617783810502,
+      "learning_rate": 1.7871834806090502e-05,
+      "loss": 0.871,
+      "step": 1577
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.411148419013395,
+      "learning_rate": 1.7868853149114966e-05,
+      "loss": 0.839,
+      "step": 1578
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5433128238880478,
+      "learning_rate": 1.7865869653974545e-05,
+      "loss": 0.9828,
+      "step": 1579
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3669869480177081,
+      "learning_rate": 1.786288432136619e-05,
+      "loss": 0.9316,
+      "step": 1580
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5452442556225698,
+      "learning_rate": 1.7859897151987276e-05,
+      "loss": 0.9209,
+      "step": 1581
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.446698103249709,
+      "learning_rate": 1.7856908146535602e-05,
+      "loss": 0.905,
+      "step": 1582
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3660707654069846,
+      "learning_rate": 1.7853917305709405e-05,
+      "loss": 0.9321,
+      "step": 1583
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5629656675943877,
+      "learning_rate": 1.785092463020734e-05,
+      "loss": 0.9017,
+      "step": 1584
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6060196692364832,
+      "learning_rate": 1.78479301207285e-05,
+      "loss": 0.8026,
+      "step": 1585
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.463758311889416,
+      "learning_rate": 1.7844933777972406e-05,
+      "loss": 0.9187,
+      "step": 1586
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9032905394259563,
+      "learning_rate": 1.7841935602638997e-05,
+      "loss": 0.9195,
+      "step": 1587
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6859178002349369,
+      "learning_rate": 1.783893559542865e-05,
+      "loss": 0.885,
+      "step": 1588
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5535702386924535,
+      "learning_rate": 1.7835933757042165e-05,
+      "loss": 0.8999,
+      "step": 1589
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4710355258912164,
+      "learning_rate": 1.7832930088180777e-05,
+      "loss": 0.8353,
+      "step": 1590
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.515526926908418,
+      "learning_rate": 1.782992458954614e-05,
+      "loss": 0.8078,
+      "step": 1591
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4536267293731475,
+      "learning_rate": 1.7826917261840337e-05,
+      "loss": 0.8524,
+      "step": 1592
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4877227876822239,
+      "learning_rate": 1.7823908105765883e-05,
+      "loss": 0.9063,
+      "step": 1593
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3361140839951366,
+      "learning_rate": 1.7820897122025717e-05,
+      "loss": 0.8861,
+      "step": 1594
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6146067722731678,
+      "learning_rate": 1.7817884311323203e-05,
+      "loss": 0.8947,
+      "step": 1595
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.415559239996958,
+      "learning_rate": 1.7814869674362133e-05,
+      "loss": 0.898,
+      "step": 1596
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8070587128079143,
+      "learning_rate": 1.781185321184673e-05,
+      "loss": 0.9469,
+      "step": 1597
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6918271919710215,
+      "learning_rate": 1.780883492448164e-05,
+      "loss": 0.8558,
+      "step": 1598
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9394417938532487,
+      "learning_rate": 1.780581481297193e-05,
+      "loss": 0.3136,
+      "step": 1599
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3710380329592233,
+      "learning_rate": 1.7802792878023108e-05,
+      "loss": 0.8846,
+      "step": 1600
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.621653634787497,
+      "learning_rate": 1.779976912034109e-05,
+      "loss": 0.9499,
+      "step": 1601
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.49535069053136,
+      "learning_rate": 1.7796743540632226e-05,
+      "loss": 0.914,
+      "step": 1602
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5287861793456374,
+      "learning_rate": 1.7793716139603297e-05,
+      "loss": 0.8583,
+      "step": 1603
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6938236693429272,
+      "learning_rate": 1.7790686917961498e-05,
+      "loss": 0.8335,
+      "step": 1604
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5387393199407082,
+      "learning_rate": 1.7787655876414463e-05,
+      "loss": 0.844,
+      "step": 1605
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5769405086202581,
+      "learning_rate": 1.7784623015670237e-05,
+      "loss": 0.9482,
+      "step": 1606
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.422886086138693,
+      "learning_rate": 1.77815883364373e-05,
+      "loss": 0.9369,
+      "step": 1607
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9304298204162212,
+      "learning_rate": 1.777855183942455e-05,
+      "loss": 0.3799,
+      "step": 1608
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6124495672824373,
+      "learning_rate": 1.7775513525341318e-05,
+      "loss": 0.827,
+      "step": 1609
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7060739694083267,
+      "learning_rate": 1.777247339489735e-05,
+      "loss": 0.8808,
+      "step": 1610
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5229384717595726,
+      "learning_rate": 1.7769431448802824e-05,
+      "loss": 0.9077,
+      "step": 1611
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5395189579010597,
+      "learning_rate": 1.7766387687768338e-05,
+      "loss": 0.8656,
+      "step": 1612
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4780489060209367,
+      "learning_rate": 1.7763342112504913e-05,
+      "loss": 0.942,
+      "step": 1613
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4799745384004357,
+      "learning_rate": 1.7760294723724e-05,
+      "loss": 0.8457,
+      "step": 1614
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4684627472723826,
+      "learning_rate": 1.775724552213746e-05,
+      "loss": 0.8746,
+      "step": 1615
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6040790935063485,
+      "learning_rate": 1.7754194508457602e-05,
+      "loss": 0.9138,
+      "step": 1616
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.35682520205269,
+      "learning_rate": 1.7751141683397128e-05,
+      "loss": 0.8522,
+      "step": 1617
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3353106341381324,
+      "learning_rate": 1.7748087047669186e-05,
+      "loss": 0.9063,
+      "step": 1618
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8100525305630744,
+      "learning_rate": 1.7745030601987338e-05,
+      "loss": 0.9223,
+      "step": 1619
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6092339721248268,
+      "learning_rate": 1.774197234706557e-05,
+      "loss": 0.9258,
+      "step": 1620
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7458819446495246,
+      "learning_rate": 1.7738912283618293e-05,
+      "loss": 0.9343,
+      "step": 1621
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7811148199017655,
+      "learning_rate": 1.7735850412360332e-05,
+      "loss": 0.8773,
+      "step": 1622
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3201410256999138,
+      "learning_rate": 1.7732786734006947e-05,
+      "loss": 0.8987,
+      "step": 1623
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.506969757466883,
+      "learning_rate": 1.772972124927381e-05,
+      "loss": 0.951,
+      "step": 1624
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5858392367904162,
+      "learning_rate": 1.7726653958877016e-05,
+      "loss": 0.874,
+      "step": 1625
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6850626361525405,
+      "learning_rate": 1.772358486353309e-05,
+      "loss": 0.867,
+      "step": 1626
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5132178985557543,
+      "learning_rate": 1.772051396395897e-05,
+      "loss": 0.9403,
+      "step": 1627
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5875971923008991,
+      "learning_rate": 1.771744126087202e-05,
+      "loss": 0.9398,
+      "step": 1628
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3874449433591847,
+      "learning_rate": 1.771436675499002e-05,
+      "loss": 0.9517,
+      "step": 1629
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6778689098484927,
+      "learning_rate": 1.771129044703118e-05,
+      "loss": 0.9133,
+      "step": 1630
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5965843953852386,
+      "learning_rate": 1.770821233771412e-05,
+      "loss": 0.9315,
+      "step": 1631
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.480797115990424,
+      "learning_rate": 1.7705132427757895e-05,
+      "loss": 0.9489,
+      "step": 1632
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8642756347111773,
+      "learning_rate": 1.7702050717881966e-05,
+      "loss": 0.3473,
+      "step": 1633
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4990152302868358,
+      "learning_rate": 1.7698967208806216e-05,
+      "loss": 0.9048,
+      "step": 1634
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.485270652222611,
+      "learning_rate": 1.7695881901250966e-05,
+      "loss": 0.9014,
+      "step": 1635
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5214858759038965,
+      "learning_rate": 1.7692794795936933e-05,
+      "loss": 0.896,
+      "step": 1636
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4474856682989883,
+      "learning_rate": 1.7689705893585273e-05,
+      "loss": 0.9221,
+      "step": 1637
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5294260479584558,
+      "learning_rate": 1.768661519491755e-05,
+      "loss": 0.9252,
+      "step": 1638
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5849465529940199,
+      "learning_rate": 1.7683522700655748e-05,
+      "loss": 0.9858,
+      "step": 1639
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7816169669656882,
+      "learning_rate": 1.768042841152228e-05,
+      "loss": 0.3248,
+      "step": 1640
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.378251448439808,
+      "learning_rate": 1.767733232823997e-05,
+      "loss": 0.9562,
+      "step": 1641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.76066055704607,
+      "learning_rate": 1.7674234451532065e-05,
+      "loss": 0.308,
+      "step": 1642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8072615083001393,
+      "learning_rate": 1.7671134782122225e-05,
+      "loss": 0.3364,
+      "step": 1643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4158020503018405,
+      "learning_rate": 1.7668033320734536e-05,
+      "loss": 0.8818,
+      "step": 1644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5579198244318886,
+      "learning_rate": 1.76649300680935e-05,
+      "loss": 0.9414,
+      "step": 1645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4821532688374293,
+      "learning_rate": 1.7661825024924035e-05,
+      "loss": 0.9064,
+      "step": 1646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.532267404015606,
+      "learning_rate": 1.7658718191951483e-05,
+      "loss": 0.9663,
+      "step": 1647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4236472895400747,
+      "learning_rate": 1.76556095699016e-05,
+      "loss": 0.9092,
+      "step": 1648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4394290112011618,
+      "learning_rate": 1.7652499159500554e-05,
+      "loss": 0.9262,
+      "step": 1649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9082894509829694,
+      "learning_rate": 1.7649386961474944e-05,
+      "loss": 0.3479,
+      "step": 1650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5694811646372455,
+      "learning_rate": 1.764627297655178e-05,
+      "loss": 0.8595,
+      "step": 1651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.778224908991867,
+      "learning_rate": 1.7643157205458483e-05,
+      "loss": 0.9557,
+      "step": 1652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7570159431849186,
+      "learning_rate": 1.7640039648922905e-05,
+      "loss": 0.9355,
+      "step": 1653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.535934897475885,
+      "learning_rate": 1.7636920307673305e-05,
+      "loss": 0.9163,
+      "step": 1654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.500786112142508,
+      "learning_rate": 1.7633799182438355e-05,
+      "loss": 0.8763,
+      "step": 1655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7958521536960721,
+      "learning_rate": 1.7630676273947157e-05,
+      "loss": 0.3603,
+      "step": 1656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5405759425893217,
+      "learning_rate": 1.7627551582929223e-05,
+      "loss": 0.8867,
+      "step": 1657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4105335679188502,
+      "learning_rate": 1.762442511011448e-05,
+      "loss": 0.8636,
+      "step": 1658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6396145007703673,
+      "learning_rate": 1.762129685623327e-05,
+      "loss": 0.8969,
+      "step": 1659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6454153765062922,
+      "learning_rate": 1.7618166822016358e-05,
+      "loss": 0.9028,
+      "step": 1660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6597883658356147,
+      "learning_rate": 1.7615035008194914e-05,
+      "loss": 0.852,
+      "step": 1661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4238215944744783,
+      "learning_rate": 1.7611901415500536e-05,
+      "loss": 0.8634,
+      "step": 1662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3317438945524687,
+      "learning_rate": 1.7608766044665227e-05,
+      "loss": 0.8193,
+      "step": 1663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5834953879779157,
+      "learning_rate": 1.7605628896421412e-05,
+      "loss": 0.8995,
+      "step": 1664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6297456879070857,
+      "learning_rate": 1.760248997150193e-05,
+      "loss": 0.9008,
+      "step": 1665
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3606229375868286,
+      "learning_rate": 1.7599349270640034e-05,
+      "loss": 0.9694,
+      "step": 1666
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.428091273337352,
+      "learning_rate": 1.759620679456939e-05,
+      "loss": 0.8583,
+      "step": 1667
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5387743968272987,
+      "learning_rate": 1.7593062544024084e-05,
+      "loss": 0.8277,
+      "step": 1668
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4367561605643178,
+      "learning_rate": 1.758991651973861e-05,
+      "loss": 0.8267,
+      "step": 1669
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5360877563358466,
+      "learning_rate": 1.758676872244788e-05,
+      "loss": 0.946,
+      "step": 1670
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6575018452365091,
+      "learning_rate": 1.7583619152887222e-05,
+      "loss": 0.8916,
+      "step": 1671
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5189025333321005,
+      "learning_rate": 1.7580467811792374e-05,
+      "loss": 0.8831,
+      "step": 1672
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4889245276310539,
+      "learning_rate": 1.7577314699899486e-05,
+      "loss": 0.962,
+      "step": 1673
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5920227440503765,
+      "learning_rate": 1.7574159817945134e-05,
+      "loss": 0.8194,
+      "step": 1674
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6663380352263728,
+      "learning_rate": 1.7571003166666295e-05,
+      "loss": 0.9153,
+      "step": 1675
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5568655430691456,
+      "learning_rate": 1.756784474680036e-05,
+      "loss": 0.9206,
+      "step": 1676
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5313420778927918,
+      "learning_rate": 1.7564684559085138e-05,
+      "loss": 0.9114,
+      "step": 1677
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7955562493684992,
+      "learning_rate": 1.756152260425885e-05,
+      "loss": 0.3385,
+      "step": 1678
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.791077961143688,
+      "learning_rate": 1.7558358883060128e-05,
+      "loss": 0.9402,
+      "step": 1679
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6186353518599614,
+      "learning_rate": 1.7555193396228015e-05,
+      "loss": 0.8572,
+      "step": 1680
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5337082328468437,
+      "learning_rate": 1.7552026144501976e-05,
+      "loss": 1.0438,
+      "step": 1681
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.946323407870686,
+      "learning_rate": 1.7548857128621878e-05,
+      "loss": 0.8694,
+      "step": 1682
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6462468216871893,
+      "learning_rate": 1.7545686349328e-05,
+      "loss": 0.9585,
+      "step": 1683
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5771995772346616,
+      "learning_rate": 1.754251380736104e-05,
+      "loss": 0.9554,
+      "step": 1684
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5107802119336817,
+      "learning_rate": 1.7539339503462103e-05,
+      "loss": 0.8278,
+      "step": 1685
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5948314643925587,
+      "learning_rate": 1.753616343837271e-05,
+      "loss": 0.8782,
+      "step": 1686
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5517116597383998,
+      "learning_rate": 1.753298561283478e-05,
+      "loss": 0.9333,
+      "step": 1687
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7926694624360804,
+      "learning_rate": 1.7529806027590668e-05,
+      "loss": 0.9011,
+      "step": 1688
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4242999599319386,
+      "learning_rate": 1.7526624683383114e-05,
+      "loss": 1.0023,
+      "step": 1689
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4756871055595069,
+      "learning_rate": 1.752344158095528e-05,
+      "loss": 0.866,
+      "step": 1690
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5269401155973021,
+      "learning_rate": 1.752025672105075e-05,
+      "loss": 0.9106,
+      "step": 1691
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6514765659791704,
+      "learning_rate": 1.7517070104413497e-05,
+      "loss": 0.8834,
+      "step": 1692
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.466126278503492,
+      "learning_rate": 1.7513881731787924e-05,
+      "loss": 0.95,
+      "step": 1693
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4524000407796849,
+      "learning_rate": 1.7510691603918825e-05,
+      "loss": 0.8784,
+      "step": 1694
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7410398325393686,
+      "learning_rate": 1.750749972155142e-05,
+      "loss": 0.845,
+      "step": 1695
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5777950166125367,
+      "learning_rate": 1.7504306085431334e-05,
+      "loss": 0.965,
+      "step": 1696
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5838872906840074,
+      "learning_rate": 1.7501110696304598e-05,
+      "loss": 0.9294,
+      "step": 1697
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5369772478631567,
+      "learning_rate": 1.7497913554917656e-05,
+      "loss": 0.9132,
+      "step": 1698
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3601257098393902,
+      "learning_rate": 1.749471466201736e-05,
+      "loss": 0.8909,
+      "step": 1699
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5347499479608235,
+      "learning_rate": 1.7491514018350974e-05,
+      "loss": 0.8639,
+      "step": 1700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5490947412133416,
+      "learning_rate": 1.7488311624666165e-05,
+      "loss": 0.8829,
+      "step": 1701
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5543988693147053,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.941,
+      "step": 1702
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6548435316823147,
+      "learning_rate": 1.748190159023401e-05,
+      "loss": 1.0323,
+      "step": 1703
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.339462331638418,
+      "learning_rate": 1.747869395098405e-05,
+      "loss": 0.9051,
+      "step": 1704
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8201796335838167,
+      "learning_rate": 1.7475484564710437e-05,
+      "loss": 0.8872,
+      "step": 1705
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6880989562572777,
+      "learning_rate": 1.7472273432162886e-05,
+      "loss": 0.8475,
+      "step": 1706
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.611526301363743,
+      "learning_rate": 1.7469060554091518e-05,
+      "loss": 0.8852,
+      "step": 1707
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6037891104770856,
+      "learning_rate": 1.7465845931246858e-05,
+      "loss": 0.8353,
+      "step": 1708
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5586701340947895,
+      "learning_rate": 1.7462629564379846e-05,
+      "loss": 0.8964,
+      "step": 1709
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4803400744931496,
+      "learning_rate": 1.7459411454241822e-05,
+      "loss": 0.8311,
+      "step": 1710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4170972934968988,
+      "learning_rate": 1.7456191601584544e-05,
+      "loss": 0.8556,
+      "step": 1711
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7313153185296979,
+      "learning_rate": 1.745297000716016e-05,
+      "loss": 0.866,
+      "step": 1712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5203219014060387,
+      "learning_rate": 1.7449746671721243e-05,
+      "loss": 0.8972,
+      "step": 1713
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4457192583170824,
+      "learning_rate": 1.744652159602076e-05,
+      "loss": 0.9248,
+      "step": 1714
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3942818973038391,
+      "learning_rate": 1.744329478081209e-05,
+      "loss": 0.9204,
+      "step": 1715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4531075562435056,
+      "learning_rate": 1.744006622684902e-05,
+      "loss": 0.8871,
+      "step": 1716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5789970700730995,
+      "learning_rate": 1.7436835934885735e-05,
+      "loss": 0.9262,
+      "step": 1717
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5611116320765956,
+      "learning_rate": 1.743360390567684e-05,
+      "loss": 0.8636,
+      "step": 1718
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.467310440937347,
+      "learning_rate": 1.7430370139977327e-05,
+      "loss": 0.8532,
+      "step": 1719
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4069213853948344,
+      "learning_rate": 1.7427134638542612e-05,
+      "loss": 0.9562,
+      "step": 1720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8629732208129236,
+      "learning_rate": 1.7423897402128505e-05,
+      "loss": 0.2969,
+      "step": 1721
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5794908589828864,
+      "learning_rate": 1.7420658431491224e-05,
+      "loss": 0.9728,
+      "step": 1722
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4555251810863923,
+      "learning_rate": 1.7417417727387392e-05,
+      "loss": 0.9158,
+      "step": 1723
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4490216871807111,
+      "learning_rate": 1.7414175290574044e-05,
+      "loss": 0.885,
+      "step": 1724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0688220651181375,
+      "learning_rate": 1.741093112180861e-05,
+      "loss": 0.9506,
+      "step": 1725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4449212317917777,
+      "learning_rate": 1.7407685221848925e-05,
+      "loss": 0.8994,
+      "step": 1726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4599573524456597,
+      "learning_rate": 1.7404437591453237e-05,
+      "loss": 0.9289,
+      "step": 1727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5625676068938543,
+      "learning_rate": 1.7401188231380185e-05,
+      "loss": 0.973,
+      "step": 1728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4693554672713998,
+      "learning_rate": 1.739793714238883e-05,
+      "loss": 0.9168,
+      "step": 1729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6503806164680692,
+      "learning_rate": 1.7394684325238616e-05,
+      "loss": 0.8105,
+      "step": 1730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5745296703674048,
+      "learning_rate": 1.7391429780689414e-05,
+      "loss": 0.9836,
+      "step": 1731
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4485110292525867,
+      "learning_rate": 1.7388173509501475e-05,
+      "loss": 0.9437,
+      "step": 1732
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5343799885243359,
+      "learning_rate": 1.7384915512435466e-05,
+      "loss": 0.9329,
+      "step": 1733
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6859775492495856,
+      "learning_rate": 1.738165579025246e-05,
+      "loss": 0.8766,
+      "step": 1734
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7636578528690656,
+      "learning_rate": 1.737839434371393e-05,
+      "loss": 0.884,
+      "step": 1735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5276155041013695,
+      "learning_rate": 1.737513117358174e-05,
+      "loss": 0.915,
+      "step": 1736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.792478974482549,
+      "learning_rate": 1.7371866280618176e-05,
+      "loss": 0.326,
+      "step": 1737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4961026604787953,
+      "learning_rate": 1.7368599665585916e-05,
+      "loss": 0.9006,
+      "step": 1738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5274209999344202,
+      "learning_rate": 1.7365331329248035e-05,
+      "loss": 0.8725,
+      "step": 1739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.351068948198467,
+      "learning_rate": 1.7362061272368026e-05,
+      "loss": 0.9436,
+      "step": 1740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5360903912923594,
+      "learning_rate": 1.735878949570977e-05,
+      "loss": 0.8497,
+      "step": 1741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3497630823993179,
+      "learning_rate": 1.7355516000037555e-05,
+      "loss": 0.9142,
+      "step": 1742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5809404969285956,
+      "learning_rate": 1.7352240786116068e-05,
+      "loss": 0.8947,
+      "step": 1743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4846738357643812,
+      "learning_rate": 1.73489638547104e-05,
+      "loss": 0.9025,
+      "step": 1744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.624635000914181,
+      "learning_rate": 1.7345685206586045e-05,
+      "loss": 0.8762,
+      "step": 1745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.830149285837252,
+      "learning_rate": 1.7342404842508896e-05,
+      "loss": 0.8836,
+      "step": 1746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.461486520725363,
+      "learning_rate": 1.733912276324524e-05,
+      "loss": 0.8534,
+      "step": 1747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.434998799493417,
+      "learning_rate": 1.7335838969561777e-05,
+      "loss": 0.9205,
+      "step": 1748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6244167337949473,
+      "learning_rate": 1.7332553462225604e-05,
+      "loss": 0.8886,
+      "step": 1749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6222076398074747,
+      "learning_rate": 1.7329266242004205e-05,
+      "loss": 0.8626,
+      "step": 1750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5200145285107727,
+      "learning_rate": 1.7325977309665485e-05,
+      "loss": 1.0063,
+      "step": 1751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.451699862945971,
+      "learning_rate": 1.7322686665977738e-05,
+      "loss": 0.8288,
+      "step": 1752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9622903219057697,
+      "learning_rate": 1.7319394311709655e-05,
+      "loss": 0.9016,
+      "step": 1753
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3631012382364607,
+      "learning_rate": 1.731610024763033e-05,
+      "loss": 0.943,
+      "step": 1754
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.69183243256369,
+      "learning_rate": 1.731280447450926e-05,
+      "loss": 0.8885,
+      "step": 1755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5841491510032268,
+      "learning_rate": 1.7309506993116333e-05,
+      "loss": 0.8766,
+      "step": 1756
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5842091051694704,
+      "learning_rate": 1.7306207804221845e-05,
+      "loss": 1.013,
+      "step": 1757
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5566617085980659,
+      "learning_rate": 1.7302906908596487e-05,
+      "loss": 0.8763,
+      "step": 1758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4113051668924068,
+      "learning_rate": 1.729960430701135e-05,
+      "loss": 0.8902,
+      "step": 1759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5081173429273016,
+      "learning_rate": 1.7296300000237917e-05,
+      "loss": 0.997,
+      "step": 1760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4691870494415773,
+      "learning_rate": 1.7292993989048076e-05,
+      "loss": 0.8829,
+      "step": 1761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5797824647838894,
+      "learning_rate": 1.7289686274214116e-05,
+      "loss": 0.928,
+      "step": 1762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.43389339394806,
+      "learning_rate": 1.7286376856508714e-05,
+      "loss": 0.9845,
+      "step": 1763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4309202791520272,
+      "learning_rate": 1.7283065736704956e-05,
+      "loss": 0.9778,
+      "step": 1764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5539428890256648,
+      "learning_rate": 1.7279752915576312e-05,
+      "loss": 0.852,
+      "step": 1765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4520281464163747,
+      "learning_rate": 1.7276438393896663e-05,
+      "loss": 0.8648,
+      "step": 1766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6531081314163723,
+      "learning_rate": 1.727312217244028e-05,
+      "loss": 0.8553,
+      "step": 1767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4559817625290705,
+      "learning_rate": 1.7269804251981835e-05,
+      "loss": 0.9928,
+      "step": 1768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.521372074174493,
+      "learning_rate": 1.726648463329639e-05,
+      "loss": 0.9311,
+      "step": 1769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.39292238696397,
+      "learning_rate": 1.726316331715941e-05,
+      "loss": 0.9482,
+      "step": 1770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6238781226882557,
+      "learning_rate": 1.7259840304346757e-05,
+      "loss": 0.866,
+      "step": 1771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.427991969722996,
+      "learning_rate": 1.7256515595634688e-05,
+      "loss": 0.9149,
+      "step": 1772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3335333993191922,
+      "learning_rate": 1.7253189191799853e-05,
+      "loss": 0.9076,
+      "step": 1773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.472499218013164,
+      "learning_rate": 1.7249861093619298e-05,
+      "loss": 0.8997,
+      "step": 1774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6890628308746691,
+      "learning_rate": 1.7246531301870467e-05,
+      "loss": 0.9985,
+      "step": 1775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7011530663800565,
+      "learning_rate": 1.7243199817331207e-05,
+      "loss": 1.0012,
+      "step": 1776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9561611437955488,
+      "learning_rate": 1.7239866640779745e-05,
+      "loss": 0.9127,
+      "step": 1777
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4494502468755988,
+      "learning_rate": 1.7236531772994714e-05,
+      "loss": 0.9929,
+      "step": 1778
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5584496393241531,
+      "learning_rate": 1.723319521475514e-05,
+      "loss": 0.9213,
+      "step": 1779
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4555365582732622,
+      "learning_rate": 1.7229856966840444e-05,
+      "loss": 0.8719,
+      "step": 1780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5727131597575448,
+      "learning_rate": 1.7226517030030444e-05,
+      "loss": 0.9302,
+      "step": 1781
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.495720422096291,
+      "learning_rate": 1.722317540510534e-05,
+      "loss": 0.8064,
+      "step": 1782
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.415398908027525,
+      "learning_rate": 1.7219832092845746e-05,
+      "loss": 0.9454,
+      "step": 1783
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4238637709112882,
+      "learning_rate": 1.7216487094032653e-05,
+      "loss": 0.9063,
+      "step": 1784
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4713464307722328,
+      "learning_rate": 1.7213140409447455e-05,
+      "loss": 0.8112,
+      "step": 1785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3735514428630609,
+      "learning_rate": 1.7209792039871942e-05,
+      "loss": 0.8631,
+      "step": 1786
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4875295284990784,
+      "learning_rate": 1.720644198608829e-05,
+      "loss": 0.8771,
+      "step": 1787
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.508326494733138,
+      "learning_rate": 1.720309024887907e-05,
+      "loss": 0.9088,
+      "step": 1788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6146606431930937,
+      "learning_rate": 1.7199736829027252e-05,
+      "loss": 0.9049,
+      "step": 1789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.516122841920055,
+      "learning_rate": 1.7196381727316192e-05,
+      "loss": 0.8811,
+      "step": 1790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4806277282829798,
+      "learning_rate": 1.7193024944529647e-05,
+      "loss": 0.8732,
+      "step": 1791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4559408235450584,
+      "learning_rate": 1.7189666481451755e-05,
+      "loss": 0.9023,
+      "step": 1792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4453281936566724,
+      "learning_rate": 1.7186306338867055e-05,
+      "loss": 0.9596,
+      "step": 1793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.498584277906921,
+      "learning_rate": 1.7182944517560483e-05,
+      "loss": 0.8495,
+      "step": 1794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.816898789845504,
+      "learning_rate": 1.7179581018317354e-05,
+      "loss": 0.8659,
+      "step": 1795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7775574133563197,
+      "learning_rate": 1.7176215841923385e-05,
+      "loss": 0.8375,
+      "step": 1796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4353467710993535,
+      "learning_rate": 1.717284898916468e-05,
+      "loss": 0.8893,
+      "step": 1797
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4930765015281215,
+      "learning_rate": 1.7169480460827734e-05,
+      "loss": 0.9407,
+      "step": 1798
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5119016940782954,
+      "learning_rate": 1.716611025769944e-05,
+      "loss": 0.8741,
+      "step": 1799
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5374560607421481,
+      "learning_rate": 1.7162738380567077e-05,
+      "loss": 0.9104,
+      "step": 1800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3886831010828815,
+      "learning_rate": 1.7159364830218312e-05,
+      "loss": 0.8395,
+      "step": 1801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.448245797696183,
+      "learning_rate": 1.715598960744121e-05,
+      "loss": 0.8416,
+      "step": 1802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5111294590928952,
+      "learning_rate": 1.7152612713024226e-05,
+      "loss": 0.8573,
+      "step": 1803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7186909367426648,
+      "learning_rate": 1.71492341477562e-05,
+      "loss": 0.9487,
+      "step": 1804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6378023900766485,
+      "learning_rate": 1.714585391242636e-05,
+      "loss": 0.9359,
+      "step": 1805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5038280119918932,
+      "learning_rate": 1.714247200782434e-05,
+      "loss": 0.906,
+      "step": 1806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7788525810740219,
+      "learning_rate": 1.7139088434740142e-05,
+      "loss": 0.8566,
+      "step": 1807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5776725594957457,
+      "learning_rate": 1.7135703193964176e-05,
+      "loss": 0.9065,
+      "step": 1808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.453841505756265,
+      "learning_rate": 1.7132316286287235e-05,
+      "loss": 0.9159,
+      "step": 1809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.46112002415715,
+      "learning_rate": 1.71289277125005e-05,
+      "loss": 0.9018,
+      "step": 1810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.632216149580276,
+      "learning_rate": 1.712553747339554e-05,
+      "loss": 0.9288,
+      "step": 1811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2984613259168827,
+      "learning_rate": 1.712214556976431e-05,
+      "loss": 0.9663,
+      "step": 1812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6099174011408217,
+      "learning_rate": 1.711875200239917e-05,
+      "loss": 0.9112,
+      "step": 1813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.605283506059939,
+      "learning_rate": 1.7115356772092858e-05,
+      "loss": 0.9828,
+      "step": 1814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6315331928711274,
+      "learning_rate": 1.711195987963849e-05,
+      "loss": 0.9046,
+      "step": 1815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6107117450770427,
+      "learning_rate": 1.7108561325829584e-05,
+      "loss": 0.9446,
+      "step": 1816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6408891204130671,
+      "learning_rate": 1.7105161111460046e-05,
+      "loss": 0.9636,
+      "step": 1817
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5407244258507045,
+      "learning_rate": 1.7101759237324165e-05,
+      "loss": 0.9403,
+      "step": 1818
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4569115346658608,
+      "learning_rate": 1.7098355704216622e-05,
+      "loss": 0.9117,
+      "step": 1819
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3826627279790076,
+      "learning_rate": 1.7094950512932475e-05,
+      "loss": 0.8837,
+      "step": 1820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5932423897031658,
+      "learning_rate": 1.7091543664267183e-05,
+      "loss": 0.9623,
+      "step": 1821
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5126250911344894,
+      "learning_rate": 1.7088135159016584e-05,
+      "loss": 0.9499,
+      "step": 1822
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4927210889833857,
+      "learning_rate": 1.7084724997976903e-05,
+      "loss": 0.8769,
+      "step": 1823
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7444005091219006,
+      "learning_rate": 1.708131318194476e-05,
+      "loss": 0.8957,
+      "step": 1824
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4384235727115369,
+      "learning_rate": 1.7077899711717152e-05,
+      "loss": 0.9226,
+      "step": 1825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5573586969929953,
+      "learning_rate": 1.7074484588091465e-05,
+      "loss": 0.8396,
+      "step": 1826
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6943024623763063,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.9327,
+      "step": 1827
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3264942269004616,
+      "learning_rate": 1.706764938383734e-05,
+      "loss": 0.9455,
+      "step": 1828
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5792481985995255,
+      "learning_rate": 1.7064229304805607e-05,
+      "loss": 0.8752,
+      "step": 1829
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.905335752327466,
+      "learning_rate": 1.70608075755692e-05,
+      "loss": 0.8723,
+      "step": 1830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8859667078128329,
+      "learning_rate": 1.705738419692744e-05,
+      "loss": 0.3358,
+      "step": 1831
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5627065563762872,
+      "learning_rate": 1.7053959169680033e-05,
+      "loss": 0.8605,
+      "step": 1832
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6044406348575977,
+      "learning_rate": 1.7050532494627058e-05,
+      "loss": 0.9248,
+      "step": 1833
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5970990414792006,
+      "learning_rate": 1.704710417256899e-05,
+      "loss": 0.9733,
+      "step": 1834
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5942881087129774,
+      "learning_rate": 1.7043674204306688e-05,
+      "loss": 0.93,
+      "step": 1835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.466825606220935,
+      "learning_rate": 1.7040242590641385e-05,
+      "loss": 0.8589,
+      "step": 1836
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3292173388448874,
+      "learning_rate": 1.7036809332374713e-05,
+      "loss": 0.9582,
+      "step": 1837
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4421528178939231,
+      "learning_rate": 1.7033374430308683e-05,
+      "loss": 0.8817,
+      "step": 1838
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5129807210984627,
+      "learning_rate": 1.7029937885245682e-05,
+      "loss": 0.8774,
+      "step": 1839
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5214130895415245,
+      "learning_rate": 1.7026499697988496e-05,
+      "loss": 0.882,
+      "step": 1840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6976364727070776,
+      "learning_rate": 1.7023059869340276e-05,
+      "loss": 0.956,
+      "step": 1841
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3286347161428194,
+      "learning_rate": 1.7019618400104572e-05,
+      "loss": 0.8316,
+      "step": 1842
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5186825909980486,
+      "learning_rate": 1.7016175291085308e-05,
+      "loss": 0.9076,
+      "step": 1843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6401109304442076,
+      "learning_rate": 1.7012730543086798e-05,
+      "loss": 0.9993,
+      "step": 1844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5035105642985296,
+      "learning_rate": 1.7009284156913737e-05,
+      "loss": 0.7926,
+      "step": 1845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4650515021185209,
+      "learning_rate": 1.70058361333712e-05,
+      "loss": 0.9737,
+      "step": 1846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7570150367833464,
+      "learning_rate": 1.700238647326464e-05,
+      "loss": 0.8754,
+      "step": 1847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6313988044756296,
+      "learning_rate": 1.6998935177399904e-05,
+      "loss": 0.8843,
+      "step": 1848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4428389337508905,
+      "learning_rate": 1.6995482246583215e-05,
+      "loss": 0.8757,
+      "step": 1849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.428385144633288,
+      "learning_rate": 1.699202768162117e-05,
+      "loss": 0.9636,
+      "step": 1850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7392542687475825,
+      "learning_rate": 1.6988571483320767e-05,
+      "loss": 0.9013,
+      "step": 1851
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.472283720333911,
+      "learning_rate": 1.6985113652489374e-05,
+      "loss": 0.9108,
+      "step": 1852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6345929022663972,
+      "learning_rate": 1.698165418993473e-05,
+      "loss": 0.8426,
+      "step": 1853
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.545556794572305,
+      "learning_rate": 1.697819309646497e-05,
+      "loss": 0.9063,
+      "step": 1854
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9339432221890696,
+      "learning_rate": 1.697473037288861e-05,
+      "loss": 0.3542,
+      "step": 1855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.393912084417808,
+      "learning_rate": 1.697126602001454e-05,
+      "loss": 0.9832,
+      "step": 1856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6237529786345433,
+      "learning_rate": 1.6967800038652035e-05,
+      "loss": 0.9168,
+      "step": 1857
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4129856900534066,
+      "learning_rate": 1.6964332429610747e-05,
+      "loss": 0.9548,
+      "step": 1858
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5444319571559653,
+      "learning_rate": 1.696086319370071e-05,
+      "loss": 0.9455,
+      "step": 1859
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.397181378998314,
+      "learning_rate": 1.695739233173233e-05,
+      "loss": 0.9599,
+      "step": 1860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3144805233666212,
+      "learning_rate": 1.6953919844516415e-05,
+      "loss": 0.8988,
+      "step": 1861
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4844170313455844,
+      "learning_rate": 1.695044573286413e-05,
+      "loss": 0.9151,
+      "step": 1862
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6429506575694284,
+      "learning_rate": 1.694696999758703e-05,
+      "loss": 0.9013,
+      "step": 1863
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.462834925554896,
+      "learning_rate": 1.6943492639497044e-05,
+      "loss": 0.8297,
+      "step": 1864
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3801884498303576,
+      "learning_rate": 1.6940013659406492e-05,
+      "loss": 0.9058,
+      "step": 1865
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.486277465984611,
+      "learning_rate": 1.693653305812805e-05,
+      "loss": 0.8542,
+      "step": 1866
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7762073614813474,
+      "learning_rate": 1.69330508364748e-05,
+      "loss": 0.9188,
+      "step": 1867
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3483277679563481,
+      "learning_rate": 1.6929566995260184e-05,
+      "loss": 0.8309,
+      "step": 1868
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5490639111449134,
+      "learning_rate": 1.692608153529802e-05,
+      "loss": 0.9155,
+      "step": 1869
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4626550077248273,
+      "learning_rate": 1.6922594457402528e-05,
+      "loss": 0.9275,
+      "step": 1870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6728514344526555,
+      "learning_rate": 1.691910576238828e-05,
+      "loss": 0.9352,
+      "step": 1871
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6905296423949694,
+      "learning_rate": 1.6915615451070234e-05,
+      "loss": 0.8731,
+      "step": 1872
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4445955044540189,
+      "learning_rate": 1.691212352426373e-05,
+      "loss": 0.9092,
+      "step": 1873
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5379695811991725,
+      "learning_rate": 1.690862998278448e-05,
+      "loss": 0.8341,
+      "step": 1874
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.483547024947764,
+      "learning_rate": 1.690513482744858e-05,
+      "loss": 0.8795,
+      "step": 1875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6764891403089985,
+      "learning_rate": 1.69016380590725e-05,
+      "loss": 0.8803,
+      "step": 1876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4931952259312042,
+      "learning_rate": 1.689813967847308e-05,
+      "loss": 0.904,
+      "step": 1877
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3735944278809964,
+      "learning_rate": 1.689463968646754e-05,
+      "loss": 0.9055,
+      "step": 1878
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6314148550799485,
+      "learning_rate": 1.6891138083873486e-05,
+      "loss": 0.8594,
+      "step": 1879
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4852944237066374,
+      "learning_rate": 1.688763487150889e-05,
+      "loss": 0.8542,
+      "step": 1880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.543406465469015,
+      "learning_rate": 1.6884130050192098e-05,
+      "loss": 0.9442,
+      "step": 1881
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5159666927316846,
+      "learning_rate": 1.6880623620741843e-05,
+      "loss": 0.9213,
+      "step": 1882
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4719725711640013,
+      "learning_rate": 1.6877115583977225e-05,
+      "loss": 0.9367,
+      "step": 1883
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7933370133062927,
+      "learning_rate": 1.687360594071772e-05,
+      "loss": 0.8699,
+      "step": 1884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.040836183674288,
+      "learning_rate": 1.6870094691783182e-05,
+      "loss": 0.3144,
+      "step": 1885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5923834117952378,
+      "learning_rate": 1.6866581837993842e-05,
+      "loss": 0.9466,
+      "step": 1886
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4085024814425162,
+      "learning_rate": 1.68630673801703e-05,
+      "loss": 0.8911,
+      "step": 1887
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.635650363451425,
+      "learning_rate": 1.6859551319133534e-05,
+      "loss": 0.884,
+      "step": 1888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.56261987464738,
+      "learning_rate": 1.6856033655704894e-05,
+      "loss": 0.8434,
+      "step": 1889
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.428811357245329,
+      "learning_rate": 1.685251439070611e-05,
+      "loss": 0.8943,
+      "step": 1890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.464265757652771,
+      "learning_rate": 1.6848993524959286e-05,
+      "loss": 0.9075,
+      "step": 1891
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6909022505379656,
+      "learning_rate": 1.684547105928689e-05,
+      "loss": 0.9051,
+      "step": 1892
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6575890573411753,
+      "learning_rate": 1.684194699451177e-05,
+      "loss": 0.9504,
+      "step": 1893
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4531623210011089,
+      "learning_rate": 1.6838421331457154e-05,
+      "loss": 0.9892,
+      "step": 1894
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6826652280886958,
+      "learning_rate": 1.683489407094663e-05,
+      "loss": 0.9738,
+      "step": 1895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5130114363566207,
+      "learning_rate": 1.683136521380417e-05,
+      "loss": 0.9192,
+      "step": 1896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4557444032686035,
+      "learning_rate": 1.682783476085412e-05,
+      "loss": 0.9928,
+      "step": 1897
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6474761436452572,
+      "learning_rate": 1.6824302712921187e-05,
+      "loss": 0.8659,
+      "step": 1898
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4207912548841013,
+      "learning_rate": 1.682076907083046e-05,
+      "loss": 0.8705,
+      "step": 1899
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3852045086029119,
+      "learning_rate": 1.68172338354074e-05,
+      "loss": 0.9266,
+      "step": 1900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.437722974696888,
+      "learning_rate": 1.6813697007477837e-05,
+      "loss": 0.9057,
+      "step": 1901
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.380899724815938,
+      "learning_rate": 1.6810158587867973e-05,
+      "loss": 0.9064,
+      "step": 1902
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.468761736995389,
+      "learning_rate": 1.6806618577404385e-05,
+      "loss": 0.9327,
+      "step": 1903
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5475257428326372,
+      "learning_rate": 1.6803076976914018e-05,
+      "loss": 0.9255,
+      "step": 1904
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4974052044671982,
+      "learning_rate": 1.6799533787224192e-05,
+      "loss": 0.9057,
+      "step": 1905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5713979886021696,
+      "learning_rate": 1.67959890091626e-05,
+      "loss": 0.9424,
+      "step": 1906
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2893939032962567,
+      "learning_rate": 1.679244264355729e-05,
+      "loss": 0.9293,
+      "step": 1907
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4765156418061116,
+      "learning_rate": 1.678889469123671e-05,
+      "loss": 0.8976,
+      "step": 1908
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.509181420996771,
+      "learning_rate": 1.6785345153029648e-05,
+      "loss": 0.365,
+      "step": 1909
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.656713904896557,
+      "learning_rate": 1.678179402976529e-05,
+      "loss": 0.8678,
+      "step": 1910
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3995329506226941,
+      "learning_rate": 1.6778241322273163e-05,
+      "loss": 0.8212,
+      "step": 1911
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5147949580754596,
+      "learning_rate": 1.677468703138319e-05,
+      "loss": 0.8517,
+      "step": 1912
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4262386346100058,
+      "learning_rate": 1.677113115792565e-05,
+      "loss": 0.8621,
+      "step": 1913
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5044534319447298,
+      "learning_rate": 1.6767573702731203e-05,
+      "loss": 0.8743,
+      "step": 1914
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3800901404342816,
+      "learning_rate": 1.676401466663086e-05,
+      "loss": 0.891,
+      "step": 1915
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5210445169789544,
+      "learning_rate": 1.676045405045602e-05,
+      "loss": 0.8876,
+      "step": 1916
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4724340581103998,
+      "learning_rate": 1.6756891855038436e-05,
+      "loss": 0.9291,
+      "step": 1917
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4764706883320164,
+      "learning_rate": 1.6753328081210244e-05,
+      "loss": 0.8795,
+      "step": 1918
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4618439501485787,
+      "learning_rate": 1.6749762729803943e-05,
+      "loss": 0.8933,
+      "step": 1919
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3946471980855777,
+      "learning_rate": 1.6746195801652393e-05,
+      "loss": 0.882,
+      "step": 1920
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5934232950774414,
+      "learning_rate": 1.674262729758883e-05,
+      "loss": 0.9255,
+      "step": 1921
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.591363392062247,
+      "learning_rate": 1.673905721844686e-05,
+      "loss": 0.898,
+      "step": 1922
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7081082566979897,
+      "learning_rate": 1.673548556506045e-05,
+      "loss": 0.904,
+      "step": 1923
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.776634681656157,
+      "learning_rate": 1.6731912338263943e-05,
+      "loss": 0.9289,
+      "step": 1924
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.489714214242666,
+      "learning_rate": 1.6728337538892043e-05,
+      "loss": 0.8945,
+      "step": 1925
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4288390220179836,
+      "learning_rate": 1.6724761167779825e-05,
+      "loss": 0.9229,
+      "step": 1926
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6201175877374947,
+      "learning_rate": 1.6721183225762726e-05,
+      "loss": 0.8427,
+      "step": 1927
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4639948877642863,
+      "learning_rate": 1.6717603713676557e-05,
+      "loss": 0.9087,
+      "step": 1928
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3851772323434917,
+      "learning_rate": 1.6714022632357495e-05,
+      "loss": 0.9229,
+      "step": 1929
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4888210963311344,
+      "learning_rate": 1.671043998264207e-05,
+      "loss": 0.9124,
+      "step": 1930
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.573404158169826,
+      "learning_rate": 1.6706855765367202e-05,
+      "loss": 0.9532,
+      "step": 1931
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3049727359363632,
+      "learning_rate": 1.670326998137016e-05,
+      "loss": 0.9143,
+      "step": 1932
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4639621599473185,
+      "learning_rate": 1.6699682631488578e-05,
+      "loss": 0.8668,
+      "step": 1933
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5249654095422258,
+      "learning_rate": 1.6696093716560466e-05,
+      "loss": 0.924,
+      "step": 1934
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4494831937704733,
+      "learning_rate": 1.6692503237424197e-05,
+      "loss": 0.9259,
+      "step": 1935
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5834105883487717,
+      "learning_rate": 1.6688911194918506e-05,
+      "loss": 0.9022,
+      "step": 1936
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5730013044324458,
+      "learning_rate": 1.668531758988249e-05,
+      "loss": 0.9607,
+      "step": 1937
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4828482593911496,
+      "learning_rate": 1.668172242315562e-05,
+      "loss": 0.9112,
+      "step": 1938
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5490312659547139,
+      "learning_rate": 1.667812569557773e-05,
+      "loss": 0.9085,
+      "step": 1939
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4382818676865894,
+      "learning_rate": 1.667452740798901e-05,
+      "loss": 0.8739,
+      "step": 1940
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3282378534446506,
+      "learning_rate": 1.6670927561230018e-05,
+      "loss": 0.8595,
+      "step": 1941
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4965124751942585,
+      "learning_rate": 1.666732615614169e-05,
+      "loss": 0.85,
+      "step": 1942
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4468147885809837,
+      "learning_rate": 1.6663723193565308e-05,
+      "loss": 0.8229,
+      "step": 1943
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7302487989187791,
+      "learning_rate": 1.666011867434252e-05,
+      "loss": 1.0199,
+      "step": 1944
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4320928092384664,
+      "learning_rate": 1.6656512599315348e-05,
+      "loss": 0.875,
+      "step": 1945
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6760163575753484,
+      "learning_rate": 1.6652904969326167e-05,
+      "loss": 0.92,
+      "step": 1946
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5269023069511656,
+      "learning_rate": 1.6649295785217722e-05,
+      "loss": 0.904,
+      "step": 1947
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3718225900332046,
+      "learning_rate": 1.6645685047833124e-05,
+      "loss": 0.7822,
+      "step": 1948
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5824897596177077,
+      "learning_rate": 1.6642072758015834e-05,
+      "loss": 0.9779,
+      "step": 1949
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5332858506742733,
+      "learning_rate": 1.6638458916609685e-05,
+      "loss": 0.9419,
+      "step": 1950
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5125142402215015,
+      "learning_rate": 1.6634843524458874e-05,
+      "loss": 0.9616,
+      "step": 1951
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5669749706776952,
+      "learning_rate": 1.6631226582407954e-05,
+      "loss": 0.9496,
+      "step": 1952
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.189677675314092,
+      "learning_rate": 1.6627608091301842e-05,
+      "loss": 0.3662,
+      "step": 1953
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3373281438103897,
+      "learning_rate": 1.6623988051985823e-05,
+      "loss": 0.8457,
+      "step": 1954
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5230997420295853,
+      "learning_rate": 1.6620366465305533e-05,
+      "loss": 0.9122,
+      "step": 1955
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.377707305958102,
+      "learning_rate": 1.6616743332106976e-05,
+      "loss": 0.8796,
+      "step": 1956
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.551055903099349,
+      "learning_rate": 1.661311865323652e-05,
+      "loss": 0.8674,
+      "step": 1957
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6513766351087809,
+      "learning_rate": 1.660949242954089e-05,
+      "loss": 0.9343,
+      "step": 1958
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4462516742486833,
+      "learning_rate": 1.6605864661867165e-05,
+      "loss": 0.9126,
+      "step": 1959
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6598395864626054,
+      "learning_rate": 1.6602235351062797e-05,
+      "loss": 0.9657,
+      "step": 1960
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.43272870194693,
+      "learning_rate": 1.6598604497975598e-05,
+      "loss": 0.8766,
+      "step": 1961
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.685255448154483,
+      "learning_rate": 1.6594972103453727e-05,
+      "loss": 0.8856,
+      "step": 1962
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5203105048786965,
+      "learning_rate": 1.6591338168345713e-05,
+      "loss": 0.9465,
+      "step": 1963
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.430726407681374,
+      "learning_rate": 1.6587702693500452e-05,
+      "loss": 0.9138,
+      "step": 1964
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.413927142240461,
+      "learning_rate": 1.6584065679767186e-05,
+      "loss": 0.9046,
+      "step": 1965
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.582377967988086,
+      "learning_rate": 1.6580427127995516e-05,
+      "loss": 0.904,
+      "step": 1966
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9117864281076555,
+      "learning_rate": 1.6576787039035417e-05,
+      "loss": 0.3333,
+      "step": 1967
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9696745480283004,
+      "learning_rate": 1.657314541373721e-05,
+      "loss": 0.8769,
+      "step": 1968
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7704686283082576,
+      "learning_rate": 1.656950225295158e-05,
+      "loss": 0.9051,
+      "step": 1969
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4378779539002324,
+      "learning_rate": 1.6565857557529567e-05,
+      "loss": 0.8737,
+      "step": 1970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4285185743723405,
+      "learning_rate": 1.6562211328322576e-05,
+      "loss": 0.8525,
+      "step": 1971
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.204406636983857,
+      "learning_rate": 1.6558563566182365e-05,
+      "loss": 0.9461,
+      "step": 1972
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3545335228408402,
+      "learning_rate": 1.6554914271961047e-05,
+      "loss": 0.8799,
+      "step": 1973
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4680363425345486,
+      "learning_rate": 1.655126344651111e-05,
+      "loss": 0.8849,
+      "step": 1974
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2834420286597472,
+      "learning_rate": 1.6547611090685378e-05,
+      "loss": 0.9061,
+      "step": 1975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4435875432256455,
+      "learning_rate": 1.6543957205337034e-05,
+      "loss": 0.9571,
+      "step": 1976
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6255764958791408,
+      "learning_rate": 1.6540301791319647e-05,
+      "loss": 0.7883,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7920660029703597,
+      "learning_rate": 1.6536644849487104e-05,
+      "loss": 0.9691,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4952383434558507,
+      "learning_rate": 1.6532986380693673e-05,
+      "loss": 0.8549,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2664267817179569,
+      "learning_rate": 1.6529326385793972e-05,
+      "loss": 0.9409,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3638323603114864,
+      "learning_rate": 1.6525664865642978e-05,
+      "loss": 0.9198,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6703378460444627,
+      "learning_rate": 1.652200182109602e-05,
+      "loss": 0.9389,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6574802103414707,
+      "learning_rate": 1.651833725300879e-05,
+      "loss": 0.8769,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3356220049895915,
+      "learning_rate": 1.6514671162237327e-05,
+      "loss": 0.9444,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.493508440800496,
+      "learning_rate": 1.651100354963803e-05,
+      "loss": 0.9429,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6090742543143652,
+      "learning_rate": 1.6507334416067656e-05,
+      "loss": 0.8426,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.603834803476885,
+      "learning_rate": 1.6503663762383312e-05,
+      "loss": 0.8728,
+      "step": 1987
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6670319955248305,
+      "learning_rate": 1.649999158944247e-05,
+      "loss": 0.8964,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4012671152463971,
+      "learning_rate": 1.6496317898102942e-05,
+      "loss": 0.9463,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4078083483616775,
+      "learning_rate": 1.649264268922291e-05,
+      "loss": 0.8795,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3180993007284014,
+      "learning_rate": 1.6488965963660892e-05,
+      "loss": 0.8524,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5562919779079722,
+      "learning_rate": 1.6485287722275783e-05,
+      "loss": 0.8819,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3797329904329805,
+      "learning_rate": 1.6481607965926812e-05,
+      "loss": 0.921,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4271421671983215,
+      "learning_rate": 1.647792669547358e-05,
+      "loss": 0.8914,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.636940227013142,
+      "learning_rate": 1.6474243911776026e-05,
+      "loss": 0.8372,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.387700690917425,
+      "learning_rate": 1.6470559615694445e-05,
+      "loss": 0.8901,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6221434503794132,
+      "learning_rate": 1.6466873808089496e-05,
+      "loss": 0.9761,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4959770809528985,
+      "learning_rate": 1.646318648982218e-05,
+      "loss": 0.8258,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4325709715801538,
+      "learning_rate": 1.6459497661753857e-05,
+      "loss": 0.9125,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0230319533656353,
+      "learning_rate": 1.6455807324746237e-05,
+      "loss": 0.3341,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3686508616923576,
+      "learning_rate": 1.645211547966138e-05,
+      "loss": 0.9118,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4022161138780582,
+      "learning_rate": 1.6448422127361707e-05,
+      "loss": 0.9522,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3899670541906008,
+      "learning_rate": 1.6444727268709984e-05,
+      "loss": 0.9214,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.45606700206336,
+      "learning_rate": 1.6441030904569327e-05,
+      "loss": 0.9148,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3448561195289104,
+      "learning_rate": 1.6437333035803208e-05,
+      "loss": 0.8743,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5715648819197445,
+      "learning_rate": 1.6433633663275453e-05,
+      "loss": 0.9288,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5561307050375306,
+      "learning_rate": 1.642993278785023e-05,
+      "loss": 0.8729,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.540332567866295,
+      "learning_rate": 1.642623041039207e-05,
+      "loss": 0.866,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6223774894138585,
+      "learning_rate": 1.6422526531765846e-05,
+      "loss": 0.8024,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3839531769548294,
+      "learning_rate": 1.6418821152836782e-05,
+      "loss": 0.9578,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3932875682646482,
+      "learning_rate": 1.641511427447046e-05,
+      "loss": 0.8685,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.472219488403293,
+      "learning_rate": 1.64114058975328e-05,
+      "loss": 0.9228,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6950888211491755,
+      "learning_rate": 1.640769602289009e-05,
+      "loss": 0.9212,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4453771525176202,
+      "learning_rate": 1.6403984651408947e-05,
+      "loss": 0.9402,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4901681519089356,
+      "learning_rate": 1.6400271783956352e-05,
+      "loss": 0.9915,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2876539539908727,
+      "learning_rate": 1.6396557421399634e-05,
+      "loss": 0.8771,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9135175062444185,
+      "learning_rate": 1.639284156460646e-05,
+      "loss": 0.3526,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4623818518678788,
+      "learning_rate": 1.638912421444486e-05,
+      "loss": 0.9509,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6777871454441866,
+      "learning_rate": 1.638540537178321e-05,
+      "loss": 0.963,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2438738026296454,
+      "learning_rate": 1.6381685037490225e-05,
+      "loss": 0.9235,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5426780390603687,
+      "learning_rate": 1.6377963212434982e-05,
+      "loss": 0.8005,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.50490739494114,
+      "learning_rate": 1.63742398974869e-05,
+      "loss": 0.8412,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.44441824195559,
+      "learning_rate": 1.637051509351574e-05,
+      "loss": 0.9227,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6153868732398677,
+      "learning_rate": 1.6366788801391618e-05,
+      "loss": 0.9126,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4472502766823665,
+      "learning_rate": 1.6363061021984997e-05,
+      "loss": 0.8415,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5193502531641168,
+      "learning_rate": 1.6359331756166694e-05,
+      "loss": 0.8252,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4973435945659064,
+      "learning_rate": 1.6355601004807856e-05,
+      "loss": 0.88,
+      "step": 2027
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4215879189794591,
+      "learning_rate": 1.635186876877999e-05,
+      "loss": 0.8839,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4100515721030893,
+      "learning_rate": 1.6348135048954943e-05,
+      "loss": 0.9207,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3975214042084307,
+      "learning_rate": 1.6344399846204918e-05,
+      "loss": 0.7465,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7076184737981555,
+      "learning_rate": 1.634066316140246e-05,
+      "loss": 0.855,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6985969392291949,
+      "learning_rate": 1.6336924995420453e-05,
+      "loss": 0.9418,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3544998120600074,
+      "learning_rate": 1.6333185349132138e-05,
+      "loss": 0.8925,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7049199999956408,
+      "learning_rate": 1.632944422341109e-05,
+      "loss": 0.8878,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4448598436383564,
+      "learning_rate": 1.6325701619131246e-05,
+      "loss": 0.9263,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5326367079112837,
+      "learning_rate": 1.632195753716687e-05,
+      "loss": 0.8386,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.66287064852823,
+      "learning_rate": 1.6318211978392588e-05,
+      "loss": 0.9434,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3421603172866972,
+      "learning_rate": 1.6314464943683353e-05,
+      "loss": 0.8913,
+      "step": 2038
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5574219046124602,
+      "learning_rate": 1.631071643391448e-05,
+      "loss": 0.8685,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.486758911574603,
+      "learning_rate": 1.6306966449961623e-05,
+      "loss": 0.932,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3878371369587263,
+      "learning_rate": 1.6303214992700773e-05,
+      "loss": 0.9122,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4605312394526726,
+      "learning_rate": 1.6299462063008272e-05,
+      "loss": 0.8802,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2705456378416817,
+      "learning_rate": 1.6295707661760804e-05,
+      "loss": 0.8642,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5219381216498384,
+      "learning_rate": 1.62919517898354e-05,
+      "loss": 0.8679,
+      "step": 2044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.874751290350141,
+      "learning_rate": 1.6288194448109433e-05,
+      "loss": 0.3563,
+      "step": 2045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.628572545871969,
+      "learning_rate": 1.6284435637460613e-05,
+      "loss": 0.8588,
+      "step": 2046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4918264999654482,
+      "learning_rate": 1.6280675358767005e-05,
+      "loss": 0.9225,
+      "step": 2047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2983771595321323,
+      "learning_rate": 1.6276913612907005e-05,
+      "loss": 0.8829,
+      "step": 2048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4042975521218652,
+      "learning_rate": 1.6273150400759363e-05,
+      "loss": 0.9553,
+      "step": 2049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3670530191766468,
+      "learning_rate": 1.626938572320316e-05,
+      "loss": 0.9586,
+      "step": 2050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7118950046842918,
+      "learning_rate": 1.6265619581117827e-05,
+      "loss": 0.8918,
+      "step": 2051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6059810860474364,
+      "learning_rate": 1.626185197538314e-05,
+      "loss": 0.8534,
+      "step": 2052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4542421346009562,
+      "learning_rate": 1.6258082906879203e-05,
+      "loss": 0.8438,
+      "step": 2053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3872075777411936,
+      "learning_rate": 1.6254312376486478e-05,
+      "loss": 0.8785,
+      "step": 2054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5172539098330657,
+      "learning_rate": 1.6250540385085754e-05,
+      "loss": 0.8844,
+      "step": 2055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3844475037746036,
+      "learning_rate": 1.624676693355818e-05,
+      "loss": 0.9405,
+      "step": 2056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7606453702335263,
+      "learning_rate": 1.6242992022785225e-05,
+      "loss": 0.8863,
+      "step": 2057
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7003422432064619,
+      "learning_rate": 1.623921565364871e-05,
+      "loss": 0.8944,
+      "step": 2058
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7130310203318548,
+      "learning_rate": 1.62354378270308e-05,
+      "loss": 0.8927,
+      "step": 2059
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5905434793199673,
+      "learning_rate": 1.6231658543813994e-05,
+      "loss": 0.8592,
+      "step": 2060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2917759289757775,
+      "learning_rate": 1.6227877804881126e-05,
+      "loss": 0.8089,
+      "step": 2061
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.439521879782407,
+      "learning_rate": 1.6224095611115385e-05,
+      "loss": 0.8094,
+      "step": 2062
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3578887301294011,
+      "learning_rate": 1.622031196340029e-05,
+      "loss": 0.9578,
+      "step": 2063
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4755677595098133,
+      "learning_rate": 1.62165268626197e-05,
+      "loss": 0.9084,
+      "step": 2064
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.377149178365839,
+      "learning_rate": 1.6212740309657814e-05,
+      "loss": 0.9298,
+      "step": 2065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4730742693498102,
+      "learning_rate": 1.6208952305399175e-05,
+      "loss": 0.9347,
+      "step": 2066
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5585505293109423,
+      "learning_rate": 1.620516285072866e-05,
+      "loss": 1.028,
+      "step": 2067
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4727086391793578,
+      "learning_rate": 1.6201371946531483e-05,
+      "loss": 0.8647,
+      "step": 2068
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5238672200605055,
+      "learning_rate": 1.6197579593693197e-05,
+      "loss": 0.8957,
+      "step": 2069
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.433432135970078,
+      "learning_rate": 1.6193785793099706e-05,
+      "loss": 0.9113,
+      "step": 2070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3850307768961654,
+      "learning_rate": 1.6189990545637234e-05,
+      "loss": 0.87,
+      "step": 2071
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6087341363590808,
+      "learning_rate": 1.6186193852192356e-05,
+      "loss": 0.8862,
+      "step": 2072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5855787381158655,
+      "learning_rate": 1.618239571365198e-05,
+      "loss": 0.8915,
+      "step": 2073
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5917248873434566,
+      "learning_rate": 1.6178596130903345e-05,
+      "loss": 0.84,
+      "step": 2074
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5559401210334554,
+      "learning_rate": 1.6174795104834042e-05,
+      "loss": 0.9011,
+      "step": 2075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4147010363941737,
+      "learning_rate": 1.6170992636331983e-05,
+      "loss": 0.9283,
+      "step": 2076
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.468813100940576,
+      "learning_rate": 1.6167188726285433e-05,
+      "loss": 0.8916,
+      "step": 2077
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.346034853427754,
+      "learning_rate": 1.6163383375582983e-05,
+      "loss": 0.8515,
+      "step": 2078
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6959678601645014,
+      "learning_rate": 1.6159576585113556e-05,
+      "loss": 0.8427,
+      "step": 2079
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.494265976179155,
+      "learning_rate": 1.615576835576643e-05,
+      "loss": 0.9348,
+      "step": 2080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5296318118583636,
+      "learning_rate": 1.6151958688431204e-05,
+      "loss": 0.8514,
+      "step": 2081
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5943274760743165,
+      "learning_rate": 1.6148147583997813e-05,
+      "loss": 0.9528,
+      "step": 2082
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4207516912340847,
+      "learning_rate": 1.6144335043356533e-05,
+      "loss": 0.8594,
+      "step": 2083
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6194334490109523,
+      "learning_rate": 1.6140521067397978e-05,
+      "loss": 0.8966,
+      "step": 2084
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7309903549193382,
+      "learning_rate": 1.613670565701309e-05,
+      "loss": 0.9132,
+      "step": 2085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4828829054704131,
+      "learning_rate": 1.6132888813093147e-05,
+      "loss": 0.8547,
+      "step": 2086
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4658897899348529,
+      "learning_rate": 1.6129070536529767e-05,
+      "loss": 0.8575,
+      "step": 2087
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5753911565320524,
+      "learning_rate": 1.6125250828214897e-05,
+      "loss": 0.9427,
+      "step": 2088
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8613786550429863,
+      "learning_rate": 1.6121429689040825e-05,
+      "loss": 0.9581,
+      "step": 2089
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.536593823331853,
+      "learning_rate": 1.6117607119900172e-05,
+      "loss": 0.841,
+      "step": 2090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.548912047268822,
+      "learning_rate": 1.6113783121685883e-05,
+      "loss": 0.8924,
+      "step": 2091
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.39243288540068,
+      "learning_rate": 1.6109957695291246e-05,
+      "loss": 0.9371,
+      "step": 2092
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3766790753082239,
+      "learning_rate": 1.6106130841609883e-05,
+      "loss": 0.9236,
+      "step": 2093
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5224816490364808,
+      "learning_rate": 1.6102302561535748e-05,
+      "loss": 0.9086,
+      "step": 2094
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3929197673454525,
+      "learning_rate": 1.6098472855963126e-05,
+      "loss": 0.9315,
+      "step": 2095
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5478797837822456,
+      "learning_rate": 1.609464172578664e-05,
+      "loss": 0.9094,
+      "step": 2096
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8001229216489654,
+      "learning_rate": 1.6090809171901237e-05,
+      "loss": 0.9044,
+      "step": 2097
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4322805067552384,
+      "learning_rate": 1.6086975195202207e-05,
+      "loss": 0.8694,
+      "step": 2098
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5616684870720325,
+      "learning_rate": 1.608313979658516e-05,
+      "loss": 0.8225,
+      "step": 2099
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6133765464974652,
+      "learning_rate": 1.6079302976946055e-05,
+      "loss": 0.8523,
+      "step": 2100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4668609215307034,
+      "learning_rate": 1.607546473718117e-05,
+      "loss": 0.8671,
+      "step": 2101
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6183739055890012,
+      "learning_rate": 1.6071625078187113e-05,
+      "loss": 0.8829,
+      "step": 2102
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5109289204325285,
+      "learning_rate": 1.6067784000860838e-05,
+      "loss": 0.8995,
+      "step": 2103
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4755414530665127,
+      "learning_rate": 1.606394150609961e-05,
+      "loss": 0.8562,
+      "step": 2104
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4392717478803918,
+      "learning_rate": 1.6060097594801044e-05,
+      "loss": 0.8945,
+      "step": 2105
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4674829321178218,
+      "learning_rate": 1.605625226786308e-05,
+      "loss": 0.8958,
+      "step": 2106
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.471457028520947,
+      "learning_rate": 1.605240552618398e-05,
+      "loss": 0.9222,
+      "step": 2107
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5716153743086634,
+      "learning_rate": 1.6048557370662346e-05,
+      "loss": 0.9905,
+      "step": 2108
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6040361578266755,
+      "learning_rate": 1.6044707802197106e-05,
+      "loss": 0.9642,
+      "step": 2109
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.592093355850832,
+      "learning_rate": 1.6040856821687523e-05,
+      "loss": 0.8493,
+      "step": 2110
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4991844010058457,
+      "learning_rate": 1.6037004430033186e-05,
+      "loss": 0.8945,
+      "step": 2111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4009151194782519,
+      "learning_rate": 1.603315062813401e-05,
+      "loss": 0.8817,
+      "step": 2112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5971079634027765,
+      "learning_rate": 1.602929541689025e-05,
+      "loss": 0.8554,
+      "step": 2113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.392574563764216,
+      "learning_rate": 1.6025438797202478e-05,
+      "loss": 0.8991,
+      "step": 2114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4263051285524602,
+      "learning_rate": 1.6021580769971602e-05,
+      "loss": 0.855,
+      "step": 2115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8538281078218202,
+      "learning_rate": 1.601772133609886e-05,
+      "loss": 0.346,
+      "step": 2116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.509314583177348,
+      "learning_rate": 1.601386049648581e-05,
+      "loss": 0.8159,
+      "step": 2117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3356096816378258,
+      "learning_rate": 1.6009998252034354e-05,
+      "loss": 0.9456,
+      "step": 2118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4564987844734394,
+      "learning_rate": 1.6006134603646706e-05,
+      "loss": 0.8351,
+      "step": 2119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.610159923284599,
+      "learning_rate": 1.6002269552225413e-05,
+      "loss": 0.9145,
+      "step": 2120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3321514865440012,
+      "learning_rate": 1.599840309867336e-05,
+      "loss": 0.915,
+      "step": 2121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4104962995637664,
+      "learning_rate": 1.5994535243893742e-05,
+      "loss": 0.9151,
+      "step": 2122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7997541167092745,
+      "learning_rate": 1.5990665988790094e-05,
+      "loss": 0.3586,
+      "step": 2123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.446492106823676,
+      "learning_rate": 1.5986795334266276e-05,
+      "loss": 0.8006,
+      "step": 2124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3127292684003837,
+      "learning_rate": 1.598292328122647e-05,
+      "loss": 0.8736,
+      "step": 2125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4516478241910866,
+      "learning_rate": 1.597904983057519e-05,
+      "loss": 0.8844,
+      "step": 2126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.459318812083033,
+      "learning_rate": 1.5975174983217273e-05,
+      "loss": 0.9549,
+      "step": 2127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3856462911876546,
+      "learning_rate": 1.5971298740057885e-05,
+      "loss": 0.9146,
+      "step": 2128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5500762751966088,
+      "learning_rate": 1.596742110200252e-05,
+      "loss": 0.9535,
+      "step": 2129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4467626410621395,
+      "learning_rate": 1.596354206995699e-05,
+      "loss": 0.9051,
+      "step": 2130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5381992292308826,
+      "learning_rate": 1.5959661644827432e-05,
+      "loss": 0.8696,
+      "step": 2131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.406221321926157,
+      "learning_rate": 1.5955779827520327e-05,
+      "loss": 0.8703,
+      "step": 2132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3309496763287714,
+      "learning_rate": 1.595189661894246e-05,
+      "loss": 0.9135,
+      "step": 2133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3602130857750188,
+      "learning_rate": 1.5948012020000948e-05,
+      "loss": 0.9501,
+      "step": 2134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.384241937255165,
+      "learning_rate": 1.5944126031603236e-05,
+      "loss": 0.8658,
+      "step": 2135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.383323932484943,
+      "learning_rate": 1.594023865465709e-05,
+      "loss": 0.9245,
+      "step": 2136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4038502416652323,
+      "learning_rate": 1.5936349890070602e-05,
+      "loss": 0.9337,
+      "step": 2137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5538719137012444,
+      "learning_rate": 1.5932459738752194e-05,
+      "loss": 0.828,
+      "step": 2138
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5704005392266531,
+      "learning_rate": 1.5928568201610593e-05,
+      "loss": 0.8224,
+      "step": 2139
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.326320690406751,
+      "learning_rate": 1.5924675279554874e-05,
+      "loss": 0.8088,
+      "step": 2140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4552450717930803,
+      "learning_rate": 1.5920780973494418e-05,
+      "loss": 0.8675,
+      "step": 2141
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5900269726142606,
+      "learning_rate": 1.5916885284338937e-05,
+      "loss": 0.8273,
+      "step": 2142
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5908164661845712,
+      "learning_rate": 1.591298821299846e-05,
+      "loss": 0.9057,
+      "step": 2143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4337178515683364,
+      "learning_rate": 1.5909089760383354e-05,
+      "loss": 0.8492,
+      "step": 2144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.523661560307685,
+      "learning_rate": 1.590518992740429e-05,
+      "loss": 0.8446,
+      "step": 2145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5152220608239626,
+      "learning_rate": 1.5901288714972268e-05,
+      "loss": 0.9108,
+      "step": 2146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6261242223827643,
+      "learning_rate": 1.5897386123998613e-05,
+      "loss": 0.9179,
+      "step": 2147
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3758853999925584,
+      "learning_rate": 1.5893482155394978e-05,
+      "loss": 0.9105,
+      "step": 2148
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8682760707058219,
+      "learning_rate": 1.588957681007332e-05,
+      "loss": 0.3252,
+      "step": 2149
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5432027084094335,
+      "learning_rate": 1.588567008894593e-05,
+      "loss": 0.8688,
+      "step": 2150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4288536946959483,
+      "learning_rate": 1.5881761992925425e-05,
+      "loss": 0.931,
+      "step": 2151
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2984264276046191,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.8444,
+      "step": 2152
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7091369549894575,
+      "learning_rate": 1.5873941679857107e-05,
+      "loss": 0.9633,
+      "step": 2153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3813193929510699,
+      "learning_rate": 1.5870029464636113e-05,
+      "loss": 0.8091,
+      "step": 2154
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3125478609475392,
+      "learning_rate": 1.586611587817566e-05,
+      "loss": 0.891,
+      "step": 2155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4716689249264063,
+      "learning_rate": 1.5862200921389947e-05,
+      "loss": 0.9326,
+      "step": 2156
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4523612945773285,
+      "learning_rate": 1.5858284595193514e-05,
+      "loss": 0.8674,
+      "step": 2157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3363962333961232,
+      "learning_rate": 1.585436690050122e-05,
+      "loss": 0.9189,
+      "step": 2158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8834030562403844,
+      "learning_rate": 1.5850447838228235e-05,
+      "loss": 0.3528,
+      "step": 2159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.421970990367768,
+      "learning_rate": 1.584652740929005e-05,
+      "loss": 0.9381,
+      "step": 2160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5336996161463692,
+      "learning_rate": 1.5842605614602482e-05,
+      "loss": 0.9464,
+      "step": 2161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2243795027048836,
+      "learning_rate": 1.5838682455081657e-05,
+      "loss": 0.8881,
+      "step": 2162
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4865764981850325,
+      "learning_rate": 1.583475793164403e-05,
+      "loss": 0.9292,
+      "step": 2163
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4753815050477015,
+      "learning_rate": 1.583083204520637e-05,
+      "loss": 0.8603,
+      "step": 2164
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2964147388713199,
+      "learning_rate": 1.5826904796685763e-05,
+      "loss": 0.8703,
+      "step": 2165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5292446857136408,
+      "learning_rate": 1.582297618699961e-05,
+      "loss": 0.7957,
+      "step": 2166
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5148288562784187,
+      "learning_rate": 1.581904621706565e-05,
+      "loss": 0.8397,
+      "step": 2167
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6078954557551346,
+      "learning_rate": 1.581511488780191e-05,
+      "loss": 0.8936,
+      "step": 2168
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5709729762820952,
+      "learning_rate": 1.581118220012675e-05,
+      "loss": 0.8943,
+      "step": 2169
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6692647232106506,
+      "learning_rate": 1.5807248154958848e-05,
+      "loss": 0.8169,
+      "step": 2170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4049933196100746,
+      "learning_rate": 1.5803312753217202e-05,
+      "loss": 0.9228,
+      "step": 2171
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.635474221215026,
+      "learning_rate": 1.5799375995821116e-05,
+      "loss": 0.898,
+      "step": 2172
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.553659932170287,
+      "learning_rate": 1.5795437883690225e-05,
+      "loss": 0.9361,
+      "step": 2173
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4474668240688067,
+      "learning_rate": 1.5791498417744463e-05,
+      "loss": 0.9461,
+      "step": 2174
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.455261302785284,
+      "learning_rate": 1.578755759890409e-05,
+      "loss": 0.8827,
+      "step": 2175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5158065252140318,
+      "learning_rate": 1.5783615428089693e-05,
+      "loss": 0.8082,
+      "step": 2176
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2723095006700331,
+      "learning_rate": 1.577967190622215e-05,
+      "loss": 0.8922,
+      "step": 2177
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4095759479018382,
+      "learning_rate": 1.5775727034222675e-05,
+      "loss": 0.8597,
+      "step": 2178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6099465915272615,
+      "learning_rate": 1.5771780813012793e-05,
+      "loss": 0.908,
+      "step": 2179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5208257123568947,
+      "learning_rate": 1.5767833243514337e-05,
+      "loss": 0.8655,
+      "step": 2180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4611612259562878,
+      "learning_rate": 1.576388432664946e-05,
+      "loss": 1.0259,
+      "step": 2181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.30880284133487,
+      "learning_rate": 1.5759934063340627e-05,
+      "loss": 0.8593,
+      "step": 2182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5608508287888272,
+      "learning_rate": 1.5755982454510626e-05,
+      "loss": 0.8777,
+      "step": 2183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5452633961299929,
+      "learning_rate": 1.5752029501082547e-05,
+      "loss": 0.9205,
+      "step": 2184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3680647052245383,
+      "learning_rate": 1.574807520397981e-05,
+      "loss": 0.9257,
+      "step": 2185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4420257570652117,
+      "learning_rate": 1.5744119564126127e-05,
+      "loss": 0.8994,
+      "step": 2186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4966876977454537,
+      "learning_rate": 1.5740162582445545e-05,
+      "loss": 0.9569,
+      "step": 2187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4967860519963248,
+      "learning_rate": 1.573620425986241e-05,
+      "loss": 0.8831,
+      "step": 2188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9470941766944733,
+      "learning_rate": 1.573224459730139e-05,
+      "loss": 0.3304,
+      "step": 2189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4879996316147293,
+      "learning_rate": 1.572828359568746e-05,
+      "loss": 0.8798,
+      "step": 2190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3005612225926124,
+      "learning_rate": 1.572432125594591e-05,
+      "loss": 0.8512,
+      "step": 2191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3872635563657048,
+      "learning_rate": 1.5720357579002346e-05,
+      "loss": 0.8609,
+      "step": 2192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6451510509568592,
+      "learning_rate": 1.5716392565782683e-05,
+      "loss": 0.8718,
+      "step": 2193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5026512594517971,
+      "learning_rate": 1.5712426217213143e-05,
+      "loss": 0.8823,
+      "step": 2194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3529289599428111,
+      "learning_rate": 1.5708458534220274e-05,
+      "loss": 0.7962,
+      "step": 2195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5069515847694586,
+      "learning_rate": 1.570448951773092e-05,
+      "loss": 0.9035,
+      "step": 2196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4580898785602376,
+      "learning_rate": 1.5700519168672248e-05,
+      "loss": 0.9372,
+      "step": 2197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.51821727009186,
+      "learning_rate": 1.5696547487971727e-05,
+      "loss": 0.8667,
+      "step": 2198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4271390550471268,
+      "learning_rate": 1.5692574476557147e-05,
+      "loss": 0.837,
+      "step": 2199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6197580648864112,
+      "learning_rate": 1.5688600135356608e-05,
+      "loss": 0.8572,
+      "step": 2200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4561272634408733,
+      "learning_rate": 1.5684624465298503e-05,
+      "loss": 0.7864,
+      "step": 2201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3937402333037126,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.9116,
+      "step": 2202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3484215139256983,
+      "learning_rate": 1.5676669142324802e-05,
+      "loss": 0.8893,
+      "step": 2203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3091942548048083,
+      "learning_rate": 1.567268949126757e-05,
+      "loss": 0.804,
+      "step": 2204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3148493615935646,
+      "learning_rate": 1.5668708515069503e-05,
+      "loss": 0.843,
+      "step": 2205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3483526856004353,
+      "learning_rate": 1.5664726214660562e-05,
+      "loss": 0.8502,
+      "step": 2206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.63158011129037,
+      "learning_rate": 1.5660742590971014e-05,
+      "loss": 0.8952,
+      "step": 2207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3337432739792972,
+      "learning_rate": 1.5656757644931433e-05,
+      "loss": 0.9488,
+      "step": 2208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5399512766054197,
+      "learning_rate": 1.5652771377472702e-05,
+      "loss": 0.8717,
+      "step": 2209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4468026604756365,
+      "learning_rate": 1.5648783789526015e-05,
+      "loss": 0.9012,
+      "step": 2210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7067606327369804,
+      "learning_rate": 1.5644794882022875e-05,
+      "loss": 0.902,
+      "step": 2211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4971688468939985,
+      "learning_rate": 1.5640804655895086e-05,
+      "loss": 0.9287,
+      "step": 2212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.354877618546282,
+      "learning_rate": 1.5636813112074766e-05,
+      "loss": 0.9474,
+      "step": 2213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8363170606810937,
+      "learning_rate": 1.5632820251494343e-05,
+      "loss": 0.9499,
+      "step": 2214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.600171282753231,
+      "learning_rate": 1.5628826075086546e-05,
+      "loss": 0.9178,
+      "step": 2215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5056276452344919,
+      "learning_rate": 1.562483058378442e-05,
+      "loss": 0.8927,
+      "step": 2216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.945562162809114,
+      "learning_rate": 1.5620833778521306e-05,
+      "loss": 0.8398,
+      "step": 2217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4186141618112442,
+      "learning_rate": 1.5616835660230864e-05,
+      "loss": 0.8215,
+      "step": 2218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4689542192154341,
+      "learning_rate": 1.561283622984705e-05,
+      "loss": 0.9097,
+      "step": 2219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.534684008479006,
+      "learning_rate": 1.5608835488304138e-05,
+      "loss": 0.8979,
+      "step": 2220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3876388036294593,
+      "learning_rate": 1.5604833436536692e-05,
+      "loss": 0.9596,
+      "step": 2221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5487391312805103,
+      "learning_rate": 1.5600830075479604e-05,
+      "loss": 0.8965,
+      "step": 2222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.391755006595629,
+      "learning_rate": 1.5596825406068043e-05,
+      "loss": 0.9275,
+      "step": 2223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5384997558976328,
+      "learning_rate": 1.5592819429237517e-05,
+      "loss": 0.9673,
+      "step": 2224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6906691237842095,
+      "learning_rate": 1.5588812145923812e-05,
+      "loss": 0.8544,
+      "step": 2225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.480162351048646,
+      "learning_rate": 1.5584803557063034e-05,
+      "loss": 0.9415,
+      "step": 2226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5164365175077197,
+      "learning_rate": 1.5580793663591583e-05,
+      "loss": 0.9039,
+      "step": 2227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4675588666684414,
+      "learning_rate": 1.5576782466446184e-05,
+      "loss": 0.8955,
+      "step": 2228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6289353665372766,
+      "learning_rate": 1.5572769966563844e-05,
+      "loss": 0.7649,
+      "step": 2229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4166898562628878,
+      "learning_rate": 1.556875616488188e-05,
+      "loss": 0.9154,
+      "step": 2230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5940118713608051,
+      "learning_rate": 1.5564741062337928e-05,
+      "loss": 0.8957,
+      "step": 2231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6883631922453783,
+      "learning_rate": 1.5560724659869905e-05,
+      "loss": 0.9627,
+      "step": 2232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5894771816451598,
+      "learning_rate": 1.5556706958416044e-05,
+      "loss": 0.8595,
+      "step": 2233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5662607698591635,
+      "learning_rate": 1.5552687958914892e-05,
+      "loss": 0.9358,
+      "step": 2234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5258953948083736,
+      "learning_rate": 1.5548667662305275e-05,
+      "loss": 0.8574,
+      "step": 2235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4040338219456894,
+      "learning_rate": 1.554464606952634e-05,
+      "loss": 0.8979,
+      "step": 2236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5369400594007905,
+      "learning_rate": 1.5540623181517532e-05,
+      "loss": 0.9366,
+      "step": 2237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8806919665176648,
+      "learning_rate": 1.5536598999218592e-05,
+      "loss": 0.3439,
+      "step": 2238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6199453123598875,
+      "learning_rate": 1.553257352356958e-05,
+      "loss": 0.8684,
+      "step": 2239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2520656215275296,
+      "learning_rate": 1.5528546755510842e-05,
+      "loss": 0.9159,
+      "step": 2240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4701516446023108,
+      "learning_rate": 1.5524518695983025e-05,
+      "loss": 0.9249,
+      "step": 2241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4088294135179658,
+      "learning_rate": 1.5520489345927095e-05,
+      "loss": 0.888,
+      "step": 2242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3486884780323214,
+      "learning_rate": 1.5516458706284306e-05,
+      "loss": 0.9258,
+      "step": 2243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4183918713098838,
+      "learning_rate": 1.5512426777996206e-05,
+      "loss": 0.92,
+      "step": 2244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4185166981634143,
+      "learning_rate": 1.550839356200467e-05,
+      "loss": 0.8933,
+      "step": 2245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6146733037174301,
+      "learning_rate": 1.5504359059251843e-05,
+      "loss": 0.7927,
+      "step": 2246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5621420378726925,
+      "learning_rate": 1.5500323270680194e-05,
+      "loss": 0.7321,
+      "step": 2247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9487127370154067,
+      "learning_rate": 1.5496286197232483e-05,
+      "loss": 0.3571,
+      "step": 2248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.35658916070422,
+      "learning_rate": 1.5492247839851767e-05,
+      "loss": 0.9208,
+      "step": 2249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4108634909517979,
+      "learning_rate": 1.5488208199481406e-05,
+      "loss": 0.8842,
+      "step": 2250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4262516081982304,
+      "learning_rate": 1.5484167277065066e-05,
+      "loss": 0.8548,
+      "step": 2251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5429323759499571,
+      "learning_rate": 1.5480125073546705e-05,
+      "loss": 0.9319,
+      "step": 2252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3841626087791337,
+      "learning_rate": 1.5476081589870574e-05,
+      "loss": 0.9014,
+      "step": 2253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.649983356944591,
+      "learning_rate": 1.5472036826981243e-05,
+      "loss": 0.9695,
+      "step": 2254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4482985567220856,
+      "learning_rate": 1.5467990785823562e-05,
+      "loss": 0.9038,
+      "step": 2255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3362472367525746,
+      "learning_rate": 1.5463943467342694e-05,
+      "loss": 0.9004,
+      "step": 2256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4007567854073044,
+      "learning_rate": 1.5459894872484083e-05,
+      "loss": 0.9684,
+      "step": 2257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.61796984660997,
+      "learning_rate": 1.5455845002193485e-05,
+      "loss": 0.8807,
+      "step": 2258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.458057620532137,
+      "learning_rate": 1.5451793857416955e-05,
+      "loss": 0.874,
+      "step": 2259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5021369806670417,
+      "learning_rate": 1.5447741439100833e-05,
+      "loss": 0.787,
+      "step": 2260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7381014168330053,
+      "learning_rate": 1.5443687748191772e-05,
+      "loss": 0.9132,
+      "step": 2261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3540898415126852,
+      "learning_rate": 1.5439632785636707e-05,
+      "loss": 0.9611,
+      "step": 2262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4220905723485562,
+      "learning_rate": 1.5435576552382883e-05,
+      "loss": 0.8881,
+      "step": 2263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3431553377314935,
+      "learning_rate": 1.5431519049377835e-05,
+      "loss": 0.9087,
+      "step": 2264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2961370734239621,
+      "learning_rate": 1.54274602775694e-05,
+      "loss": 0.863,
+      "step": 2265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3762272186152134,
+      "learning_rate": 1.5423400237905695e-05,
+      "loss": 0.903,
+      "step": 2266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5387414546573555,
+      "learning_rate": 1.5419338931335155e-05,
+      "loss": 0.9676,
+      "step": 2267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6189088284489723,
+      "learning_rate": 1.541527635880651e-05,
+      "loss": 0.832,
+      "step": 2268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6331762913586843,
+      "learning_rate": 1.541121252126876e-05,
+      "loss": 0.871,
+      "step": 2269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5171575393036398,
+      "learning_rate": 1.5407147419671227e-05,
+      "loss": 0.8905,
+      "step": 2270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4316603683942186,
+      "learning_rate": 1.5403081054963524e-05,
+      "loss": 0.8599,
+      "step": 2271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3474805696668777,
+      "learning_rate": 1.539901342809554e-05,
+      "loss": 0.9241,
+      "step": 2272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4477942990205288,
+      "learning_rate": 1.5394944540017484e-05,
+      "loss": 0.785,
+      "step": 2273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.375109046437524,
+      "learning_rate": 1.539087439167985e-05,
+      "loss": 0.8774,
+      "step": 2274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3627256481363823,
+      "learning_rate": 1.5386802984033417e-05,
+      "loss": 0.8875,
+      "step": 2275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2827976869478268,
+      "learning_rate": 1.538273031802927e-05,
+      "loss": 0.9094,
+      "step": 2276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5484833652332188,
+      "learning_rate": 1.5378656394618788e-05,
+      "loss": 0.8757,
+      "step": 2277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8288346129902062,
+      "learning_rate": 1.537458121475363e-05,
+      "loss": 0.9001,
+      "step": 2278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.37712553108305,
+      "learning_rate": 1.537050477938577e-05,
+      "loss": 0.9418,
+      "step": 2279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.548431791344098,
+      "learning_rate": 1.5366427089467458e-05,
+      "loss": 0.8796,
+      "step": 2280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.605941113595781,
+      "learning_rate": 1.5362348145951242e-05,
+      "loss": 0.8877,
+      "step": 2281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3944602729194844,
+      "learning_rate": 1.5358267949789968e-05,
+      "loss": 0.9196,
+      "step": 2282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.629319528483046,
+      "learning_rate": 1.5354186501936764e-05,
+      "loss": 0.9201,
+      "step": 2283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.420243089173775,
+      "learning_rate": 1.535010380334506e-05,
+      "loss": 0.7979,
+      "step": 2284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7455501606322166,
+      "learning_rate": 1.5346019854968576e-05,
+      "loss": 0.9369,
+      "step": 2285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4095648267675434,
+      "learning_rate": 1.534193465776132e-05,
+      "loss": 0.9682,
+      "step": 2286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3625991844792567,
+      "learning_rate": 1.53378482126776e-05,
+      "loss": 0.9045,
+      "step": 2287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2975952126331425,
+      "learning_rate": 1.5333760520672e-05,
+      "loss": 0.9706,
+      "step": 2288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.705059630390903,
+      "learning_rate": 1.532967158269941e-05,
+      "loss": 0.9164,
+      "step": 2289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4009019764024053,
+      "learning_rate": 1.5325581399715013e-05,
+      "loss": 0.9911,
+      "step": 2290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4617561861562665,
+      "learning_rate": 1.5321489972674268e-05,
+      "loss": 0.8413,
+      "step": 2291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7649872965902604,
+      "learning_rate": 1.5317397302532933e-05,
+      "loss": 0.9461,
+      "step": 2292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.945462129313132,
+      "learning_rate": 1.5313303390247062e-05,
+      "loss": 0.325,
+      "step": 2293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6442386849458337,
+      "learning_rate": 1.5309208236772988e-05,
+      "loss": 0.8823,
+      "step": 2294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.351108691206905,
+      "learning_rate": 1.5305111843067343e-05,
+      "loss": 0.8636,
+      "step": 2295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.629401278689014,
+      "learning_rate": 1.530101421008704e-05,
+      "loss": 0.8586,
+      "step": 2296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5085386710431767,
+      "learning_rate": 1.529691533878929e-05,
+      "loss": 0.9112,
+      "step": 2297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4307107546741473,
+      "learning_rate": 1.5292815230131592e-05,
+      "loss": 0.8613,
+      "step": 2298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4314818482870362,
+      "learning_rate": 1.5288713885071723e-05,
+      "loss": 0.8195,
+      "step": 2299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7662229928387017,
+      "learning_rate": 1.5284611304567772e-05,
+      "loss": 0.877,
+      "step": 2300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5834178446475238,
+      "learning_rate": 1.5280507489578087e-05,
+      "loss": 0.852,
+      "step": 2301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3132249371325437,
+      "learning_rate": 1.527640244106133e-05,
+      "loss": 0.8881,
+      "step": 2302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3965578693444147,
+      "learning_rate": 1.5272296159976438e-05,
+      "loss": 0.9597,
+      "step": 2303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4857132475717196,
+      "learning_rate": 1.5268188647282633e-05,
+      "loss": 0.8924,
+      "step": 2304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8222442474419664,
+      "learning_rate": 1.5264079903939437e-05,
+      "loss": 0.8943,
+      "step": 2305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.552734078701159,
+      "learning_rate": 1.5259969930906653e-05,
+      "loss": 0.8836,
+      "step": 2306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8007006927146774,
+      "learning_rate": 1.5255858729144368e-05,
+      "loss": 0.8999,
+      "step": 2307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2328321360461432,
+      "learning_rate": 1.5251746299612959e-05,
+      "loss": 0.8642,
+      "step": 2308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3393351052321958,
+      "learning_rate": 1.5247632643273092e-05,
+      "loss": 0.9278,
+      "step": 2309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.532450030571645,
+      "learning_rate": 1.5243517761085717e-05,
+      "loss": 0.8624,
+      "step": 2310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.696772369816577,
+      "learning_rate": 1.5239401654012072e-05,
+      "loss": 0.8041,
+      "step": 2311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3992107034629708,
+      "learning_rate": 1.5235284323013674e-05,
+      "loss": 0.8638,
+      "step": 2312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4534537125595492,
+      "learning_rate": 1.5231165769052343e-05,
+      "loss": 0.8273,
+      "step": 2313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5868384381394676,
+      "learning_rate": 1.5227045993090164e-05,
+      "loss": 0.9711,
+      "step": 2314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.578135235441649,
+      "learning_rate": 1.5222924996089518e-05,
+      "loss": 0.9509,
+      "step": 2315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5827739214502767,
+      "learning_rate": 1.5218802779013077e-05,
+      "loss": 0.962,
+      "step": 2316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3964253733248453,
+      "learning_rate": 1.5214679342823786e-05,
+      "loss": 0.9139,
+      "step": 2317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.420897818103144,
+      "learning_rate": 1.521055468848488e-05,
+      "loss": 0.9577,
+      "step": 2318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.366209770458451,
+      "learning_rate": 1.520642881695988e-05,
+      "loss": 0.8727,
+      "step": 2319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2955331050244765,
+      "learning_rate": 1.520230172921259e-05,
+      "loss": 0.8803,
+      "step": 2320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4311093005368292,
+      "learning_rate": 1.5198173426207095e-05,
+      "loss": 0.9317,
+      "step": 2321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4325823685380945,
+      "learning_rate": 1.5194043908907774e-05,
+      "loss": 0.8965,
+      "step": 2322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3878217239573705,
+      "learning_rate": 1.5189913178279277e-05,
+      "loss": 0.8678,
+      "step": 2323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6083570635997588,
+      "learning_rate": 1.5185781235286543e-05,
+      "loss": 0.9364,
+      "step": 2324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7317265595725782,
+      "learning_rate": 1.5181648080894796e-05,
+      "loss": 0.8544,
+      "step": 2325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3870363254643336,
+      "learning_rate": 1.5177513716069543e-05,
+      "loss": 0.8949,
+      "step": 2326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4668381535325745,
+      "learning_rate": 1.5173378141776569e-05,
+      "loss": 0.8749,
+      "step": 2327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9537819485161397,
+      "learning_rate": 1.5169241358981946e-05,
+      "loss": 0.3455,
+      "step": 2328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4274521247339735,
+      "learning_rate": 1.5165103368652028e-05,
+      "loss": 0.8641,
+      "step": 2329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6050826685124675,
+      "learning_rate": 1.5160964171753444e-05,
+      "loss": 0.863,
+      "step": 2330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.429337965600591,
+      "learning_rate": 1.515682376925312e-05,
+      "loss": 0.9224,
+      "step": 2331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4831736873410963,
+      "learning_rate": 1.515268216211825e-05,
+      "loss": 0.8651,
+      "step": 2332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3691947412067333,
+      "learning_rate": 1.514853935131631e-05,
+      "loss": 0.8515,
+      "step": 2333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.578314145480583,
+      "learning_rate": 1.5144395337815066e-05,
+      "loss": 0.8925,
+      "step": 2334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3003209770357662,
+      "learning_rate": 1.514025012258256e-05,
+      "loss": 0.8414,
+      "step": 2335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4716077443987046,
+      "learning_rate": 1.5136103706587111e-05,
+      "loss": 0.9372,
+      "step": 2336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5880096315877341,
+      "learning_rate": 1.5131956090797326e-05,
+      "loss": 0.8109,
+      "step": 2337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4644640773832716,
+      "learning_rate": 1.5127807276182084e-05,
+      "loss": 0.8915,
+      "step": 2338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4236492161009207,
+      "learning_rate": 1.5123657263710558e-05,
+      "loss": 0.9772,
+      "step": 2339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.433436242163942,
+      "learning_rate": 1.5119506054352178e-05,
+      "loss": 0.8215,
+      "step": 2340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6063318369015331,
+      "learning_rate": 1.5115353649076676e-05,
+      "loss": 0.8725,
+      "step": 2341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.839816332165306,
+      "learning_rate": 1.5111200048854055e-05,
+      "loss": 0.8826,
+      "step": 2342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5216059868754872,
+      "learning_rate": 1.5107045254654588e-05,
+      "loss": 0.8669,
+      "step": 2343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5300404609259486,
+      "learning_rate": 1.5102889267448846e-05,
+      "loss": 0.8356,
+      "step": 2344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4313848808295826,
+      "learning_rate": 1.5098732088207662e-05,
+      "loss": 0.886,
+      "step": 2345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5299634473431738,
+      "learning_rate": 1.509457371790215e-05,
+      "loss": 0.8882,
+      "step": 2346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4963645801974077,
+      "learning_rate": 1.5090414157503715e-05,
+      "loss": 0.8834,
+      "step": 2347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4424008056458453,
+      "learning_rate": 1.5086253407984024e-05,
+      "loss": 0.9216,
+      "step": 2348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5844090382834033,
+      "learning_rate": 1.5082091470315028e-05,
+      "loss": 0.9112,
+      "step": 2349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4526264421115531,
+      "learning_rate": 1.5077928345468959e-05,
+      "loss": 0.9302,
+      "step": 2350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3418442052853163,
+      "learning_rate": 1.5073764034418326e-05,
+      "loss": 0.8916,
+      "step": 2351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6891761376321324,
+      "learning_rate": 1.5069598538135905e-05,
+      "loss": 0.8041,
+      "step": 2352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6325851389366492,
+      "learning_rate": 1.5065431857594762e-05,
+      "loss": 0.9762,
+      "step": 2353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4238778885306893,
+      "learning_rate": 1.5061263993768234e-05,
+      "loss": 0.9239,
+      "step": 2354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4156220777115123,
+      "learning_rate": 1.5057094947629928e-05,
+      "loss": 0.8043,
+      "step": 2355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5654947364937417,
+      "learning_rate": 1.5052924720153743e-05,
+      "loss": 0.8883,
+      "step": 2356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.500895562527217,
+      "learning_rate": 1.504875331231384e-05,
+      "loss": 0.8911,
+      "step": 2357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9238417436146753,
+      "learning_rate": 1.5044580725084659e-05,
+      "loss": 0.8234,
+      "step": 2358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4575158021709844,
+      "learning_rate": 1.504040695944092e-05,
+      "loss": 0.8808,
+      "step": 2359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5683316267511338,
+      "learning_rate": 1.503623201635761e-05,
+      "loss": 0.8476,
+      "step": 2360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9450791116160474,
+      "learning_rate": 1.503205589681e-05,
+      "loss": 0.3525,
+      "step": 2361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5042833864780634,
+      "learning_rate": 1.5027878601773633e-05,
+      "loss": 0.9452,
+      "step": 2362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4131609923776498,
+      "learning_rate": 1.5023700132224321e-05,
+      "loss": 0.9178,
+      "step": 2363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4386296226455535,
+      "learning_rate": 1.501952048913816e-05,
+      "loss": 0.9758,
+      "step": 2364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6133745053380633,
+      "learning_rate": 1.5015339673491514e-05,
+      "loss": 0.9059,
+      "step": 2365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4305133297514012,
+      "learning_rate": 1.5011157686261015e-05,
+      "loss": 0.9652,
+      "step": 2366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3992892034589874,
+      "learning_rate": 1.5006974528423585e-05,
+      "loss": 0.8351,
+      "step": 2367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4661697775481777,
+      "learning_rate": 1.5002790200956408e-05,
+      "loss": 0.957,
+      "step": 2368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.83615936348965,
+      "learning_rate": 1.4998604704836939e-05,
+      "loss": 0.8929,
+      "step": 2369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4292429261452897,
+      "learning_rate": 1.4994418041042913e-05,
+      "loss": 0.8038,
+      "step": 2370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5957900003809427,
+      "learning_rate": 1.499023021055234e-05,
+      "loss": 0.8624,
+      "step": 2371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.566093941659696,
+      "learning_rate": 1.4986041214343487e-05,
+      "loss": 0.8765,
+      "step": 2372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4259109803843022,
+      "learning_rate": 1.498185105339491e-05,
+      "loss": 0.8846,
+      "step": 2373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.331796638042093,
+      "learning_rate": 1.4977659728685435e-05,
+      "loss": 0.8723,
+      "step": 2374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5317344057461537,
+      "learning_rate": 1.4973467241194146e-05,
+      "loss": 0.8441,
+      "step": 2375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4247032159462272,
+      "learning_rate": 1.4969273591900415e-05,
+      "loss": 0.8656,
+      "step": 2376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4195801482441541,
+      "learning_rate": 1.4965078781783882e-05,
+      "loss": 0.8739,
+      "step": 2377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5502697740156313,
+      "learning_rate": 1.4960882811824446e-05,
+      "loss": 0.8784,
+      "step": 2378
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4750552652752638,
+      "learning_rate": 1.4956685683002292e-05,
+      "loss": 0.9735,
+      "step": 2379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.522049382723157,
+      "learning_rate": 1.4952487396297872e-05,
+      "loss": 0.8709,
+      "step": 2380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4467281355070856,
+      "learning_rate": 1.49482879526919e-05,
+      "loss": 0.9195,
+      "step": 2381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9023408022745709,
+      "learning_rate": 1.494408735316537e-05,
+      "loss": 0.3312,
+      "step": 2382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4877313363881013,
+      "learning_rate": 1.493988559869954e-05,
+      "loss": 0.8392,
+      "step": 2383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3222375779438422,
+      "learning_rate": 1.4935682690275945e-05,
+      "loss": 0.8781,
+      "step": 2384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9539362745691424,
+      "learning_rate": 1.4931478628876382e-05,
+      "loss": 0.357,
+      "step": 2385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5534801347691423,
+      "learning_rate": 1.4927273415482916e-05,
+      "loss": 0.8494,
+      "step": 2386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5393226604550732,
+      "learning_rate": 1.4923067051077893e-05,
+      "loss": 0.8438,
+      "step": 2387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2562688905863733,
+      "learning_rate": 1.4918859536643915e-05,
+      "loss": 0.8121,
+      "step": 2388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4484418447155691,
+      "learning_rate": 1.491465087316386e-05,
+      "loss": 0.8762,
+      "step": 2389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6561482033845658,
+      "learning_rate": 1.4910441061620872e-05,
+      "loss": 0.8421,
+      "step": 2390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1526764033938166,
+      "learning_rate": 1.490623010299836e-05,
+      "loss": 0.9026,
+      "step": 2391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.001240085062583,
+      "learning_rate": 1.490201799828001e-05,
+      "loss": 0.8276,
+      "step": 2392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6877704895867491,
+      "learning_rate": 1.4897804748449767e-05,
+      "loss": 0.8119,
+      "step": 2393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.349063557279118,
+      "learning_rate": 1.4893590354491845e-05,
+      "loss": 0.8973,
+      "step": 2394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3497819721141477,
+      "learning_rate": 1.488937481739073e-05,
+      "loss": 0.8139,
+      "step": 2395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.427016609449477,
+      "learning_rate": 1.4885158138131171e-05,
+      "loss": 0.9004,
+      "step": 2396
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6930834728031425,
+      "learning_rate": 1.4880940317698182e-05,
+      "loss": 0.8265,
+      "step": 2397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5074730513249242,
+      "learning_rate": 1.4876721357077051e-05,
+      "loss": 0.9503,
+      "step": 2398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4632392717400273,
+      "learning_rate": 1.4872501257253325e-05,
+      "loss": 0.8291,
+      "step": 2399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5221970631070543,
+      "learning_rate": 1.486828001921282e-05,
+      "loss": 0.96,
+      "step": 2400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4653147169219638,
+      "learning_rate": 1.4864057643941616e-05,
+      "loss": 0.8475,
+      "step": 2401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2854470399396005,
+      "learning_rate": 1.485983413242606e-05,
+      "loss": 0.8813,
+      "step": 2402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6584598356029185,
+      "learning_rate": 1.4855609485652771e-05,
+      "loss": 0.846,
+      "step": 2403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3443065865674046,
+      "learning_rate": 1.4851383704608621e-05,
+      "loss": 0.9016,
+      "step": 2404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4248373800478973,
+      "learning_rate": 1.4847156790280753e-05,
+      "loss": 0.8781,
+      "step": 2405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6376030252757539,
+      "learning_rate": 1.4842928743656577e-05,
+      "loss": 0.8687,
+      "step": 2406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.440488834136608,
+      "learning_rate": 1.4838699565723764e-05,
+      "loss": 0.8349,
+      "step": 2407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.389871529118748,
+      "learning_rate": 1.483446925747025e-05,
+      "loss": 0.8012,
+      "step": 2408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4600342212260655,
+      "learning_rate": 1.4830237819884237e-05,
+      "loss": 0.9294,
+      "step": 2409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5131935483661065,
+      "learning_rate": 1.4826005253954185e-05,
+      "loss": 0.8992,
+      "step": 2410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4004860423845367,
+      "learning_rate": 1.4821771560668828e-05,
+      "loss": 0.8442,
+      "step": 2411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4613737835668572,
+      "learning_rate": 1.4817536741017153e-05,
+      "loss": 0.8467,
+      "step": 2412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4061947401103583,
+      "learning_rate": 1.4813300795988417e-05,
+      "loss": 0.8241,
+      "step": 2413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2683224601296244,
+      "learning_rate": 1.4809063726572132e-05,
+      "loss": 0.8952,
+      "step": 2414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5535422652571538,
+      "learning_rate": 1.4804825533758082e-05,
+      "loss": 0.9388,
+      "step": 2415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6076200230999473,
+      "learning_rate": 1.4800586218536315e-05,
+      "loss": 0.9591,
+      "step": 2416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.400041287235936,
+      "learning_rate": 1.479634578189712e-05,
+      "loss": 0.8987,
+      "step": 2417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.377021043874476,
+      "learning_rate": 1.4792104224831079e-05,
+      "loss": 0.9753,
+      "step": 2418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.335754231380377,
+      "learning_rate": 1.4787861548329012e-05,
+      "loss": 0.89,
+      "step": 2419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5095214720356913,
+      "learning_rate": 1.478361775338201e-05,
+      "loss": 0.932,
+      "step": 2420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3076321598873812,
+      "learning_rate": 1.4779372840981425e-05,
+      "loss": 0.956,
+      "step": 2421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.369567585488684,
+      "learning_rate": 1.4775126812118865e-05,
+      "loss": 0.8672,
+      "step": 2422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.804175563342299,
+      "learning_rate": 1.477087966778621e-05,
+      "loss": 0.8879,
+      "step": 2423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.689058301798696,
+      "learning_rate": 1.4766631408975586e-05,
+      "loss": 0.8523,
+      "step": 2424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.312661117076535,
+      "learning_rate": 1.4762382036679393e-05,
+      "loss": 0.8567,
+      "step": 2425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3588466412209272,
+      "learning_rate": 1.475813155189028e-05,
+      "loss": 0.8619,
+      "step": 2426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3263685508280627,
+      "learning_rate": 1.4753879955601162e-05,
+      "loss": 0.9336,
+      "step": 2427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.456260335945933,
+      "learning_rate": 1.4749627248805214e-05,
+      "loss": 0.8655,
+      "step": 2428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6345935181748061,
+      "learning_rate": 1.4745373432495869e-05,
+      "loss": 0.8579,
+      "step": 2429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.63078982838802,
+      "learning_rate": 1.4741118507666815e-05,
+      "loss": 0.9137,
+      "step": 2430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5178031392493103,
+      "learning_rate": 1.4736862475312005e-05,
+      "loss": 0.9154,
+      "step": 2431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6975881382058196,
+      "learning_rate": 1.4732605336425651e-05,
+      "loss": 0.9785,
+      "step": 2432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4917404999460437,
+      "learning_rate": 1.4728347092002218e-05,
+      "loss": 0.8574,
+      "step": 2433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.430996562609836,
+      "learning_rate": 1.4724087743036432e-05,
+      "loss": 0.8788,
+      "step": 2434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4830372818362922,
+      "learning_rate": 1.4719827290523281e-05,
+      "loss": 0.8599,
+      "step": 2435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4356241257311562,
+      "learning_rate": 1.4715565735458004e-05,
+      "loss": 0.8991,
+      "step": 2436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4217908102042733,
+      "learning_rate": 1.4711303078836098e-05,
+      "loss": 0.8859,
+      "step": 2437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4335111469512989,
+      "learning_rate": 1.470703932165333e-05,
+      "loss": 0.8343,
+      "step": 2438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4389369322442183,
+      "learning_rate": 1.4702774464905703e-05,
+      "loss": 0.8744,
+      "step": 2439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4257248383950092,
+      "learning_rate": 1.469850850958949e-05,
+      "loss": 0.867,
+      "step": 2440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.489892645435096,
+      "learning_rate": 1.4694241456701227e-05,
+      "loss": 0.8774,
+      "step": 2441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2522499599960846,
+      "learning_rate": 1.4689973307237687e-05,
+      "loss": 0.836,
+      "step": 2442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.552581436494605,
+      "learning_rate": 1.468570406219592e-05,
+      "loss": 0.8857,
+      "step": 2443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9034813464894758,
+      "learning_rate": 1.4681433722573212e-05,
+      "loss": 0.8745,
+      "step": 2444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7042494838096385,
+      "learning_rate": 1.4677162289367126e-05,
+      "loss": 0.8993,
+      "step": 2445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0272032107655216,
+      "learning_rate": 1.467288976357546e-05,
+      "loss": 0.3507,
+      "step": 2446
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3359220640506526,
+      "learning_rate": 1.466861614619628e-05,
+      "loss": 0.8886,
+      "step": 2447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.427324392080972,
+      "learning_rate": 1.4664341438227903e-05,
+      "loss": 0.8058,
+      "step": 2448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.654523611012364,
+      "learning_rate": 1.46600656406689e-05,
+      "loss": 0.8601,
+      "step": 2449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4648545995946267,
+      "learning_rate": 1.4655788754518101e-05,
+      "loss": 0.7682,
+      "step": 2450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8865049667434419,
+      "learning_rate": 1.4651510780774585e-05,
+      "loss": 0.3383,
+      "step": 2451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3405881105529465,
+      "learning_rate": 1.4647231720437687e-05,
+      "loss": 0.9416,
+      "step": 2452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4222131755338903,
+      "learning_rate": 1.4642951574506996e-05,
+      "loss": 0.8263,
+      "step": 2453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8321441396951367,
+      "learning_rate": 1.4638670343982356e-05,
+      "loss": 0.3209,
+      "step": 2454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.392619106979098,
+      "learning_rate": 1.463438802986386e-05,
+      "loss": 0.833,
+      "step": 2455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4416212994695166,
+      "learning_rate": 1.4630104633151858e-05,
+      "loss": 0.8464,
+      "step": 2456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8940930834255698,
+      "learning_rate": 1.4625820154846953e-05,
+      "loss": 0.3257,
+      "step": 2457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4518221858725542,
+      "learning_rate": 1.462153459595e-05,
+      "loss": 0.7631,
+      "step": 2458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.297559268233778,
+      "learning_rate": 1.4617247957462105e-05,
+      "loss": 0.8622,
+      "step": 2459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5874471655167313,
+      "learning_rate": 1.4612960240384624e-05,
+      "loss": 0.8689,
+      "step": 2460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.765896841660207,
+      "learning_rate": 1.4608671445719176e-05,
+      "loss": 0.9403,
+      "step": 2461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3266408678098056,
+      "learning_rate": 1.4604381574467616e-05,
+      "loss": 0.9079,
+      "step": 2462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9644536893054212,
+      "learning_rate": 1.4600090627632061e-05,
+      "loss": 0.3515,
+      "step": 2463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5591183272000826,
+      "learning_rate": 1.4595798606214882e-05,
+      "loss": 0.8543,
+      "step": 2464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9083395414548567,
+      "learning_rate": 1.4591505511218685e-05,
+      "loss": 0.3563,
+      "step": 2465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5305677409675236,
+      "learning_rate": 1.4587211343646345e-05,
+      "loss": 0.8811,
+      "step": 2466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3240987503909145,
+      "learning_rate": 1.4582916104500977e-05,
+      "loss": 0.9405,
+      "step": 2467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.214697916471032,
+      "learning_rate": 1.4578619794785956e-05,
+      "loss": 0.9009,
+      "step": 2468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4146701537207016,
+      "learning_rate": 1.457432241550489e-05,
+      "loss": 0.8477,
+      "step": 2469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4341981273880275,
+      "learning_rate": 1.4570023967661651e-05,
+      "loss": 0.8904,
+      "step": 2470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5617096666431618,
+      "learning_rate": 1.4565724452260361e-05,
+      "loss": 0.9562,
+      "step": 2471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2771080496656244,
+      "learning_rate": 1.4561423870305383e-05,
+      "loss": 0.8642,
+      "step": 2472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4269934224052399,
+      "learning_rate": 1.4557122222801332e-05,
+      "loss": 0.8667,
+      "step": 2473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6239354579443843,
+      "learning_rate": 1.455281951075308e-05,
+      "loss": 0.8982,
+      "step": 2474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2853315938161574,
+      "learning_rate": 1.4548515735165735e-05,
+      "loss": 0.8794,
+      "step": 2475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4950762431034248,
+      "learning_rate": 1.4544210897044662e-05,
+      "loss": 0.8418,
+      "step": 2476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5800141497228144,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 0.922,
+      "step": 2477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6768337114826615,
+      "learning_rate": 1.4535598037224016e-05,
+      "loss": 0.9138,
+      "step": 2478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4183248717676997,
+      "learning_rate": 1.453129001753641e-05,
+      "loss": 0.8768,
+      "step": 2479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2319062328588768,
+      "learning_rate": 1.4526980939339008e-05,
+      "loss": 0.8361,
+      "step": 2480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4220273760221085,
+      "learning_rate": 1.4522670803638403e-05,
+      "loss": 0.8686,
+      "step": 2481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5702338495533112,
+      "learning_rate": 1.4518359611441452e-05,
+      "loss": 0.8722,
+      "step": 2482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5212147942761973,
+      "learning_rate": 1.4514047363755244e-05,
+      "loss": 0.9242,
+      "step": 2483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.619409485916836,
+      "learning_rate": 1.450973406158712e-05,
+      "loss": 0.937,
+      "step": 2484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4174108984231562,
+      "learning_rate": 1.4505419705944672e-05,
+      "loss": 0.9332,
+      "step": 2485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5244552049577198,
+      "learning_rate": 1.450110429783573e-05,
+      "loss": 0.8876,
+      "step": 2486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.001233135928483,
+      "learning_rate": 1.4496787838268378e-05,
+      "loss": 0.3707,
+      "step": 2487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3222148961601148,
+      "learning_rate": 1.4492470328250937e-05,
+      "loss": 0.9267,
+      "step": 2488
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.321131517811123,
+      "learning_rate": 1.4488151768791981e-05,
+      "loss": 0.8573,
+      "step": 2489
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3838353273818826,
+      "learning_rate": 1.4483832160900326e-05,
+      "loss": 0.7848,
+      "step": 2490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5302241371602243,
+      "learning_rate": 1.447951150558503e-05,
+      "loss": 0.8495,
+      "step": 2491
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.383114049112275,
+      "learning_rate": 1.4475189803855399e-05,
+      "loss": 0.8219,
+      "step": 2492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.379191325250921,
+      "learning_rate": 1.4470867056720986e-05,
+      "loss": 0.8633,
+      "step": 2493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.374261044533996,
+      "learning_rate": 1.4466543265191581e-05,
+      "loss": 0.8753,
+      "step": 2494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5698929814633877,
+      "learning_rate": 1.4462218430277224e-05,
+      "loss": 0.9101,
+      "step": 2495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.754659180994369,
+      "learning_rate": 1.44578925529882e-05,
+      "loss": 0.8941,
+      "step": 2496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3555804118684756,
+      "learning_rate": 1.445356563433503e-05,
+      "loss": 0.8446,
+      "step": 2497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5655428107230935,
+      "learning_rate": 1.4449237675328483e-05,
+      "loss": 0.9254,
+      "step": 2498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4179944024248319,
+      "learning_rate": 1.4444908676979572e-05,
+      "loss": 0.7922,
+      "step": 2499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4306349878318732,
+      "learning_rate": 1.4440578640299554e-05,
+      "loss": 0.9124,
+      "step": 2500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4057487375892963,
+      "learning_rate": 1.443624756629992e-05,
+      "loss": 0.7937,
+      "step": 2501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.959703629237683,
+      "learning_rate": 1.4431915455992416e-05,
+      "loss": 0.3069,
+      "step": 2502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3794660485556631,
+      "learning_rate": 1.442758231038902e-05,
+      "loss": 0.878,
+      "step": 2503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.615842428578242,
+      "learning_rate": 1.442324813050195e-05,
+      "loss": 0.9043,
+      "step": 2504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4424541974853513,
+      "learning_rate": 1.441891291734368e-05,
+      "loss": 0.8685,
+      "step": 2505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8013155439881876,
+      "learning_rate": 1.4414576671926912e-05,
+      "loss": 0.3216,
+      "step": 2506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7248805028446959,
+      "learning_rate": 1.4410239395264594e-05,
+      "loss": 0.8993,
+      "step": 2507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6101262083541032,
+      "learning_rate": 1.4405901088369912e-05,
+      "loss": 0.852,
+      "step": 2508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.40907005430812,
+      "learning_rate": 1.4401561752256299e-05,
+      "loss": 0.9285,
+      "step": 2509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4433310083949351,
+      "learning_rate": 1.4397221387937421e-05,
+      "loss": 0.9448,
+      "step": 2510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.464338085634876,
+      "learning_rate": 1.4392879996427187e-05,
+      "loss": 0.8269,
+      "step": 2511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4392826537091399,
+      "learning_rate": 1.438853757873975e-05,
+      "loss": 0.8649,
+      "step": 2512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3955350783763023,
+      "learning_rate": 1.43841941358895e-05,
+      "loss": 0.9145,
+      "step": 2513
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.47521857063144,
+      "learning_rate": 1.437984966889106e-05,
+      "loss": 0.9298,
+      "step": 2514
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.637925496356776,
+      "learning_rate": 1.4375504178759301e-05,
+      "loss": 0.8291,
+      "step": 2515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3408259942475451,
+      "learning_rate": 1.437115766650933e-05,
+      "loss": 0.9389,
+      "step": 2516
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.556762069550069,
+      "learning_rate": 1.4366810133156495e-05,
+      "loss": 0.8537,
+      "step": 2517
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7535168017856988,
+      "learning_rate": 1.4362461579716373e-05,
+      "loss": 0.8531,
+      "step": 2518
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2898085147763103,
+      "learning_rate": 1.4358112007204796e-05,
+      "loss": 0.8193,
+      "step": 2519
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4160184005183938,
+      "learning_rate": 1.4353761416637819e-05,
+      "loss": 0.9242,
+      "step": 2520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4505937992117637,
+      "learning_rate": 1.434940980903174e-05,
+      "loss": 0.8706,
+      "step": 2521
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3897966418449819,
+      "learning_rate": 1.43450571854031e-05,
+      "loss": 0.9258,
+      "step": 2522
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.449961250374203,
+      "learning_rate": 1.4340703546768664e-05,
+      "loss": 0.9013,
+      "step": 2523
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3299021569474143,
+      "learning_rate": 1.4336348894145448e-05,
+      "loss": 0.8777,
+      "step": 2524
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3679749891181332,
+      "learning_rate": 1.43319932285507e-05,
+      "loss": 0.8531,
+      "step": 2525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.225261738820702,
+      "learning_rate": 1.4327636551001902e-05,
+      "loss": 0.9164,
+      "step": 2526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4582549720527263,
+      "learning_rate": 1.4323278862516774e-05,
+      "loss": 0.9147,
+      "step": 2527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3273892573626784,
+      "learning_rate": 1.431892016411327e-05,
+      "loss": 0.812,
+      "step": 2528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5271798384926805,
+      "learning_rate": 1.4314560456809592e-05,
+      "loss": 0.8963,
+      "step": 2529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4835462082220499,
+      "learning_rate": 1.4310199741624157e-05,
+      "loss": 0.9723,
+      "step": 2530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5159125964632234,
+      "learning_rate": 1.4305838019575633e-05,
+      "loss": 0.9049,
+      "step": 2531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4289726661630802,
+      "learning_rate": 1.430147529168292e-05,
+      "loss": 0.8788,
+      "step": 2532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4576685281376076,
+      "learning_rate": 1.429711155896515e-05,
+      "loss": 0.961,
+      "step": 2533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4067202397745056,
+      "learning_rate": 1.4292746822441692e-05,
+      "loss": 0.9013,
+      "step": 2534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4856733654476078,
+      "learning_rate": 1.4288381083132149e-05,
+      "loss": 0.8861,
+      "step": 2535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3843936070726652,
+      "learning_rate": 1.4284014342056355e-05,
+      "loss": 0.8702,
+      "step": 2536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4576533423873292,
+      "learning_rate": 1.4279646600234388e-05,
+      "loss": 0.7968,
+      "step": 2537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2553881477494369,
+      "learning_rate": 1.4275277858686548e-05,
+      "loss": 0.8991,
+      "step": 2538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5183692303713805,
+      "learning_rate": 1.4270908118433371e-05,
+      "loss": 0.9356,
+      "step": 2539
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5235294558820247,
+      "learning_rate": 1.4266537380495634e-05,
+      "loss": 0.8765,
+      "step": 2540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4493190018662951,
+      "learning_rate": 1.4262165645894342e-05,
+      "loss": 0.9116,
+      "step": 2541
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3218951946039377,
+      "learning_rate": 1.4257792915650728e-05,
+      "loss": 0.9227,
+      "step": 2542
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.440647004108404,
+      "learning_rate": 1.4253419190786264e-05,
+      "loss": 0.8283,
+      "step": 2543
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6219193826420353,
+      "learning_rate": 1.4249044472322657e-05,
+      "loss": 0.9475,
+      "step": 2544
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6038560622084053,
+      "learning_rate": 1.4244668761281834e-05,
+      "loss": 0.9073,
+      "step": 2545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.416289752865661,
+      "learning_rate": 1.424029205868597e-05,
+      "loss": 0.9425,
+      "step": 2546
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5738836681992094,
+      "learning_rate": 1.4235914365557455e-05,
+      "loss": 0.8851,
+      "step": 2547
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5041053702079985,
+      "learning_rate": 1.4231535682918923e-05,
+      "loss": 0.8746,
+      "step": 2548
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4937466239043116,
+      "learning_rate": 1.4227156011793234e-05,
+      "loss": 1.0052,
+      "step": 2549
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.41035156006992,
+      "learning_rate": 1.422277535320348e-05,
+      "loss": 0.856,
+      "step": 2550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4068458526592393,
+      "learning_rate": 1.4218393708172982e-05,
+      "loss": 0.8929,
+      "step": 2551
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.59927223481385,
+      "learning_rate": 1.4214011077725293e-05,
+      "loss": 0.8949,
+      "step": 2552
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6144970233726026,
+      "learning_rate": 1.4209627462884198e-05,
+      "loss": 0.8937,
+      "step": 2553
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5561086901696273,
+      "learning_rate": 1.4205242864673707e-05,
+      "loss": 0.8422,
+      "step": 2554
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5266038359752994,
+      "learning_rate": 1.4200857284118067e-05,
+      "loss": 0.8968,
+      "step": 2555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3990174674567508,
+      "learning_rate": 1.4196470722241741e-05,
+      "loss": 0.8887,
+      "step": 2556
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.515273129527405,
+      "learning_rate": 1.4192083180069441e-05,
+      "loss": 0.8738,
+      "step": 2557
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4476279430689292,
+      "learning_rate": 1.4187694658626092e-05,
+      "loss": 0.9118,
+      "step": 2558
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4417288956678438,
+      "learning_rate": 1.4183305158936851e-05,
+      "loss": 0.9134,
+      "step": 2559
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6423951703669446,
+      "learning_rate": 1.417891468202711e-05,
+      "loss": 0.9373,
+      "step": 2560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3216423385662226,
+      "learning_rate": 1.4174523228922486e-05,
+      "loss": 0.8462,
+      "step": 2561
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5985498727245373,
+      "learning_rate": 1.4170130800648814e-05,
+      "loss": 0.9322,
+      "step": 2562
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3634116341344829,
+      "learning_rate": 1.4165737398232173e-05,
+      "loss": 0.8917,
+      "step": 2563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.332052343235081,
+      "learning_rate": 1.4161343022698864e-05,
+      "loss": 0.8639,
+      "step": 2564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3308551013173646,
+      "learning_rate": 1.4156947675075408e-05,
+      "loss": 0.9535,
+      "step": 2565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.403820455749816,
+      "learning_rate": 1.415255135638856e-05,
+      "loss": 0.8465,
+      "step": 2566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.466444409991185,
+      "learning_rate": 1.4148154067665305e-05,
+      "loss": 0.9115,
+      "step": 2567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3660410522513904,
+      "learning_rate": 1.4143755809932843e-05,
+      "loss": 0.8438,
+      "step": 2568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5002762631980608,
+      "learning_rate": 1.4139356584218614e-05,
+      "loss": 0.9316,
+      "step": 2569
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4724129386017795,
+      "learning_rate": 1.4134956391550269e-05,
+      "loss": 0.898,
+      "step": 2570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5837350430380717,
+      "learning_rate": 1.4130555232955706e-05,
+      "loss": 0.8268,
+      "step": 2571
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.628259001926628,
+      "learning_rate": 1.4126153109463025e-05,
+      "loss": 0.8868,
+      "step": 2572
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4678294792314015,
+      "learning_rate": 1.4121750022100566e-05,
+      "loss": 0.848,
+      "step": 2573
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.517896945215126,
+      "learning_rate": 1.4117345971896894e-05,
+      "loss": 0.8622,
+      "step": 2574
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5606168366028863,
+      "learning_rate": 1.4112940959880791e-05,
+      "loss": 0.9071,
+      "step": 2575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4410772581355227,
+      "learning_rate": 1.4108534987081273e-05,
+      "loss": 0.8863,
+      "step": 2576
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.569678484617207,
+      "learning_rate": 1.410412805452757e-05,
+      "loss": 0.7814,
+      "step": 2577
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4276863571473113,
+      "learning_rate": 1.4099720163249144e-05,
+      "loss": 0.9419,
+      "step": 2578
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3564115997969606,
+      "learning_rate": 1.4095311314275678e-05,
+      "loss": 0.9409,
+      "step": 2579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6187014634172197,
+      "learning_rate": 1.4090901508637087e-05,
+      "loss": 0.8859,
+      "step": 2580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7330198997558404,
+      "learning_rate": 1.4086490747363492e-05,
+      "loss": 0.8812,
+      "step": 2581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3933799356067045,
+      "learning_rate": 1.4082079031485253e-05,
+      "loss": 0.7615,
+      "step": 2582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4326675200770536,
+      "learning_rate": 1.4077666362032942e-05,
+      "loss": 0.8636,
+      "step": 2583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3985332690643688,
+      "learning_rate": 1.4073252740037367e-05,
+      "loss": 0.8511,
+      "step": 2584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5035865493422815,
+      "learning_rate": 1.4068838166529544e-05,
+      "loss": 0.9004,
+      "step": 2585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6097473077303504,
+      "learning_rate": 1.4064422642540719e-05,
+      "loss": 0.9268,
+      "step": 2586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3760840283967968,
+      "learning_rate": 1.4060006169102363e-05,
+      "loss": 0.822,
+      "step": 2587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5573301319859965,
+      "learning_rate": 1.4055588747246158e-05,
+      "loss": 0.824,
+      "step": 2588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4205612641202485,
+      "learning_rate": 1.4051170378004018e-05,
+      "loss": 0.8696,
+      "step": 2589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4251927402301132,
+      "learning_rate": 1.4046751062408076e-05,
+      "loss": 0.857,
+      "step": 2590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4526906742365633,
+      "learning_rate": 1.4042330801490682e-05,
+      "loss": 0.9346,
+      "step": 2591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6838225100847595,
+      "learning_rate": 1.403790959628441e-05,
+      "loss": 0.9334,
+      "step": 2592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5125729601874176,
+      "learning_rate": 1.4033487447822053e-05,
+      "loss": 0.8971,
+      "step": 2593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4759423376509302,
+      "learning_rate": 1.4029064357136628e-05,
+      "loss": 0.9267,
+      "step": 2594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5474866226262085,
+      "learning_rate": 1.4024640325261367e-05,
+      "loss": 0.8046,
+      "step": 2595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3968421444399874,
+      "learning_rate": 1.4020215353229726e-05,
+      "loss": 0.8745,
+      "step": 2596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4674226566319317,
+      "learning_rate": 1.4015789442075376e-05,
+      "loss": 0.8347,
+      "step": 2597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4305886668626155,
+      "learning_rate": 1.4011362592832214e-05,
+      "loss": 0.8437,
+      "step": 2598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6321575251986986,
+      "learning_rate": 1.400693480653435e-05,
+      "loss": 0.9749,
+      "step": 2599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4181709736470574,
+      "learning_rate": 1.4002506084216117e-05,
+      "loss": 0.9266,
+      "step": 2600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7947765453393132,
+      "learning_rate": 1.3998076426912061e-05,
+      "loss": 0.9106,
+      "step": 2601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5158074044306045,
+      "learning_rate": 1.3993645835656955e-05,
+      "loss": 0.7601,
+      "step": 2602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5324432439867566,
+      "learning_rate": 1.3989214311485787e-05,
+      "loss": 0.9008,
+      "step": 2603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6704085549907486,
+      "learning_rate": 1.3984781855433756e-05,
+      "loss": 0.8627,
+      "step": 2604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6257902754643552,
+      "learning_rate": 1.3980348468536287e-05,
+      "loss": 0.8764,
+      "step": 2605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.239993048640492,
+      "learning_rate": 1.3975914151829017e-05,
+      "loss": 0.9321,
+      "step": 2606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8529626254514404,
+      "learning_rate": 1.3971478906347806e-05,
+      "loss": 0.8682,
+      "step": 2607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8448329417548159,
+      "learning_rate": 1.396704273312873e-05,
+      "loss": 0.3175,
+      "step": 2608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.451862067106698,
+      "learning_rate": 1.3962605633208073e-05,
+      "loss": 0.8604,
+      "step": 2609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4702687185520813,
+      "learning_rate": 1.3958167607622347e-05,
+      "loss": 0.9232,
+      "step": 2610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6512774612847272,
+      "learning_rate": 1.3953728657408272e-05,
+      "loss": 0.8766,
+      "step": 2611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.325563291145746,
+      "learning_rate": 1.394928878360279e-05,
+      "loss": 0.7663,
+      "step": 2612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6850175391042796,
+      "learning_rate": 1.3944847987243056e-05,
+      "loss": 0.8568,
+      "step": 2613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5044241626499406,
+      "learning_rate": 1.3940406269366436e-05,
+      "loss": 0.8196,
+      "step": 2614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5835866252322985,
+      "learning_rate": 1.393596363101052e-05,
+      "loss": 0.8259,
+      "step": 2615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4622932992060935,
+      "learning_rate": 1.393152007321311e-05,
+      "loss": 0.88,
+      "step": 2616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3047974677197762,
+      "learning_rate": 1.3927075597012215e-05,
+      "loss": 0.8773,
+      "step": 2617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.401477291021865,
+      "learning_rate": 1.3922630203446072e-05,
+      "loss": 0.8362,
+      "step": 2618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5605621014879874,
+      "learning_rate": 1.3918183893553123e-05,
+      "loss": 0.7719,
+      "step": 2619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3756526135208338,
+      "learning_rate": 1.3913736668372027e-05,
+      "loss": 0.9267,
+      "step": 2620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.361235066055886,
+      "learning_rate": 1.3909288528941653e-05,
+      "loss": 0.8024,
+      "step": 2621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5027505211375485,
+      "learning_rate": 1.3904839476301091e-05,
+      "loss": 0.8662,
+      "step": 2622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.52035786998657,
+      "learning_rate": 1.3900389511489639e-05,
+      "loss": 0.8855,
+      "step": 2623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.276641356566796,
+      "learning_rate": 1.3895938635546804e-05,
+      "loss": 0.8484,
+      "step": 2624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.404555107323176,
+      "learning_rate": 1.3891486849512322e-05,
+      "loss": 0.8914,
+      "step": 2625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4265557576215502,
+      "learning_rate": 1.388703415442612e-05,
+      "loss": 0.9581,
+      "step": 2626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3749721826372947,
+      "learning_rate": 1.388258055132835e-05,
+      "loss": 0.8745,
+      "step": 2627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4407921583939474,
+      "learning_rate": 1.3878126041259381e-05,
+      "loss": 0.9058,
+      "step": 2628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.412824725886247,
+      "learning_rate": 1.3873670625259781e-05,
+      "loss": 0.8879,
+      "step": 2629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.395833299824254,
+      "learning_rate": 1.3869214304370338e-05,
+      "loss": 0.9118,
+      "step": 2630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4325023038185696,
+      "learning_rate": 1.3864757079632045e-05,
+      "loss": 0.9031,
+      "step": 2631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4400560275844982,
+      "learning_rate": 1.3860298952086118e-05,
+      "loss": 0.937,
+      "step": 2632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6120185644381455,
+      "learning_rate": 1.3855839922773968e-05,
+      "loss": 0.9211,
+      "step": 2633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8594615979345239,
+      "learning_rate": 1.3851379992737229e-05,
+      "loss": 0.3214,
+      "step": 2634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6059541067234224,
+      "learning_rate": 1.3846919163017738e-05,
+      "loss": 0.915,
+      "step": 2635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5222039571677914,
+      "learning_rate": 1.3842457434657548e-05,
+      "loss": 0.8153,
+      "step": 2636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4289934813714156,
+      "learning_rate": 1.383799480869892e-05,
+      "loss": 0.9169,
+      "step": 2637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4580096278655192,
+      "learning_rate": 1.3833531286184324e-05,
+      "loss": 0.8961,
+      "step": 2638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.431852256100158,
+      "learning_rate": 1.3829066868156433e-05,
+      "loss": 0.9564,
+      "step": 2639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.533394514882259,
+      "learning_rate": 1.382460155565814e-05,
+      "loss": 0.9108,
+      "step": 2640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5078492296024473,
+      "learning_rate": 1.3820135349732546e-05,
+      "loss": 0.8217,
+      "step": 2641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.547073269670143,
+      "learning_rate": 1.3815668251422953e-05,
+      "loss": 0.9108,
+      "step": 2642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.844719619697523,
+      "learning_rate": 1.3811200261772877e-05,
+      "loss": 0.8304,
+      "step": 2643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7898835927873684,
+      "learning_rate": 1.380673138182604e-05,
+      "loss": 0.8516,
+      "step": 2644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5937448126343847,
+      "learning_rate": 1.3802261612626372e-05,
+      "loss": 0.83,
+      "step": 2645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4429464407674535,
+      "learning_rate": 1.3797790955218014e-05,
+      "loss": 0.9524,
+      "step": 2646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3986477047889547,
+      "learning_rate": 1.3793319410645307e-05,
+      "loss": 0.9305,
+      "step": 2647
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5353157031019835,
+      "learning_rate": 1.378884697995281e-05,
+      "loss": 0.8977,
+      "step": 2648
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3383483201282307,
+      "learning_rate": 1.3784373664185282e-05,
+      "loss": 0.9332,
+      "step": 2649
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4277754893821355,
+      "learning_rate": 1.3779899464387688e-05,
+      "loss": 0.8646,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5481888184023798,
+      "learning_rate": 1.3775424381605205e-05,
+      "loss": 0.9324,
+      "step": 2651
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4714686356946542,
+      "learning_rate": 1.3770948416883205e-05,
+      "loss": 0.8804,
+      "step": 2652
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4669755383672847,
+      "learning_rate": 1.3766471571267284e-05,
+      "loss": 0.885,
+      "step": 2653
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3552008324215652,
+      "learning_rate": 1.3761993845803225e-05,
+      "loss": 0.9643,
+      "step": 2654
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8468283420463011,
+      "learning_rate": 1.3757515241537031e-05,
+      "loss": 0.3217,
+      "step": 2655
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2920960941758886,
+      "learning_rate": 1.3753035759514904e-05,
+      "loss": 0.9343,
+      "step": 2656
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5091857740453016,
+      "learning_rate": 1.3748555400783245e-05,
+      "loss": 0.9207,
+      "step": 2657
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4299632735398338,
+      "learning_rate": 1.3744074166388677e-05,
+      "loss": 0.8937,
+      "step": 2658
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3905957960962714,
+      "learning_rate": 1.3739592057378005e-05,
+      "loss": 0.8504,
+      "step": 2659
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5850265228465883,
+      "learning_rate": 1.3735109074798259e-05,
+      "loss": 0.8316,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3664090733419905,
+      "learning_rate": 1.373062521969666e-05,
+      "loss": 0.8936,
+      "step": 2661
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.595661939045099,
+      "learning_rate": 1.3726140493120639e-05,
+      "loss": 0.8979,
+      "step": 2662
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3444722197550636,
+      "learning_rate": 1.3721654896117826e-05,
+      "loss": 0.9154,
+      "step": 2663
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5877124479957379,
+      "learning_rate": 1.3717168429736061e-05,
+      "loss": 0.9277,
+      "step": 2664
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4310795171006792,
+      "learning_rate": 1.3712681095023376e-05,
+      "loss": 0.8793,
+      "step": 2665
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.35458440404843,
+      "learning_rate": 1.370819289302802e-05,
+      "loss": 0.9304,
+      "step": 2666
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4810619876746554,
+      "learning_rate": 1.3703703824798438e-05,
+      "loss": 0.856,
+      "step": 2667
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4018382832408174,
+      "learning_rate": 1.369921389138327e-05,
+      "loss": 0.8895,
+      "step": 2668
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5792172271039533,
+      "learning_rate": 1.369472309383137e-05,
+      "loss": 0.9842,
+      "step": 2669
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4482192489752068,
+      "learning_rate": 1.3690231433191787e-05,
+      "loss": 0.8758,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4456435069376843,
+      "learning_rate": 1.3685738910513771e-05,
+      "loss": 0.8542,
+      "step": 2671
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.336789119056566,
+      "learning_rate": 1.3681245526846782e-05,
+      "loss": 0.8027,
+      "step": 2672
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4665755950736112,
+      "learning_rate": 1.3676751283240469e-05,
+      "loss": 0.8532,
+      "step": 2673
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5622988571399659,
+      "learning_rate": 1.367225618074469e-05,
+      "loss": 0.8718,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4680124032471622,
+      "learning_rate": 1.3667760220409503e-05,
+      "loss": 0.8207,
+      "step": 2675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.429503164498111,
+      "learning_rate": 1.366326340328516e-05,
+      "loss": 0.898,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5106450008308003,
+      "learning_rate": 1.3658765730422126e-05,
+      "loss": 0.9109,
+      "step": 2677
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4102071597031536,
+      "learning_rate": 1.3654267202871047e-05,
+      "loss": 0.9364,
+      "step": 2678
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5346638991677883,
+      "learning_rate": 1.3649767821682788e-05,
+      "loss": 0.8921,
+      "step": 2679
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4788022563460717,
+      "learning_rate": 1.3645267587908404e-05,
+      "loss": 0.9039,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3789313512630113,
+      "learning_rate": 1.3640766502599148e-05,
+      "loss": 0.8535,
+      "step": 2681
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3797766056750267,
+      "learning_rate": 1.3636264566806473e-05,
+      "loss": 0.8912,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5794891260456347,
+      "learning_rate": 1.3631761781582031e-05,
+      "loss": 0.7905,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3700951909578696,
+      "learning_rate": 1.3627258147977678e-05,
+      "loss": 0.8504,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4665715543588154,
+      "learning_rate": 1.3622753667045459e-05,
+      "loss": 0.8409,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4286278324375734,
+      "learning_rate": 1.361824833983762e-05,
+      "loss": 0.8186,
+      "step": 2686
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2618747997035793,
+      "learning_rate": 1.3613742167406614e-05,
+      "loss": 0.8608,
+      "step": 2687
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4133239208251391,
+      "learning_rate": 1.3609235150805074e-05,
+      "loss": 0.8741,
+      "step": 2688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5103995101601069,
+      "learning_rate": 1.3604727291085845e-05,
+      "loss": 0.8436,
+      "step": 2689
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.490327124752854,
+      "learning_rate": 1.3600218589301962e-05,
+      "loss": 0.8769,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2291126900239977,
+      "learning_rate": 1.3595709046506656e-05,
+      "loss": 0.9348,
+      "step": 2691
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3683233073134076,
+      "learning_rate": 1.3591198663753358e-05,
+      "loss": 0.8823,
+      "step": 2692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9280492408896337,
+      "learning_rate": 1.3586687442095697e-05,
+      "loss": 0.3374,
+      "step": 2693
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5655953664078606,
+      "learning_rate": 1.3582175382587491e-05,
+      "loss": 0.7969,
+      "step": 2694
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7751709568192449,
+      "learning_rate": 1.3577662486282757e-05,
+      "loss": 0.9202,
+      "step": 2695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3914000997776887,
+      "learning_rate": 1.3573148754235712e-05,
+      "loss": 0.8776,
+      "step": 2696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.292619574496971,
+      "learning_rate": 1.3568634187500762e-05,
+      "loss": 0.8441,
+      "step": 2697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6721705934102595,
+      "learning_rate": 1.3564118787132507e-05,
+      "loss": 0.8796,
+      "step": 2698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8622105164694549,
+      "learning_rate": 1.3559602554185751e-05,
+      "loss": 0.3033,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3139195180990364,
+      "learning_rate": 1.3555085489715487e-05,
+      "loss": 0.7475,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4395956890155168,
+      "learning_rate": 1.3550567594776893e-05,
+      "loss": 0.8933,
+      "step": 2701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.393412083504995,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.8661,
+      "step": 2702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3390615281866014,
+      "learning_rate": 1.3541529317716453e-05,
+      "loss": 0.8396,
+      "step": 2703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4944494302311004,
+      "learning_rate": 1.3537008937705947e-05,
+      "loss": 0.8495,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4588533914641837,
+      "learning_rate": 1.35324877314498e-05,
+      "loss": 0.8481,
+      "step": 2705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7951588388786142,
+      "learning_rate": 1.3527965700004168e-05,
+      "loss": 0.3305,
+      "step": 2706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4278610475969367,
+      "learning_rate": 1.3523442844425393e-05,
+      "loss": 0.8847,
+      "step": 2707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2566448917144541,
+      "learning_rate": 1.3518919165770023e-05,
+      "loss": 0.8673,
+      "step": 2708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3868269139403437,
+      "learning_rate": 1.3514394665094786e-05,
+      "loss": 0.886,
+      "step": 2709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.23748714423486,
+      "learning_rate": 1.3509869343456603e-05,
+      "loss": 0.8737,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4958673902288788,
+      "learning_rate": 1.350534320191259e-05,
+      "loss": 0.9207,
+      "step": 2711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3222347816577644,
+      "learning_rate": 1.3500816241520059e-05,
+      "loss": 0.8255,
+      "step": 2712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.442461306089769,
+      "learning_rate": 1.3496288463336504e-05,
+      "loss": 0.9017,
+      "step": 2713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4070497794893102,
+      "learning_rate": 1.3491759868419616e-05,
+      "loss": 0.8347,
+      "step": 2714
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4676779559228024,
+      "learning_rate": 1.3487230457827273e-05,
+      "loss": 0.8356,
+      "step": 2715
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5534142834991447,
+      "learning_rate": 1.3482700232617552e-05,
+      "loss": 0.9119,
+      "step": 2716
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6512695364839567,
+      "learning_rate": 1.3478169193848705e-05,
+      "loss": 0.8854,
+      "step": 2717
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9404515292546731,
+      "learning_rate": 1.3473637342579191e-05,
+      "loss": 0.8552,
+      "step": 2718
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.638153216880537,
+      "learning_rate": 1.3469104679867646e-05,
+      "loss": 0.9331,
+      "step": 2719
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4718691590622501,
+      "learning_rate": 1.3464571206772903e-05,
+      "loss": 0.8541,
+      "step": 2720
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5376357567174224,
+      "learning_rate": 1.346003692435398e-05,
+      "loss": 0.8681,
+      "step": 2721
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6025152297633412,
+      "learning_rate": 1.3455501833670089e-05,
+      "loss": 0.8213,
+      "step": 2722
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4857630332175427,
+      "learning_rate": 1.3450965935780622e-05,
+      "loss": 0.8519,
+      "step": 2723
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.35435945937307,
+      "learning_rate": 1.344642923174517e-05,
+      "loss": 0.82,
+      "step": 2724
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5257038937719627,
+      "learning_rate": 1.3441891722623507e-05,
+      "loss": 0.8516,
+      "step": 2725
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3854099625470668,
+      "learning_rate": 1.3437353409475596e-05,
+      "loss": 0.9315,
+      "step": 2726
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.02330935101869,
+      "learning_rate": 1.3432814293361585e-05,
+      "loss": 0.8323,
+      "step": 2727
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3516827189445648,
+      "learning_rate": 1.3428274375341812e-05,
+      "loss": 0.8364,
+      "step": 2728
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.298260055448875,
+      "learning_rate": 1.3423733656476806e-05,
+      "loss": 0.875,
+      "step": 2729
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9402579985671157,
+      "learning_rate": 1.341919213782727e-05,
+      "loss": 0.3317,
+      "step": 2730
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.404877551225526,
+      "learning_rate": 1.3414649820454118e-05,
+      "loss": 0.8369,
+      "step": 2731
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4569397263506838,
+      "learning_rate": 1.3410106705418424e-05,
+      "loss": 0.9473,
+      "step": 2732
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4835074386579736,
+      "learning_rate": 1.3405562793781463e-05,
+      "loss": 0.8485,
+      "step": 2733
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8521980191219083,
+      "learning_rate": 1.3401018086604698e-05,
+      "loss": 0.3556,
+      "step": 2734
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3997092684338868,
+      "learning_rate": 1.3396472584949765e-05,
+      "loss": 0.7977,
+      "step": 2735
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8093156834033487,
+      "learning_rate": 1.33919262898785e-05,
+      "loss": 0.3449,
+      "step": 2736
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2633286338676704,
+      "learning_rate": 1.3387379202452917e-05,
+      "loss": 0.8237,
+      "step": 2737
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7981384334420416,
+      "learning_rate": 1.3382831323735213e-05,
+      "loss": 0.9182,
+      "step": 2738
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.463791416464488,
+      "learning_rate": 1.3378282654787774e-05,
+      "loss": 0.8949,
+      "step": 2739
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5331657957890301,
+      "learning_rate": 1.337373319667317e-05,
+      "loss": 0.8081,
+      "step": 2740
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.513597267527198,
+      "learning_rate": 1.3369182950454155e-05,
+      "loss": 0.9781,
+      "step": 2741
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.569337832978486,
+      "learning_rate": 1.3364631917193671e-05,
+      "loss": 0.8654,
+      "step": 2742
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5340515000606945,
+      "learning_rate": 1.3360080097954833e-05,
+      "loss": 0.8375,
+      "step": 2743
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9859001605063671,
+      "learning_rate": 1.3355527493800948e-05,
+      "loss": 0.3091,
+      "step": 2744
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3190392980272423,
+      "learning_rate": 1.3350974105795511e-05,
+      "loss": 0.8391,
+      "step": 2745
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3109891098432258,
+      "learning_rate": 1.3346419935002186e-05,
+      "loss": 0.8052,
+      "step": 2746
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4070143125945398,
+      "learning_rate": 1.3341864982484828e-05,
+      "loss": 0.8624,
+      "step": 2747
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4746437384536173,
+      "learning_rate": 1.3337309249307482e-05,
+      "loss": 0.8636,
+      "step": 2748
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4646471349372367,
+      "learning_rate": 1.333275273653436e-05,
+      "loss": 0.9017,
+      "step": 2749
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3825523231380634,
+      "learning_rate": 1.3328195445229869e-05,
+      "loss": 0.9316,
+      "step": 2750
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4205471168177572,
+      "learning_rate": 1.332363737645859e-05,
+      "loss": 0.8697,
+      "step": 2751
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4143053367920866,
+      "learning_rate": 1.3319078531285286e-05,
+      "loss": 0.8981,
+      "step": 2752
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5796261861355516,
+      "learning_rate": 1.331451891077491e-05,
+      "loss": 0.958,
+      "step": 2753
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3784716545349904,
+      "learning_rate": 1.3309958515992585e-05,
+      "loss": 0.7835,
+      "step": 2754
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3714778354024928,
+      "learning_rate": 1.3305397348003618e-05,
+      "loss": 0.8958,
+      "step": 2755
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5493638686283826,
+      "learning_rate": 1.3300835407873505e-05,
+      "loss": 0.8977,
+      "step": 2756
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4634259757006258,
+      "learning_rate": 1.329627269666791e-05,
+      "loss": 0.8462,
+      "step": 2757
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3447874882345294,
+      "learning_rate": 1.3291709215452686e-05,
+      "loss": 0.8637,
+      "step": 2758
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.46631777605198,
+      "learning_rate": 1.3287144965293858e-05,
+      "loss": 0.8598,
+      "step": 2759
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5271707153842025,
+      "learning_rate": 1.3282579947257643e-05,
+      "loss": 0.949,
+      "step": 2760
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6294027619215443,
+      "learning_rate": 1.3278014162410425e-05,
+      "loss": 0.7969,
+      "step": 2761
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4982752458060427,
+      "learning_rate": 1.3273447611818768e-05,
+      "loss": 0.9121,
+      "step": 2762
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2917002890970246,
+      "learning_rate": 1.3268880296549424e-05,
+      "loss": 0.8878,
+      "step": 2763
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4771702313003898,
+      "learning_rate": 1.3264312217669321e-05,
+      "loss": 0.9529,
+      "step": 2764
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.374160582171779,
+      "learning_rate": 1.3259743376245556e-05,
+      "loss": 0.8503,
+      "step": 2765
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4679637843885989,
+      "learning_rate": 1.3255173773345413e-05,
+      "loss": 0.9244,
+      "step": 2766
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2875655846187317,
+      "learning_rate": 1.3250603410036356e-05,
+      "loss": 0.8389,
+      "step": 2767
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.499564603731924,
+      "learning_rate": 1.3246032287386015e-05,
+      "loss": 0.9563,
+      "step": 2768
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6576115628009187,
+      "learning_rate": 1.3241460406462208e-05,
+      "loss": 0.7855,
+      "step": 2769
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.517047036766871,
+      "learning_rate": 1.3236887768332927e-05,
+      "loss": 0.8999,
+      "step": 2770
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3336785770236914,
+      "learning_rate": 1.3232314374066343e-05,
+      "loss": 0.9127,
+      "step": 2771
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3143820552119605,
+      "learning_rate": 1.3227740224730799e-05,
+      "loss": 0.8981,
+      "step": 2772
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4757263298579144,
+      "learning_rate": 1.3223165321394814e-05,
+      "loss": 0.8437,
+      "step": 2773
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4442835108983245,
+      "learning_rate": 1.3218589665127093e-05,
+      "loss": 0.8314,
+      "step": 2774
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3315542218495533,
+      "learning_rate": 1.3214013256996503e-05,
+      "loss": 0.9225,
+      "step": 2775
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4025539317722029,
+      "learning_rate": 1.3209436098072095e-05,
+      "loss": 0.9129,
+      "step": 2776
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5196167301562642,
+      "learning_rate": 1.3204858189423097e-05,
+      "loss": 0.9796,
+      "step": 2777
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.346062685541074,
+      "learning_rate": 1.3200279532118907e-05,
+      "loss": 0.8542,
+      "step": 2778
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4084339521429838,
+      "learning_rate": 1.3195700127229097e-05,
+      "loss": 0.8982,
+      "step": 2779
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.346306080269962,
+      "learning_rate": 1.3191119975823421e-05,
+      "loss": 0.8565,
+      "step": 2780
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9174900282828216,
+      "learning_rate": 1.31865390789718e-05,
+      "loss": 0.3242,
+      "step": 2781
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4261095707568456,
+      "learning_rate": 1.3181957437744333e-05,
+      "loss": 0.9327,
+      "step": 2782
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.435811314263392,
+      "learning_rate": 1.3177375053211293e-05,
+      "loss": 0.8541,
+      "step": 2783
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9191109663040323,
+      "learning_rate": 1.3172791926443122e-05,
+      "loss": 0.3497,
+      "step": 2784
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3376890253361224,
+      "learning_rate": 1.3168208058510441e-05,
+      "loss": 0.9509,
+      "step": 2785
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8808116339541431,
+      "learning_rate": 1.316362345048404e-05,
+      "loss": 0.3041,
+      "step": 2786
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5675887242949134,
+      "learning_rate": 1.3159038103434889e-05,
+      "loss": 0.8123,
+      "step": 2787
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.479852674020534,
+      "learning_rate": 1.3154452018434123e-05,
+      "loss": 0.8988,
+      "step": 2788
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2273852472738545,
+      "learning_rate": 1.3149865196553049e-05,
+      "loss": 0.8744,
+      "step": 2789
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3194277644584498,
+      "learning_rate": 1.3145277638863152e-05,
+      "loss": 0.8952,
+      "step": 2790
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.492389169766071,
+      "learning_rate": 1.3140689346436083e-05,
+      "loss": 0.8553,
+      "step": 2791
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3349625985805669,
+      "learning_rate": 1.3136100320343674e-05,
+      "loss": 0.9191,
+      "step": 2792
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5256027924008635,
+      "learning_rate": 1.3131510561657917e-05,
+      "loss": 1.0017,
+      "step": 2793
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.584737001536214,
+      "learning_rate": 1.3126920071450977e-05,
+      "loss": 0.8844,
+      "step": 2794
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.31418489338073,
+      "learning_rate": 1.3122328850795203e-05,
+      "loss": 0.891,
+      "step": 2795
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5273655605710132,
+      "learning_rate": 1.3117736900763091e-05,
+      "loss": 0.8901,
+      "step": 2796
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4642726611478112,
+      "learning_rate": 1.3113144222427334e-05,
+      "loss": 0.8697,
+      "step": 2797
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.65752365964945,
+      "learning_rate": 1.3108550816860777e-05,
+      "loss": 0.9416,
+      "step": 2798
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4458756735026272,
+      "learning_rate": 1.3103956685136435e-05,
+      "loss": 0.9093,
+      "step": 2799
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.620923480318348,
+      "learning_rate": 1.3099361828327506e-05,
+      "loss": 0.9545,
+      "step": 2800
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9369605205954596,
+      "learning_rate": 1.3094766247507343e-05,
+      "loss": 0.315,
+      "step": 2801
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6217311640079706,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.9124,
+      "step": 2802
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.292977066531907,
+      "learning_rate": 1.3085572918127605e-05,
+      "loss": 0.8573,
+      "step": 2803
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.224009892632372,
+      "learning_rate": 1.3080975171715592e-05,
+      "loss": 0.8639,
+      "step": 2804
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5531301749697768,
+      "learning_rate": 1.3076376705587468e-05,
+      "loss": 0.8572,
+      "step": 2805
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4922651450442632,
+      "learning_rate": 1.3071777520817443e-05,
+      "loss": 0.9102,
+      "step": 2806
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3198714535383125,
+      "learning_rate": 1.3067177618479883e-05,
+      "loss": 0.8952,
+      "step": 2807
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6565717820909742,
+      "learning_rate": 1.3062576999649323e-05,
+      "loss": 0.8719,
+      "step": 2808
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7905667661889144,
+      "learning_rate": 1.3057975665400475e-05,
+      "loss": 0.8703,
+      "step": 2809
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5267722964469028,
+      "learning_rate": 1.3053373616808202e-05,
+      "loss": 0.8753,
+      "step": 2810
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3703134209967969,
+      "learning_rate": 1.3048770854947553e-05,
+      "loss": 0.9015,
+      "step": 2811
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.330131053571653,
+      "learning_rate": 1.3044167380893726e-05,
+      "loss": 0.7464,
+      "step": 2812
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4087981548885895,
+      "learning_rate": 1.3039563195722096e-05,
+      "loss": 0.8774,
+      "step": 2813
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.43548309807309,
+      "learning_rate": 1.3034958300508197e-05,
+      "loss": 0.8509,
+      "step": 2814
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4876310374941417,
+      "learning_rate": 1.3030352696327741e-05,
+      "loss": 0.8381,
+      "step": 2815
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5924307386096148,
+      "learning_rate": 1.3025746384256595e-05,
+      "loss": 0.9437,
+      "step": 2816
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3512327999331264,
+      "learning_rate": 1.3021139365370787e-05,
+      "loss": 0.8617,
+      "step": 2817
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5995053768701828,
+      "learning_rate": 1.3016531640746524e-05,
+      "loss": 0.8566,
+      "step": 2818
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4391657530784563,
+      "learning_rate": 1.301192321146017e-05,
+      "loss": 0.9008,
+      "step": 2819
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4061270183193015,
+      "learning_rate": 1.3007314078588255e-05,
+      "loss": 0.8893,
+      "step": 2820
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5721682773094316,
+      "learning_rate": 1.3002704243207468e-05,
+      "loss": 0.8499,
+      "step": 2821
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4406997272365956,
+      "learning_rate": 1.2998093706394676e-05,
+      "loss": 0.8869,
+      "step": 2822
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5687787218618148,
+      "learning_rate": 1.2993482469226892e-05,
+      "loss": 0.9351,
+      "step": 2823
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0412201349513743,
+      "learning_rate": 1.2988870532781306e-05,
+      "loss": 0.3252,
+      "step": 2824
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3557584573901178,
+      "learning_rate": 1.2984257898135266e-05,
+      "loss": 0.848,
+      "step": 2825
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5764654127289022,
+      "learning_rate": 1.2979644566366284e-05,
+      "loss": 0.891,
+      "step": 2826
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2571764931420957,
+      "learning_rate": 1.297503053855203e-05,
+      "loss": 0.7778,
+      "step": 2827
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6396797851964806,
+      "learning_rate": 1.297041581577035e-05,
+      "loss": 0.892,
+      "step": 2828
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.600777381261169,
+      "learning_rate": 1.296580039909924e-05,
+      "loss": 0.8459,
+      "step": 2829
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2570862131863656,
+      "learning_rate": 1.296118428961686e-05,
+      "loss": 0.8713,
+      "step": 2830
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4757539967756321,
+      "learning_rate": 1.2956567488401534e-05,
+      "loss": 0.7618,
+      "step": 2831
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5752293889899953,
+      "learning_rate": 1.295194999653175e-05,
+      "loss": 0.8065,
+      "step": 2832
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3536762696806495,
+      "learning_rate": 1.2947331815086153e-05,
+      "loss": 0.8186,
+      "step": 2833
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.420220437465423,
+      "learning_rate": 1.2942712945143547e-05,
+      "loss": 0.8313,
+      "step": 2834
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.535317788591522,
+      "learning_rate": 1.2938093387782908e-05,
+      "loss": 0.7851,
+      "step": 2835
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6624921969718232,
+      "learning_rate": 1.2933473144083359e-05,
+      "loss": 0.8577,
+      "step": 2836
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.597251109143332,
+      "learning_rate": 1.292885221512419e-05,
+      "loss": 0.8359,
+      "step": 2837
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.422663088394019,
+      "learning_rate": 1.2924230601984855e-05,
+      "loss": 0.7572,
+      "step": 2838
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8755452222120067,
+      "learning_rate": 1.291960830574496e-05,
+      "loss": 0.3252,
+      "step": 2839
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5175529002887995,
+      "learning_rate": 1.291498532748427e-05,
+      "loss": 0.8386,
+      "step": 2840
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6590066845415288,
+      "learning_rate": 1.2910361668282718e-05,
+      "loss": 0.8527,
+      "step": 2841
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3547232739467012,
+      "learning_rate": 1.2905737329220394e-05,
+      "loss": 0.8828,
+      "step": 2842
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9352992541283203,
+      "learning_rate": 1.2901112311377536e-05,
+      "loss": 0.8556,
+      "step": 2843
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6668354758563957,
+      "learning_rate": 1.2896486615834557e-05,
+      "loss": 0.3409,
+      "step": 2844
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5135897318793683,
+      "learning_rate": 1.2891860243672014e-05,
+      "loss": 0.8589,
+      "step": 2845
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6238132944134112,
+      "learning_rate": 1.288723319597063e-05,
+      "loss": 0.8711,
+      "step": 2846
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5117542499247825,
+      "learning_rate": 1.2882605473811282e-05,
+      "loss": 0.8952,
+      "step": 2847
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5140260835722537,
+      "learning_rate": 1.2877977078275012e-05,
+      "loss": 0.9237,
+      "step": 2848
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.547050139839024,
+      "learning_rate": 1.2873348010443007e-05,
+      "loss": 0.8401,
+      "step": 2849
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4240158896533306,
+      "learning_rate": 1.2868718271396622e-05,
+      "loss": 0.8486,
+      "step": 2850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5421106270063638,
+      "learning_rate": 1.2864087862217365e-05,
+      "loss": 0.8987,
+      "step": 2851
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5260434383950567,
+      "learning_rate": 1.2859456783986892e-05,
+      "loss": 0.8559,
+      "step": 2852
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.555891695627861,
+      "learning_rate": 1.2854825037787034e-05,
+      "loss": 0.9277,
+      "step": 2853
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3669859451498059,
+      "learning_rate": 1.2850192624699762e-05,
+      "loss": 0.9069,
+      "step": 2854
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4260370286369646,
+      "learning_rate": 1.2845559545807208e-05,
+      "loss": 0.8505,
+      "step": 2855
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8744328951602257,
+      "learning_rate": 1.2840925802191661e-05,
+      "loss": 0.33,
+      "step": 2856
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4492934225745673,
+      "learning_rate": 1.2836291394935568e-05,
+      "loss": 0.8974,
+      "step": 2857
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5499553871560139,
+      "learning_rate": 1.2831656325121516e-05,
+      "loss": 0.8693,
+      "step": 2858
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4159297838171299,
+      "learning_rate": 1.2827020593832269e-05,
+      "loss": 0.8539,
+      "step": 2859
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4912202200260303,
+      "learning_rate": 1.2822384202150726e-05,
+      "loss": 0.854,
+      "step": 2860
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3863013814731202,
+      "learning_rate": 1.2817747151159954e-05,
+      "loss": 0.8878,
+      "step": 2861
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3929211995111637,
+      "learning_rate": 1.2813109441943166e-05,
+      "loss": 0.8858,
+      "step": 2862
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3976120270651626,
+      "learning_rate": 1.2808471075583733e-05,
+      "loss": 0.8636,
+      "step": 2863
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.389595806146513,
+      "learning_rate": 1.2803832053165177e-05,
+      "loss": 0.9057,
+      "step": 2864
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6247254125893544,
+      "learning_rate": 1.2799192375771172e-05,
+      "loss": 0.8099,
+      "step": 2865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5931098782630584,
+      "learning_rate": 1.279455204448555e-05,
+      "loss": 0.8777,
+      "step": 2866
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.332669443634771,
+      "learning_rate": 1.2789911060392295e-05,
+      "loss": 0.7954,
+      "step": 2867
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3761425750681961,
+      "learning_rate": 1.2785269424575537e-05,
+      "loss": 0.7842,
+      "step": 2868
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4800211587143304,
+      "learning_rate": 1.278062713811956e-05,
+      "loss": 0.7922,
+      "step": 2869
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4300191942674076,
+      "learning_rate": 1.2775984202108811e-05,
+      "loss": 0.866,
+      "step": 2870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2961036876046856,
+      "learning_rate": 1.2771340617627877e-05,
+      "loss": 0.8125,
+      "step": 2871
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3955828991517687,
+      "learning_rate": 1.2766696385761494e-05,
+      "loss": 0.8926,
+      "step": 2872
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6516070346818763,
+      "learning_rate": 1.2762051507594562e-05,
+      "loss": 0.8499,
+      "step": 2873
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6192950271731728,
+      "learning_rate": 1.2757405984212123e-05,
+      "loss": 0.9199,
+      "step": 2874
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.503052796297095,
+      "learning_rate": 1.275275981669937e-05,
+      "loss": 0.7629,
+      "step": 2875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4524095944350233,
+      "learning_rate": 1.274811300614165e-05,
+      "loss": 0.7698,
+      "step": 2876
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4262779507230168,
+      "learning_rate": 1.274346555362446e-05,
+      "loss": 0.8276,
+      "step": 2877
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4357493731345239,
+      "learning_rate": 1.2738817460233442e-05,
+      "loss": 0.7778,
+      "step": 2878
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4112609580702526,
+      "learning_rate": 1.273416872705439e-05,
+      "loss": 0.8642,
+      "step": 2879
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4835508744129458,
+      "learning_rate": 1.2729519355173254e-05,
+      "loss": 0.8629,
+      "step": 2880
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.367291082136817,
+      "learning_rate": 1.2724869345676125e-05,
+      "loss": 0.9205,
+      "step": 2881
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4035915246735866,
+      "learning_rate": 1.2720218699649243e-05,
+      "loss": 0.8473,
+      "step": 2882
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4245481122805437,
+      "learning_rate": 1.2715567418179004e-05,
+      "loss": 0.9134,
+      "step": 2883
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3359999056511975,
+      "learning_rate": 1.2710915502351944e-05,
+      "loss": 0.8786,
+      "step": 2884
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.543552965310203,
+      "learning_rate": 1.2706262953254753e-05,
+      "loss": 0.8447,
+      "step": 2885
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5624320277688122,
+      "learning_rate": 1.2701609771974266e-05,
+      "loss": 0.8168,
+      "step": 2886
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4960428919003828,
+      "learning_rate": 1.269695595959747e-05,
+      "loss": 0.8791,
+      "step": 2887
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4859725637139685,
+      "learning_rate": 1.2692301517211489e-05,
+      "loss": 0.9044,
+      "step": 2888
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6363436763058141,
+      "learning_rate": 1.2687646445903608e-05,
+      "loss": 0.8768,
+      "step": 2889
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.487628766204558,
+      "learning_rate": 1.268299074676125e-05,
+      "loss": 0.8297,
+      "step": 2890
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4636866411737361,
+      "learning_rate": 1.2678334420871984e-05,
+      "loss": 0.8336,
+      "step": 2891
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5409720678180998,
+      "learning_rate": 1.2673677469323532e-05,
+      "loss": 0.8385,
+      "step": 2892
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.503300617936997,
+      "learning_rate": 1.2669019893203758e-05,
+      "loss": 0.854,
+      "step": 2893
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.558709754546662,
+      "learning_rate": 1.266436169360067e-05,
+      "loss": 0.8503,
+      "step": 2894
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.748108289041869,
+      "learning_rate": 1.2659702871602423e-05,
+      "loss": 0.8621,
+      "step": 2895
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4674572658043554,
+      "learning_rate": 1.2655043428297322e-05,
+      "loss": 0.8414,
+      "step": 2896
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3357567100164935,
+      "learning_rate": 1.2650383364773812e-05,
+      "loss": 0.9621,
+      "step": 2897
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3223168498418596,
+      "learning_rate": 1.2645722682120483e-05,
+      "loss": 0.8976,
+      "step": 2898
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5668371266720529,
+      "learning_rate": 1.2641061381426072e-05,
+      "loss": 0.9014,
+      "step": 2899
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.440731960843974,
+      "learning_rate": 1.2636399463779458e-05,
+      "loss": 0.8922,
+      "step": 2900
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6289682090648314,
+      "learning_rate": 1.2631736930269669e-05,
+      "loss": 0.8364,
+      "step": 2901
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6478553700038558,
+      "learning_rate": 1.262707378198587e-05,
+      "loss": 0.9544,
+      "step": 2902
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4663725733256079,
+      "learning_rate": 1.2622410020017374e-05,
+      "loss": 0.8472,
+      "step": 2903
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6485851665770868,
+      "learning_rate": 1.2617745645453637e-05,
+      "loss": 0.8724,
+      "step": 2904
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.622954768793349,
+      "learning_rate": 1.2613080659384253e-05,
+      "loss": 0.8446,
+      "step": 2905
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4334313225264799,
+      "learning_rate": 1.2608415062898971e-05,
+      "loss": 0.8179,
+      "step": 2906
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2360768657963812,
+      "learning_rate": 1.2603748857087668e-05,
+      "loss": 0.8575,
+      "step": 2907
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4295615788283238,
+      "learning_rate": 1.2599082043040372e-05,
+      "loss": 0.8954,
+      "step": 2908
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4531978690170937,
+      "learning_rate": 1.2594414621847255e-05,
+      "loss": 0.888,
+      "step": 2909
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3773406930728338,
+      "learning_rate": 1.258974659459862e-05,
+      "loss": 0.9444,
+      "step": 2910
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.46623077211472,
+      "learning_rate": 1.2585077962384924e-05,
+      "loss": 0.8639,
+      "step": 2911
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3953303530587875,
+      "learning_rate": 1.258040872629676e-05,
+      "loss": 0.8458,
+      "step": 2912
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3597271527516432,
+      "learning_rate": 1.2575738887424858e-05,
+      "loss": 0.8017,
+      "step": 2913
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4115092082432688,
+      "learning_rate": 1.2571068446860095e-05,
+      "loss": 0.9054,
+      "step": 2914
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6165288906301793,
+      "learning_rate": 1.2566397405693492e-05,
+      "loss": 0.8772,
+      "step": 2915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5197170948046925,
+      "learning_rate": 1.2561725765016192e-05,
+      "loss": 1.003,
+      "step": 2916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.471604164934742,
+      "learning_rate": 1.2557053525919503e-05,
+      "loss": 0.9166,
+      "step": 2917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.478769754215771,
+      "learning_rate": 1.2552380689494857e-05,
+      "loss": 0.8545,
+      "step": 2918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3551497915510557,
+      "learning_rate": 1.2547707256833823e-05,
+      "loss": 0.8997,
+      "step": 2919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5662929685398246,
+      "learning_rate": 1.254303322902812e-05,
+      "loss": 0.8499,
+      "step": 2920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.660737843459019,
+      "learning_rate": 1.2538358607169605e-05,
+      "loss": 0.9351,
+      "step": 2921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4782547714334449,
+      "learning_rate": 1.2533683392350264e-05,
+      "loss": 0.9234,
+      "step": 2922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.380743854848808,
+      "learning_rate": 1.2529007585662235e-05,
+      "loss": 0.8727,
+      "step": 2923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4669722866160526,
+      "learning_rate": 1.2524331188197776e-05,
+      "loss": 0.8531,
+      "step": 2924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8843042136817216,
+      "learning_rate": 1.25196542010493e-05,
+      "loss": 0.3118,
+      "step": 2925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4103489952766648,
+      "learning_rate": 1.2514976625309357e-05,
+      "loss": 0.8847,
+      "step": 2926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5607810912942268,
+      "learning_rate": 1.2510298462070619e-05,
+      "loss": 0.8251,
+      "step": 2927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5495810073888892,
+      "learning_rate": 1.2505619712425912e-05,
+      "loss": 0.7579,
+      "step": 2928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.284160925724493,
+      "learning_rate": 1.2500940377468188e-05,
+      "loss": 0.8677,
+      "step": 2929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2522987946338024,
+      "learning_rate": 1.2496260458290545e-05,
+      "loss": 0.852,
+      "step": 2930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4964180263746187,
+      "learning_rate": 1.249157995598621e-05,
+      "loss": 0.9062,
+      "step": 2931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4671293293346372,
+      "learning_rate": 1.2486898871648552e-05,
+      "loss": 0.8764,
+      "step": 2932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.436913718000181,
+      "learning_rate": 1.2482217206371065e-05,
+      "loss": 0.8216,
+      "step": 2933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7265448450678933,
+      "learning_rate": 1.2477534961247393e-05,
+      "loss": 0.9017,
+      "step": 2934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4570819172218483,
+      "learning_rate": 1.247285213737131e-05,
+      "loss": 0.9173,
+      "step": 2935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4395717619838253,
+      "learning_rate": 1.2468168735836716e-05,
+      "loss": 0.8328,
+      "step": 2936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6316807020716233,
+      "learning_rate": 1.2463484757737663e-05,
+      "loss": 0.9196,
+      "step": 2937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4842598632028106,
+      "learning_rate": 1.2458800204168324e-05,
+      "loss": 0.859,
+      "step": 2938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.339004209088753,
+      "learning_rate": 1.2454115076223012e-05,
+      "loss": 0.8488,
+      "step": 2939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4390710921553949,
+      "learning_rate": 1.2449429374996176e-05,
+      "loss": 0.8926,
+      "step": 2940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5354211775212254,
+      "learning_rate": 1.2444743101582392e-05,
+      "loss": 0.9376,
+      "step": 2941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3997777983897326,
+      "learning_rate": 1.2440056257076376e-05,
+      "loss": 0.91,
+      "step": 2942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4190529850055054,
+      "learning_rate": 1.2435368842572975e-05,
+      "loss": 0.8968,
+      "step": 2943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4147911394280994,
+      "learning_rate": 1.243068085916717e-05,
+      "loss": 0.789,
+      "step": 2944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4818082828739265,
+      "learning_rate": 1.2425992307954075e-05,
+      "loss": 0.8539,
+      "step": 2945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5425223924156828,
+      "learning_rate": 1.2421303190028935e-05,
+      "loss": 0.9645,
+      "step": 2946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3171606731963654,
+      "learning_rate": 1.241661350648713e-05,
+      "loss": 0.8771,
+      "step": 2947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4668479674338866,
+      "learning_rate": 1.2411923258424167e-05,
+      "loss": 0.9488,
+      "step": 2948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.48443304437788,
+      "learning_rate": 1.240723244693569e-05,
+      "loss": 0.856,
+      "step": 2949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4663900656968,
+      "learning_rate": 1.2402541073117475e-05,
+      "loss": 0.8797,
+      "step": 2950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4446205274041983,
+      "learning_rate": 1.2397849138065428e-05,
+      "loss": 0.8935,
+      "step": 2951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4001794613294685,
+      "learning_rate": 1.2393156642875579e-05,
+      "loss": 0.8156,
+      "step": 2952
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.373598546837832,
+      "learning_rate": 1.2388463588644102e-05,
+      "loss": 0.8819,
+      "step": 2953
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3595147718722762,
+      "learning_rate": 1.2383769976467295e-05,
+      "loss": 0.8764,
+      "step": 2954
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.493008891813113,
+      "learning_rate": 1.237907580744158e-05,
+      "loss": 0.8964,
+      "step": 2955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6334055492059742,
+      "learning_rate": 1.237438108266352e-05,
+      "loss": 0.8859,
+      "step": 2956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5206122659564747,
+      "learning_rate": 1.2369685803229802e-05,
+      "loss": 0.913,
+      "step": 2957
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.384897602393876,
+      "learning_rate": 1.236498997023725e-05,
+      "loss": 0.9501,
+      "step": 2958
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3548174703348206,
+      "learning_rate": 1.2360293584782799e-05,
+      "loss": 0.9317,
+      "step": 2959
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.511302509347131,
+      "learning_rate": 1.2355596647963533e-05,
+      "loss": 0.8618,
+      "step": 2960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.574174360922581,
+      "learning_rate": 1.2350899160876657e-05,
+      "loss": 0.8419,
+      "step": 2961
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3873586722837887,
+      "learning_rate": 1.2346201124619502e-05,
+      "loss": 0.823,
+      "step": 2962
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5674222067757635,
+      "learning_rate": 1.234150254028953e-05,
+      "loss": 0.8077,
+      "step": 2963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.392208042164477,
+      "learning_rate": 1.2336803408984333e-05,
+      "loss": 0.7711,
+      "step": 2964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3437310497478019,
+      "learning_rate": 1.2332103731801626e-05,
+      "loss": 0.8969,
+      "step": 2965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4877327236075626,
+      "learning_rate": 1.2327403509839253e-05,
+      "loss": 0.8666,
+      "step": 2966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6759832300494302,
+      "learning_rate": 1.2322702744195192e-05,
+      "loss": 0.8277,
+      "step": 2967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3762100155699448,
+      "learning_rate": 1.2318001435967535e-05,
+      "loss": 0.8168,
+      "step": 2968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4008679387862262,
+      "learning_rate": 1.2313299586254512e-05,
+      "loss": 0.9323,
+      "step": 2969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.337826624961473,
+      "learning_rate": 1.230859719615448e-05,
+      "loss": 0.8607,
+      "step": 2970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2592926335081416,
+      "learning_rate": 1.2303894266765908e-05,
+      "loss": 0.821,
+      "step": 2971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3355145189126434,
+      "learning_rate": 1.2299190799187405e-05,
+      "loss": 0.8837,
+      "step": 2972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4291473069063707,
+      "learning_rate": 1.2294486794517703e-05,
+      "loss": 0.8396,
+      "step": 2973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5630904679409707,
+      "learning_rate": 1.228978225385566e-05,
+      "loss": 0.8794,
+      "step": 2974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.547518354345178,
+      "learning_rate": 1.2285077178300253e-05,
+      "loss": 0.8506,
+      "step": 2975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.546085980064757,
+      "learning_rate": 1.2280371568950588e-05,
+      "loss": 0.9193,
+      "step": 2976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2997699477828981,
+      "learning_rate": 1.22756654269059e-05,
+      "loss": 0.8235,
+      "step": 2977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5694756686013684,
+      "learning_rate": 1.2270958753265541e-05,
+      "loss": 0.8518,
+      "step": 2978
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5683788521627309,
+      "learning_rate": 1.2266251549128989e-05,
+      "loss": 0.939,
+      "step": 2979
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.507400188255447,
+      "learning_rate": 1.2261543815595853e-05,
+      "loss": 0.9251,
+      "step": 2980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.711953762745263,
+      "learning_rate": 1.2256835553765853e-05,
+      "loss": 0.9525,
+      "step": 2981
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8813211183435995,
+      "learning_rate": 1.2252126764738845e-05,
+      "loss": 0.3422,
+      "step": 2982
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4273060196502532,
+      "learning_rate": 1.2247417449614801e-05,
+      "loss": 0.8891,
+      "step": 2983
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4557649393213015,
+      "learning_rate": 1.2242707609493814e-05,
+      "loss": 0.8817,
+      "step": 2984
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4776730423092455,
+      "learning_rate": 1.2237997245476108e-05,
+      "loss": 0.9373,
+      "step": 2985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5023754800906353,
+      "learning_rate": 1.2233286358662019e-05,
+      "loss": 0.8223,
+      "step": 2986
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5557183090397682,
+      "learning_rate": 1.222857495015202e-05,
+      "loss": 0.8227,
+      "step": 2987
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3720835125108328,
+      "learning_rate": 1.2223863021046687e-05,
+      "loss": 0.925,
+      "step": 2988
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4295231797681311,
+      "learning_rate": 1.2219150572446729e-05,
+      "loss": 0.893,
+      "step": 2989
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.348506772437856,
+      "learning_rate": 1.221443760545298e-05,
+      "loss": 0.9044,
+      "step": 2990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7005311786479078,
+      "learning_rate": 1.2209724121166384e-05,
+      "loss": 0.8561,
+      "step": 2991
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.357581686999907,
+      "learning_rate": 1.2205010120688012e-05,
+      "loss": 0.9322,
+      "step": 2992
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4693097908211747,
+      "learning_rate": 1.2200295605119059e-05,
+      "loss": 0.9076,
+      "step": 2993
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.312130254367557,
+      "learning_rate": 1.2195580575560833e-05,
+      "loss": 0.8045,
+      "step": 2994
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5142933535676377,
+      "learning_rate": 1.2190865033114763e-05,
+      "loss": 0.8597,
+      "step": 2995
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4014620260386157,
+      "learning_rate": 1.2186148978882406e-05,
+      "loss": 0.9623,
+      "step": 2996
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5077486148207053,
+      "learning_rate": 1.2181432413965428e-05,
+      "loss": 0.8405,
+      "step": 2997
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4031928879949722,
+      "learning_rate": 1.217671533946562e-05,
+      "loss": 0.8714,
+      "step": 2998
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5586041202773804,
+      "learning_rate": 1.2171997756484895e-05,
+      "loss": 0.7691,
+      "step": 2999
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5050302311989334,
+      "learning_rate": 1.2167279666125275e-05,
+      "loss": 0.7938,
+      "step": 3000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4605864317221,
+      "learning_rate": 1.216256106948891e-05,
+      "loss": 0.8903,
+      "step": 3001
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7399280537520456,
+      "learning_rate": 1.2157841967678064e-05,
+      "loss": 0.8701,
+      "step": 3002
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.201863932314353,
+      "learning_rate": 1.215312236179512e-05,
+      "loss": 0.9012,
+      "step": 3003
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3230824785873494,
+      "learning_rate": 1.2148402252942575e-05,
+      "loss": 0.8315,
+      "step": 3004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.292362376353151,
+      "learning_rate": 1.214368164222305e-05,
+      "loss": 0.8847,
+      "step": 3005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.315792264578551,
+      "learning_rate": 1.2138960530739283e-05,
+      "loss": 0.8831,
+      "step": 3006
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7180431006832764,
+      "learning_rate": 1.2134238919594122e-05,
+      "loss": 0.9555,
+      "step": 3007
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.499204376806803,
+      "learning_rate": 1.2129516809890536e-05,
+      "loss": 0.9263,
+      "step": 3008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.497233422980793,
+      "learning_rate": 1.2124794202731611e-05,
+      "loss": 0.835,
+      "step": 3009
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4675135023821613,
+      "learning_rate": 1.212007109922055e-05,
+      "loss": 0.8658,
+      "step": 3010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2599068858175837,
+      "learning_rate": 1.2115347500460666e-05,
+      "loss": 0.8713,
+      "step": 3011
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2566553199473303,
+      "learning_rate": 1.2110623407555398e-05,
+      "loss": 0.8825,
+      "step": 3012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4885573950716942,
+      "learning_rate": 1.2105898821608291e-05,
+      "loss": 0.8454,
+      "step": 3013
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.358555798359703,
+      "learning_rate": 1.2101173743723007e-05,
+      "loss": 0.8169,
+      "step": 3014
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2938516791672219,
+      "learning_rate": 1.2096448175003329e-05,
+      "loss": 0.7889,
+      "step": 3015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.473109130130557,
+      "learning_rate": 1.2091722116553148e-05,
+      "loss": 0.8483,
+      "step": 3016
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4099539865326929,
+      "learning_rate": 1.2086995569476474e-05,
+      "loss": 0.8039,
+      "step": 3017
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4280624006301712,
+      "learning_rate": 1.2082268534877425e-05,
+      "loss": 0.9476,
+      "step": 3018
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.331332074968661,
+      "learning_rate": 1.207754101386024e-05,
+      "loss": 0.8361,
+      "step": 3019
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.340391751819343,
+      "learning_rate": 1.2072813007529267e-05,
+      "loss": 0.856,
+      "step": 3020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.452627894270564,
+      "learning_rate": 1.206808451698897e-05,
+      "loss": 0.8399,
+      "step": 3021
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2531290631573224,
+      "learning_rate": 1.2063355543343925e-05,
+      "loss": 0.8818,
+      "step": 3022
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6892794326843585,
+      "learning_rate": 1.2058626087698814e-05,
+      "loss": 0.8702,
+      "step": 3023
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.333843669200861,
+      "learning_rate": 1.2053896151158446e-05,
+      "loss": 0.874,
+      "step": 3024
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.46007175981807,
+      "learning_rate": 1.2049165734827737e-05,
+      "loss": 0.8874,
+      "step": 3025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3511418621030704,
+      "learning_rate": 1.2044434839811702e-05,
+      "loss": 0.7554,
+      "step": 3026
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4641088358794427,
+      "learning_rate": 1.2039703467215489e-05,
+      "loss": 0.8504,
+      "step": 3027
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3961767014444657,
+      "learning_rate": 1.203497161814434e-05,
+      "loss": 0.8116,
+      "step": 3028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5106527626286321,
+      "learning_rate": 1.203023929370362e-05,
+      "loss": 0.7546,
+      "step": 3029
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3812305748714984,
+      "learning_rate": 1.2025506494998797e-05,
+      "loss": 0.9022,
+      "step": 3030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4536285156128268,
+      "learning_rate": 1.2020773223135458e-05,
+      "loss": 0.8524,
+      "step": 3031
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4920000194439378,
+      "learning_rate": 1.2016039479219293e-05,
+      "loss": 0.8502,
+      "step": 3032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4418751740827567,
+      "learning_rate": 1.2011305264356101e-05,
+      "loss": 0.8617,
+      "step": 3033
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4090200454050512,
+      "learning_rate": 1.2006570579651799e-05,
+      "loss": 0.919,
+      "step": 3034
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5198068883611635,
+      "learning_rate": 1.2001835426212413e-05,
+      "loss": 0.8118,
+      "step": 3035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.516676265353004,
+      "learning_rate": 1.1997099805144071e-05,
+      "loss": 0.8816,
+      "step": 3036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2685939964065611,
+      "learning_rate": 1.1992363717553015e-05,
+      "loss": 0.8116,
+      "step": 3037
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3990909685267978,
+      "learning_rate": 1.1987627164545597e-05,
+      "loss": 0.8341,
+      "step": 3038
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6251760803476545,
+      "learning_rate": 1.1982890147228275e-05,
+      "loss": 0.8656,
+      "step": 3039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.424315107003013,
+      "learning_rate": 1.1978152666707614e-05,
+      "loss": 0.8372,
+      "step": 3040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9660462928937239,
+      "learning_rate": 1.1973414724090297e-05,
+      "loss": 0.3461,
+      "step": 3041
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.654321301042364,
+      "learning_rate": 1.1968676320483103e-05,
+      "loss": 0.9203,
+      "step": 3042
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4530487848106561,
+      "learning_rate": 1.1963937456992922e-05,
+      "loss": 0.9107,
+      "step": 3043
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4630946308693005,
+      "learning_rate": 1.195919813472676e-05,
+      "loss": 0.9254,
+      "step": 3044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4754084266364316,
+      "learning_rate": 1.1954458354791716e-05,
+      "loss": 0.8686,
+      "step": 3045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.611348266475679,
+      "learning_rate": 1.1949718118295006e-05,
+      "loss": 0.8129,
+      "step": 3046
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0019443745816585,
+      "learning_rate": 1.194497742634395e-05,
+      "loss": 0.3489,
+      "step": 3047
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4659915400914505,
+      "learning_rate": 1.1940236280045968e-05,
+      "loss": 0.8748,
+      "step": 3048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.71761967706553,
+      "learning_rate": 1.1935494680508606e-05,
+      "loss": 0.8715,
+      "step": 3049
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5164726322881468,
+      "learning_rate": 1.1930752628839487e-05,
+      "loss": 0.7458,
+      "step": 3050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7200021138514376,
+      "learning_rate": 1.1926010126146368e-05,
+      "loss": 0.8808,
+      "step": 3051
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.393900459474437,
+      "learning_rate": 1.1921267173537085e-05,
+      "loss": 0.8703,
+      "step": 3052
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4933222998083704,
+      "learning_rate": 1.1916523772119603e-05,
+      "loss": 0.8706,
+      "step": 3053
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5730561123281428,
+      "learning_rate": 1.1911779923001976e-05,
+      "loss": 0.8166,
+      "step": 3054
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2444854292962508,
+      "learning_rate": 1.1907035627292367e-05,
+      "loss": 0.8736,
+      "step": 3055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.431234221392283,
+      "learning_rate": 1.1902290886099048e-05,
+      "loss": 0.9329,
+      "step": 3056
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5751700224859408,
+      "learning_rate": 1.1897545700530387e-05,
+      "loss": 0.8516,
+      "step": 3057
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5363727280373627,
+      "learning_rate": 1.1892800071694862e-05,
+      "loss": 0.9249,
+      "step": 3058
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5859313598288451,
+      "learning_rate": 1.188805400070105e-05,
+      "loss": 0.8546,
+      "step": 3059
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.012910401994648,
+      "learning_rate": 1.1883307488657637e-05,
+      "loss": 0.8183,
+      "step": 3060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5128460889604634,
+      "learning_rate": 1.1878560536673407e-05,
+      "loss": 0.829,
+      "step": 3061
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4679675104005057,
+      "learning_rate": 1.187381314585725e-05,
+      "loss": 0.889,
+      "step": 3062
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4133970324165064,
+      "learning_rate": 1.1869065317318151e-05,
+      "loss": 0.7788,
+      "step": 3063
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3780765989080113,
+      "learning_rate": 1.1864317052165213e-05,
+      "loss": 0.8574,
+      "step": 3064
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.437439415784921,
+      "learning_rate": 1.1859568351507623e-05,
+      "loss": 0.8575,
+      "step": 3065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6865035355746034,
+      "learning_rate": 1.1854819216454678e-05,
+      "loss": 0.8581,
+      "step": 3066
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.319392966838635,
+      "learning_rate": 1.1850069648115785e-05,
+      "loss": 0.8765,
+      "step": 3067
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.455221215195415,
+      "learning_rate": 1.1845319647600433e-05,
+      "loss": 0.8913,
+      "step": 3068
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.602324623112788,
+      "learning_rate": 1.184056921601823e-05,
+      "loss": 0.8388,
+      "step": 3069
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4966083679868898,
+      "learning_rate": 1.1835818354478876e-05,
+      "loss": 0.9157,
+      "step": 3070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8192445688038177,
+      "learning_rate": 1.183106706409217e-05,
+      "loss": 0.9695,
+      "step": 3071
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3695512627201891,
+      "learning_rate": 1.1826315345968014e-05,
+      "loss": 0.8006,
+      "step": 3072
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2715114936339134,
+      "learning_rate": 1.1821563201216413e-05,
+      "loss": 0.9115,
+      "step": 3073
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4951401845158783,
+      "learning_rate": 1.1816810630947466e-05,
+      "loss": 0.9116,
+      "step": 3074
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.566681414382169,
+      "learning_rate": 1.1812057636271374e-05,
+      "loss": 0.7301,
+      "step": 3075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4581646359031806,
+      "learning_rate": 1.1807304218298438e-05,
+      "loss": 0.8599,
+      "step": 3076
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4795194138643568,
+      "learning_rate": 1.180255037813906e-05,
+      "loss": 0.9039,
+      "step": 3077
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4435725086997526,
+      "learning_rate": 1.179779611690373e-05,
+      "loss": 0.7618,
+      "step": 3078
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.583157657588781,
+      "learning_rate": 1.1793041435703048e-05,
+      "loss": 0.8929,
+      "step": 3079
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8743703877817455,
+      "learning_rate": 1.1788286335647712e-05,
+      "loss": 0.8314,
+      "step": 3080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1796359925678022,
+      "learning_rate": 1.1783530817848505e-05,
+      "loss": 0.8939,
+      "step": 3081
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3686494082127807,
+      "learning_rate": 1.1778774883416325e-05,
+      "loss": 0.7968,
+      "step": 3082
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.363617184109668,
+      "learning_rate": 1.1774018533462152e-05,
+      "loss": 0.9151,
+      "step": 3083
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3846582820377824,
+      "learning_rate": 1.1769261769097076e-05,
+      "loss": 0.8732,
+      "step": 3084
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3779437884252015,
+      "learning_rate": 1.1764504591432271e-05,
+      "loss": 0.9035,
+      "step": 3085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7283154646505554,
+      "learning_rate": 1.1759747001579019e-05,
+      "loss": 0.9436,
+      "step": 3086
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3891214093803212,
+      "learning_rate": 1.1754989000648693e-05,
+      "loss": 0.7946,
+      "step": 3087
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.506600063575142,
+      "learning_rate": 1.1750230589752763e-05,
+      "loss": 0.8634,
+      "step": 3088
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4777622886298265,
+      "learning_rate": 1.174547177000279e-05,
+      "loss": 0.9042,
+      "step": 3089
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5611376712322091,
+      "learning_rate": 1.1740712542510439e-05,
+      "loss": 0.8122,
+      "step": 3090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4889340514664842,
+      "learning_rate": 1.1735952908387463e-05,
+      "loss": 0.8116,
+      "step": 3091
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.347914404261615,
+      "learning_rate": 1.1731192868745716e-05,
+      "loss": 0.8829,
+      "step": 3092
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4937668557243533,
+      "learning_rate": 1.1726432424697144e-05,
+      "loss": 0.824,
+      "step": 3093
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4041903724659708,
+      "learning_rate": 1.1721671577353783e-05,
+      "loss": 0.9096,
+      "step": 3094
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5318421184838733,
+      "learning_rate": 1.171691032782777e-05,
+      "loss": 0.8226,
+      "step": 3095
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7790736698045904,
+      "learning_rate": 1.1712148677231336e-05,
+      "loss": 0.8268,
+      "step": 3096
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5518442757622766,
+      "learning_rate": 1.1707386626676798e-05,
+      "loss": 0.7812,
+      "step": 3097
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7627156813863276,
+      "learning_rate": 1.1702624177276574e-05,
+      "loss": 0.8417,
+      "step": 3098
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3227947069693133,
+      "learning_rate": 1.1697861330143174e-05,
+      "loss": 0.8252,
+      "step": 3099
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7663971765311302,
+      "learning_rate": 1.1693098086389198e-05,
+      "loss": 0.8531,
+      "step": 3100
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7351294060287639,
+      "learning_rate": 1.1688334447127338e-05,
+      "loss": 0.759,
+      "step": 3101
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.722082010454196,
+      "learning_rate": 1.1683570413470384e-05,
+      "loss": 0.9057,
+      "step": 3102
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7910847676532884,
+      "learning_rate": 1.1678805986531216e-05,
+      "loss": 0.8273,
+      "step": 3103
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5471257526725348,
+      "learning_rate": 1.16740411674228e-05,
+      "loss": 0.8945,
+      "step": 3104
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4183108085337643,
+      "learning_rate": 1.1669275957258204e-05,
+      "loss": 0.9055,
+      "step": 3105
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.421261728518497,
+      "learning_rate": 1.1664510357150575e-05,
+      "loss": 0.815,
+      "step": 3106
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4434060823473944,
+      "learning_rate": 1.1659744368213159e-05,
+      "loss": 0.9223,
+      "step": 3107
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4725094666423635,
+      "learning_rate": 1.1654977991559297e-05,
+      "loss": 0.8076,
+      "step": 3108
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4774040398734796,
+      "learning_rate": 1.165021122830241e-05,
+      "loss": 0.8453,
+      "step": 3109
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6651641815473082,
+      "learning_rate": 1.1645444079556017e-05,
+      "loss": 0.8856,
+      "step": 3110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5036688084614938,
+      "learning_rate": 1.164067654643372e-05,
+      "loss": 0.8824,
+      "step": 3111
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5822542527877423,
+      "learning_rate": 1.163590863004922e-05,
+      "loss": 0.9043,
+      "step": 3112
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2481746630549335,
+      "learning_rate": 1.1631140331516302e-05,
+      "loss": 0.8621,
+      "step": 3113
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5057241172506834,
+      "learning_rate": 1.1626371651948839e-05,
+      "loss": 0.8129,
+      "step": 3114
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3543105735944232,
+      "learning_rate": 1.1621602592460794e-05,
+      "loss": 0.9079,
+      "step": 3115
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.604022830985037,
+      "learning_rate": 1.1616833154166224e-05,
+      "loss": 0.8384,
+      "step": 3116
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4886478794749518,
+      "learning_rate": 1.1612063338179269e-05,
+      "loss": 0.8097,
+      "step": 3117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6295960415754123,
+      "learning_rate": 1.1607293145614156e-05,
+      "loss": 0.7728,
+      "step": 3118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4522191539033344,
+      "learning_rate": 1.1602522577585207e-05,
+      "loss": 0.9451,
+      "step": 3119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6034491521738696,
+      "learning_rate": 1.159775163520682e-05,
+      "loss": 0.9183,
+      "step": 3120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.449268700723357,
+      "learning_rate": 1.1592980319593492e-05,
+      "loss": 0.8214,
+      "step": 3121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6526954977646768,
+      "learning_rate": 1.1588208631859808e-05,
+      "loss": 0.8086,
+      "step": 3122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3262290317258014,
+      "learning_rate": 1.1583436573120424e-05,
+      "loss": 0.9211,
+      "step": 3123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3867548870697202,
+      "learning_rate": 1.1578664144490099e-05,
+      "loss": 0.8379,
+      "step": 3124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5175965871387511,
+      "learning_rate": 1.1573891347083676e-05,
+      "loss": 0.8695,
+      "step": 3125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4524827974980379,
+      "learning_rate": 1.1569118182016074e-05,
+      "loss": 0.8104,
+      "step": 3126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3809905900351476,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 0.8211,
+      "step": 3127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3353853939223754,
+      "learning_rate": 1.1559570753357481e-05,
+      "loss": 0.8801,
+      "step": 3128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3752891848728184,
+      "learning_rate": 1.1554796491996767e-05,
+      "loss": 0.8375,
+      "step": 3129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6699279205034372,
+      "learning_rate": 1.1550021867435437e-05,
+      "loss": 0.846,
+      "step": 3130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.297098362456666,
+      "learning_rate": 1.1545246880788845e-05,
+      "loss": 0.9009,
+      "step": 3131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5235535895303025,
+      "learning_rate": 1.154047153317243e-05,
+      "loss": 0.8915,
+      "step": 3132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3636984364423204,
+      "learning_rate": 1.153569582570171e-05,
+      "loss": 0.8818,
+      "step": 3133
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3220122875059277,
+      "learning_rate": 1.1530919759492289e-05,
+      "loss": 0.969,
+      "step": 3134
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8002835706233236,
+      "learning_rate": 1.1526143335659866e-05,
+      "loss": 0.781,
+      "step": 3135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5655544663195764,
+      "learning_rate": 1.15213665553202e-05,
+      "loss": 0.9365,
+      "step": 3136
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.400438407657353,
+      "learning_rate": 1.1516589419589159e-05,
+      "loss": 0.969,
+      "step": 3137
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5885045517561907,
+      "learning_rate": 1.1511811929582682e-05,
+      "loss": 0.909,
+      "step": 3138
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4630718208113698,
+      "learning_rate": 1.1507034086416781e-05,
+      "loss": 0.9308,
+      "step": 3139
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4686223980331212,
+      "learning_rate": 1.1502255891207572e-05,
+      "loss": 0.8355,
+      "step": 3140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9354897055549466,
+      "learning_rate": 1.1497477345071238e-05,
+      "loss": 0.3223,
+      "step": 3141
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3394201911811199,
+      "learning_rate": 1.1492698449124042e-05,
+      "loss": 0.8666,
+      "step": 3142
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4512550720352924,
+      "learning_rate": 1.1487919204482343e-05,
+      "loss": 0.9129,
+      "step": 3143
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4555437021618638,
+      "learning_rate": 1.1483139612262569e-05,
+      "loss": 0.9393,
+      "step": 3144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3724218359431628,
+      "learning_rate": 1.1478359673581235e-05,
+      "loss": 0.8974,
+      "step": 3145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4362551434068322,
+      "learning_rate": 1.147357938955493e-05,
+      "loss": 0.8402,
+      "step": 3146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4967591247823635,
+      "learning_rate": 1.1468798761300335e-05,
+      "loss": 0.8436,
+      "step": 3147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4510910568733975,
+      "learning_rate": 1.1464017789934206e-05,
+      "loss": 0.8404,
+      "step": 3148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.531675922995948,
+      "learning_rate": 1.1459236476573373e-05,
+      "loss": 0.8667,
+      "step": 3149
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6720610569654422,
+      "learning_rate": 1.1454454822334753e-05,
+      "loss": 0.8412,
+      "step": 3150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5006690271807168,
+      "learning_rate": 1.1449672828335344e-05,
+      "loss": 0.871,
+      "step": 3151
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3715690049951308,
+      "learning_rate": 1.1444890495692214e-05,
+      "loss": 0.9099,
+      "step": 3152
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5133068709058257,
+      "learning_rate": 1.1440107825522522e-05,
+      "loss": 0.9019,
+      "step": 3153
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3814938563260044,
+      "learning_rate": 1.1435324818943501e-05,
+      "loss": 0.8867,
+      "step": 3154
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3785832373916405,
+      "learning_rate": 1.1430541477072457e-05,
+      "loss": 0.8274,
+      "step": 3155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4264784401677657,
+      "learning_rate": 1.142575780102678e-05,
+      "loss": 0.8389,
+      "step": 3156
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3109979329427974,
+      "learning_rate": 1.1420973791923941e-05,
+      "loss": 0.8932,
+      "step": 3157
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3402078122803545,
+      "learning_rate": 1.1416189450881483e-05,
+      "loss": 0.9515,
+      "step": 3158
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4710769890530828,
+      "learning_rate": 1.1411404779017026e-05,
+      "loss": 0.7858,
+      "step": 3159
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4478997998118535,
+      "learning_rate": 1.1406619777448271e-05,
+      "loss": 0.7494,
+      "step": 3160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.782353692942064,
+      "learning_rate": 1.1401834447293001e-05,
+      "loss": 0.8911,
+      "step": 3161
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.418182572355784,
+      "learning_rate": 1.1397048789669061e-05,
+      "loss": 0.8355,
+      "step": 3162
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6036213457799025,
+      "learning_rate": 1.1392262805694382e-05,
+      "loss": 0.8555,
+      "step": 3163
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3161952446736378,
+      "learning_rate": 1.138747649648698e-05,
+      "loss": 0.7779,
+      "step": 3164
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5656688421834137,
+      "learning_rate": 1.1382689863164924e-05,
+      "loss": 0.9487,
+      "step": 3165
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4391459780451075,
+      "learning_rate": 1.137790290684638e-05,
+      "loss": 0.9187,
+      "step": 3166
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3957428030261596,
+      "learning_rate": 1.1373115628649582e-05,
+      "loss": 0.7765,
+      "step": 3167
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4744429134366377,
+      "learning_rate": 1.1368328029692834e-05,
+      "loss": 0.796,
+      "step": 3168
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4289570007884824,
+      "learning_rate": 1.1363540111094524e-05,
+      "loss": 0.8491,
+      "step": 3169
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5375573003663248,
+      "learning_rate": 1.1358751873973106e-05,
+      "loss": 0.9315,
+      "step": 3170
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3442175905496911,
+      "learning_rate": 1.1353963319447114e-05,
+      "loss": 0.8699,
+      "step": 3171
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3556248055585878,
+      "learning_rate": 1.1349174448635158e-05,
+      "loss": 0.7842,
+      "step": 3172
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4499807545858407,
+      "learning_rate": 1.1344385262655915e-05,
+      "loss": 0.9286,
+      "step": 3173
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4326422350461436,
+      "learning_rate": 1.1339595762628144e-05,
+      "loss": 0.8862,
+      "step": 3174
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4043441964772698,
+      "learning_rate": 1.1334805949670666e-05,
+      "loss": 0.8887,
+      "step": 3175
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3564087586115487,
+      "learning_rate": 1.1330015824902385e-05,
+      "loss": 0.8986,
+      "step": 3176
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4187450713962702,
+      "learning_rate": 1.1325225389442278e-05,
+      "loss": 0.8738,
+      "step": 3177
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.500901250743024,
+      "learning_rate": 1.1320434644409384e-05,
+      "loss": 0.9082,
+      "step": 3178
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3701879941190935,
+      "learning_rate": 1.1315643590922827e-05,
+      "loss": 0.9028,
+      "step": 3179
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2702735022157703,
+      "learning_rate": 1.1310852230101797e-05,
+      "loss": 0.846,
+      "step": 3180
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5143694107192218,
+      "learning_rate": 1.1306060563065556e-05,
+      "loss": 0.9128,
+      "step": 3181
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5025758729603544,
+      "learning_rate": 1.1301268590933434e-05,
+      "loss": 0.8676,
+      "step": 3182
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5516499433241973,
+      "learning_rate": 1.1296476314824842e-05,
+      "loss": 0.8963,
+      "step": 3183
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2619508228695464,
+      "learning_rate": 1.1291683735859254e-05,
+      "loss": 0.8583,
+      "step": 3184
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4600969959255856,
+      "learning_rate": 1.1286890855156215e-05,
+      "loss": 0.8135,
+      "step": 3185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3648612463451428,
+      "learning_rate": 1.1282097673835343e-05,
+      "loss": 0.964,
+      "step": 3186
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.408230953232585,
+      "learning_rate": 1.1277304193016332e-05,
+      "loss": 0.9216,
+      "step": 3187
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3614985203381647,
+      "learning_rate": 1.1272510413818929e-05,
+      "loss": 0.8879,
+      "step": 3188
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3341023461384207,
+      "learning_rate": 1.1267716337362968e-05,
+      "loss": 0.8861,
+      "step": 3189
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5598086328497889,
+      "learning_rate": 1.1262921964768348e-05,
+      "loss": 0.891,
+      "step": 3190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3691015843363636,
+      "learning_rate": 1.1258127297155027e-05,
+      "loss": 0.8791,
+      "step": 3191
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3519206199662215,
+      "learning_rate": 1.1253332335643043e-05,
+      "loss": 0.8901,
+      "step": 3192
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4265451925602446,
+      "learning_rate": 1.1248537081352504e-05,
+      "loss": 0.8382,
+      "step": 3193
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3397813537793721,
+      "learning_rate": 1.1243741535403576e-05,
+      "loss": 0.8338,
+      "step": 3194
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4797015082966112,
+      "learning_rate": 1.1238945698916504e-05,
+      "loss": 0.8629,
+      "step": 3195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.50219390288469,
+      "learning_rate": 1.1234149573011592e-05,
+      "loss": 0.7937,
+      "step": 3196
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5411575793503773,
+      "learning_rate": 1.1229353158809216e-05,
+      "loss": 0.8635,
+      "step": 3197
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5902635438371475,
+      "learning_rate": 1.1224556457429818e-05,
+      "loss": 0.9358,
+      "step": 3198
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3068818566745855,
+      "learning_rate": 1.1219759469993914e-05,
+      "loss": 0.8411,
+      "step": 3199
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.547690589548854,
+      "learning_rate": 1.1214962197622075e-05,
+      "loss": 0.8158,
+      "step": 3200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5045698021077403,
+      "learning_rate": 1.1210164641434942e-05,
+      "loss": 0.8607,
+      "step": 3201
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5073812601846175,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.8351,
+      "step": 3202
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3994553853367406,
+      "learning_rate": 1.1200568682097716e-05,
+      "loss": 0.9068,
+      "step": 3203
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5101653762853424,
+      "learning_rate": 1.1195770281189236e-05,
+      "loss": 0.8502,
+      "step": 3204
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4947101342885902,
+      "learning_rate": 1.11909716009487e-05,
+      "loss": 0.9071,
+      "step": 3205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3762661268457008,
+      "learning_rate": 1.1186172642497077e-05,
+      "loss": 0.8973,
+      "step": 3206
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4062281812193775,
+      "learning_rate": 1.118137340695541e-05,
+      "loss": 0.9108,
+      "step": 3207
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.480754004502827,
+      "learning_rate": 1.1176573895444794e-05,
+      "loss": 0.866,
+      "step": 3208
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6581946171698438,
+      "learning_rate": 1.1171774109086401e-05,
+      "loss": 0.8574,
+      "step": 3209
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.439093803734207,
+      "learning_rate": 1.1166974049001458e-05,
+      "loss": 0.8622,
+      "step": 3210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5948127961539507,
+      "learning_rate": 1.116217371631126e-05,
+      "loss": 0.8294,
+      "step": 3211
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4182513870895743,
+      "learning_rate": 1.1157373112137171e-05,
+      "loss": 0.9321,
+      "step": 3212
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4575381394417286,
+      "learning_rate": 1.1152572237600603e-05,
+      "loss": 0.8344,
+      "step": 3213
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4660880418034354,
+      "learning_rate": 1.1147771093823045e-05,
+      "loss": 0.8666,
+      "step": 3214
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5638955797953278,
+      "learning_rate": 1.1142969681926048e-05,
+      "loss": 0.8724,
+      "step": 3215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3846935527861206,
+      "learning_rate": 1.113816800303122e-05,
+      "loss": 0.8614,
+      "step": 3216
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.70949218780247,
+      "learning_rate": 1.1133366058260232e-05,
+      "loss": 0.8525,
+      "step": 3217
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4560236323093925,
+      "learning_rate": 1.1128563848734817e-05,
+      "loss": 0.8842,
+      "step": 3218
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5451365163840847,
+      "learning_rate": 1.1123761375576779e-05,
+      "loss": 0.9024,
+      "step": 3219
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3867848716338,
+      "learning_rate": 1.1118958639907969e-05,
+      "loss": 0.8899,
+      "step": 3220
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4996118614571643,
+      "learning_rate": 1.1114155642850308e-05,
+      "loss": 0.8742,
+      "step": 3221
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9349425689585481,
+      "learning_rate": 1.1109352385525782e-05,
+      "loss": 0.3365,
+      "step": 3222
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4489256544447984,
+      "learning_rate": 1.1104548869056424e-05,
+      "loss": 0.9305,
+      "step": 3223
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.413013319958576,
+      "learning_rate": 1.1099745094564342e-05,
+      "loss": 0.878,
+      "step": 3224
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5069643993095365,
+      "learning_rate": 1.1094941063171699e-05,
+      "loss": 0.8433,
+      "step": 3225
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.336620279026069,
+      "learning_rate": 1.1090136776000711e-05,
+      "loss": 0.8056,
+      "step": 3226
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.395681893587308,
+      "learning_rate": 1.1085332234173664e-05,
+      "loss": 0.8558,
+      "step": 3227
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6747714648274186,
+      "learning_rate": 1.10805274388129e-05,
+      "loss": 0.9456,
+      "step": 3228
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9726801164602124,
+      "learning_rate": 1.1075722391040817e-05,
+      "loss": 0.353,
+      "step": 3229
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5198087610464737,
+      "learning_rate": 1.1070917091979878e-05,
+      "loss": 0.8582,
+      "step": 3230
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5968587500814786,
+      "learning_rate": 1.10661115427526e-05,
+      "loss": 0.8366,
+      "step": 3231
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5396198773811056,
+      "learning_rate": 1.106130574448156e-05,
+      "loss": 0.8698,
+      "step": 3232
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4506663395572341,
+      "learning_rate": 1.1056499698289392e-05,
+      "loss": 0.8829,
+      "step": 3233
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2969783718780825,
+      "learning_rate": 1.1051693405298788e-05,
+      "loss": 0.8799,
+      "step": 3234
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3892835532090662,
+      "learning_rate": 1.1046886866632498e-05,
+      "loss": 0.8323,
+      "step": 3235
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8891087370013573,
+      "learning_rate": 1.1042080083413336e-05,
+      "loss": 0.3364,
+      "step": 3236
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3153465403309512,
+      "learning_rate": 1.1037273056764157e-05,
+      "loss": 0.8142,
+      "step": 3237
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3507943271888676,
+      "learning_rate": 1.1032465787807893e-05,
+      "loss": 0.8866,
+      "step": 3238
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4397305437355834,
+      "learning_rate": 1.1027658277667518e-05,
+      "loss": 0.796,
+      "step": 3239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5110699434330976,
+      "learning_rate": 1.1022850527466065e-05,
+      "loss": 0.9167,
+      "step": 3240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7748015321334307,
+      "learning_rate": 1.101804253832663e-05,
+      "loss": 0.8163,
+      "step": 3241
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5136345243460296,
+      "learning_rate": 1.1013234311372353e-05,
+      "loss": 0.8523,
+      "step": 3242
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5896348734022807,
+      "learning_rate": 1.100842584772644e-05,
+      "loss": 0.887,
+      "step": 3243
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4423979175373318,
+      "learning_rate": 1.1003617148512149e-05,
+      "loss": 0.8395,
+      "step": 3244
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5055996352937633,
+      "learning_rate": 1.0998808214852796e-05,
+      "loss": 0.863,
+      "step": 3245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2719839333530392,
+      "learning_rate": 1.099399904787174e-05,
+      "loss": 0.846,
+      "step": 3246
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.522533795495694,
+      "learning_rate": 1.0989189648692408e-05,
+      "loss": 0.8546,
+      "step": 3247
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3238194392468925,
+      "learning_rate": 1.0984380018438279e-05,
+      "loss": 0.8768,
+      "step": 3248
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5209300741743716,
+      "learning_rate": 1.0979570158232875e-05,
+      "loss": 0.7585,
+      "step": 3249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7556471192466117,
+      "learning_rate": 1.0974760069199786e-05,
+      "loss": 0.8433,
+      "step": 3250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3946055598188356,
+      "learning_rate": 1.096994975246265e-05,
+      "loss": 0.8441,
+      "step": 3251
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.918498783720375,
+      "learning_rate": 1.0965139209145153e-05,
+      "loss": 0.8632,
+      "step": 3252
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4030675786573157,
+      "learning_rate": 1.0960328440371039e-05,
+      "loss": 0.8807,
+      "step": 3253
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.400948138170017,
+      "learning_rate": 1.095551744726411e-05,
+      "loss": 0.8308,
+      "step": 3254
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.517048585648085,
+      "learning_rate": 1.0950706230948207e-05,
+      "loss": 0.861,
+      "step": 3255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3050948602070962,
+      "learning_rate": 1.0945894792547234e-05,
+      "loss": 0.836,
+      "step": 3256
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.488506457793547,
+      "learning_rate": 1.0941083133185146e-05,
+      "loss": 0.8529,
+      "step": 3257
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.433369094022262,
+      "learning_rate": 1.0936271253985941e-05,
+      "loss": 0.8661,
+      "step": 3258
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3875842053297514,
+      "learning_rate": 1.0931459156073679e-05,
+      "loss": 0.9041,
+      "step": 3259
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.412282257128803,
+      "learning_rate": 1.0926646840572463e-05,
+      "loss": 0.8359,
+      "step": 3260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4542792476207704,
+      "learning_rate": 1.0921834308606458e-05,
+      "loss": 0.9112,
+      "step": 3261
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.500897298716677,
+      "learning_rate": 1.0917021561299864e-05,
+      "loss": 0.9332,
+      "step": 3262
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4275058486473355,
+      "learning_rate": 1.0912208599776939e-05,
+      "loss": 0.8995,
+      "step": 3263
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4211037254822234,
+      "learning_rate": 1.0907395425161999e-05,
+      "loss": 0.9139,
+      "step": 3264
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4050995213436084,
+      "learning_rate": 1.0902582038579395e-05,
+      "loss": 0.8354,
+      "step": 3265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4974475479870746,
+      "learning_rate": 1.0897768441153536e-05,
+      "loss": 0.8153,
+      "step": 3266
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.412041135513596,
+      "learning_rate": 1.089295463400888e-05,
+      "loss": 0.8894,
+      "step": 3267
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.491443608387105,
+      "learning_rate": 1.0888140618269934e-05,
+      "loss": 0.8433,
+      "step": 3268
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4527504897243158,
+      "learning_rate": 1.088332639506125e-05,
+      "loss": 0.8622,
+      "step": 3269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4525848667924097,
+      "learning_rate": 1.0878511965507435e-05,
+      "loss": 0.8803,
+      "step": 3270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.369125952485571,
+      "learning_rate": 1.0873697330733132e-05,
+      "loss": 0.9162,
+      "step": 3271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2961907707447895,
+      "learning_rate": 1.0868882491863048e-05,
+      "loss": 0.8074,
+      "step": 3272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9195850696265195,
+      "learning_rate": 1.0864067450021926e-05,
+      "loss": 0.3642,
+      "step": 3273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3388902528641635,
+      "learning_rate": 1.0859252206334568e-05,
+      "loss": 0.7924,
+      "step": 3274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5429729540079526,
+      "learning_rate": 1.0854436761925802e-05,
+      "loss": 0.8295,
+      "step": 3275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3513487622046556,
+      "learning_rate": 1.0849621117920526e-05,
+      "loss": 0.8604,
+      "step": 3276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3383886231852877,
+      "learning_rate": 1.0844805275443673e-05,
+      "loss": 0.8704,
+      "step": 3277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5683372396234836,
+      "learning_rate": 1.083998923562022e-05,
+      "loss": 0.7999,
+      "step": 3278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4120720743658346,
+      "learning_rate": 1.0835172999575201e-05,
+      "loss": 0.8071,
+      "step": 3279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4219719407026294,
+      "learning_rate": 1.0830356568433686e-05,
+      "loss": 0.905,
+      "step": 3280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.550804339796694,
+      "learning_rate": 1.0825539943320793e-05,
+      "loss": 0.9136,
+      "step": 3281
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6178041494130784,
+      "learning_rate": 1.0820723125361685e-05,
+      "loss": 0.9173,
+      "step": 3282
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3657765792113936,
+      "learning_rate": 1.0815906115681579e-05,
+      "loss": 0.8622,
+      "step": 3283
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4167990490444722,
+      "learning_rate": 1.0811088915405717e-05,
+      "loss": 0.8967,
+      "step": 3284
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3804517532817808,
+      "learning_rate": 1.0806271525659403e-05,
+      "loss": 0.7558,
+      "step": 3285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4673309972777784,
+      "learning_rate": 1.0801453947567985e-05,
+      "loss": 0.7906,
+      "step": 3286
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4064551064799597,
+      "learning_rate": 1.0796636182256846e-05,
+      "loss": 0.8757,
+      "step": 3287
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4860978205002118,
+      "learning_rate": 1.079181823085141e-05,
+      "loss": 0.833,
+      "step": 3288
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6913933863602182,
+      "learning_rate": 1.0787000094477157e-05,
+      "loss": 0.9043,
+      "step": 3289
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.475357657168501,
+      "learning_rate": 1.0782181774259608e-05,
+      "loss": 0.8384,
+      "step": 3290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5505328183602032,
+      "learning_rate": 1.0777363271324318e-05,
+      "loss": 0.8601,
+      "step": 3291
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5008862852349154,
+      "learning_rate": 1.077254458679689e-05,
+      "loss": 0.8765,
+      "step": 3292
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3576032682014516,
+      "learning_rate": 1.0767725721802967e-05,
+      "loss": 0.8405,
+      "step": 3293
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4799977373596405,
+      "learning_rate": 1.0762906677468238e-05,
+      "loss": 0.8246,
+      "step": 3294
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.261867349099804,
+      "learning_rate": 1.0758087454918437e-05,
+      "loss": 0.8698,
+      "step": 3295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4721664472986458,
+      "learning_rate": 1.0753268055279328e-05,
+      "loss": 0.9105,
+      "step": 3296
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.405084957210123,
+      "learning_rate": 1.074844847967673e-05,
+      "loss": 0.8782,
+      "step": 3297
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5766235505618889,
+      "learning_rate": 1.0743628729236488e-05,
+      "loss": 0.8583,
+      "step": 3298
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4574825780583431,
+      "learning_rate": 1.0738808805084503e-05,
+      "loss": 0.8743,
+      "step": 3299
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4879152606397799,
+      "learning_rate": 1.0733988708346708e-05,
+      "loss": 0.8943,
+      "step": 3300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6945035499755132,
+      "learning_rate": 1.0729168440149077e-05,
+      "loss": 0.8587,
+      "step": 3301
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7116493873523388,
+      "learning_rate": 1.0724348001617626e-05,
+      "loss": 0.8792,
+      "step": 3302
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4195991664935301,
+      "learning_rate": 1.071952739387841e-05,
+      "loss": 0.8005,
+      "step": 3303
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6186570502625592,
+      "learning_rate": 1.0714706618057521e-05,
+      "loss": 0.8612,
+      "step": 3304
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5725985512190277,
+      "learning_rate": 1.0709885675281096e-05,
+      "loss": 0.8169,
+      "step": 3305
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4697268875730034,
+      "learning_rate": 1.070506456667531e-05,
+      "loss": 0.7828,
+      "step": 3306
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3733405623928798,
+      "learning_rate": 1.0700243293366365e-05,
+      "loss": 0.7682,
+      "step": 3307
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.385707077861223,
+      "learning_rate": 1.0695421856480519e-05,
+      "loss": 0.8699,
+      "step": 3308
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3671698432872128,
+      "learning_rate": 1.0690600257144062e-05,
+      "loss": 0.8566,
+      "step": 3309
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2562807853328557,
+      "learning_rate": 1.0685778496483312e-05,
+      "loss": 0.7756,
+      "step": 3310
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.432256248787592,
+      "learning_rate": 1.0680956575624637e-05,
+      "loss": 0.8291,
+      "step": 3311
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6006305204260534,
+      "learning_rate": 1.0676134495694439e-05,
+      "loss": 0.8341,
+      "step": 3312
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4602116538548533,
+      "learning_rate": 1.0671312257819155e-05,
+      "loss": 0.7872,
+      "step": 3313
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4536833933617224,
+      "learning_rate": 1.066648986312526e-05,
+      "loss": 0.8948,
+      "step": 3314
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4740122084886806,
+      "learning_rate": 1.066166731273927e-05,
+      "loss": 0.8412,
+      "step": 3315
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7936136752172585,
+      "learning_rate": 1.0656844607787727e-05,
+      "loss": 0.892,
+      "step": 3316
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2913633749641416,
+      "learning_rate": 1.0652021749397216e-05,
+      "loss": 0.8016,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.524952638171077,
+      "learning_rate": 1.0647198738694362e-05,
+      "loss": 0.859,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4305612156065701,
+      "learning_rate": 1.0642375576805822e-05,
+      "loss": 0.8886,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8176563368147927,
+      "learning_rate": 1.0637552264858278e-05,
+      "loss": 0.3428,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5608220591829975,
+      "learning_rate": 1.063272880397846e-05,
+      "loss": 0.8365,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4073317665150227,
+      "learning_rate": 1.0627905195293135e-05,
+      "loss": 0.9181,
+      "step": 3322
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3750375808922841,
+      "learning_rate": 1.0623081439929092e-05,
+      "loss": 0.8391,
+      "step": 3323
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3502911948056306,
+      "learning_rate": 1.0618257539013162e-05,
+      "loss": 0.8246,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4364394598401726,
+      "learning_rate": 1.0613433493672212e-05,
+      "loss": 0.8636,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4119382077863705,
+      "learning_rate": 1.0608609305033132e-05,
+      "loss": 0.8503,
+      "step": 3326
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4285975402761721,
+      "learning_rate": 1.0603784974222862e-05,
+      "loss": 0.7625,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4806941925151347,
+      "learning_rate": 1.059896050236836e-05,
+      "loss": 0.8629,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5974230351250054,
+      "learning_rate": 1.0594135890596626e-05,
+      "loss": 0.8308,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4199316360264063,
+      "learning_rate": 1.0589311140034687e-05,
+      "loss": 0.8772,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9374677279997454,
+      "learning_rate": 1.0584486251809607e-05,
+      "loss": 0.3231,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4092125416267967,
+      "learning_rate": 1.0579661227048484e-05,
+      "loss": 0.8906,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3912957052775976,
+      "learning_rate": 1.0574836066878436e-05,
+      "loss": 0.7822,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5214334566315058,
+      "learning_rate": 1.0570010772426627e-05,
+      "loss": 0.7969,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3584620174993083,
+      "learning_rate": 1.0565185344820248e-05,
+      "loss": 0.7973,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4008528666105484,
+      "learning_rate": 1.056035978518651e-05,
+      "loss": 0.7914,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5890886958906585,
+      "learning_rate": 1.0555534094652675e-05,
+      "loss": 0.8714,
+      "step": 3337
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4497686240042948,
+      "learning_rate": 1.055070827434602e-05,
+      "loss": 0.901,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4720263835329306,
+      "learning_rate": 1.0545882325393855e-05,
+      "loss": 0.7266,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5557475453766123,
+      "learning_rate": 1.0541056248923525e-05,
+      "loss": 0.8511,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8157541800058901,
+      "learning_rate": 1.0536230046062403e-05,
+      "loss": 0.3379,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4507454852829846,
+      "learning_rate": 1.0531403717937888e-05,
+      "loss": 0.9001,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5582387969360811,
+      "learning_rate": 1.052657726567741e-05,
+      "loss": 0.9101,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4455404887934284,
+      "learning_rate": 1.0521750690408434e-05,
+      "loss": 0.8998,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4149830177218141,
+      "learning_rate": 1.0516923993258441e-05,
+      "loss": 0.8047,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5139444481096482,
+      "learning_rate": 1.0512097175354952e-05,
+      "loss": 0.8879,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.564061075251842,
+      "learning_rate": 1.0507270237825513e-05,
+      "loss": 0.8589,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.600578731104028,
+      "learning_rate": 1.0502443181797696e-05,
+      "loss": 0.8265,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.529991932229045,
+      "learning_rate": 1.04976160083991e-05,
+      "loss": 0.8825,
+      "step": 3349
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4257194717949433,
+      "learning_rate": 1.0492788718757356e-05,
+      "loss": 0.859,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5481766813570168,
+      "learning_rate": 1.0487961314000121e-05,
+      "loss": 0.9265,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7954439866732235,
+      "learning_rate": 1.0483133795255072e-05,
+      "loss": 0.3158,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4419756671776065,
+      "learning_rate": 1.0478306163649919e-05,
+      "loss": 0.8269,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5925319577454673,
+      "learning_rate": 1.0473478420312403e-05,
+      "loss": 0.8426,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.454752769834404,
+      "learning_rate": 1.0468650566370276e-05,
+      "loss": 0.8296,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3799821564103434,
+      "learning_rate": 1.0463822602951332e-05,
+      "loss": 0.8385,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2911316044777288,
+      "learning_rate": 1.045899453118338e-05,
+      "loss": 0.9213,
+      "step": 3357
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5856373401875128,
+      "learning_rate": 1.045416635219426e-05,
+      "loss": 0.8897,
+      "step": 3358
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4944949851843312,
+      "learning_rate": 1.0449338067111839e-05,
+      "loss": 0.8466,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4438510052329614,
+      "learning_rate": 1.0444509677063997e-05,
+      "loss": 0.8537,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4253224575315486,
+      "learning_rate": 1.043968118317865e-05,
+      "loss": 0.8797,
+      "step": 3361
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4237900762876985,
+      "learning_rate": 1.0434852586583737e-05,
+      "loss": 0.9016,
+      "step": 3362
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.385505287050664,
+      "learning_rate": 1.0430023888407215e-05,
+      "loss": 0.8269,
+      "step": 3363
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3824235044374373,
+      "learning_rate": 1.0425195089777072e-05,
+      "loss": 0.8486,
+      "step": 3364
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3363181946515676,
+      "learning_rate": 1.042036619182131e-05,
+      "loss": 0.8656,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3769306059274253,
+      "learning_rate": 1.0415537195667963e-05,
+      "loss": 0.7979,
+      "step": 3366
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3437704057724318,
+      "learning_rate": 1.0410708102445091e-05,
+      "loss": 0.8795,
+      "step": 3367
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5801029138542073,
+      "learning_rate": 1.0405878913280762e-05,
+      "loss": 0.8885,
+      "step": 3368
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4037382211046627,
+      "learning_rate": 1.0401049629303077e-05,
+      "loss": 0.8228,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3928843358190088,
+      "learning_rate": 1.0396220251640163e-05,
+      "loss": 0.8134,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4154174106697666,
+      "learning_rate": 1.0391390781420157e-05,
+      "loss": 0.8066,
+      "step": 3371
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9039117454388134,
+      "learning_rate": 1.0386561219771222e-05,
+      "loss": 0.3323,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6176655472818309,
+      "learning_rate": 1.038173156782155e-05,
+      "loss": 0.8305,
+      "step": 3373
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.572117586452326,
+      "learning_rate": 1.0376901826699349e-05,
+      "loss": 0.9186,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4538043418438247,
+      "learning_rate": 1.037207199753284e-05,
+      "loss": 0.8045,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2011663359227436,
+      "learning_rate": 1.0367242081450274e-05,
+      "loss": 0.7749,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.67280307416685,
+      "learning_rate": 1.0362412079579925e-05,
+      "loss": 0.933,
+      "step": 3377
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5232488876128898,
+      "learning_rate": 1.0357581993050076e-05,
+      "loss": 0.8128,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4373897206983042,
+      "learning_rate": 1.0352751822989037e-05,
+      "loss": 0.8645,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7123411655510947,
+      "learning_rate": 1.0347921570525139e-05,
+      "loss": 0.8188,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5148923647970316,
+      "learning_rate": 1.0343091236786727e-05,
+      "loss": 0.8678,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5304738542180532,
+      "learning_rate": 1.0338260822902166e-05,
+      "loss": 0.8519,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2966406686620444,
+      "learning_rate": 1.0333430329999847e-05,
+      "loss": 0.8247,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4055012166427823,
+      "learning_rate": 1.0328599759208167e-05,
+      "loss": 0.9245,
+      "step": 3384
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3812381044259643,
+      "learning_rate": 1.0323769111655549e-05,
+      "loss": 0.906,
+      "step": 3385
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5038095405992304,
+      "learning_rate": 1.0318938388470439e-05,
+      "loss": 0.8307,
+      "step": 3386
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7944063065900513,
+      "learning_rate": 1.0314107590781284e-05,
+      "loss": 0.3517,
+      "step": 3387
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6270276999354705,
+      "learning_rate": 1.0309276719716567e-05,
+      "loss": 0.8457,
+      "step": 3388
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5765300754959457,
+      "learning_rate": 1.0304445776404778e-05,
+      "loss": 0.8124,
+      "step": 3389
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3348605223522607,
+      "learning_rate": 1.0299614761974426e-05,
+      "loss": 0.9029,
+      "step": 3390
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6366106990026816,
+      "learning_rate": 1.0294783677554035e-05,
+      "loss": 0.8696,
+      "step": 3391
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4577689502156046,
+      "learning_rate": 1.0289952524272147e-05,
+      "loss": 0.8429,
+      "step": 3392
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7120799633270154,
+      "learning_rate": 1.0285121303257321e-05,
+      "loss": 0.8248,
+      "step": 3393
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9089973986061121,
+      "learning_rate": 1.0280290015638129e-05,
+      "loss": 0.3627,
+      "step": 3394
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3561757233821814,
+      "learning_rate": 1.0275458662543161e-05,
+      "loss": 0.8559,
+      "step": 3395
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5449940372835376,
+      "learning_rate": 1.0270627245101026e-05,
+      "loss": 0.8215,
+      "step": 3396
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2878587383584195,
+      "learning_rate": 1.0265795764440335e-05,
+      "loss": 0.8348,
+      "step": 3397
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5177199102321068,
+      "learning_rate": 1.0260964221689729e-05,
+      "loss": 0.9529,
+      "step": 3398
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.563450037500708,
+      "learning_rate": 1.0256132617977856e-05,
+      "loss": 0.8202,
+      "step": 3399
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.422944037103568,
+      "learning_rate": 1.0251300954433377e-05,
+      "loss": 0.929,
+      "step": 3400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3383386822306005,
+      "learning_rate": 1.0246469232184968e-05,
+      "loss": 0.8325,
+      "step": 3401
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5252639056039872,
+      "learning_rate": 1.0241637452361323e-05,
+      "loss": 0.8408,
+      "step": 3402
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2936133135124792,
+      "learning_rate": 1.0236805616091148e-05,
+      "loss": 0.7926,
+      "step": 3403
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.823261602405843,
+      "learning_rate": 1.0231973724503152e-05,
+      "loss": 0.3524,
+      "step": 3404
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6340610325664995,
+      "learning_rate": 1.0227141778726075e-05,
+      "loss": 0.8217,
+      "step": 3405
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8259746336356296,
+      "learning_rate": 1.0222309779888656e-05,
+      "loss": 0.3607,
+      "step": 3406
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4087914601565654,
+      "learning_rate": 1.0217477729119648e-05,
+      "loss": 0.8592,
+      "step": 3407
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4812567583968104,
+      "learning_rate": 1.0212645627547821e-05,
+      "loss": 0.8718,
+      "step": 3408
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0356564590795982,
+      "learning_rate": 1.0207813476301955e-05,
+      "loss": 0.3653,
+      "step": 3409
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2516975238618395,
+      "learning_rate": 1.0202981276510841e-05,
+      "loss": 0.8742,
+      "step": 3410
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3432517326923943,
+      "learning_rate": 1.019814902930328e-05,
+      "loss": 0.8582,
+      "step": 3411
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4166719592716657,
+      "learning_rate": 1.0193316735808085e-05,
+      "loss": 0.8335,
+      "step": 3412
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.266556815487124,
+      "learning_rate": 1.0188484397154083e-05,
+      "loss": 0.8493,
+      "step": 3413
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4260501790843365,
+      "learning_rate": 1.0183652014470105e-05,
+      "loss": 0.8731,
+      "step": 3414
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.469974719290576,
+      "learning_rate": 1.0178819588885001e-05,
+      "loss": 0.9264,
+      "step": 3415
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.925301636828885,
+      "learning_rate": 1.0173987121527619e-05,
+      "loss": 0.327,
+      "step": 3416
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4926511046840938,
+      "learning_rate": 1.0169154613526831e-05,
+      "loss": 0.8761,
+      "step": 3417
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6290364264882893,
+      "learning_rate": 1.0164322066011509e-05,
+      "loss": 0.9029,
+      "step": 3418
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2345519225771537,
+      "learning_rate": 1.0159489480110536e-05,
+      "loss": 0.8194,
+      "step": 3419
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5341449784614702,
+      "learning_rate": 1.0154656856952805e-05,
+      "loss": 0.852,
+      "step": 3420
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5686703370107327,
+      "learning_rate": 1.0149824197667213e-05,
+      "loss": 0.7736,
+      "step": 3421
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3919099131495478,
+      "learning_rate": 1.0144991503382676e-05,
+      "loss": 0.8704,
+      "step": 3422
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.342730158715274,
+      "learning_rate": 1.0140158775228111e-05,
+      "loss": 0.7477,
+      "step": 3423
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.433967463664584,
+      "learning_rate": 1.013532601433244e-05,
+      "loss": 0.7749,
+      "step": 3424
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3450195129146365,
+      "learning_rate": 1.0130493221824598e-05,
+      "loss": 0.8148,
+      "step": 3425
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3576115365940458,
+      "learning_rate": 1.0125660398833528e-05,
+      "loss": 0.8098,
+      "step": 3426
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.682810772803721,
+      "learning_rate": 1.0120827546488175e-05,
+      "loss": 0.8974,
+      "step": 3427
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5815080295529864,
+      "learning_rate": 1.0115994665917497e-05,
+      "loss": 0.8791,
+      "step": 3428
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3273680268796113,
+      "learning_rate": 1.0111161758250451e-05,
+      "loss": 0.833,
+      "step": 3429
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5861881485544678,
+      "learning_rate": 1.010632882461601e-05,
+      "loss": 0.9005,
+      "step": 3430
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.486615505332043,
+      "learning_rate": 1.0101495866143143e-05,
+      "loss": 0.7811,
+      "step": 3431
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4187267303133957,
+      "learning_rate": 1.0096662883960833e-05,
+      "loss": 0.9513,
+      "step": 3432
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3946573350043077,
+      "learning_rate": 1.0091829879198061e-05,
+      "loss": 0.8659,
+      "step": 3433
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.397473477161376,
+      "learning_rate": 1.0086996852983822e-05,
+      "loss": 0.9039,
+      "step": 3434
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9320843507599951,
+      "learning_rate": 1.008216380644711e-05,
+      "loss": 0.2924,
+      "step": 3435
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4572545460855695,
+      "learning_rate": 1.0077330740716922e-05,
+      "loss": 0.8394,
+      "step": 3436
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.304388279638198,
+      "learning_rate": 1.0072497656922266e-05,
+      "loss": 0.8702,
+      "step": 3437
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3150299090643778,
+      "learning_rate": 1.0067664556192154e-05,
+      "loss": 0.8178,
+      "step": 3438
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5357642158967026,
+      "learning_rate": 1.0062831439655591e-05,
+      "loss": 0.7881,
+      "step": 3439
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.588994093213062,
+      "learning_rate": 1.0057998308441598e-05,
+      "loss": 0.9227,
+      "step": 3440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3774811739963508,
+      "learning_rate": 1.0053165163679196e-05,
+      "loss": 0.8628,
+      "step": 3441
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5394552971302613,
+      "learning_rate": 1.0048332006497406e-05,
+      "loss": 0.9178,
+      "step": 3442
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4357760000003168,
+      "learning_rate": 1.0043498838025252e-05,
+      "loss": 0.873,
+      "step": 3443
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3667153769434364,
+      "learning_rate": 1.0038665659391768e-05,
+      "loss": 0.8536,
+      "step": 3444
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4129285981820916,
+      "learning_rate": 1.0033832471725982e-05,
+      "loss": 0.8652,
+      "step": 3445
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3788770005255109,
+      "learning_rate": 1.0028999276156926e-05,
+      "loss": 0.9494,
+      "step": 3446
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.437959527600928,
+      "learning_rate": 1.0024166073813634e-05,
+      "loss": 0.851,
+      "step": 3447
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3953992388955814,
+      "learning_rate": 1.0019332865825149e-05,
+      "loss": 0.8021,
+      "step": 3448
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2450964443498471,
+      "learning_rate": 1.0014499653320504e-05,
+      "loss": 0.8689,
+      "step": 3449
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8268230647239653,
+      "learning_rate": 1.0009666437428736e-05,
+      "loss": 0.3281,
+      "step": 3450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6439971532235766,
+      "learning_rate": 1.000483321927889e-05,
+      "loss": 0.7811,
+      "step": 3451
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5491931444265257,
+      "learning_rate": 1e-05,
+      "loss": 0.8971,
+      "step": 3452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3962384987833836,
+      "learning_rate": 9.995166780721112e-06,
+      "loss": 0.8105,
+      "step": 3453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4999578781720608,
+      "learning_rate": 9.990333562571266e-06,
+      "loss": 0.7888,
+      "step": 3454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5713598082496674,
+      "learning_rate": 9.985500346679497e-06,
+      "loss": 0.8393,
+      "step": 3455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5588921724933924,
+      "learning_rate": 9.980667134174854e-06,
+      "loss": 0.8367,
+      "step": 3456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.409938850974922,
+      "learning_rate": 9.975833926186367e-06,
+      "loss": 0.9221,
+      "step": 3457
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6915889480538027,
+      "learning_rate": 9.971000723843077e-06,
+      "loss": 0.881,
+      "step": 3458
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2717013195965097,
+      "learning_rate": 9.966167528274021e-06,
+      "loss": 0.8117,
+      "step": 3459
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6657230773270417,
+      "learning_rate": 9.961334340608233e-06,
+      "loss": 0.9457,
+      "step": 3460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.476567160508203,
+      "learning_rate": 9.956501161974747e-06,
+      "loss": 0.8671,
+      "step": 3461
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.398525152043447,
+      "learning_rate": 9.951667993502599e-06,
+      "loss": 0.8317,
+      "step": 3462
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.409430063915483,
+      "learning_rate": 9.946834836320809e-06,
+      "loss": 0.7966,
+      "step": 3463
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3155647348693218,
+      "learning_rate": 9.942001691558405e-06,
+      "loss": 0.8294,
+      "step": 3464
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4181669138795794,
+      "learning_rate": 9.937168560344412e-06,
+      "loss": 0.8945,
+      "step": 3465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4805933068300412,
+      "learning_rate": 9.93233544380785e-06,
+      "loss": 0.8231,
+      "step": 3466
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4679960030080992,
+      "learning_rate": 9.927502343077732e-06,
+      "loss": 0.8718,
+      "step": 3467
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4759980906730832,
+      "learning_rate": 9.922669259283078e-06,
+      "loss": 0.8738,
+      "step": 3468
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4435891341511244,
+      "learning_rate": 9.917836193552895e-06,
+      "loss": 0.7917,
+      "step": 3469
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6149539257186423,
+      "learning_rate": 9.913003147016181e-06,
+      "loss": 0.8177,
+      "step": 3470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2646296481353163,
+      "learning_rate": 9.908170120801942e-06,
+      "loss": 0.903,
+      "step": 3471
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6149703568830507,
+      "learning_rate": 9.903337116039172e-06,
+      "loss": 0.8874,
+      "step": 3472
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5138555218440664,
+      "learning_rate": 9.898504133856858e-06,
+      "loss": 0.8392,
+      "step": 3473
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3751495839434065,
+      "learning_rate": 9.893671175383995e-06,
+      "loss": 0.8933,
+      "step": 3474
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3562143958654773,
+      "learning_rate": 9.888838241749552e-06,
+      "loss": 0.8951,
+      "step": 3475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3420581396688758,
+      "learning_rate": 9.884005334082508e-06,
+      "loss": 0.8905,
+      "step": 3476
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6688196872686993,
+      "learning_rate": 9.879172453511827e-06,
+      "loss": 0.9132,
+      "step": 3477
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4965535508101526,
+      "learning_rate": 9.874339601166474e-06,
+      "loss": 0.7874,
+      "step": 3478
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5107073172630312,
+      "learning_rate": 9.869506778175405e-06,
+      "loss": 0.8915,
+      "step": 3479
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3490376296511224,
+      "learning_rate": 9.864673985667563e-06,
+      "loss": 0.865,
+      "step": 3480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8606336133466326,
+      "learning_rate": 9.859841224771892e-06,
+      "loss": 0.3232,
+      "step": 3481
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6294404704402772,
+      "learning_rate": 9.855008496617326e-06,
+      "loss": 0.8966,
+      "step": 3482
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4277643586753277,
+      "learning_rate": 9.850175802332788e-06,
+      "loss": 0.8575,
+      "step": 3483
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4295334133640925,
+      "learning_rate": 9.845343143047198e-06,
+      "loss": 0.8474,
+      "step": 3484
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3416235595279287,
+      "learning_rate": 9.840510519889467e-06,
+      "loss": 0.797,
+      "step": 3485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3743923663939088,
+      "learning_rate": 9.835677933988493e-06,
+      "loss": 0.8545,
+      "step": 3486
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4360387653563598,
+      "learning_rate": 9.830845386473169e-06,
+      "loss": 0.8848,
+      "step": 3487
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4465719670119639,
+      "learning_rate": 9.826012878472383e-06,
+      "loss": 0.7971,
+      "step": 3488
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4395948997565176,
+      "learning_rate": 9.821180411115002e-06,
+      "loss": 0.7928,
+      "step": 3489
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3503594356826982,
+      "learning_rate": 9.816347985529898e-06,
+      "loss": 0.8364,
+      "step": 3490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3977523449230942,
+      "learning_rate": 9.81151560284592e-06,
+      "loss": 0.8894,
+      "step": 3491
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.427630707751418,
+      "learning_rate": 9.806683264191916e-06,
+      "loss": 0.8475,
+      "step": 3492
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.390789281458476,
+      "learning_rate": 9.801850970696722e-06,
+      "loss": 0.8817,
+      "step": 3493
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4045923045630875,
+      "learning_rate": 9.797018723489162e-06,
+      "loss": 0.8881,
+      "step": 3494
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5868689862231609,
+      "learning_rate": 9.792186523698048e-06,
+      "loss": 0.897,
+      "step": 3495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.304606199586758,
+      "learning_rate": 9.787354372452182e-06,
+      "loss": 0.8632,
+      "step": 3496
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3613617515241991,
+      "learning_rate": 9.782522270880354e-06,
+      "loss": 0.8005,
+      "step": 3497
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.474100117188263,
+      "learning_rate": 9.777690220111348e-06,
+      "loss": 0.8629,
+      "step": 3498
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4886728595038716,
+      "learning_rate": 9.772858221273926e-06,
+      "loss": 0.8898,
+      "step": 3499
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3751076024932953,
+      "learning_rate": 9.768026275496848e-06,
+      "loss": 0.8846,
+      "step": 3500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5461795175423334,
+      "learning_rate": 9.763194383908857e-06,
+      "loss": 0.9162,
+      "step": 3501
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5141999923105947,
+      "learning_rate": 9.75836254763868e-06,
+      "loss": 0.9415,
+      "step": 3502
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4606037115570538,
+      "learning_rate": 9.753530767815036e-06,
+      "loss": 0.8304,
+      "step": 3503
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6429926386650577,
+      "learning_rate": 9.748699045566626e-06,
+      "loss": 0.84,
+      "step": 3504
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5817934299018297,
+      "learning_rate": 9.743867382022147e-06,
+      "loss": 0.8337,
+      "step": 3505
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6776480807980998,
+      "learning_rate": 9.739035778310273e-06,
+      "loss": 0.8274,
+      "step": 3506
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4683466063134807,
+      "learning_rate": 9.73420423555967e-06,
+      "loss": 0.8072,
+      "step": 3507
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4015235262069392,
+      "learning_rate": 9.729372754898979e-06,
+      "loss": 0.8712,
+      "step": 3508
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2740110088831653,
+      "learning_rate": 9.72454133745684e-06,
+      "loss": 0.8242,
+      "step": 3509
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.455902727462656,
+      "learning_rate": 9.719709984361873e-06,
+      "loss": 0.8048,
+      "step": 3510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3958298369679383,
+      "learning_rate": 9.714878696742682e-06,
+      "loss": 0.8955,
+      "step": 3511
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3503633296382895,
+      "learning_rate": 9.710047475727854e-06,
+      "loss": 0.8193,
+      "step": 3512
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4416193274786029,
+      "learning_rate": 9.705216322445967e-06,
+      "loss": 0.8766,
+      "step": 3513
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.429921863190966,
+      "learning_rate": 9.700385238025579e-06,
+      "loss": 0.7381,
+      "step": 3514
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4837315532550757,
+      "learning_rate": 9.695554223595224e-06,
+      "loss": 0.8643,
+      "step": 3515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4185058933318586,
+      "learning_rate": 9.690723280283436e-06,
+      "loss": 0.7869,
+      "step": 3516
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4232037133997082,
+      "learning_rate": 9.685892409218718e-06,
+      "loss": 0.9165,
+      "step": 3517
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6296266888238622,
+      "learning_rate": 9.681061611529566e-06,
+      "loss": 0.8628,
+      "step": 3518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3830645878320222,
+      "learning_rate": 9.676230888344451e-06,
+      "loss": 0.8816,
+      "step": 3519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3966046074814396,
+      "learning_rate": 9.67140024079184e-06,
+      "loss": 0.8832,
+      "step": 3520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3218236889727253,
+      "learning_rate": 9.66656967000016e-06,
+      "loss": 0.7842,
+      "step": 3521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4563715946151394,
+      "learning_rate": 9.661739177097836e-06,
+      "loss": 0.8157,
+      "step": 3522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4276943501238542,
+      "learning_rate": 9.656908763213276e-06,
+      "loss": 0.8315,
+      "step": 3523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5598785833817714,
+      "learning_rate": 9.652078429474863e-06,
+      "loss": 0.8444,
+      "step": 3524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4390880666578192,
+      "learning_rate": 9.647248177010964e-06,
+      "loss": 0.9087,
+      "step": 3525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5631165635038766,
+      "learning_rate": 9.642418006949926e-06,
+      "loss": 0.9012,
+      "step": 3526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.401429788340974,
+      "learning_rate": 9.63758792042008e-06,
+      "loss": 0.8189,
+      "step": 3527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4051232884020528,
+      "learning_rate": 9.632757918549729e-06,
+      "loss": 0.8885,
+      "step": 3528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4049373593315253,
+      "learning_rate": 9.627928002467164e-06,
+      "loss": 0.8789,
+      "step": 3529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7769730338513625,
+      "learning_rate": 9.623098173300655e-06,
+      "loss": 0.7836,
+      "step": 3530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7560963613957583,
+      "learning_rate": 9.618268432178451e-06,
+      "loss": 0.9626,
+      "step": 3531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4968967710416943,
+      "learning_rate": 9.613438780228777e-06,
+      "loss": 0.863,
+      "step": 3532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2567442006122844,
+      "learning_rate": 9.60860921857985e-06,
+      "loss": 0.8815,
+      "step": 3533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4565100383201046,
+      "learning_rate": 9.603779748359842e-06,
+      "loss": 0.8918,
+      "step": 3534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3208910327756287,
+      "learning_rate": 9.598950370696924e-06,
+      "loss": 0.8856,
+      "step": 3535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5268204279075828,
+      "learning_rate": 9.594121086719241e-06,
+      "loss": 0.9415,
+      "step": 3536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3787442009454582,
+      "learning_rate": 9.589291897554912e-06,
+      "loss": 0.8071,
+      "step": 3537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3772955610702697,
+      "learning_rate": 9.584462804332038e-06,
+      "loss": 0.8394,
+      "step": 3538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4024396819717755,
+      "learning_rate": 9.579633808178693e-06,
+      "loss": 0.8812,
+      "step": 3539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4484462324206373,
+      "learning_rate": 9.574804910222934e-06,
+      "loss": 0.8493,
+      "step": 3540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.588517954263737,
+      "learning_rate": 9.569976111592789e-06,
+      "loss": 0.8839,
+      "step": 3541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.47160428960858,
+      "learning_rate": 9.565147413416266e-06,
+      "loss": 0.838,
+      "step": 3542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3582415294730648,
+      "learning_rate": 9.560318816821354e-06,
+      "loss": 0.8813,
+      "step": 3543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4717764889042155,
+      "learning_rate": 9.555490322936007e-06,
+      "loss": 0.8361,
+      "step": 3544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.420587255872903,
+      "learning_rate": 9.550661932888164e-06,
+      "loss": 0.9022,
+      "step": 3545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5727905664727972,
+      "learning_rate": 9.545833647805743e-06,
+      "loss": 0.8461,
+      "step": 3546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4340290427362725,
+      "learning_rate": 9.541005468816622e-06,
+      "loss": 0.8367,
+      "step": 3547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4525572631647652,
+      "learning_rate": 9.53617739704867e-06,
+      "loss": 0.8262,
+      "step": 3548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3596503866711438,
+      "learning_rate": 9.531349433629729e-06,
+      "loss": 0.8865,
+      "step": 3549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.328003759319903,
+      "learning_rate": 9.526521579687603e-06,
+      "loss": 0.8414,
+      "step": 3550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4274244800583733,
+      "learning_rate": 9.521693836350083e-06,
+      "loss": 0.8624,
+      "step": 3551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2721743081537307,
+      "learning_rate": 9.516866204744932e-06,
+      "loss": 0.798,
+      "step": 3552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.423403909592571,
+      "learning_rate": 9.512038685999882e-06,
+      "loss": 0.8182,
+      "step": 3553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3349668325476816,
+      "learning_rate": 9.507211281242646e-06,
+      "loss": 0.7936,
+      "step": 3554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2729769829172104,
+      "learning_rate": 9.502383991600901e-06,
+      "loss": 0.8028,
+      "step": 3555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3212568408846261,
+      "learning_rate": 9.497556818202306e-06,
+      "loss": 0.8854,
+      "step": 3556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3679916047148348,
+      "learning_rate": 9.492729762174489e-06,
+      "loss": 0.8882,
+      "step": 3557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6020324192081918,
+      "learning_rate": 9.487902824645048e-06,
+      "loss": 0.8382,
+      "step": 3558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4011749069206587,
+      "learning_rate": 9.483076006741564e-06,
+      "loss": 0.914,
+      "step": 3559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5141374591847243,
+      "learning_rate": 9.478249309591571e-06,
+      "loss": 0.7728,
+      "step": 3560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.440791733711582,
+      "learning_rate": 9.473422734322593e-06,
+      "loss": 0.936,
+      "step": 3561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8757881665786142,
+      "learning_rate": 9.468596282062114e-06,
+      "loss": 0.3428,
+      "step": 3562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4844334352435307,
+      "learning_rate": 9.4637699539376e-06,
+      "loss": 0.8313,
+      "step": 3563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4568034114088537,
+      "learning_rate": 9.458943751076475e-06,
+      "loss": 0.7888,
+      "step": 3564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.483814626877465,
+      "learning_rate": 9.45411767460615e-06,
+      "loss": 0.7983,
+      "step": 3565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3152010626237944,
+      "learning_rate": 9.449291725653985e-06,
+      "loss": 0.856,
+      "step": 3566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5257691708588121,
+      "learning_rate": 9.444465905347327e-06,
+      "loss": 0.8419,
+      "step": 3567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7190177871427816,
+      "learning_rate": 9.439640214813491e-06,
+      "loss": 0.8877,
+      "step": 3568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4985641707877073,
+      "learning_rate": 9.434814655179756e-06,
+      "loss": 0.8346,
+      "step": 3569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4425306392945307,
+      "learning_rate": 9.429989227573373e-06,
+      "loss": 0.9176,
+      "step": 3570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3746720838755477,
+      "learning_rate": 9.425163933121564e-06,
+      "loss": 0.836,
+      "step": 3571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5429985562596362,
+      "learning_rate": 9.420338772951521e-06,
+      "loss": 0.7823,
+      "step": 3572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4744078909042018,
+      "learning_rate": 9.415513748190396e-06,
+      "loss": 0.7614,
+      "step": 3573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7209365591126629,
+      "learning_rate": 9.410688859965316e-06,
+      "loss": 0.8301,
+      "step": 3574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3494378842155879,
+      "learning_rate": 9.405864109403378e-06,
+      "loss": 0.8302,
+      "step": 3575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3813497247339959,
+      "learning_rate": 9.401039497631642e-06,
+      "loss": 0.866,
+      "step": 3576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2198492695495537,
+      "learning_rate": 9.39621502577714e-06,
+      "loss": 0.8478,
+      "step": 3577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.61010211426017,
+      "learning_rate": 9.391390694966871e-06,
+      "loss": 0.8767,
+      "step": 3578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.704134001873991,
+      "learning_rate": 9.386566506327793e-06,
+      "loss": 0.8722,
+      "step": 3579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6219507680910457,
+      "learning_rate": 9.38174246098684e-06,
+      "loss": 0.8499,
+      "step": 3580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4237076096099812,
+      "learning_rate": 9.37691856007091e-06,
+      "loss": 0.8444,
+      "step": 3581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4439272757009594,
+      "learning_rate": 9.372094804706867e-06,
+      "loss": 0.8567,
+      "step": 3582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3840393884520266,
+      "learning_rate": 9.367271196021539e-06,
+      "loss": 0.7988,
+      "step": 3583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3749031519106785,
+      "learning_rate": 9.362447735141723e-06,
+      "loss": 0.7628,
+      "step": 3584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3315978589720228,
+      "learning_rate": 9.357624423194185e-06,
+      "loss": 0.7265,
+      "step": 3585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.408940903040862,
+      "learning_rate": 9.35280126130564e-06,
+      "loss": 0.8107,
+      "step": 3586
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3050787861493243,
+      "learning_rate": 9.347978250602786e-06,
+      "loss": 0.8877,
+      "step": 3587
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.808001497731327,
+      "learning_rate": 9.343155392212276e-06,
+      "loss": 0.339,
+      "step": 3588
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.271314135667705,
+      "learning_rate": 9.338332687260733e-06,
+      "loss": 0.8631,
+      "step": 3589
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4720510180123563,
+      "learning_rate": 9.333510136874741e-06,
+      "loss": 0.8051,
+      "step": 3590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5133052468058892,
+      "learning_rate": 9.32868774218085e-06,
+      "loss": 0.7821,
+      "step": 3591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.46234917010658,
+      "learning_rate": 9.323865504305566e-06,
+      "loss": 0.9526,
+      "step": 3592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3227504731588213,
+      "learning_rate": 9.319043424375366e-06,
+      "loss": 0.85,
+      "step": 3593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5174377442756495,
+      "learning_rate": 9.314221503516691e-06,
+      "loss": 0.8886,
+      "step": 3594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4949720014446959,
+      "learning_rate": 9.309399742855943e-06,
+      "loss": 0.9539,
+      "step": 3595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4120682945185337,
+      "learning_rate": 9.304578143519481e-06,
+      "loss": 0.8741,
+      "step": 3596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3994489706707465,
+      "learning_rate": 9.299756706633636e-06,
+      "loss": 0.7803,
+      "step": 3597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3259161483491646,
+      "learning_rate": 9.294935433324697e-06,
+      "loss": 0.8158,
+      "step": 3598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5547741968815594,
+      "learning_rate": 9.290114324718906e-06,
+      "loss": 0.8629,
+      "step": 3599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.352249375454564,
+      "learning_rate": 9.285293381942482e-06,
+      "loss": 0.8699,
+      "step": 3600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4072800711492426,
+      "learning_rate": 9.280472606121594e-06,
+      "loss": 0.8532,
+      "step": 3601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.654395452937887,
+      "learning_rate": 9.275651998382377e-06,
+      "loss": 0.8206,
+      "step": 3602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5882043159144885,
+      "learning_rate": 9.270831559850925e-06,
+      "loss": 0.8294,
+      "step": 3603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4364936962977737,
+      "learning_rate": 9.266011291653296e-06,
+      "loss": 0.8994,
+      "step": 3604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3703816811813574,
+      "learning_rate": 9.261191194915499e-06,
+      "loss": 0.8437,
+      "step": 3605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4929826774726052,
+      "learning_rate": 9.256371270763515e-06,
+      "loss": 0.9054,
+      "step": 3606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3279618485600522,
+      "learning_rate": 9.251551520323273e-06,
+      "loss": 0.8317,
+      "step": 3607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3984147521518384,
+      "learning_rate": 9.246731944720675e-06,
+      "loss": 0.9185,
+      "step": 3608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.58201288518725,
+      "learning_rate": 9.241912545081566e-06,
+      "loss": 0.9057,
+      "step": 3609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.567417912075355,
+      "learning_rate": 9.237093322531765e-06,
+      "loss": 0.9229,
+      "step": 3610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4436698180955618,
+      "learning_rate": 9.232274278197038e-06,
+      "loss": 0.8431,
+      "step": 3611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4854795700482486,
+      "learning_rate": 9.227455413203115e-06,
+      "loss": 0.9073,
+      "step": 3612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4293128197513787,
+      "learning_rate": 9.222636728675687e-06,
+      "loss": 0.8548,
+      "step": 3613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4813444641378652,
+      "learning_rate": 9.217818225740394e-06,
+      "loss": 0.8235,
+      "step": 3614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4726405471753816,
+      "learning_rate": 9.212999905522841e-06,
+      "loss": 0.9532,
+      "step": 3615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.35946535969364,
+      "learning_rate": 9.20818176914859e-06,
+      "loss": 0.754,
+      "step": 3616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3769326309951566,
+      "learning_rate": 9.203363817743159e-06,
+      "loss": 0.8283,
+      "step": 3617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.490114739691744,
+      "learning_rate": 9.198546052432018e-06,
+      "loss": 0.8051,
+      "step": 3618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5055819452526438,
+      "learning_rate": 9.193728474340598e-06,
+      "loss": 0.8298,
+      "step": 3619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4324767421589004,
+      "learning_rate": 9.188911084594286e-06,
+      "loss": 0.9232,
+      "step": 3620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4386311336013364,
+      "learning_rate": 9.184093884318426e-06,
+      "loss": 0.893,
+      "step": 3621
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4429187011610316,
+      "learning_rate": 9.179276874638315e-06,
+      "loss": 0.8807,
+      "step": 3622
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.467235808090102,
+      "learning_rate": 9.174460056679214e-06,
+      "loss": 0.8067,
+      "step": 3623
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3880377809290287,
+      "learning_rate": 9.16964343156632e-06,
+      "loss": 0.8749,
+      "step": 3624
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3900396944950153,
+      "learning_rate": 9.164827000424804e-06,
+      "loss": 0.8239,
+      "step": 3625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3999335049692658,
+      "learning_rate": 9.160010764379782e-06,
+      "loss": 0.8855,
+      "step": 3626
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7598920865615002,
+      "learning_rate": 9.15519472455633e-06,
+      "loss": 0.9147,
+      "step": 3627
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3450064482122197,
+      "learning_rate": 9.150378882079475e-06,
+      "loss": 0.8669,
+      "step": 3628
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.322550110521103,
+      "learning_rate": 9.145563238074198e-06,
+      "loss": 0.8886,
+      "step": 3629
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5315689416487994,
+      "learning_rate": 9.140747793665439e-06,
+      "loss": 0.8772,
+      "step": 3630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.337534278554035,
+      "learning_rate": 9.135932549978075e-06,
+      "loss": 0.8542,
+      "step": 3631
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3680346160196941,
+      "learning_rate": 9.131117508136953e-06,
+      "loss": 0.7951,
+      "step": 3632
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3785027441750946,
+      "learning_rate": 9.12630266926687e-06,
+      "loss": 0.8613,
+      "step": 3633
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.92137407376053,
+      "learning_rate": 9.121488034492569e-06,
+      "loss": 0.3458,
+      "step": 3634
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3616645129198242,
+      "learning_rate": 9.11667360493875e-06,
+      "loss": 0.8985,
+      "step": 3635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5686851438664762,
+      "learning_rate": 9.111859381730071e-06,
+      "loss": 0.8012,
+      "step": 3636
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.502591740200375,
+      "learning_rate": 9.107045365991123e-06,
+      "loss": 0.9713,
+      "step": 3637
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3178483834558912,
+      "learning_rate": 9.102231558846467e-06,
+      "loss": 0.8959,
+      "step": 3638
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2851938456685843,
+      "learning_rate": 9.097417961420608e-06,
+      "loss": 0.785,
+      "step": 3639
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0918073484726687,
+      "learning_rate": 9.092604574838004e-06,
+      "loss": 0.9197,
+      "step": 3640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.476397571990377,
+      "learning_rate": 9.08779140022306e-06,
+      "loss": 0.8163,
+      "step": 3641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.32157160394685,
+      "learning_rate": 9.082978438700138e-06,
+      "loss": 0.8903,
+      "step": 3642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5251531141297938,
+      "learning_rate": 9.078165691393548e-06,
+      "loss": 0.7586,
+      "step": 3643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.63149351194706,
+      "learning_rate": 9.073353159427538e-06,
+      "loss": 0.8153,
+      "step": 3644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5791357906031458,
+      "learning_rate": 9.068540843926324e-06,
+      "loss": 0.9268,
+      "step": 3645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5209111822193677,
+      "learning_rate": 9.06372874601406e-06,
+      "loss": 0.8482,
+      "step": 3646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1645795734612956,
+      "learning_rate": 9.058916866814857e-06,
+      "loss": 0.6842,
+      "step": 3647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5763810319166138,
+      "learning_rate": 9.054105207452766e-06,
+      "loss": 0.8708,
+      "step": 3648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.409372429544046,
+      "learning_rate": 9.049293769051797e-06,
+      "loss": 0.7862,
+      "step": 3649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3887514696501861,
+      "learning_rate": 9.044482552735895e-06,
+      "loss": 0.8234,
+      "step": 3650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4237640161039977,
+      "learning_rate": 9.039671559628963e-06,
+      "loss": 0.8167,
+      "step": 3651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4797547626715986,
+      "learning_rate": 9.034860790854848e-06,
+      "loss": 0.8635,
+      "step": 3652
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6376077381756147,
+      "learning_rate": 9.030050247537352e-06,
+      "loss": 0.821,
+      "step": 3653
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5123309009875996,
+      "learning_rate": 9.025239930800214e-06,
+      "loss": 0.8598,
+      "step": 3654
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4201874478057306,
+      "learning_rate": 9.020429841767125e-06,
+      "loss": 0.7664,
+      "step": 3655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4104704506833088,
+      "learning_rate": 9.015619981561726e-06,
+      "loss": 0.8372,
+      "step": 3656
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4274461457888306,
+      "learning_rate": 9.010810351307593e-06,
+      "loss": 0.8264,
+      "step": 3657
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.50307120945164,
+      "learning_rate": 9.006000952128262e-06,
+      "loss": 0.8302,
+      "step": 3658
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3303321163143917,
+      "learning_rate": 9.001191785147207e-06,
+      "loss": 0.7848,
+      "step": 3659
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5566102315694863,
+      "learning_rate": 8.996382851487851e-06,
+      "loss": 0.8337,
+      "step": 3660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.350540387232985,
+      "learning_rate": 8.991574152273562e-06,
+      "loss": 0.7944,
+      "step": 3661
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4692775566401828,
+      "learning_rate": 8.986765688627652e-06,
+      "loss": 0.9261,
+      "step": 3662
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4754467752753881,
+      "learning_rate": 8.981957461673375e-06,
+      "loss": 0.9122,
+      "step": 3663
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4751818556180016,
+      "learning_rate": 8.977149472533938e-06,
+      "loss": 0.853,
+      "step": 3664
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4982121418702838,
+      "learning_rate": 8.972341722332485e-06,
+      "loss": 0.8164,
+      "step": 3665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3554341572214257,
+      "learning_rate": 8.96753421219211e-06,
+      "loss": 0.8351,
+      "step": 3666
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4273524040853462,
+      "learning_rate": 8.962726943235845e-06,
+      "loss": 0.7495,
+      "step": 3667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3278289661342635,
+      "learning_rate": 8.957919916586668e-06,
+      "loss": 0.8635,
+      "step": 3668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.415638814315592,
+      "learning_rate": 8.953113133367506e-06,
+      "loss": 0.8099,
+      "step": 3669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6108040992121972,
+      "learning_rate": 8.948306594701216e-06,
+      "loss": 0.8644,
+      "step": 3670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3928125205605932,
+      "learning_rate": 8.943500301710612e-06,
+      "loss": 0.849,
+      "step": 3671
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4263875090410174,
+      "learning_rate": 8.938694255518444e-06,
+      "loss": 0.9428,
+      "step": 3672
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3068800465943815,
+      "learning_rate": 8.933888457247402e-06,
+      "loss": 0.9137,
+      "step": 3673
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.484848507848754,
+      "learning_rate": 8.929082908020122e-06,
+      "loss": 0.8755,
+      "step": 3674
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.739395187581526,
+      "learning_rate": 8.924277608959185e-06,
+      "loss": 0.8508,
+      "step": 3675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4402809816314825,
+      "learning_rate": 8.919472561187102e-06,
+      "loss": 0.8728,
+      "step": 3676
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4125680983002775,
+      "learning_rate": 8.91466776582634e-06,
+      "loss": 0.7384,
+      "step": 3677
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2714720247843756,
+      "learning_rate": 8.909863223999292e-06,
+      "loss": 0.864,
+      "step": 3678
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6191381942215402,
+      "learning_rate": 8.905058936828305e-06,
+      "loss": 0.8752,
+      "step": 3679
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5091538232162136,
+      "learning_rate": 8.900254905435658e-06,
+      "loss": 0.9165,
+      "step": 3680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5629433569131972,
+      "learning_rate": 8.895451130943578e-06,
+      "loss": 0.8256,
+      "step": 3681
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4238194156636768,
+      "learning_rate": 8.890647614474223e-06,
+      "loss": 0.8212,
+      "step": 3682
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8088611912616787,
+      "learning_rate": 8.885844357149695e-06,
+      "loss": 0.3902,
+      "step": 3683
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9599801679607902,
+      "learning_rate": 8.881041360092035e-06,
+      "loss": 0.3242,
+      "step": 3684
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.60980110175648,
+      "learning_rate": 8.876238624423224e-06,
+      "loss": 0.8785,
+      "step": 3685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4839687368679044,
+      "learning_rate": 8.871436151265183e-06,
+      "loss": 0.8752,
+      "step": 3686
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4882005803580445,
+      "learning_rate": 8.86663394173977e-06,
+      "loss": 0.8671,
+      "step": 3687
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4178082136173882,
+      "learning_rate": 8.861831996968785e-06,
+      "loss": 0.8665,
+      "step": 3688
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.574815976418105,
+      "learning_rate": 8.857030318073955e-06,
+      "loss": 0.8933,
+      "step": 3689
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4431843315959276,
+      "learning_rate": 8.852228906176958e-06,
+      "loss": 0.8461,
+      "step": 3690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4804551506424388,
+      "learning_rate": 8.8474277623994e-06,
+      "loss": 0.8183,
+      "step": 3691
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.621543558176056,
+      "learning_rate": 8.842626887862832e-06,
+      "loss": 0.9262,
+      "step": 3692
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2599794494671617,
+      "learning_rate": 8.83782628368874e-06,
+      "loss": 0.8692,
+      "step": 3693
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3462655631952707,
+      "learning_rate": 8.833025950998547e-06,
+      "loss": 0.8694,
+      "step": 3694
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4281421022439762,
+      "learning_rate": 8.828225890913604e-06,
+      "loss": 0.9251,
+      "step": 3695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2492095750408565,
+      "learning_rate": 8.82342610455521e-06,
+      "loss": 0.8945,
+      "step": 3696
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5149663865692948,
+      "learning_rate": 8.818626593044595e-06,
+      "loss": 0.8192,
+      "step": 3697
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4909150180348938,
+      "learning_rate": 8.813827357502926e-06,
+      "loss": 0.869,
+      "step": 3698
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9020382175022832,
+      "learning_rate": 8.809028399051302e-06,
+      "loss": 0.3518,
+      "step": 3699
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4520313431220402,
+      "learning_rate": 8.804229718810765e-06,
+      "loss": 0.9426,
+      "step": 3700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3110802967675608,
+      "learning_rate": 8.799431317902289e-06,
+      "loss": 0.8607,
+      "step": 3701
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7065063319823648,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.7935,
+      "step": 3702
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2540867443840078,
+      "learning_rate": 8.78983535856506e-06,
+      "loss": 0.8654,
+      "step": 3703
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3607794628838974,
+      "learning_rate": 8.785037802377929e-06,
+      "loss": 0.9087,
+      "step": 3704
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4927415355990383,
+      "learning_rate": 8.780240530006088e-06,
+      "loss": 0.8236,
+      "step": 3705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.043188025608745,
+      "learning_rate": 8.775443542570182e-06,
+      "loss": 0.3716,
+      "step": 3706
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.414927917232927,
+      "learning_rate": 8.77064684119079e-06,
+      "loss": 0.8667,
+      "step": 3707
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3673059854936396,
+      "learning_rate": 8.765850426988414e-06,
+      "loss": 0.8297,
+      "step": 3708
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5143694303989128,
+      "learning_rate": 8.7610543010835e-06,
+      "loss": 0.9296,
+      "step": 3709
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9946903070291274,
+      "learning_rate": 8.756258464596426e-06,
+      "loss": 0.3683,
+      "step": 3710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5999828561687655,
+      "learning_rate": 8.7514629186475e-06,
+      "loss": 0.8469,
+      "step": 3711
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2716777884402222,
+      "learning_rate": 8.746667664356957e-06,
+      "loss": 0.8119,
+      "step": 3712
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.344547015956051,
+      "learning_rate": 8.741872702844975e-06,
+      "loss": 0.87,
+      "step": 3713
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4579613046452948,
+      "learning_rate": 8.737078035231659e-06,
+      "loss": 0.868,
+      "step": 3714
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4720169685824185,
+      "learning_rate": 8.732283662637034e-06,
+      "loss": 0.8837,
+      "step": 3715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2252768844980197,
+      "learning_rate": 8.727489586181073e-06,
+      "loss": 0.8548,
+      "step": 3716
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3151230578592215,
+      "learning_rate": 8.722695806983673e-06,
+      "loss": 0.8467,
+      "step": 3717
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5932891000445994,
+      "learning_rate": 8.717902326164657e-06,
+      "loss": 0.9325,
+      "step": 3718
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.47362186258443,
+      "learning_rate": 8.713109144843785e-06,
+      "loss": 0.8729,
+      "step": 3719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5056953254232075,
+      "learning_rate": 8.708316264140751e-06,
+      "loss": 0.8864,
+      "step": 3720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4934534543403102,
+      "learning_rate": 8.703523685175161e-06,
+      "loss": 0.9236,
+      "step": 3721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3906336526923,
+      "learning_rate": 8.69873140906657e-06,
+      "loss": 0.8048,
+      "step": 3722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4372350899576056,
+      "learning_rate": 8.69393943693445e-06,
+      "loss": 0.8234,
+      "step": 3723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5466391896638483,
+      "learning_rate": 8.689147769898205e-06,
+      "loss": 0.856,
+      "step": 3724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5769944602986026,
+      "learning_rate": 8.684356409077177e-06,
+      "loss": 0.854,
+      "step": 3725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5608698292307723,
+      "learning_rate": 8.67956535559062e-06,
+      "loss": 0.8282,
+      "step": 3726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6257226419219133,
+      "learning_rate": 8.674774610557728e-06,
+      "loss": 0.7686,
+      "step": 3727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4473572784365238,
+      "learning_rate": 8.669984175097617e-06,
+      "loss": 0.828,
+      "step": 3728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5766672847205068,
+      "learning_rate": 8.665194050329337e-06,
+      "loss": 0.8642,
+      "step": 3729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4354587534372352,
+      "learning_rate": 8.66040423737186e-06,
+      "loss": 0.8939,
+      "step": 3730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4958470579399212,
+      "learning_rate": 8.655614737344087e-06,
+      "loss": 0.9224,
+      "step": 3731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5866593262083413,
+      "learning_rate": 8.650825551364844e-06,
+      "loss": 0.8107,
+      "step": 3732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4286557817274976,
+      "learning_rate": 8.646036680552889e-06,
+      "loss": 0.7898,
+      "step": 3733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.49711372800259,
+      "learning_rate": 8.641248126026897e-06,
+      "loss": 0.8839,
+      "step": 3734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.475445424476935,
+      "learning_rate": 8.63645988890548e-06,
+      "loss": 0.8853,
+      "step": 3735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6377280532848835,
+      "learning_rate": 8.63167197030717e-06,
+      "loss": 0.8435,
+      "step": 3736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5656859330352237,
+      "learning_rate": 8.626884371350421e-06,
+      "loss": 0.8265,
+      "step": 3737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3987400018119063,
+      "learning_rate": 8.62209709315362e-06,
+      "loss": 0.8263,
+      "step": 3738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.394969069383033,
+      "learning_rate": 8.617310136835078e-06,
+      "loss": 0.7883,
+      "step": 3739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7140430414561054,
+      "learning_rate": 8.612523503513024e-06,
+      "loss": 0.7338,
+      "step": 3740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4398814971817229,
+      "learning_rate": 8.60773719430562e-06,
+      "loss": 0.9348,
+      "step": 3741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4729099352410537,
+      "learning_rate": 8.602951210330942e-06,
+      "loss": 0.9038,
+      "step": 3742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3374320239659074,
+      "learning_rate": 8.598165552707002e-06,
+      "loss": 0.7976,
+      "step": 3743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4929592199136175,
+      "learning_rate": 8.593380222551729e-06,
+      "loss": 0.8883,
+      "step": 3744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4218434321127036,
+      "learning_rate": 8.588595220982975e-06,
+      "loss": 0.8428,
+      "step": 3745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.32917261179005,
+      "learning_rate": 8.583810549118524e-06,
+      "loss": 0.8247,
+      "step": 3746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4693009225936944,
+      "learning_rate": 8.579026208076064e-06,
+      "loss": 0.9103,
+      "step": 3747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3354877975442685,
+      "learning_rate": 8.574242198973223e-06,
+      "loss": 0.7976,
+      "step": 3748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.484015506579505,
+      "learning_rate": 8.569458522927547e-06,
+      "loss": 0.977,
+      "step": 3749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8765068607300504,
+      "learning_rate": 8.564675181056502e-06,
+      "loss": 0.3536,
+      "step": 3750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4653358840764616,
+      "learning_rate": 8.559892174477478e-06,
+      "loss": 0.8365,
+      "step": 3751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2953205048862486,
+      "learning_rate": 8.55510950430779e-06,
+      "loss": 0.8583,
+      "step": 3752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.422696732872559,
+      "learning_rate": 8.550327171664661e-06,
+      "loss": 0.8618,
+      "step": 3753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4063541989664001,
+      "learning_rate": 8.545545177665252e-06,
+      "loss": 0.768,
+      "step": 3754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4509323019911897,
+      "learning_rate": 8.54076352342663e-06,
+      "loss": 0.8402,
+      "step": 3755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4747712482252835,
+      "learning_rate": 8.535982210065797e-06,
+      "loss": 0.9576,
+      "step": 3756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4200240889415054,
+      "learning_rate": 8.531201238699665e-06,
+      "loss": 0.8261,
+      "step": 3757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3360012008582538,
+      "learning_rate": 8.52642061044507e-06,
+      "loss": 0.8501,
+      "step": 3758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5472446190309326,
+      "learning_rate": 8.52164032641877e-06,
+      "loss": 0.8644,
+      "step": 3759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.469815897840417,
+      "learning_rate": 8.516860387737436e-06,
+      "loss": 0.7903,
+      "step": 3760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6280436930700162,
+      "learning_rate": 8.512080795517662e-06,
+      "loss": 0.9012,
+      "step": 3761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2568879364982588,
+      "learning_rate": 8.50730155087596e-06,
+      "loss": 0.8578,
+      "step": 3762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.720820360703925,
+      "learning_rate": 8.502522654928767e-06,
+      "loss": 0.8634,
+      "step": 3763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.544101591965494,
+      "learning_rate": 8.49774410879243e-06,
+      "loss": 0.842,
+      "step": 3764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.575814262104861,
+      "learning_rate": 8.492965913583222e-06,
+      "loss": 0.798,
+      "step": 3765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4435210621339085,
+      "learning_rate": 8.488188070417323e-06,
+      "loss": 0.8032,
+      "step": 3766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.916004344681442,
+      "learning_rate": 8.483410580410843e-06,
+      "loss": 0.8702,
+      "step": 3767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5064238258529847,
+      "learning_rate": 8.478633444679801e-06,
+      "loss": 0.7975,
+      "step": 3768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4736029177617054,
+      "learning_rate": 8.47385666434014e-06,
+      "loss": 0.8767,
+      "step": 3769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5350354992839987,
+      "learning_rate": 8.469080240507711e-06,
+      "loss": 0.883,
+      "step": 3770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4472707936224234,
+      "learning_rate": 8.464304174298292e-06,
+      "loss": 0.8234,
+      "step": 3771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.386760113327573,
+      "learning_rate": 8.459528466827576e-06,
+      "loss": 0.884,
+      "step": 3772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.301808046486341,
+      "learning_rate": 8.454753119211158e-06,
+      "loss": 0.8764,
+      "step": 3773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6143737389027555,
+      "learning_rate": 8.449978132564565e-06,
+      "loss": 0.8372,
+      "step": 3774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.461787736285232,
+      "learning_rate": 8.445203508003236e-06,
+      "loss": 0.8934,
+      "step": 3775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4281086424614915,
+      "learning_rate": 8.44042924664252e-06,
+      "loss": 0.8477,
+      "step": 3776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4311179030356793,
+      "learning_rate": 8.43565534959769e-06,
+      "loss": 0.8676,
+      "step": 3777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3602770711709342,
+      "learning_rate": 8.430881817983931e-06,
+      "loss": 0.8781,
+      "step": 3778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4737895802559111,
+      "learning_rate": 8.426108652916329e-06,
+      "loss": 0.8753,
+      "step": 3779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4894731459733923,
+      "learning_rate": 8.421335855509903e-06,
+      "loss": 0.824,
+      "step": 3780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2854324749803256,
+      "learning_rate": 8.41656342687958e-06,
+      "loss": 0.8785,
+      "step": 3781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.339716958144086,
+      "learning_rate": 8.411791368140197e-06,
+      "loss": 0.895,
+      "step": 3782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5556465095252934,
+      "learning_rate": 8.407019680406508e-06,
+      "loss": 0.8271,
+      "step": 3783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4461301745933728,
+      "learning_rate": 8.402248364793182e-06,
+      "loss": 0.7326,
+      "step": 3784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4795888966719792,
+      "learning_rate": 8.3974774224148e-06,
+      "loss": 0.823,
+      "step": 3785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4046169627155651,
+      "learning_rate": 8.392706854385847e-06,
+      "loss": 0.9134,
+      "step": 3786
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.481080069935016,
+      "learning_rate": 8.387936661820733e-06,
+      "loss": 0.8109,
+      "step": 3787
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3783459734112478,
+      "learning_rate": 8.383166845833777e-06,
+      "loss": 0.9482,
+      "step": 3788
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5053364690658528,
+      "learning_rate": 8.378397407539209e-06,
+      "loss": 0.8239,
+      "step": 3789
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1832163660341024,
+      "learning_rate": 8.373628348051165e-06,
+      "loss": 0.8668,
+      "step": 3790
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.381206535582202,
+      "learning_rate": 8.368859668483703e-06,
+      "loss": 0.8552,
+      "step": 3791
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3847250528942607,
+      "learning_rate": 8.364091369950783e-06,
+      "loss": 0.9023,
+      "step": 3792
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6363542060886245,
+      "learning_rate": 8.359323453566283e-06,
+      "loss": 0.7639,
+      "step": 3793
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3242330944218341,
+      "learning_rate": 8.354555920443987e-06,
+      "loss": 0.7948,
+      "step": 3794
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.256241261646971,
+      "learning_rate": 8.349788771697593e-06,
+      "loss": 0.8451,
+      "step": 3795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5169178257742937,
+      "learning_rate": 8.345022008440704e-06,
+      "loss": 0.8129,
+      "step": 3796
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5172240919489481,
+      "learning_rate": 8.340255631786843e-06,
+      "loss": 0.7435,
+      "step": 3797
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3758617553573378,
+      "learning_rate": 8.335489642849428e-06,
+      "loss": 0.8201,
+      "step": 3798
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4432807999052983,
+      "learning_rate": 8.3307240427418e-06,
+      "loss": 0.8063,
+      "step": 3799
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6474266970139628,
+      "learning_rate": 8.325958832577201e-06,
+      "loss": 0.8054,
+      "step": 3800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6749639223464494,
+      "learning_rate": 8.321194013468786e-06,
+      "loss": 0.8441,
+      "step": 3801
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4026745257334892,
+      "learning_rate": 8.316429586529616e-06,
+      "loss": 0.8244,
+      "step": 3802
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2923184009681465,
+      "learning_rate": 8.311665552872662e-06,
+      "loss": 0.8592,
+      "step": 3803
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5419515788001992,
+      "learning_rate": 8.306901913610805e-06,
+      "loss": 0.8691,
+      "step": 3804
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4586798230035831,
+      "learning_rate": 8.30213866985683e-06,
+      "loss": 0.7941,
+      "step": 3805
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5242093380706372,
+      "learning_rate": 8.297375822723428e-06,
+      "loss": 0.8163,
+      "step": 3806
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3665116969556927,
+      "learning_rate": 8.292613373323203e-06,
+      "loss": 0.858,
+      "step": 3807
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6026898126638671,
+      "learning_rate": 8.287851322768666e-06,
+      "loss": 0.8313,
+      "step": 3808
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3705275254872518,
+      "learning_rate": 8.283089672172232e-06,
+      "loss": 0.8161,
+      "step": 3809
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4737242303126872,
+      "learning_rate": 8.278328422646222e-06,
+      "loss": 0.8358,
+      "step": 3810
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.397857335888745,
+      "learning_rate": 8.273567575302861e-06,
+      "loss": 0.8586,
+      "step": 3811
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.500025643437456,
+      "learning_rate": 8.268807131254288e-06,
+      "loss": 0.9056,
+      "step": 3812
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4417751545744042,
+      "learning_rate": 8.264047091612538e-06,
+      "loss": 0.8565,
+      "step": 3813
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4740418857953064,
+      "learning_rate": 8.259287457489564e-06,
+      "loss": 0.8447,
+      "step": 3814
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3443466784374998,
+      "learning_rate": 8.254528229997211e-06,
+      "loss": 0.8153,
+      "step": 3815
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.493733502689942,
+      "learning_rate": 8.249769410247239e-06,
+      "loss": 0.8315,
+      "step": 3816
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4889474464071153,
+      "learning_rate": 8.24501099935131e-06,
+      "loss": 0.8845,
+      "step": 3817
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6813774002563011,
+      "learning_rate": 8.240252998420983e-06,
+      "loss": 0.8625,
+      "step": 3818
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.468863988825738,
+      "learning_rate": 8.235495408567732e-06,
+      "loss": 0.7962,
+      "step": 3819
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6231877464369744,
+      "learning_rate": 8.230738230902928e-06,
+      "loss": 0.9005,
+      "step": 3820
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4096450955762019,
+      "learning_rate": 8.22598146653785e-06,
+      "loss": 0.7747,
+      "step": 3821
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6081042979083204,
+      "learning_rate": 8.221225116583677e-06,
+      "loss": 0.8344,
+      "step": 3822
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5459868375091343,
+      "learning_rate": 8.216469182151498e-06,
+      "loss": 0.8066,
+      "step": 3823
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3622485268888176,
+      "learning_rate": 8.211713664352293e-06,
+      "loss": 0.7873,
+      "step": 3824
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.49143883013621,
+      "learning_rate": 8.206958564296954e-06,
+      "loss": 0.9116,
+      "step": 3825
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5818685530373284,
+      "learning_rate": 8.202203883096272e-06,
+      "loss": 0.7721,
+      "step": 3826
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.923033014250726,
+      "learning_rate": 8.197449621860944e-06,
+      "loss": 0.3334,
+      "step": 3827
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5076111111776078,
+      "learning_rate": 8.192695781701562e-06,
+      "loss": 0.8287,
+      "step": 3828
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6755224196968321,
+      "learning_rate": 8.187942363728626e-06,
+      "loss": 0.8249,
+      "step": 3829
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6398240592332278,
+      "learning_rate": 8.183189369052538e-06,
+      "loss": 0.8649,
+      "step": 3830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2389481994167582,
+      "learning_rate": 8.17843679878359e-06,
+      "loss": 0.8193,
+      "step": 3831
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4016039837669882,
+      "learning_rate": 8.17368465403199e-06,
+      "loss": 0.9018,
+      "step": 3832
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3596032529242756,
+      "learning_rate": 8.168932935907833e-06,
+      "loss": 0.8173,
+      "step": 3833
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3099226278553266,
+      "learning_rate": 8.164181645521126e-06,
+      "loss": 0.8756,
+      "step": 3834
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5078606881955319,
+      "learning_rate": 8.15943078398177e-06,
+      "loss": 0.8613,
+      "step": 3835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6099830781875908,
+      "learning_rate": 8.15468035239957e-06,
+      "loss": 0.869,
+      "step": 3836
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5064763402148404,
+      "learning_rate": 8.14993035188422e-06,
+      "loss": 0.9439,
+      "step": 3837
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.347387602193213,
+      "learning_rate": 8.145180783545324e-06,
+      "loss": 0.8585,
+      "step": 3838
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4095366720398286,
+      "learning_rate": 8.140431648492382e-06,
+      "loss": 0.8703,
+      "step": 3839
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.227683166218185,
+      "learning_rate": 8.13568294783479e-06,
+      "loss": 0.8817,
+      "step": 3840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4318849893685222,
+      "learning_rate": 8.130934682681849e-06,
+      "loss": 0.8773,
+      "step": 3841
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4975102867392485,
+      "learning_rate": 8.126186854142752e-06,
+      "loss": 0.8536,
+      "step": 3842
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4178935633973675,
+      "learning_rate": 8.121439463326597e-06,
+      "loss": 0.8583,
+      "step": 3843
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.853345345505174,
+      "learning_rate": 8.116692511342365e-06,
+      "loss": 0.3621,
+      "step": 3844
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4643583078283737,
+      "learning_rate": 8.111945999298952e-06,
+      "loss": 0.8396,
+      "step": 3845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5141879192662218,
+      "learning_rate": 8.10719992830514e-06,
+      "loss": 0.8872,
+      "step": 3846
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.679919863308799,
+      "learning_rate": 8.102454299469615e-06,
+      "loss": 0.8201,
+      "step": 3847
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.442904645268295,
+      "learning_rate": 8.097709113900956e-06,
+      "loss": 0.8832,
+      "step": 3848
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3972234830774664,
+      "learning_rate": 8.092964372707636e-06,
+      "loss": 0.9463,
+      "step": 3849
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4631665234466493,
+      "learning_rate": 8.088220076998029e-06,
+      "loss": 0.8648,
+      "step": 3850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2591285943123254,
+      "learning_rate": 8.0834762278804e-06,
+      "loss": 0.8343,
+      "step": 3851
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4070522549833657,
+      "learning_rate": 8.078732826462917e-06,
+      "loss": 0.9144,
+      "step": 3852
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4661778160816774,
+      "learning_rate": 8.073989873853637e-06,
+      "loss": 0.8708,
+      "step": 3853
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5160750342810743,
+      "learning_rate": 8.069247371160514e-06,
+      "loss": 0.7942,
+      "step": 3854
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6119367166875718,
+      "learning_rate": 8.064505319491398e-06,
+      "loss": 0.8553,
+      "step": 3855
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.367028429450525,
+      "learning_rate": 8.059763719954033e-06,
+      "loss": 0.8285,
+      "step": 3856
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4701219521135274,
+      "learning_rate": 8.055022573656055e-06,
+      "loss": 0.8257,
+      "step": 3857
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3831092656115833,
+      "learning_rate": 8.050281881704997e-06,
+      "loss": 0.8723,
+      "step": 3858
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5745489540895334,
+      "learning_rate": 8.045541645208288e-06,
+      "loss": 0.8047,
+      "step": 3859
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.58253378951427,
+      "learning_rate": 8.040801865273243e-06,
+      "loss": 0.8515,
+      "step": 3860
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3477314444309891,
+      "learning_rate": 8.036062543007076e-06,
+      "loss": 0.8398,
+      "step": 3861
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.366991053500511,
+      "learning_rate": 8.0313236795169e-06,
+      "loss": 0.9,
+      "step": 3862
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.529866341519627,
+      "learning_rate": 8.026585275909704e-06,
+      "loss": 0.8634,
+      "step": 3863
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.357561736825964,
+      "learning_rate": 8.021847333292388e-06,
+      "loss": 0.7932,
+      "step": 3864
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3669840204861803,
+      "learning_rate": 8.017109852771729e-06,
+      "loss": 0.8846,
+      "step": 3865
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0147996328907256,
+      "learning_rate": 8.012372835454406e-06,
+      "loss": 0.9027,
+      "step": 3866
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2348257898542403,
+      "learning_rate": 8.007636282446986e-06,
+      "loss": 0.8429,
+      "step": 3867
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3546989375261893,
+      "learning_rate": 8.00290019485593e-06,
+      "loss": 0.8412,
+      "step": 3868
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.525857690282193,
+      "learning_rate": 7.998164573787592e-06,
+      "loss": 0.8739,
+      "step": 3869
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6262260629347693,
+      "learning_rate": 7.993429420348203e-06,
+      "loss": 0.8596,
+      "step": 3870
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.51772982282246,
+      "learning_rate": 7.988694735643902e-06,
+      "loss": 0.8345,
+      "step": 3871
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4294823658326008,
+      "learning_rate": 7.983960520780712e-06,
+      "loss": 0.8575,
+      "step": 3872
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4107911669823776,
+      "learning_rate": 7.979226776864544e-06,
+      "loss": 0.7989,
+      "step": 3873
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4753793891234486,
+      "learning_rate": 7.974493505001201e-06,
+      "loss": 0.8148,
+      "step": 3874
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.555913526812343,
+      "learning_rate": 7.969760706296385e-06,
+      "loss": 0.8568,
+      "step": 3875
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4406752774635811,
+      "learning_rate": 7.965028381855664e-06,
+      "loss": 0.8511,
+      "step": 3876
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.438675105992132,
+      "learning_rate": 7.960296532784515e-06,
+      "loss": 0.8406,
+      "step": 3877
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5014848760493564,
+      "learning_rate": 7.9555651601883e-06,
+      "loss": 0.7237,
+      "step": 3878
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8863640571563507,
+      "learning_rate": 7.950834265172266e-06,
+      "loss": 0.3502,
+      "step": 3879
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2787713951768571,
+      "learning_rate": 7.946103848841554e-06,
+      "loss": 0.8035,
+      "step": 3880
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4279112764970951,
+      "learning_rate": 7.94137391230119e-06,
+      "loss": 0.8433,
+      "step": 3881
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5247138667577969,
+      "learning_rate": 7.936644456656082e-06,
+      "loss": 0.8462,
+      "step": 3882
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8057065507493604,
+      "learning_rate": 7.931915483011035e-06,
+      "loss": 0.3251,
+      "step": 3883
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8444376501041313,
+      "learning_rate": 7.927186992470736e-06,
+      "loss": 0.3292,
+      "step": 3884
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5524881935496142,
+      "learning_rate": 7.922458986139761e-06,
+      "loss": 0.8193,
+      "step": 3885
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5702087453470714,
+      "learning_rate": 7.917731465122576e-06,
+      "loss": 0.7837,
+      "step": 3886
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5440342483626506,
+      "learning_rate": 7.913004430523526e-06,
+      "loss": 0.7835,
+      "step": 3887
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4085996726858885,
+      "learning_rate": 7.908277883446855e-06,
+      "loss": 0.8379,
+      "step": 3888
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2960327900950117,
+      "learning_rate": 7.903551824996675e-06,
+      "loss": 0.8207,
+      "step": 3889
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3646501135471174,
+      "learning_rate": 7.898826256276995e-06,
+      "loss": 0.7851,
+      "step": 3890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5404777440279533,
+      "learning_rate": 7.894101178391714e-06,
+      "loss": 0.8001,
+      "step": 3891
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4879105442914047,
+      "learning_rate": 7.889376592444605e-06,
+      "loss": 0.8203,
+      "step": 3892
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4130601742181295,
+      "learning_rate": 7.884652499539335e-06,
+      "loss": 0.8284,
+      "step": 3893
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5208452924547775,
+      "learning_rate": 7.879928900779457e-06,
+      "loss": 0.8893,
+      "step": 3894
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4347533744779224,
+      "learning_rate": 7.875205797268394e-06,
+      "loss": 0.8337,
+      "step": 3895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5382185355053928,
+      "learning_rate": 7.870483190109468e-06,
+      "loss": 0.8698,
+      "step": 3896
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3558823877819808,
+      "learning_rate": 7.865761080405882e-06,
+      "loss": 0.8562,
+      "step": 3897
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4726484662768524,
+      "learning_rate": 7.86103946926072e-06,
+      "loss": 0.8643,
+      "step": 3898
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4108247968582,
+      "learning_rate": 7.85631835777695e-06,
+      "loss": 0.8056,
+      "step": 3899
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5382425113486704,
+      "learning_rate": 7.851597747057426e-06,
+      "loss": 0.8896,
+      "step": 3900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.63086058108348,
+      "learning_rate": 7.846877638204887e-06,
+      "loss": 0.8275,
+      "step": 3901
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.553380579206785,
+      "learning_rate": 7.84215803232194e-06,
+      "loss": 0.8167,
+      "step": 3902
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5278258587184337,
+      "learning_rate": 7.837438930511093e-06,
+      "loss": 0.8325,
+      "step": 3903
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5868841567522767,
+      "learning_rate": 7.832720333874728e-06,
+      "loss": 0.8061,
+      "step": 3904
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4167538244565792,
+      "learning_rate": 7.828002243515107e-06,
+      "loss": 0.8431,
+      "step": 3905
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5835765438246021,
+      "learning_rate": 7.823284660534381e-06,
+      "loss": 0.8556,
+      "step": 3906
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5105778408107289,
+      "learning_rate": 7.818567586034578e-06,
+      "loss": 0.853,
+      "step": 3907
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3622683790283752,
+      "learning_rate": 7.8138510211176e-06,
+      "loss": 0.8336,
+      "step": 3908
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.477763111073078,
+      "learning_rate": 7.80913496688524e-06,
+      "loss": 0.828,
+      "step": 3909
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.560333593066661,
+      "learning_rate": 7.804419424439172e-06,
+      "loss": 0.8737,
+      "step": 3910
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.46646249403625,
+      "learning_rate": 7.799704394880943e-06,
+      "loss": 0.8469,
+      "step": 3911
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3335479389617615,
+      "learning_rate": 7.794989879311991e-06,
+      "loss": 0.8321,
+      "step": 3912
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4148086244382008,
+      "learning_rate": 7.790275878833619e-06,
+      "loss": 0.85,
+      "step": 3913
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5599994546728069,
+      "learning_rate": 7.785562394547023e-06,
+      "loss": 0.8901,
+      "step": 3914
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.416786162438713,
+      "learning_rate": 7.780849427553273e-06,
+      "loss": 0.8368,
+      "step": 3915
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2555798750528289,
+      "learning_rate": 7.776136978953316e-06,
+      "loss": 0.8113,
+      "step": 3916
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5308106775074777,
+      "learning_rate": 7.771425049847984e-06,
+      "loss": 0.8299,
+      "step": 3917
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.357819158748458,
+      "learning_rate": 7.766713641337983e-06,
+      "loss": 0.8391,
+      "step": 3918
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.593746994053119,
+      "learning_rate": 7.762002754523897e-06,
+      "loss": 0.853,
+      "step": 3919
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4946031353238858,
+      "learning_rate": 7.757292390506191e-06,
+      "loss": 0.8329,
+      "step": 3920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4636789713683203,
+      "learning_rate": 7.752582550385204e-06,
+      "loss": 0.8565,
+      "step": 3921
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.424063490420883,
+      "learning_rate": 7.747873235261157e-06,
+      "loss": 0.8976,
+      "step": 3922
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5471326909578724,
+      "learning_rate": 7.743164446234149e-06,
+      "loss": 0.7977,
+      "step": 3923
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3450696226406185,
+      "learning_rate": 7.73845618440415e-06,
+      "loss": 0.8387,
+      "step": 3924
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9266863073507474,
+      "learning_rate": 7.733748450871011e-06,
+      "loss": 0.8583,
+      "step": 3925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5139731654905673,
+      "learning_rate": 7.729041246734462e-06,
+      "loss": 0.8228,
+      "step": 3926
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6790758017143192,
+      "learning_rate": 7.724334573094101e-06,
+      "loss": 0.8303,
+      "step": 3927
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3130255687124601,
+      "learning_rate": 7.719628431049413e-06,
+      "loss": 0.9192,
+      "step": 3928
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6282751755683016,
+      "learning_rate": 7.71492282169975e-06,
+      "loss": 0.9017,
+      "step": 3929
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4981723034110144,
+      "learning_rate": 7.710217746144341e-06,
+      "loss": 0.801,
+      "step": 3930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.341580442189316,
+      "learning_rate": 7.705513205482297e-06,
+      "loss": 0.8048,
+      "step": 3931
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.487211009060395,
+      "learning_rate": 7.700809200812596e-06,
+      "loss": 0.7876,
+      "step": 3932
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5425565024024739,
+      "learning_rate": 7.696105733234099e-06,
+      "loss": 0.7763,
+      "step": 3933
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3098183559846637,
+      "learning_rate": 7.691402803845527e-06,
+      "loss": 0.7894,
+      "step": 3934
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3225369724392708,
+      "learning_rate": 7.68670041374549e-06,
+      "loss": 0.7881,
+      "step": 3935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4478912031098143,
+      "learning_rate": 7.681998564032467e-06,
+      "loss": 0.7877,
+      "step": 3936
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3632588166082595,
+      "learning_rate": 7.677297255804811e-06,
+      "loss": 0.8345,
+      "step": 3937
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9737144663632736,
+      "learning_rate": 7.672596490160747e-06,
+      "loss": 0.3358,
+      "step": 3938
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5469877245318275,
+      "learning_rate": 7.667896268198379e-06,
+      "loss": 0.8507,
+      "step": 3939
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.281300420176089,
+      "learning_rate": 7.66319659101567e-06,
+      "loss": 0.8628,
+      "step": 3940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8431429296120123,
+      "learning_rate": 7.658497459710472e-06,
+      "loss": 0.3126,
+      "step": 3941
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4301530895471217,
+      "learning_rate": 7.6537988753805e-06,
+      "loss": 0.9024,
+      "step": 3942
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4004654006466817,
+      "learning_rate": 7.649100839123344e-06,
+      "loss": 0.8568,
+      "step": 3943
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4403571999624494,
+      "learning_rate": 7.644403352036467e-06,
+      "loss": 0.8231,
+      "step": 3944
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3580275721095185,
+      "learning_rate": 7.639706415217201e-06,
+      "loss": 0.8773,
+      "step": 3945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5379641106208168,
+      "learning_rate": 7.635010029762755e-06,
+      "loss": 0.8644,
+      "step": 3946
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5661028612188268,
+      "learning_rate": 7.6303141967702e-06,
+      "loss": 0.8415,
+      "step": 3947
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5205939310782801,
+      "learning_rate": 7.625618917336482e-06,
+      "loss": 0.8753,
+      "step": 3948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7425247483928072,
+      "learning_rate": 7.620924192558423e-06,
+      "loss": 0.8838,
+      "step": 3949
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.44009310653146,
+      "learning_rate": 7.616230023532709e-06,
+      "loss": 0.8138,
+      "step": 3950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3992381201574247,
+      "learning_rate": 7.611536411355899e-06,
+      "loss": 0.9108,
+      "step": 3951
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4791180656960128,
+      "learning_rate": 7.606843357124426e-06,
+      "loss": 0.814,
+      "step": 3952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5128382842418315,
+      "learning_rate": 7.602150861934578e-06,
+      "loss": 0.8056,
+      "step": 3953
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8813251466069273,
+      "learning_rate": 7.597458926882529e-06,
+      "loss": 0.3489,
+      "step": 3954
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.493765949818518,
+      "learning_rate": 7.592767553064312e-06,
+      "loss": 0.7881,
+      "step": 3955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2748501101093348,
+      "learning_rate": 7.588076741575836e-06,
+      "loss": 0.8122,
+      "step": 3956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3753928739469214,
+      "learning_rate": 7.583386493512872e-06,
+      "loss": 0.7931,
+      "step": 3957
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4579932300888518,
+      "learning_rate": 7.5786968099710645e-06,
+      "loss": 0.8817,
+      "step": 3958
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2391749294932688,
+      "learning_rate": 7.574007692045928e-06,
+      "loss": 0.7476,
+      "step": 3959
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5266460303116258,
+      "learning_rate": 7.5693191408328325e-06,
+      "loss": 0.8863,
+      "step": 3960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4001846348188698,
+      "learning_rate": 7.564631157427027e-06,
+      "loss": 0.832,
+      "step": 3961
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3619921739411784,
+      "learning_rate": 7.559943742923626e-06,
+      "loss": 0.9073,
+      "step": 3962
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8096329160949032,
+      "learning_rate": 7.55525689841761e-06,
+      "loss": 0.3241,
+      "step": 3963
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4217503317475708,
+      "learning_rate": 7.550570625003826e-06,
+      "loss": 0.8777,
+      "step": 3964
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4624702641630836,
+      "learning_rate": 7.5458849237769915e-06,
+      "loss": 0.8274,
+      "step": 3965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4223706120499529,
+      "learning_rate": 7.54119979583168e-06,
+      "loss": 0.8615,
+      "step": 3966
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3349460372149062,
+      "learning_rate": 7.536515242262341e-06,
+      "loss": 0.8587,
+      "step": 3967
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5094578701967627,
+      "learning_rate": 7.531831264163286e-06,
+      "loss": 0.8722,
+      "step": 3968
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8521364852081045,
+      "learning_rate": 7.527147862628695e-06,
+      "loss": 0.3333,
+      "step": 3969
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3390398588903791,
+      "learning_rate": 7.522465038752609e-06,
+      "loss": 0.9447,
+      "step": 3970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5691904725772001,
+      "learning_rate": 7.517782793628938e-06,
+      "loss": 0.7773,
+      "step": 3971
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.437712327161783,
+      "learning_rate": 7.513101128351454e-06,
+      "loss": 0.9071,
+      "step": 3972
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4371124910153377,
+      "learning_rate": 7.508420044013793e-06,
+      "loss": 0.8583,
+      "step": 3973
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.434640774252635,
+      "learning_rate": 7.503739541709457e-06,
+      "loss": 0.8027,
+      "step": 3974
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5215308643823906,
+      "learning_rate": 7.499059622531812e-06,
+      "loss": 0.7998,
+      "step": 3975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4233905619619556,
+      "learning_rate": 7.49438028757409e-06,
+      "loss": 0.9129,
+      "step": 3976
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6152575305538228,
+      "learning_rate": 7.489701537929384e-06,
+      "loss": 0.8575,
+      "step": 3977
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4684522699866303,
+      "learning_rate": 7.48502337469065e-06,
+      "loss": 0.8257,
+      "step": 3978
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2857423173607951,
+      "learning_rate": 7.480345798950702e-06,
+      "loss": 0.8252,
+      "step": 3979
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6244608369775477,
+      "learning_rate": 7.475668811802228e-06,
+      "loss": 0.7919,
+      "step": 3980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.660382677232523,
+      "learning_rate": 7.4709924143377696e-06,
+      "loss": 0.7712,
+      "step": 3981
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4466859342528529,
+      "learning_rate": 7.4663166076497376e-06,
+      "loss": 0.8693,
+      "step": 3982
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.253153635260305,
+      "learning_rate": 7.4616413928303964e-06,
+      "loss": 0.8288,
+      "step": 3983
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.432949832347982,
+      "learning_rate": 7.456966770971882e-06,
+      "loss": 0.8611,
+      "step": 3984
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.654687288755272,
+      "learning_rate": 7.4522927431661805e-06,
+      "loss": 0.7739,
+      "step": 3985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4540036618777414,
+      "learning_rate": 7.447619310505147e-06,
+      "loss": 0.8273,
+      "step": 3986
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4622291844531243,
+      "learning_rate": 7.442946474080499e-06,
+      "loss": 0.7933,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3514932789744367,
+      "learning_rate": 7.438274234983809e-06,
+      "loss": 0.8039,
+      "step": 3988
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5514088136788373,
+      "learning_rate": 7.433602594306512e-06,
+      "loss": 0.887,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5322458868953754,
+      "learning_rate": 7.428931553139904e-06,
+      "loss": 0.8365,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.384575511972655,
+      "learning_rate": 7.4242611125751445e-06,
+      "loss": 0.8497,
+      "step": 3991
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3213976895042756,
+      "learning_rate": 7.419591273703245e-06,
+      "loss": 0.812,
+      "step": 3992
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5716519668692461,
+      "learning_rate": 7.414922037615079e-06,
+      "loss": 0.8533,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5355347459178688,
+      "learning_rate": 7.410253405401382e-06,
+      "loss": 0.8135,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.939914269984516,
+      "learning_rate": 7.405585378152749e-06,
+      "loss": 0.3249,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5646833019493598,
+      "learning_rate": 7.400917956959628e-06,
+      "loss": 0.8355,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4322248850050128,
+      "learning_rate": 7.396251142912337e-06,
+      "loss": 0.8362,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.235414863743788,
+      "learning_rate": 7.391584937101034e-06,
+      "loss": 0.8693,
+      "step": 3998
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.294460085551339,
+      "learning_rate": 7.386919340615749e-06,
+      "loss": 0.7786,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5092882034419526,
+      "learning_rate": 7.382254354546367e-06,
+      "loss": 0.845,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2960092659583327,
+      "learning_rate": 7.377589979982628e-06,
+      "loss": 0.905,
+      "step": 4001
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3597418855897803,
+      "learning_rate": 7.372926218014131e-06,
+      "loss": 0.8258,
+      "step": 4002
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4807505565862369,
+      "learning_rate": 7.368263069730331e-06,
+      "loss": 0.8712,
+      "step": 4003
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3377627081873373,
+      "learning_rate": 7.363600536220546e-06,
+      "loss": 0.8534,
+      "step": 4004
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4402833172413079,
+      "learning_rate": 7.358938618573932e-06,
+      "loss": 0.864,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2655980407527947,
+      "learning_rate": 7.3542773178795216e-06,
+      "loss": 0.8662,
+      "step": 4006
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6481355248148812,
+      "learning_rate": 7.349616635226191e-06,
+      "loss": 0.7618,
+      "step": 4007
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2883529776713392,
+      "learning_rate": 7.344956571702679e-06,
+      "loss": 0.8389,
+      "step": 4008
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4733728624458065,
+      "learning_rate": 7.340297128397577e-06,
+      "loss": 0.8307,
+      "step": 4009
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3365062163973156,
+      "learning_rate": 7.335638306399337e-06,
+      "loss": 0.8739,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3278711160455579,
+      "learning_rate": 7.330980106796247e-06,
+      "loss": 0.8348,
+      "step": 4011
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4355968194175286,
+      "learning_rate": 7.326322530676471e-06,
+      "loss": 0.9093,
+      "step": 4012
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8143306063703842,
+      "learning_rate": 7.3216655791280175e-06,
+      "loss": 0.3316,
+      "step": 4013
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.360521303542236,
+      "learning_rate": 7.317009253238752e-06,
+      "loss": 0.8919,
+      "step": 4014
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7935957872402526,
+      "learning_rate": 7.312353554096393e-06,
+      "loss": 0.3349,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4203158524655182,
+      "learning_rate": 7.30769848278851e-06,
+      "loss": 0.8608,
+      "step": 4016
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2636815900591305,
+      "learning_rate": 7.303044040402536e-06,
+      "loss": 0.8244,
+      "step": 4017
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.380026898889672,
+      "learning_rate": 7.298390228025737e-06,
+      "loss": 0.9059,
+      "step": 4018
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3105225852159377,
+      "learning_rate": 7.293737046745249e-06,
+      "loss": 0.8377,
+      "step": 4019
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5454422415586047,
+      "learning_rate": 7.289084497648058e-06,
+      "loss": 0.8803,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4179957166582344,
+      "learning_rate": 7.284432581820998e-06,
+      "loss": 0.8246,
+      "step": 4021
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2551732959068278,
+      "learning_rate": 7.279781300350758e-06,
+      "loss": 0.8108,
+      "step": 4022
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.48777467636917,
+      "learning_rate": 7.2751306543238805e-06,
+      "loss": 0.8819,
+      "step": 4023
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5658873470893908,
+      "learning_rate": 7.27048064482675e-06,
+      "loss": 0.8779,
+      "step": 4024
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4329109423795188,
+      "learning_rate": 7.265831272945612e-06,
+      "loss": 0.8751,
+      "step": 4025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3207738870079062,
+      "learning_rate": 7.261182539766563e-06,
+      "loss": 0.8604,
+      "step": 4026
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4049262717461986,
+      "learning_rate": 7.256534446375543e-06,
+      "loss": 0.8636,
+      "step": 4027
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4255930645742714,
+      "learning_rate": 7.25188699385835e-06,
+      "loss": 0.8658,
+      "step": 4028
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.485431652681221,
+      "learning_rate": 7.24724018330063e-06,
+      "loss": 0.8162,
+      "step": 4029
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2862814453139493,
+      "learning_rate": 7.242594015787883e-06,
+      "loss": 0.8189,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4160652178268391,
+      "learning_rate": 7.237948492405442e-06,
+      "loss": 0.7981,
+      "step": 4031
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.436552380672859,
+      "learning_rate": 7.23330361423851e-06,
+      "loss": 0.7666,
+      "step": 4032
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.498942457490684,
+      "learning_rate": 7.228659382372127e-06,
+      "loss": 0.9183,
+      "step": 4033
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4511328675275896,
+      "learning_rate": 7.224015797891191e-06,
+      "loss": 0.8301,
+      "step": 4034
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3815665760963811,
+      "learning_rate": 7.21937286188044e-06,
+      "loss": 0.7797,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6364873708904342,
+      "learning_rate": 7.214730575424469e-06,
+      "loss": 0.8685,
+      "step": 4036
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4193385302004342,
+      "learning_rate": 7.210088939607709e-06,
+      "loss": 0.7878,
+      "step": 4037
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5728229950466202,
+      "learning_rate": 7.205447955514451e-06,
+      "loss": 0.8154,
+      "step": 4038
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.418169133553306,
+      "learning_rate": 7.20080762422883e-06,
+      "loss": 0.8431,
+      "step": 4039
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4721707491113019,
+      "learning_rate": 7.196167946834826e-06,
+      "loss": 0.8158,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.370917171671462,
+      "learning_rate": 7.191528924416271e-06,
+      "loss": 0.87,
+      "step": 4041
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4760456896367928,
+      "learning_rate": 7.186890558056836e-06,
+      "loss": 0.8774,
+      "step": 4042
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2016534088205737,
+      "learning_rate": 7.18225284884005e-06,
+      "loss": 0.8819,
+      "step": 4043
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6612606122766522,
+      "learning_rate": 7.177615797849278e-06,
+      "loss": 0.8183,
+      "step": 4044
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4284650854433951,
+      "learning_rate": 7.172979406167735e-06,
+      "loss": 0.7784,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3178468598112467,
+      "learning_rate": 7.168343674878487e-06,
+      "loss": 0.778,
+      "step": 4046
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3576375844798225,
+      "learning_rate": 7.163708605064437e-06,
+      "loss": 0.8331,
+      "step": 4047
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3543054566112847,
+      "learning_rate": 7.15907419780834e-06,
+      "loss": 0.8668,
+      "step": 4048
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3403824885308662,
+      "learning_rate": 7.154440454192793e-06,
+      "loss": 0.9069,
+      "step": 4049
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8448849237433186,
+      "learning_rate": 7.149807375300239e-06,
+      "loss": 0.3267,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3465822224689834,
+      "learning_rate": 7.145174962212969e-06,
+      "loss": 0.8171,
+      "step": 4051
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6501537412576142,
+      "learning_rate": 7.140543216013109e-06,
+      "loss": 0.949,
+      "step": 4052
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9552282505973152,
+      "learning_rate": 7.135912137782639e-06,
+      "loss": 0.3141,
+      "step": 4053
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8572351006782111,
+      "learning_rate": 7.1312817286033784e-06,
+      "loss": 0.3467,
+      "step": 4054
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4218916250137275,
+      "learning_rate": 7.126651989556994e-06,
+      "loss": 0.8139,
+      "step": 4055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4618065691959776,
+      "learning_rate": 7.122022921724993e-06,
+      "loss": 0.8941,
+      "step": 4056
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.528894395835897,
+      "learning_rate": 7.117394526188719e-06,
+      "loss": 0.925,
+      "step": 4057
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4229688720414047,
+      "learning_rate": 7.112766804029373e-06,
+      "loss": 0.7987,
+      "step": 4058
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.603676423154499,
+      "learning_rate": 7.1081397563279885e-06,
+      "loss": 0.835,
+      "step": 4059
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4642568009968235,
+      "learning_rate": 7.103513384165446e-06,
+      "loss": 0.8365,
+      "step": 4060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5779818337718972,
+      "learning_rate": 7.0988876886224635e-06,
+      "loss": 0.7563,
+      "step": 4061
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6644244292945192,
+      "learning_rate": 7.094262670779611e-06,
+      "loss": 0.74,
+      "step": 4062
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9530389123482829,
+      "learning_rate": 7.0896383317172845e-06,
+      "loss": 0.3291,
+      "step": 4063
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4439081115502193,
+      "learning_rate": 7.085014672515733e-06,
+      "loss": 0.7958,
+      "step": 4064
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.48300673892835,
+      "learning_rate": 7.080391694255045e-06,
+      "loss": 0.8805,
+      "step": 4065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4814951208714935,
+      "learning_rate": 7.075769398015147e-06,
+      "loss": 0.9266,
+      "step": 4066
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5077635659544213,
+      "learning_rate": 7.071147784875809e-06,
+      "loss": 0.906,
+      "step": 4067
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.525242014299514,
+      "learning_rate": 7.066526855916647e-06,
+      "loss": 0.8863,
+      "step": 4068
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.595501849380148,
+      "learning_rate": 7.061906612217096e-06,
+      "loss": 0.8448,
+      "step": 4069
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5060984369799253,
+      "learning_rate": 7.057287054856455e-06,
+      "loss": 0.8336,
+      "step": 4070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3920515704772587,
+      "learning_rate": 7.052668184913851e-06,
+      "loss": 0.817,
+      "step": 4071
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.413615194197871,
+      "learning_rate": 7.048050003468252e-06,
+      "loss": 0.823,
+      "step": 4072
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.540746389736846,
+      "learning_rate": 7.043432511598467e-06,
+      "loss": 0.9207,
+      "step": 4073
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2531081865614424,
+      "learning_rate": 7.038815710383141e-06,
+      "loss": 0.8804,
+      "step": 4074
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5550505044204708,
+      "learning_rate": 7.034199600900765e-06,
+      "loss": 0.939,
+      "step": 4075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4761354330873884,
+      "learning_rate": 7.029584184229653e-06,
+      "loss": 0.8422,
+      "step": 4076
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7945827105032043,
+      "learning_rate": 7.024969461447973e-06,
+      "loss": 0.3307,
+      "step": 4077
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4962021549382294,
+      "learning_rate": 7.0203554336337206e-06,
+      "loss": 0.8578,
+      "step": 4078
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.387360633797198,
+      "learning_rate": 7.015742101864737e-06,
+      "loss": 0.8725,
+      "step": 4079
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3947455938686388,
+      "learning_rate": 7.011129467218696e-06,
+      "loss": 0.8815,
+      "step": 4080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.300912888341301,
+      "learning_rate": 7.006517530773113e-06,
+      "loss": 0.774,
+      "step": 4081
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2872764792959677,
+      "learning_rate": 7.001906293605329e-06,
+      "loss": 0.8299,
+      "step": 4082
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4972984957501676,
+      "learning_rate": 6.9972957567925325e-06,
+      "loss": 0.8413,
+      "step": 4083
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.434924057387539,
+      "learning_rate": 6.992685921411748e-06,
+      "loss": 0.8089,
+      "step": 4084
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3539369272860728,
+      "learning_rate": 6.98807678853983e-06,
+      "loss": 0.8128,
+      "step": 4085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3362311924354842,
+      "learning_rate": 6.983468359253476e-06,
+      "loss": 0.8295,
+      "step": 4086
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4891369608012521,
+      "learning_rate": 6.978860634629213e-06,
+      "loss": 0.867,
+      "step": 4087
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3185065858384972,
+      "learning_rate": 6.9742536157434114e-06,
+      "loss": 0.8652,
+      "step": 4088
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5313936477545826,
+      "learning_rate": 6.969647303672262e-06,
+      "loss": 0.8704,
+      "step": 4089
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7551649233931472,
+      "learning_rate": 6.965041699491804e-06,
+      "loss": 0.3241,
+      "step": 4090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8106437550522526,
+      "learning_rate": 6.960436804277908e-06,
+      "loss": 0.3507,
+      "step": 4091
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4268161427547736,
+      "learning_rate": 6.9558326191062775e-06,
+      "loss": 0.8902,
+      "step": 4092
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5353929359035485,
+      "learning_rate": 6.95122914505245e-06,
+      "loss": 0.7933,
+      "step": 4093
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.730421795434039,
+      "learning_rate": 6.9466263831918015e-06,
+      "loss": 0.8664,
+      "step": 4094
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7210413116591243,
+      "learning_rate": 6.942024334599531e-06,
+      "loss": 0.8472,
+      "step": 4095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.466800227459811,
+      "learning_rate": 6.93742300035068e-06,
+      "loss": 0.7794,
+      "step": 4096
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3937999293018455,
+      "learning_rate": 6.932822381520121e-06,
+      "loss": 0.7822,
+      "step": 4097
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.558196612961541,
+      "learning_rate": 6.92822247918256e-06,
+      "loss": 0.8065,
+      "step": 4098
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4331822129452696,
+      "learning_rate": 6.923623294412534e-06,
+      "loss": 0.8168,
+      "step": 4099
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4815568397903058,
+      "learning_rate": 6.919024828284413e-06,
+      "loss": 0.8234,
+      "step": 4100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5729062013456891,
+      "learning_rate": 6.914427081872401e-06,
+      "loss": 0.8991,
+      "step": 4101
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8936306186370127,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.3264,
+      "step": 4102
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5144505805052484,
+      "learning_rate": 6.90523375249266e-06,
+      "loss": 0.8432,
+      "step": 4103
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4348186371479237,
+      "learning_rate": 6.900638171672497e-06,
+      "loss": 0.8396,
+      "step": 4104
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.595521510328429,
+      "learning_rate": 6.896043314863568e-06,
+      "loss": 0.9411,
+      "step": 4105
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3618433523264837,
+      "learning_rate": 6.891449183139227e-06,
+      "loss": 0.7777,
+      "step": 4106
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.324848945010024,
+      "learning_rate": 6.88685577757267e-06,
+      "loss": 0.8853,
+      "step": 4107
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7554898133009502,
+      "learning_rate": 6.88226309923691e-06,
+      "loss": 0.8283,
+      "step": 4108
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3576531233824523,
+      "learning_rate": 6.877671149204801e-06,
+      "loss": 0.891,
+      "step": 4109
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.47677119990268,
+      "learning_rate": 6.873079928549024e-06,
+      "loss": 0.8275,
+      "step": 4110
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8488148487446681,
+      "learning_rate": 6.8684894383420865e-06,
+      "loss": 0.3182,
+      "step": 4111
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4415400651002197,
+      "learning_rate": 6.8638996796563275e-06,
+      "loss": 0.8724,
+      "step": 4112
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4167742486003656,
+      "learning_rate": 6.859310653563917e-06,
+      "loss": 0.8249,
+      "step": 4113
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4793070611232852,
+      "learning_rate": 6.854722361136851e-06,
+      "loss": 0.8581,
+      "step": 4114
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2894146698282294,
+      "learning_rate": 6.850134803446955e-06,
+      "loss": 0.8169,
+      "step": 4115
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4877944347848857,
+      "learning_rate": 6.845547981565881e-06,
+      "loss": 0.8318,
+      "step": 4116
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8799505493476157,
+      "learning_rate": 6.8409618965651125e-06,
+      "loss": 0.3599,
+      "step": 4117
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.467202015709495,
+      "learning_rate": 6.83637654951596e-06,
+      "loss": 0.7922,
+      "step": 4118
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4664981184763466,
+      "learning_rate": 6.83179194148956e-06,
+      "loss": 0.9283,
+      "step": 4119
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.399346106397274,
+      "learning_rate": 6.827208073556884e-06,
+      "loss": 0.8478,
+      "step": 4120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4130029719896253,
+      "learning_rate": 6.8226249467887115e-06,
+      "loss": 0.7969,
+      "step": 4121
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5657724943944973,
+      "learning_rate": 6.81804256225567e-06,
+      "loss": 0.8389,
+      "step": 4122
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4375012653023915,
+      "learning_rate": 6.813460921028202e-06,
+      "loss": 0.9325,
+      "step": 4123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4011686935473764,
+      "learning_rate": 6.808880024176581e-06,
+      "loss": 0.8124,
+      "step": 4124
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3735590246834153,
+      "learning_rate": 6.804299872770903e-06,
+      "loss": 0.891,
+      "step": 4125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4998532014778108,
+      "learning_rate": 6.799720467881099e-06,
+      "loss": 0.7542,
+      "step": 4126
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5551702095063675,
+      "learning_rate": 6.795141810576906e-06,
+      "loss": 0.8109,
+      "step": 4127
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5565704224573844,
+      "learning_rate": 6.790563901927907e-06,
+      "loss": 0.7405,
+      "step": 4128
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3945263173312614,
+      "learning_rate": 6.7859867430035006e-06,
+      "loss": 0.8928,
+      "step": 4129
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.593621447464884,
+      "learning_rate": 6.781410334872911e-06,
+      "loss": 0.7722,
+      "step": 4130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.238250226615948,
+      "learning_rate": 6.776834678605186e-06,
+      "loss": 0.7682,
+      "step": 4131
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3975788469241808,
+      "learning_rate": 6.772259775269203e-06,
+      "loss": 0.7892,
+      "step": 4132
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.376254726393351,
+      "learning_rate": 6.767685625933662e-06,
+      "loss": 0.8457,
+      "step": 4133
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.383936034280788,
+      "learning_rate": 6.763112231667076e-06,
+      "loss": 0.9207,
+      "step": 4134
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4866025911878868,
+      "learning_rate": 6.758539593537796e-06,
+      "loss": 0.7821,
+      "step": 4135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4194263370146019,
+      "learning_rate": 6.7539677126139894e-06,
+      "loss": 0.7913,
+      "step": 4136
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.433294867976207,
+      "learning_rate": 6.7493965899636486e-06,
+      "loss": 0.7948,
+      "step": 4137
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3901480971400966,
+      "learning_rate": 6.744826226654587e-06,
+      "loss": 0.8548,
+      "step": 4138
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5320051056295763,
+      "learning_rate": 6.740256623754448e-06,
+      "loss": 0.8665,
+      "step": 4139
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4649678273104199,
+      "learning_rate": 6.735687782330683e-06,
+      "loss": 0.8402,
+      "step": 4140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3889835104566026,
+      "learning_rate": 6.731119703450577e-06,
+      "loss": 0.7785,
+      "step": 4141
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4424155650620423,
+      "learning_rate": 6.7265523881812335e-06,
+      "loss": 0.8174,
+      "step": 4142
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5899231150451114,
+      "learning_rate": 6.7219858375895785e-06,
+      "loss": 0.7566,
+      "step": 4143
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3927449959694322,
+      "learning_rate": 6.717420052742358e-06,
+      "loss": 0.7821,
+      "step": 4144
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.480147895668604,
+      "learning_rate": 6.712855034706141e-06,
+      "loss": 0.8191,
+      "step": 4145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5231233181130175,
+      "learning_rate": 6.70829078454732e-06,
+      "loss": 0.8902,
+      "step": 4146
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6688998413141465,
+      "learning_rate": 6.703727303332094e-06,
+      "loss": 0.8547,
+      "step": 4147
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.487571873901898,
+      "learning_rate": 6.699164592126499e-06,
+      "loss": 0.8383,
+      "step": 4148
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4079169612227123,
+      "learning_rate": 6.694602651996385e-06,
+      "loss": 0.7667,
+      "step": 4149
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6025870959320427,
+      "learning_rate": 6.690041484007419e-06,
+      "loss": 0.8794,
+      "step": 4150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4147850042856225,
+      "learning_rate": 6.685481089225092e-06,
+      "loss": 0.7816,
+      "step": 4151
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2862360875889831,
+      "learning_rate": 6.680921468714718e-06,
+      "loss": 0.799,
+      "step": 4152
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4366328381256777,
+      "learning_rate": 6.676362623541415e-06,
+      "loss": 0.9017,
+      "step": 4153
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.500383038838065,
+      "learning_rate": 6.671804554770135e-06,
+      "loss": 0.8304,
+      "step": 4154
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2651402941954724,
+      "learning_rate": 6.6672472634656414e-06,
+      "loss": 0.762,
+      "step": 4155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5537863988192027,
+      "learning_rate": 6.66269075069252e-06,
+      "loss": 0.8654,
+      "step": 4156
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.367136033315608,
+      "learning_rate": 6.6581350175151715e-06,
+      "loss": 0.8236,
+      "step": 4157
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5510267242373068,
+      "learning_rate": 6.653580064997817e-06,
+      "loss": 0.8204,
+      "step": 4158
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2609935778753811,
+      "learning_rate": 6.649025894204495e-06,
+      "loss": 0.8253,
+      "step": 4159
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3191193304695545,
+      "learning_rate": 6.644472506199053e-06,
+      "loss": 0.8938,
+      "step": 4160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4343246693704728,
+      "learning_rate": 6.639919902045169e-06,
+      "loss": 0.837,
+      "step": 4161
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4235153705074768,
+      "learning_rate": 6.6353680828063306e-06,
+      "loss": 0.9105,
+      "step": 4162
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.538675941051293,
+      "learning_rate": 6.630817049545844e-06,
+      "loss": 0.8009,
+      "step": 4163
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.573990402982657,
+      "learning_rate": 6.626266803326831e-06,
+      "loss": 0.8009,
+      "step": 4164
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.349235348966594,
+      "learning_rate": 6.621717345212231e-06,
+      "loss": 0.8098,
+      "step": 4165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.375875300762048,
+      "learning_rate": 6.617168676264791e-06,
+      "loss": 0.8209,
+      "step": 4166
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4786015684629978,
+      "learning_rate": 6.612620797547087e-06,
+      "loss": 0.8196,
+      "step": 4167
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4590395089728088,
+      "learning_rate": 6.608073710121501e-06,
+      "loss": 0.9172,
+      "step": 4168
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.439167559855049,
+      "learning_rate": 6.603527415050237e-06,
+      "loss": 0.7508,
+      "step": 4169
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.317155404819832,
+      "learning_rate": 6.598981913395306e-06,
+      "loss": 0.8445,
+      "step": 4170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2301618420364648,
+      "learning_rate": 6.594437206218539e-06,
+      "loss": 0.7968,
+      "step": 4171
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4949245721240965,
+      "learning_rate": 6.589893294581579e-06,
+      "loss": 0.8821,
+      "step": 4172
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3523744723985995,
+      "learning_rate": 6.585350179545884e-06,
+      "loss": 0.7763,
+      "step": 4173
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.417083630151581,
+      "learning_rate": 6.580807862172731e-06,
+      "loss": 0.8731,
+      "step": 4174
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4346886047869207,
+      "learning_rate": 6.576266343523199e-06,
+      "loss": 0.9036,
+      "step": 4175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5340577610696056,
+      "learning_rate": 6.571725624658189e-06,
+      "loss": 0.8244,
+      "step": 4176
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5280645747612418,
+      "learning_rate": 6.567185706638417e-06,
+      "loss": 0.7882,
+      "step": 4177
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2985215900920564,
+      "learning_rate": 6.562646590524406e-06,
+      "loss": 0.7443,
+      "step": 4178
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.475359074328717,
+      "learning_rate": 6.558108277376496e-06,
+      "loss": 0.9245,
+      "step": 4179
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9721485135513757,
+      "learning_rate": 6.553570768254831e-06,
+      "loss": 0.3412,
+      "step": 4180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4147354501858538,
+      "learning_rate": 6.549034064219379e-06,
+      "loss": 0.8838,
+      "step": 4181
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6845249225195797,
+      "learning_rate": 6.5444981663299135e-06,
+      "loss": 0.8188,
+      "step": 4182
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7170103675974178,
+      "learning_rate": 6.539963075646021e-06,
+      "loss": 0.8265,
+      "step": 4183
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3756278572860934,
+      "learning_rate": 6.535428793227102e-06,
+      "loss": 0.8473,
+      "step": 4184
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5865118743651678,
+      "learning_rate": 6.530895320132358e-06,
+      "loss": 0.8326,
+      "step": 4185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2986020579822197,
+      "learning_rate": 6.526362657420813e-06,
+      "loss": 0.8223,
+      "step": 4186
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.752031447872313,
+      "learning_rate": 6.521830806151297e-06,
+      "loss": 0.79,
+      "step": 4187
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3716378762489827,
+      "learning_rate": 6.517299767382451e-06,
+      "loss": 0.8159,
+      "step": 4188
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6077610112352816,
+      "learning_rate": 6.512769542172727e-06,
+      "loss": 0.8324,
+      "step": 4189
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3794370209217903,
+      "learning_rate": 6.508240131580386e-06,
+      "loss": 0.8094,
+      "step": 4190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.27304634569455,
+      "learning_rate": 6.503711536663499e-06,
+      "loss": 0.9299,
+      "step": 4191
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7915954960445426,
+      "learning_rate": 6.499183758479944e-06,
+      "loss": 0.3219,
+      "step": 4192
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5799377264683774,
+      "learning_rate": 6.494656798087412e-06,
+      "loss": 0.8785,
+      "step": 4193
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3712298999472066,
+      "learning_rate": 6.490130656543401e-06,
+      "loss": 0.8239,
+      "step": 4194
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.852395606632373,
+      "learning_rate": 6.485605334905216e-06,
+      "loss": 0.7694,
+      "step": 4195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.75395148841272,
+      "learning_rate": 6.481080834229978e-06,
+      "loss": 0.8826,
+      "step": 4196
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4051221536796374,
+      "learning_rate": 6.47655715557461e-06,
+      "loss": 0.8934,
+      "step": 4197
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5128813235811764,
+      "learning_rate": 6.472034299995837e-06,
+      "loss": 0.8364,
+      "step": 4198
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.509019990797452,
+      "learning_rate": 6.467512268550204e-06,
+      "loss": 0.8056,
+      "step": 4199
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6635778531772962,
+      "learning_rate": 6.462991062294057e-06,
+      "loss": 0.9425,
+      "step": 4200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4492217157698564,
+      "learning_rate": 6.45847068228355e-06,
+      "loss": 0.8003,
+      "step": 4201
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4306574303795523,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 0.8952,
+      "step": 4202
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4208096689503036,
+      "learning_rate": 6.449432405223107e-06,
+      "loss": 0.8559,
+      "step": 4203
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4006992522726627,
+      "learning_rate": 6.444914510284519e-06,
+      "loss": 0.8232,
+      "step": 4204
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4007788722480299,
+      "learning_rate": 6.44039744581425e-06,
+      "loss": 0.8447,
+      "step": 4205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.492992637029976,
+      "learning_rate": 6.435881212867494e-06,
+      "loss": 0.8254,
+      "step": 4206
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5937910518782015,
+      "learning_rate": 6.431365812499242e-06,
+      "loss": 0.7467,
+      "step": 4207
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2812051394505937,
+      "learning_rate": 6.426851245764289e-06,
+      "loss": 0.8438,
+      "step": 4208
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.477217115384169,
+      "learning_rate": 6.422337513717244e-06,
+      "loss": 0.7668,
+      "step": 4209
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.470877767480189,
+      "learning_rate": 6.417824617412515e-06,
+      "loss": 0.8416,
+      "step": 4210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.396810879585123,
+      "learning_rate": 6.413312557904307e-06,
+      "loss": 0.7626,
+      "step": 4211
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4567238314620887,
+      "learning_rate": 6.408801336246645e-06,
+      "loss": 0.7778,
+      "step": 4212
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3125779328880536,
+      "learning_rate": 6.404290953493347e-06,
+      "loss": 0.8648,
+      "step": 4213
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.325498950246133,
+      "learning_rate": 6.399781410698042e-06,
+      "loss": 0.8384,
+      "step": 4214
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3568458439531355,
+      "learning_rate": 6.395272708914156e-06,
+      "loss": 0.8031,
+      "step": 4215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.410992858968764,
+      "learning_rate": 6.390764849194926e-06,
+      "loss": 0.853,
+      "step": 4216
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3981473318003852,
+      "learning_rate": 6.38625783259339e-06,
+      "loss": 0.8115,
+      "step": 4217
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3431667445451263,
+      "learning_rate": 6.3817516601623805e-06,
+      "loss": 0.8796,
+      "step": 4218
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3691504814130562,
+      "learning_rate": 6.377246332954544e-06,
+      "loss": 0.7353,
+      "step": 4219
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4072387702724756,
+      "learning_rate": 6.372741852022324e-06,
+      "loss": 0.8817,
+      "step": 4220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5120617701596062,
+      "learning_rate": 6.36823821841797e-06,
+      "loss": 0.8299,
+      "step": 4221
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6236411004168696,
+      "learning_rate": 6.36373543319353e-06,
+      "loss": 0.8628,
+      "step": 4222
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.245746242713602,
+      "learning_rate": 6.359233497400858e-06,
+      "loss": 0.8393,
+      "step": 4223
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7154947155416183,
+      "learning_rate": 6.3547324120916e-06,
+      "loss": 0.8352,
+      "step": 4224
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5713588184641445,
+      "learning_rate": 6.350232178317214e-06,
+      "loss": 0.7593,
+      "step": 4225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3004270836098588,
+      "learning_rate": 6.345732797128954e-06,
+      "loss": 0.838,
+      "step": 4226
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4283736555043585,
+      "learning_rate": 6.341234269577878e-06,
+      "loss": 0.8661,
+      "step": 4227
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.506266843090225,
+      "learning_rate": 6.336736596714842e-06,
+      "loss": 0.8497,
+      "step": 4228
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.513288925704829,
+      "learning_rate": 6.3322397795905e-06,
+      "loss": 0.7767,
+      "step": 4229
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4930537085112032,
+      "learning_rate": 6.327743819255313e-06,
+      "loss": 0.7917,
+      "step": 4230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.297647998423009,
+      "learning_rate": 6.323248716759534e-06,
+      "loss": 0.8647,
+      "step": 4231
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.332295261575304,
+      "learning_rate": 6.318754473153221e-06,
+      "loss": 0.8023,
+      "step": 4232
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4274856772169817,
+      "learning_rate": 6.314261089486231e-06,
+      "loss": 0.8488,
+      "step": 4233
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4828094376621843,
+      "learning_rate": 6.309768566808217e-06,
+      "loss": 0.847,
+      "step": 4234
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5252582978647513,
+      "learning_rate": 6.305276906168633e-06,
+      "loss": 0.8819,
+      "step": 4235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4434784516188566,
+      "learning_rate": 6.300786108616732e-06,
+      "loss": 0.7488,
+      "step": 4236
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3117261603522674,
+      "learning_rate": 6.296296175201565e-06,
+      "loss": 0.8374,
+      "step": 4237
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3891473317015244,
+      "learning_rate": 6.291807106971981e-06,
+      "loss": 0.7711,
+      "step": 4238
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.588283988633166,
+      "learning_rate": 6.287318904976625e-06,
+      "loss": 0.8558,
+      "step": 4239
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.554095954832167,
+      "learning_rate": 6.282831570263943e-06,
+      "loss": 0.8746,
+      "step": 4240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.629557306780163,
+      "learning_rate": 6.278345103882175e-06,
+      "loss": 0.7383,
+      "step": 4241
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3915828937817551,
+      "learning_rate": 6.273859506879365e-06,
+      "loss": 0.8946,
+      "step": 4242
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4434973621095728,
+      "learning_rate": 6.269374780303345e-06,
+      "loss": 0.8357,
+      "step": 4243
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.632965291782185,
+      "learning_rate": 6.264890925201745e-06,
+      "loss": 0.8347,
+      "step": 4244
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3674917536474693,
+      "learning_rate": 6.260407942621998e-06,
+      "loss": 0.7915,
+      "step": 4245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3848768036280812,
+      "learning_rate": 6.255925833611327e-06,
+      "loss": 0.943,
+      "step": 4246
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4433174927329897,
+      "learning_rate": 6.251444599216756e-06,
+      "loss": 0.7456,
+      "step": 4247
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9415590312652048,
+      "learning_rate": 6.246964240485099e-06,
+      "loss": 0.3529,
+      "step": 4248
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6894160492638008,
+      "learning_rate": 6.242484758462972e-06,
+      "loss": 0.8447,
+      "step": 4249
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3035035193386735,
+      "learning_rate": 6.238006154196779e-06,
+      "loss": 0.8487,
+      "step": 4250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.591451505627097,
+      "learning_rate": 6.23352842873272e-06,
+      "loss": 0.9135,
+      "step": 4251
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3382616280488455,
+      "learning_rate": 6.229051583116796e-06,
+      "loss": 0.7808,
+      "step": 4252
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.581072434342835,
+      "learning_rate": 6.2245756183947995e-06,
+      "loss": 0.7799,
+      "step": 4253
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4652129089585146,
+      "learning_rate": 6.220100535612313e-06,
+      "loss": 0.8485,
+      "step": 4254
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5604234062928044,
+      "learning_rate": 6.215626335814723e-06,
+      "loss": 0.8722,
+      "step": 4255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4007858286664163,
+      "learning_rate": 6.2111530200471935e-06,
+      "loss": 0.8063,
+      "step": 4256
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4459710991902535,
+      "learning_rate": 6.206680589354696e-06,
+      "loss": 0.8536,
+      "step": 4257
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4525943230605125,
+      "learning_rate": 6.202209044781991e-06,
+      "loss": 0.8338,
+      "step": 4258
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.50384393538145,
+      "learning_rate": 6.197738387373631e-06,
+      "loss": 0.8181,
+      "step": 4259
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5001895003770507,
+      "learning_rate": 6.193268618173962e-06,
+      "loss": 0.8414,
+      "step": 4260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5014986989616157,
+      "learning_rate": 6.188799738227124e-06,
+      "loss": 0.9541,
+      "step": 4261
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.468910105835831,
+      "learning_rate": 6.18433174857705e-06,
+      "loss": 0.8768,
+      "step": 4262
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5090954540467645,
+      "learning_rate": 6.179864650267457e-06,
+      "loss": 0.8795,
+      "step": 4263
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3939941070331088,
+      "learning_rate": 6.17539844434186e-06,
+      "loss": 0.9215,
+      "step": 4264
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5681736029675502,
+      "learning_rate": 6.170933131843569e-06,
+      "loss": 0.8103,
+      "step": 4265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4430785170433695,
+      "learning_rate": 6.16646871381568e-06,
+      "loss": 0.8547,
+      "step": 4266
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5176629671169857,
+      "learning_rate": 6.162005191301082e-06,
+      "loss": 0.8358,
+      "step": 4267
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7127915117738377,
+      "learning_rate": 6.1575425653424555e-06,
+      "loss": 0.8852,
+      "step": 4268
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9307209549893665,
+      "learning_rate": 6.153080836982266e-06,
+      "loss": 0.3375,
+      "step": 4269
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4111039148823523,
+      "learning_rate": 6.148620007262775e-06,
+      "loss": 0.9023,
+      "step": 4270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3554616970516378,
+      "learning_rate": 6.144160077226035e-06,
+      "loss": 0.8947,
+      "step": 4271
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6455928876714077,
+      "learning_rate": 6.139701047913885e-06,
+      "loss": 0.8517,
+      "step": 4272
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6237397855139317,
+      "learning_rate": 6.135242920367955e-06,
+      "loss": 0.8149,
+      "step": 4273
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.449193928549702,
+      "learning_rate": 6.130785695629664e-06,
+      "loss": 0.7908,
+      "step": 4274
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4797848056442116,
+      "learning_rate": 6.126329374740223e-06,
+      "loss": 0.7772,
+      "step": 4275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6584399293501135,
+      "learning_rate": 6.121873958740623e-06,
+      "loss": 0.8944,
+      "step": 4276
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4196461589578617,
+      "learning_rate": 6.117419448671651e-06,
+      "loss": 0.8511,
+      "step": 4277
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5236881282248365,
+      "learning_rate": 6.112965845573884e-06,
+      "loss": 0.9161,
+      "step": 4278
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4004184315792443,
+      "learning_rate": 6.108513150487682e-06,
+      "loss": 0.8552,
+      "step": 4279
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.397850362231667,
+      "learning_rate": 6.104061364453196e-06,
+      "loss": 0.8726,
+      "step": 4280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.505198715986615,
+      "learning_rate": 6.099610488510368e-06,
+      "loss": 0.8826,
+      "step": 4281
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4501053117509792,
+      "learning_rate": 6.095160523698913e-06,
+      "loss": 0.8147,
+      "step": 4282
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4644763880072755,
+      "learning_rate": 6.09071147105835e-06,
+      "loss": 0.8353,
+      "step": 4283
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5582500577166882,
+      "learning_rate": 6.086263331627976e-06,
+      "loss": 0.8452,
+      "step": 4284
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4587571970699578,
+      "learning_rate": 6.081816106446878e-06,
+      "loss": 0.839,
+      "step": 4285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2969889842061815,
+      "learning_rate": 6.077369796553928e-06,
+      "loss": 0.839,
+      "step": 4286
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3326597996093192,
+      "learning_rate": 6.072924402987785e-06,
+      "loss": 0.8325,
+      "step": 4287
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.386576818536693,
+      "learning_rate": 6.068479926786894e-06,
+      "loss": 0.8935,
+      "step": 4288
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3943188329209395,
+      "learning_rate": 6.064036368989482e-06,
+      "loss": 0.8013,
+      "step": 4289
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2575423049671666,
+      "learning_rate": 6.0595937306335666e-06,
+      "loss": 0.8299,
+      "step": 4290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5127120666610412,
+      "learning_rate": 6.055152012756946e-06,
+      "loss": 0.883,
+      "step": 4291
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4461158324152272,
+      "learning_rate": 6.050711216397212e-06,
+      "loss": 0.8011,
+      "step": 4292
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.255689961796309,
+      "learning_rate": 6.04627134259173e-06,
+      "loss": 0.7625,
+      "step": 4293
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6353414790572611,
+      "learning_rate": 6.0418323923776565e-06,
+      "loss": 0.8339,
+      "step": 4294
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4185120885409621,
+      "learning_rate": 6.0373943667919285e-06,
+      "loss": 0.807,
+      "step": 4295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3513936918603378,
+      "learning_rate": 6.032957266871274e-06,
+      "loss": 0.8903,
+      "step": 4296
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4366055004193379,
+      "learning_rate": 6.028521093652195e-06,
+      "loss": 0.8149,
+      "step": 4297
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3778814960274006,
+      "learning_rate": 6.0240858481709854e-06,
+      "loss": 0.8322,
+      "step": 4298
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2785585617095774,
+      "learning_rate": 6.0196515314637164e-06,
+      "loss": 0.867,
+      "step": 4299
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4186514342042469,
+      "learning_rate": 6.0152181445662485e-06,
+      "loss": 0.8652,
+      "step": 4300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4491098560951776,
+      "learning_rate": 6.010785688514216e-06,
+      "loss": 0.8721,
+      "step": 4301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4138503969357206,
+      "learning_rate": 6.006354164343047e-06,
+      "loss": 0.8754,
+      "step": 4302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4673056310662789,
+      "learning_rate": 6.0019235730879414e-06,
+      "loss": 0.8147,
+      "step": 4303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3413836663355185,
+      "learning_rate": 5.997493915783887e-06,
+      "loss": 0.7428,
+      "step": 4304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6571155751456612,
+      "learning_rate": 5.993065193465653e-06,
+      "loss": 0.8255,
+      "step": 4305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.386258314819785,
+      "learning_rate": 5.9886374071677875e-06,
+      "loss": 0.8126,
+      "step": 4306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6832059565184125,
+      "learning_rate": 5.984210557924628e-06,
+      "loss": 0.8096,
+      "step": 4307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2509020605244172,
+      "learning_rate": 5.979784646770279e-06,
+      "loss": 0.7397,
+      "step": 4308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5688376594080107,
+      "learning_rate": 5.975359674738637e-06,
+      "loss": 0.8038,
+      "step": 4309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6431356622347084,
+      "learning_rate": 5.970935642863375e-06,
+      "loss": 0.8522,
+      "step": 4310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5348489237759009,
+      "learning_rate": 5.966512552177949e-06,
+      "loss": 0.7724,
+      "step": 4311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3747303776927502,
+      "learning_rate": 5.962090403715592e-06,
+      "loss": 0.8389,
+      "step": 4312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5817183974070743,
+      "learning_rate": 5.9576691985093235e-06,
+      "loss": 0.8636,
+      "step": 4313
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4498172344233735,
+      "learning_rate": 5.953248937591929e-06,
+      "loss": 0.8532,
+      "step": 4314
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2929147142891533,
+      "learning_rate": 5.948829621995984e-06,
+      "loss": 0.8615,
+      "step": 4315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3669170548118452,
+      "learning_rate": 5.944411252753846e-06,
+      "loss": 0.8234,
+      "step": 4316
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3873774434908999,
+      "learning_rate": 5.939993830897641e-06,
+      "loss": 0.7439,
+      "step": 4317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.566217294554743,
+      "learning_rate": 5.935577357459282e-06,
+      "loss": 0.7592,
+      "step": 4318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.505575332001506,
+      "learning_rate": 5.931161833470458e-06,
+      "loss": 0.7697,
+      "step": 4319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3834612834388402,
+      "learning_rate": 5.926747259962638e-06,
+      "loss": 0.8533,
+      "step": 4320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4389951176560545,
+      "learning_rate": 5.92233363796706e-06,
+      "loss": 0.8403,
+      "step": 4321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4469450635506795,
+      "learning_rate": 5.9179209685147525e-06,
+      "loss": 0.7903,
+      "step": 4322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4416974001075,
+      "learning_rate": 5.913509252636511e-06,
+      "loss": 0.8761,
+      "step": 4323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4398427252266477,
+      "learning_rate": 5.909098491362916e-06,
+      "loss": 0.8267,
+      "step": 4324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4409529760251871,
+      "learning_rate": 5.90468868572432e-06,
+      "loss": 0.8025,
+      "step": 4325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6198685784698694,
+      "learning_rate": 5.900279836750859e-06,
+      "loss": 0.8161,
+      "step": 4326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3505587930220215,
+      "learning_rate": 5.895871945472434e-06,
+      "loss": 0.8447,
+      "step": 4327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4245812408853182,
+      "learning_rate": 5.891465012918731e-06,
+      "loss": 0.806,
+      "step": 4328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4211720966680543,
+      "learning_rate": 5.887059040119209e-06,
+      "loss": 0.8067,
+      "step": 4329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4558989699342442,
+      "learning_rate": 5.882654028103108e-06,
+      "loss": 0.8538,
+      "step": 4330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4152251654979937,
+      "learning_rate": 5.878249977899433e-06,
+      "loss": 0.8639,
+      "step": 4331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5119627119939212,
+      "learning_rate": 5.873846890536976e-06,
+      "loss": 0.8323,
+      "step": 4332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.382540626254464,
+      "learning_rate": 5.8694447670442985e-06,
+      "loss": 0.8252,
+      "step": 4333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.236847545414012,
+      "learning_rate": 5.865043608449732e-06,
+      "loss": 0.8008,
+      "step": 4334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3006345706296107,
+      "learning_rate": 5.86064341578139e-06,
+      "loss": 0.8261,
+      "step": 4335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5705140628555114,
+      "learning_rate": 5.85624419006716e-06,
+      "loss": 0.8304,
+      "step": 4336
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9560450903125399,
+      "learning_rate": 5.851845932334698e-06,
+      "loss": 0.3257,
+      "step": 4337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4345550257139037,
+      "learning_rate": 5.84744864361144e-06,
+      "loss": 0.8811,
+      "step": 4338
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4646735219591631,
+      "learning_rate": 5.843052324924596e-06,
+      "loss": 0.8505,
+      "step": 4339
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5632767784954522,
+      "learning_rate": 5.83865697730114e-06,
+      "loss": 0.826,
+      "step": 4340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3724701926561396,
+      "learning_rate": 5.834262601767828e-06,
+      "loss": 0.8357,
+      "step": 4341
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3894978788180528,
+      "learning_rate": 5.829869199351188e-06,
+      "loss": 0.7285,
+      "step": 4342
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5056022042232862,
+      "learning_rate": 5.825476771077518e-06,
+      "loss": 0.8732,
+      "step": 4343
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3594049581975511,
+      "learning_rate": 5.821085317972891e-06,
+      "loss": 0.8071,
+      "step": 4344
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.426420079127766,
+      "learning_rate": 5.816694841063149e-06,
+      "loss": 0.837,
+      "step": 4345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6557987246122676,
+      "learning_rate": 5.812305341373914e-06,
+      "loss": 0.7949,
+      "step": 4346
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5684910477866751,
+      "learning_rate": 5.807916819930563e-06,
+      "loss": 0.8455,
+      "step": 4347
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5031814990112005,
+      "learning_rate": 5.803529277758261e-06,
+      "loss": 0.8887,
+      "step": 4348
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.253253187367845,
+      "learning_rate": 5.799142715881938e-06,
+      "loss": 0.7544,
+      "step": 4349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5725081132401277,
+      "learning_rate": 5.794757135326294e-06,
+      "loss": 0.8732,
+      "step": 4350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5591659015589374,
+      "learning_rate": 5.790372537115802e-06,
+      "loss": 0.8555,
+      "step": 4351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6061185452276394,
+      "learning_rate": 5.785988922274711e-06,
+      "loss": 0.8433,
+      "step": 4352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5060769398929812,
+      "learning_rate": 5.7816062918270225e-06,
+      "loss": 0.866,
+      "step": 4353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.361922133818583,
+      "learning_rate": 5.7772246467965246e-06,
+      "loss": 0.8871,
+      "step": 4354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5091309243459323,
+      "learning_rate": 5.772843988206769e-06,
+      "loss": 0.9164,
+      "step": 4355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4106831015506072,
+      "learning_rate": 5.76846431708108e-06,
+      "loss": 0.7768,
+      "step": 4356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8663618325193143,
+      "learning_rate": 5.7640856344425465e-06,
+      "loss": 0.8162,
+      "step": 4357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2245658626461553,
+      "learning_rate": 5.759707941314032e-06,
+      "loss": 0.8264,
+      "step": 4358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.061817501495855,
+      "learning_rate": 5.7553312387181685e-06,
+      "loss": 0.8357,
+      "step": 4359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5553629165236145,
+      "learning_rate": 5.750955527677347e-06,
+      "loss": 0.8562,
+      "step": 4360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.53810718340051,
+      "learning_rate": 5.746580809213736e-06,
+      "loss": 0.8346,
+      "step": 4361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6304397408610432,
+      "learning_rate": 5.742207084349274e-06,
+      "loss": 0.9015,
+      "step": 4362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6222530206970949,
+      "learning_rate": 5.737834354105659e-06,
+      "loss": 0.8514,
+      "step": 4363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5184247350745008,
+      "learning_rate": 5.733462619504364e-06,
+      "loss": 0.844,
+      "step": 4364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5542523803151984,
+      "learning_rate": 5.729091881566631e-06,
+      "loss": 0.9175,
+      "step": 4365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3978160544236615,
+      "learning_rate": 5.724722141313456e-06,
+      "loss": 0.8409,
+      "step": 4366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4025366771302141,
+      "learning_rate": 5.720353399765615e-06,
+      "loss": 0.8381,
+      "step": 4367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5183792000593377,
+      "learning_rate": 5.715985657943644e-06,
+      "loss": 0.8227,
+      "step": 4368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.748280145836894,
+      "learning_rate": 5.711618916867851e-06,
+      "loss": 0.3147,
+      "step": 4369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.372096064858729,
+      "learning_rate": 5.707253177558308e-06,
+      "loss": 0.8043,
+      "step": 4370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5708921172932517,
+      "learning_rate": 5.702888441034853e-06,
+      "loss": 0.8723,
+      "step": 4371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4808292616788827,
+      "learning_rate": 5.698524708317082e-06,
+      "loss": 0.8946,
+      "step": 4372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3427792018419864,
+      "learning_rate": 5.694161980424368e-06,
+      "loss": 0.7711,
+      "step": 4373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3974828964949968,
+      "learning_rate": 5.689800258375844e-06,
+      "loss": 0.8335,
+      "step": 4374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2134117843502503,
+      "learning_rate": 5.685439543190409e-06,
+      "loss": 0.7655,
+      "step": 4375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4692356867718188,
+      "learning_rate": 5.681079835886727e-06,
+      "loss": 0.7673,
+      "step": 4376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5862665316608566,
+      "learning_rate": 5.676721137483226e-06,
+      "loss": 0.8585,
+      "step": 4377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4394047313614347,
+      "learning_rate": 5.672363448998102e-06,
+      "loss": 0.7739,
+      "step": 4378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3529931855338269,
+      "learning_rate": 5.668006771449302e-06,
+      "loss": 0.8324,
+      "step": 4379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4497687775364378,
+      "learning_rate": 5.6636511058545525e-06,
+      "loss": 0.9044,
+      "step": 4380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5213890853835024,
+      "learning_rate": 5.6592964532313365e-06,
+      "loss": 0.8179,
+      "step": 4381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.478262068248342,
+      "learning_rate": 5.654942814596902e-06,
+      "loss": 0.8414,
+      "step": 4382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3837117460910946,
+      "learning_rate": 5.650590190968263e-06,
+      "loss": 0.7734,
+      "step": 4383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6458157661537622,
+      "learning_rate": 5.646238583362184e-06,
+      "loss": 0.8523,
+      "step": 4384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6966816860102454,
+      "learning_rate": 5.641887992795205e-06,
+      "loss": 0.8899,
+      "step": 4385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4374113858939788,
+      "learning_rate": 5.6375384202836254e-06,
+      "loss": 0.8221,
+      "step": 4386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.428425929936069,
+      "learning_rate": 5.633189866843507e-06,
+      "loss": 0.7949,
+      "step": 4387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.484082934858807,
+      "learning_rate": 5.628842333490674e-06,
+      "loss": 0.7901,
+      "step": 4388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4991614511184272,
+      "learning_rate": 5.624495821240703e-06,
+      "loss": 0.7751,
+      "step": 4389
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3917369830178221,
+      "learning_rate": 5.620150331108943e-06,
+      "loss": 0.8411,
+      "step": 4390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.513431232766497,
+      "learning_rate": 5.615805864110504e-06,
+      "loss": 0.827,
+      "step": 4391
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5020482488305424,
+      "learning_rate": 5.611462421260251e-06,
+      "loss": 0.8172,
+      "step": 4392
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4921293946382255,
+      "learning_rate": 5.607120003572817e-06,
+      "loss": 0.8029,
+      "step": 4393
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.618901813781922,
+      "learning_rate": 5.602778612062585e-06,
+      "loss": 0.8448,
+      "step": 4394
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5388424529668256,
+      "learning_rate": 5.598438247743706e-06,
+      "loss": 0.833,
+      "step": 4395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3046444318995354,
+      "learning_rate": 5.594098911630091e-06,
+      "loss": 0.8201,
+      "step": 4396
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5203985709097454,
+      "learning_rate": 5.58976060473541e-06,
+      "loss": 0.8727,
+      "step": 4397
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4174538707438733,
+      "learning_rate": 5.585423328073094e-06,
+      "loss": 0.8481,
+      "step": 4398
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.424667041266223,
+      "learning_rate": 5.581087082656325e-06,
+      "loss": 0.8137,
+      "step": 4399
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.455661799348346,
+      "learning_rate": 5.576751869498054e-06,
+      "loss": 0.8111,
+      "step": 4400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6100683512657807,
+      "learning_rate": 5.572417689610987e-06,
+      "loss": 0.8298,
+      "step": 4401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3152238958137867,
+      "learning_rate": 5.5680845440075885e-06,
+      "loss": 0.7922,
+      "step": 4402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.386333283054433,
+      "learning_rate": 5.563752433700082e-06,
+      "loss": 0.8462,
+      "step": 4403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5842246815233396,
+      "learning_rate": 5.559421359700452e-06,
+      "loss": 0.887,
+      "step": 4404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9231606374610966,
+      "learning_rate": 5.555091323020432e-06,
+      "loss": 0.3161,
+      "step": 4405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4900159932034664,
+      "learning_rate": 5.550762324671521e-06,
+      "loss": 0.8244,
+      "step": 4406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4883715040269252,
+      "learning_rate": 5.546434365664974e-06,
+      "loss": 0.8429,
+      "step": 4407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.582801746603013,
+      "learning_rate": 5.5421074470118045e-06,
+      "loss": 0.855,
+      "step": 4408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8678966127423097,
+      "learning_rate": 5.5377815697227776e-06,
+      "loss": 0.3312,
+      "step": 4409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5638858657317964,
+      "learning_rate": 5.5334567348084265e-06,
+      "loss": 0.8955,
+      "step": 4410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6200262519006703,
+      "learning_rate": 5.529132943279021e-06,
+      "loss": 0.7759,
+      "step": 4411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3310397784712833,
+      "learning_rate": 5.5248101961446065e-06,
+      "loss": 0.7596,
+      "step": 4412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4530825909968506,
+      "learning_rate": 5.520488494414975e-06,
+      "loss": 0.7975,
+      "step": 4413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.503052018672958,
+      "learning_rate": 5.516167839099679e-06,
+      "loss": 0.856,
+      "step": 4414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3347766047130232,
+      "learning_rate": 5.511848231208022e-06,
+      "loss": 0.9411,
+      "step": 4415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.359225803442016,
+      "learning_rate": 5.507529671749064e-06,
+      "loss": 0.8778,
+      "step": 4416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7242945035870008,
+      "learning_rate": 5.503212161731628e-06,
+      "loss": 0.8246,
+      "step": 4417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5992071114038786,
+      "learning_rate": 5.498895702164274e-06,
+      "loss": 0.8331,
+      "step": 4418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5451353940733725,
+      "learning_rate": 5.494580294055333e-06,
+      "loss": 0.8012,
+      "step": 4419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.488193809764088,
+      "learning_rate": 5.490265938412883e-06,
+      "loss": 0.856,
+      "step": 4420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.531953056023603,
+      "learning_rate": 5.4859526362447605e-06,
+      "loss": 0.898,
+      "step": 4421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.466591457454563,
+      "learning_rate": 5.481640388558551e-06,
+      "loss": 0.843,
+      "step": 4422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.382143136105328,
+      "learning_rate": 5.4773291963616006e-06,
+      "loss": 0.8348,
+      "step": 4423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4595352349595985,
+      "learning_rate": 5.473019060660997e-06,
+      "loss": 0.8394,
+      "step": 4424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4918264362885365,
+      "learning_rate": 5.468709982463591e-06,
+      "loss": 0.7896,
+      "step": 4425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4049767372515047,
+      "learning_rate": 5.464401962775986e-06,
+      "loss": 0.7831,
+      "step": 4426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.429010411997647,
+      "learning_rate": 5.460095002604533e-06,
+      "loss": 0.8122,
+      "step": 4427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4145809462691723,
+      "learning_rate": 5.45578910295534e-06,
+      "loss": 0.8534,
+      "step": 4428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7319571292074991,
+      "learning_rate": 5.451484264834271e-06,
+      "loss": 0.823,
+      "step": 4429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.625213531072949,
+      "learning_rate": 5.447180489246924e-06,
+      "loss": 0.8202,
+      "step": 4430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4906172504263588,
+      "learning_rate": 5.442877777198669e-06,
+      "loss": 0.7706,
+      "step": 4431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4920874088980793,
+      "learning_rate": 5.43857612969462e-06,
+      "loss": 0.8523,
+      "step": 4432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3836969581862542,
+      "learning_rate": 5.434275547739641e-06,
+      "loss": 0.8519,
+      "step": 4433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5148861635181443,
+      "learning_rate": 5.42997603233835e-06,
+      "loss": 0.8813,
+      "step": 4434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3965685959074967,
+      "learning_rate": 5.425677584495112e-06,
+      "loss": 0.7946,
+      "step": 4435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9166251981221127,
+      "learning_rate": 5.421380205214051e-06,
+      "loss": 0.8042,
+      "step": 4436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4464172117189402,
+      "learning_rate": 5.417083895499024e-06,
+      "loss": 0.7559,
+      "step": 4437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3574169001042158,
+      "learning_rate": 5.412788656353658e-06,
+      "loss": 0.9286,
+      "step": 4438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4076995107799493,
+      "learning_rate": 5.408494488781317e-06,
+      "loss": 0.8437,
+      "step": 4439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4228475632149882,
+      "learning_rate": 5.404201393785123e-06,
+      "loss": 0.916,
+      "step": 4440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.549351414989158,
+      "learning_rate": 5.3999093723679395e-06,
+      "loss": 0.826,
+      "step": 4441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5388365085844244,
+      "learning_rate": 5.39561842553239e-06,
+      "loss": 0.8221,
+      "step": 4442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.341995071227043,
+      "learning_rate": 5.391328554280829e-06,
+      "loss": 0.7786,
+      "step": 4443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5561369663863633,
+      "learning_rate": 5.387039759615378e-06,
+      "loss": 0.8233,
+      "step": 4444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5667210426705647,
+      "learning_rate": 5.382752042537899e-06,
+      "loss": 0.7424,
+      "step": 4445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5075810988656286,
+      "learning_rate": 5.378465404050003e-06,
+      "loss": 0.827,
+      "step": 4446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4562725247024544,
+      "learning_rate": 5.374179845153048e-06,
+      "loss": 0.8379,
+      "step": 4447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8038457546512405,
+      "learning_rate": 5.369895366848144e-06,
+      "loss": 0.3358,
+      "step": 4448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4842366248182006,
+      "learning_rate": 5.365611970136145e-06,
+      "loss": 0.8117,
+      "step": 4449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6709230935057313,
+      "learning_rate": 5.361329656017649e-06,
+      "loss": 0.794,
+      "step": 4450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5387975923449324,
+      "learning_rate": 5.357048425493007e-06,
+      "loss": 0.7274,
+      "step": 4451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3531270790579684,
+      "learning_rate": 5.352768279562315e-06,
+      "loss": 0.8071,
+      "step": 4452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4566089258404527,
+      "learning_rate": 5.348489219225417e-06,
+      "loss": 0.7737,
+      "step": 4453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4342289038609468,
+      "learning_rate": 5.3442112454819e-06,
+      "loss": 0.8189,
+      "step": 4454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5053840844518434,
+      "learning_rate": 5.339934359331104e-06,
+      "loss": 0.8856,
+      "step": 4455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5080987746244463,
+      "learning_rate": 5.335658561772101e-06,
+      "loss": 0.8089,
+      "step": 4456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.839324060773937,
+      "learning_rate": 5.331383853803724e-06,
+      "loss": 0.8145,
+      "step": 4457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.498780826730005,
+      "learning_rate": 5.327110236424544e-06,
+      "loss": 0.8238,
+      "step": 4458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4611618244608902,
+      "learning_rate": 5.3228377106328775e-06,
+      "loss": 0.8497,
+      "step": 4459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4759978710926638,
+      "learning_rate": 5.3185662774267875e-06,
+      "loss": 0.8957,
+      "step": 4460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.576081780336865,
+      "learning_rate": 5.314295937804082e-06,
+      "loss": 0.8981,
+      "step": 4461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6212599301475261,
+      "learning_rate": 5.310026692762316e-06,
+      "loss": 0.8939,
+      "step": 4462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4763640019291224,
+      "learning_rate": 5.305758543298778e-06,
+      "loss": 0.8265,
+      "step": 4463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5588311561910986,
+      "learning_rate": 5.3014914904105105e-06,
+      "loss": 0.8727,
+      "step": 4464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3335454785684293,
+      "learning_rate": 5.297225535094302e-06,
+      "loss": 0.8926,
+      "step": 4465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6354198103973723,
+      "learning_rate": 5.292960678346674e-06,
+      "loss": 0.8286,
+      "step": 4466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.453256815141083,
+      "learning_rate": 5.288696921163902e-06,
+      "loss": 0.8306,
+      "step": 4467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.307165861408919,
+      "learning_rate": 5.284434264542002e-06,
+      "loss": 0.766,
+      "step": 4468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5938569359105743,
+      "learning_rate": 5.280172709476723e-06,
+      "loss": 0.7814,
+      "step": 4469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9559373413405672,
+      "learning_rate": 5.275912256963571e-06,
+      "loss": 0.3479,
+      "step": 4470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9000019690000334,
+      "learning_rate": 5.2716529079977856e-06,
+      "loss": 0.3258,
+      "step": 4471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3764502750206267,
+      "learning_rate": 5.267394663574351e-06,
+      "loss": 0.8096,
+      "step": 4472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3787878690807869,
+      "learning_rate": 5.2631375246879955e-06,
+      "loss": 0.8471,
+      "step": 4473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2535421676238763,
+      "learning_rate": 5.2588814923331854e-06,
+      "loss": 0.7559,
+      "step": 4474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5403722857322744,
+      "learning_rate": 5.254626567504135e-06,
+      "loss": 0.8154,
+      "step": 4475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7775978512926978,
+      "learning_rate": 5.250372751194788e-06,
+      "loss": 0.342,
+      "step": 4476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4310148401493037,
+      "learning_rate": 5.246120044398839e-06,
+      "loss": 0.8904,
+      "step": 4477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5003548031150493,
+      "learning_rate": 5.241868448109722e-06,
+      "loss": 0.8289,
+      "step": 4478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5076239003440834,
+      "learning_rate": 5.237617963320608e-06,
+      "loss": 0.8126,
+      "step": 4479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6390164130179972,
+      "learning_rate": 5.233368591024412e-06,
+      "loss": 0.8271,
+      "step": 4480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.533919308673761,
+      "learning_rate": 5.229120332213794e-06,
+      "loss": 0.7336,
+      "step": 4481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3311413158121979,
+      "learning_rate": 5.224873187881136e-06,
+      "loss": 0.7806,
+      "step": 4482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2932170164255992,
+      "learning_rate": 5.220627159018578e-06,
+      "loss": 0.8019,
+      "step": 4483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.531277760307563,
+      "learning_rate": 5.216382246617993e-06,
+      "loss": 0.8908,
+      "step": 4484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4348066692261237,
+      "learning_rate": 5.212138451670989e-06,
+      "loss": 0.7672,
+      "step": 4485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.366045679400478,
+      "learning_rate": 5.2078957751689206e-06,
+      "loss": 0.8047,
+      "step": 4486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.34046114732214,
+      "learning_rate": 5.20365421810288e-06,
+      "loss": 0.8945,
+      "step": 4487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4321217411563052,
+      "learning_rate": 5.199413781463689e-06,
+      "loss": 0.831,
+      "step": 4488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4477171495991878,
+      "learning_rate": 5.195174466241917e-06,
+      "loss": 0.8124,
+      "step": 4489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3967339806132228,
+      "learning_rate": 5.190936273427868e-06,
+      "loss": 0.7988,
+      "step": 4490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.805574481492177,
+      "learning_rate": 5.186699204011585e-06,
+      "loss": 0.3301,
+      "step": 4491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4266117525111504,
+      "learning_rate": 5.1824632589828465e-06,
+      "loss": 0.7418,
+      "step": 4492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4342952734683052,
+      "learning_rate": 5.17822843933117e-06,
+      "loss": 0.7672,
+      "step": 4493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3902627052428154,
+      "learning_rate": 5.173994746045816e-06,
+      "loss": 0.8372,
+      "step": 4494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2373050781526378,
+      "learning_rate": 5.169762180115765e-06,
+      "loss": 0.8381,
+      "step": 4495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3199151290423123,
+      "learning_rate": 5.16553074252975e-06,
+      "loss": 0.8824,
+      "step": 4496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5321191928800468,
+      "learning_rate": 5.161300434276237e-06,
+      "loss": 0.8242,
+      "step": 4497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3667582512440828,
+      "learning_rate": 5.157071256343422e-06,
+      "loss": 0.8292,
+      "step": 4498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.581681238709597,
+      "learning_rate": 5.152843209719246e-06,
+      "loss": 0.9121,
+      "step": 4499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4639550005205448,
+      "learning_rate": 5.148616295391382e-06,
+      "loss": 0.8378,
+      "step": 4500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3227331554859414,
+      "learning_rate": 5.1443905143472305e-06,
+      "loss": 0.7974,
+      "step": 4501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.505669022686916,
+      "learning_rate": 5.14016586757394e-06,
+      "loss": 0.8293,
+      "step": 4502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4469881152042756,
+      "learning_rate": 5.135942356058385e-06,
+      "loss": 0.7211,
+      "step": 4503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.477252731389027,
+      "learning_rate": 5.131719980787182e-06,
+      "loss": 0.7417,
+      "step": 4504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.384440988626623,
+      "learning_rate": 5.127498742746675e-06,
+      "loss": 0.9194,
+      "step": 4505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6896485232215501,
+      "learning_rate": 5.123278642922952e-06,
+      "loss": 0.9312,
+      "step": 4506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2915419307944953,
+      "learning_rate": 5.119059682301819e-06,
+      "loss": 0.8508,
+      "step": 4507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4212061270310063,
+      "learning_rate": 5.114841861868831e-06,
+      "loss": 0.8126,
+      "step": 4508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.468073100491747,
+      "learning_rate": 5.1106251826092716e-06,
+      "loss": 0.8181,
+      "step": 4509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.560872592998939,
+      "learning_rate": 5.106409645508155e-06,
+      "loss": 0.8251,
+      "step": 4510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.56810609641717,
+      "learning_rate": 5.102195251550237e-06,
+      "loss": 0.8889,
+      "step": 4511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.618062832301333,
+      "learning_rate": 5.097982001719994e-06,
+      "loss": 0.8631,
+      "step": 4512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.362190469445197,
+      "learning_rate": 5.093769897001641e-06,
+      "loss": 0.7091,
+      "step": 4513
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3833745464797038,
+      "learning_rate": 5.089558938379131e-06,
+      "loss": 0.8261,
+      "step": 4514
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4768480864864184,
+      "learning_rate": 5.085349126836141e-06,
+      "loss": 0.9492,
+      "step": 4515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5444612370928807,
+      "learning_rate": 5.081140463356089e-06,
+      "loss": 0.9123,
+      "step": 4516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5312007548237583,
+      "learning_rate": 5.076932948922111e-06,
+      "loss": 0.8498,
+      "step": 4517
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6265024417321052,
+      "learning_rate": 5.072726584517086e-06,
+      "loss": 0.8181,
+      "step": 4518
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3985544720537066,
+      "learning_rate": 5.068521371123622e-06,
+      "loss": 0.7978,
+      "step": 4519
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.503437702462011,
+      "learning_rate": 5.064317309724057e-06,
+      "loss": 0.8156,
+      "step": 4520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8016548351966756,
+      "learning_rate": 5.060114401300465e-06,
+      "loss": 0.3438,
+      "step": 4521
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.506198834281312,
+      "learning_rate": 5.0559126468346354e-06,
+      "loss": 0.81,
+      "step": 4522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5161621509729026,
+      "learning_rate": 5.051712047308104e-06,
+      "loss": 0.8009,
+      "step": 4523
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7146744535901257,
+      "learning_rate": 5.047512603702132e-06,
+      "loss": 0.8517,
+      "step": 4524
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4227024463650564,
+      "learning_rate": 5.043314316997709e-06,
+      "loss": 0.8578,
+      "step": 4525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.283323270001088,
+      "learning_rate": 5.039117188175556e-06,
+      "loss": 0.7787,
+      "step": 4526
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3692576199226896,
+      "learning_rate": 5.034921218216126e-06,
+      "loss": 0.8635,
+      "step": 4527
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4081134866816505,
+      "learning_rate": 5.030726408099589e-06,
+      "loss": 0.9368,
+      "step": 4528
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3647333884969972,
+      "learning_rate": 5.026532758805859e-06,
+      "loss": 0.8316,
+      "step": 4529
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4782823771698343,
+      "learning_rate": 5.022340271314572e-06,
+      "loss": 0.897,
+      "step": 4530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2459189380550428,
+      "learning_rate": 5.018148946605092e-06,
+      "loss": 0.8375,
+      "step": 4531
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4245408018389651,
+      "learning_rate": 5.013958785656516e-06,
+      "loss": 0.9104,
+      "step": 4532
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3663946250477728,
+      "learning_rate": 5.009769789447668e-06,
+      "loss": 0.878,
+      "step": 4533
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4889949616649736,
+      "learning_rate": 5.0055819589570904e-06,
+      "loss": 0.8248,
+      "step": 4534
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3973172131891143,
+      "learning_rate": 5.001395295163065e-06,
+      "loss": 0.7906,
+      "step": 4535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5180849992577252,
+      "learning_rate": 4.997209799043597e-06,
+      "loss": 0.7637,
+      "step": 4536
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3678956105381423,
+      "learning_rate": 4.993025471576417e-06,
+      "loss": 0.7792,
+      "step": 4537
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.386641441124462,
+      "learning_rate": 4.988842313738986e-06,
+      "loss": 0.8597,
+      "step": 4538
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2796436909860898,
+      "learning_rate": 4.9846603265084935e-06,
+      "loss": 0.8482,
+      "step": 4539
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9104113093504237,
+      "learning_rate": 4.980479510861845e-06,
+      "loss": 0.3076,
+      "step": 4540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5604483140035275,
+      "learning_rate": 4.976299867775682e-06,
+      "loss": 0.8138,
+      "step": 4541
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.461619237196551,
+      "learning_rate": 4.972121398226371e-06,
+      "loss": 0.7858,
+      "step": 4542
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4754957724440456,
+      "learning_rate": 4.967944103190002e-06,
+      "loss": 0.8772,
+      "step": 4543
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.456856612306713,
+      "learning_rate": 4.9637679836423926e-06,
+      "loss": 0.8643,
+      "step": 4544
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3935185583466247,
+      "learning_rate": 4.959593040559083e-06,
+      "loss": 0.8556,
+      "step": 4545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5275407198267879,
+      "learning_rate": 4.955419274915345e-06,
+      "loss": 0.8872,
+      "step": 4546
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.532512031594679,
+      "learning_rate": 4.951246687686164e-06,
+      "loss": 0.8451,
+      "step": 4547
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7196571795337954,
+      "learning_rate": 4.94707527984626e-06,
+      "loss": 0.8261,
+      "step": 4548
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7672563143972103,
+      "learning_rate": 4.942905052370073e-06,
+      "loss": 0.313,
+      "step": 4549
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3882736374094085,
+      "learning_rate": 4.938736006231769e-06,
+      "loss": 0.8346,
+      "step": 4550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4891459013367765,
+      "learning_rate": 4.934568142405239e-06,
+      "loss": 0.8619,
+      "step": 4551
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3190510757828653,
+      "learning_rate": 4.930401461864099e-06,
+      "loss": 0.8991,
+      "step": 4552
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5827776490335297,
+      "learning_rate": 4.926235965581679e-06,
+      "loss": 0.8309,
+      "step": 4553
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4168386435313955,
+      "learning_rate": 4.922071654531043e-06,
+      "loss": 0.849,
+      "step": 4554
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4720349417788874,
+      "learning_rate": 4.917908529684975e-06,
+      "loss": 0.9505,
+      "step": 4555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5480680079705795,
+      "learning_rate": 4.91374659201598e-06,
+      "loss": 0.8567,
+      "step": 4556
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2266388750898982,
+      "learning_rate": 4.909585842496287e-06,
+      "loss": 0.8331,
+      "step": 4557
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4410847302704344,
+      "learning_rate": 4.905426282097853e-06,
+      "loss": 0.8324,
+      "step": 4558
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3908886478737463,
+      "learning_rate": 4.9012679117923436e-06,
+      "loss": 0.8585,
+      "step": 4559
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4501652030516372,
+      "learning_rate": 4.897110732551157e-06,
+      "loss": 0.8974,
+      "step": 4560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4534998879302232,
+      "learning_rate": 4.892954745345413e-06,
+      "loss": 0.8178,
+      "step": 4561
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2703453637431656,
+      "learning_rate": 4.888799951145948e-06,
+      "loss": 0.7791,
+      "step": 4562
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.505063842539341,
+      "learning_rate": 4.884646350923323e-06,
+      "loss": 0.8761,
+      "step": 4563
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5996110634334908,
+      "learning_rate": 4.8804939456478215e-06,
+      "loss": 0.7712,
+      "step": 4564
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5736127727123124,
+      "learning_rate": 4.8763427362894475e-06,
+      "loss": 0.8745,
+      "step": 4565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5032957825871684,
+      "learning_rate": 4.872192723817917e-06,
+      "loss": 0.867,
+      "step": 4566
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9536430981218512,
+      "learning_rate": 4.868043909202678e-06,
+      "loss": 0.3731,
+      "step": 4567
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.628592184147086,
+      "learning_rate": 4.863896293412892e-06,
+      "loss": 0.8634,
+      "step": 4568
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.913613399845025,
+      "learning_rate": 4.859749877417443e-06,
+      "loss": 0.8612,
+      "step": 4569
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3476565471593558,
+      "learning_rate": 4.855604662184935e-06,
+      "loss": 0.7867,
+      "step": 4570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6611259870708521,
+      "learning_rate": 4.851460648683695e-06,
+      "loss": 0.8543,
+      "step": 4571
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.593920873266002,
+      "learning_rate": 4.847317837881757e-06,
+      "loss": 0.7428,
+      "step": 4572
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5082126683079213,
+      "learning_rate": 4.843176230746883e-06,
+      "loss": 0.8174,
+      "step": 4573
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4511265523001697,
+      "learning_rate": 4.8390358282465574e-06,
+      "loss": 0.8506,
+      "step": 4574
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7499705614440846,
+      "learning_rate": 4.834896631347975e-06,
+      "loss": 0.864,
+      "step": 4575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.650954780990009,
+      "learning_rate": 4.830758641018055e-06,
+      "loss": 0.8253,
+      "step": 4576
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5726215735600095,
+      "learning_rate": 4.826621858223431e-06,
+      "loss": 0.8757,
+      "step": 4577
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3210203005981902,
+      "learning_rate": 4.822486283930461e-06,
+      "loss": 0.9078,
+      "step": 4578
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.599523019148098,
+      "learning_rate": 4.818351919105207e-06,
+      "loss": 0.8421,
+      "step": 4579
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3370304725741007,
+      "learning_rate": 4.8142187647134595e-06,
+      "loss": 0.7919,
+      "step": 4580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5167297350877778,
+      "learning_rate": 4.810086821720726e-06,
+      "loss": 0.9118,
+      "step": 4581
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3664144435318473,
+      "learning_rate": 4.805956091092228e-06,
+      "loss": 0.8118,
+      "step": 4582
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5749322833630273,
+      "learning_rate": 4.801826573792905e-06,
+      "loss": 0.7619,
+      "step": 4583
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4436898795166926,
+      "learning_rate": 4.797698270787415e-06,
+      "loss": 0.8065,
+      "step": 4584
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2930932085923927,
+      "learning_rate": 4.793571183040124e-06,
+      "loss": 0.8655,
+      "step": 4585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.511749942467589,
+      "learning_rate": 4.789445311515123e-06,
+      "loss": 0.7977,
+      "step": 4586
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.416551855433553,
+      "learning_rate": 4.785320657176216e-06,
+      "loss": 0.8471,
+      "step": 4587
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4991309708287353,
+      "learning_rate": 4.7811972209869235e-06,
+      "loss": 0.8712,
+      "step": 4588
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4053003170737826,
+      "learning_rate": 4.77707500391048e-06,
+      "loss": 0.8534,
+      "step": 4589
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.442337568757676,
+      "learning_rate": 4.772954006909837e-06,
+      "loss": 0.9233,
+      "step": 4590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3742264582042067,
+      "learning_rate": 4.768834230947661e-06,
+      "loss": 0.8291,
+      "step": 4591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4333504917387017,
+      "learning_rate": 4.764715676986327e-06,
+      "loss": 0.772,
+      "step": 4592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4574227689128374,
+      "learning_rate": 4.760598345987931e-06,
+      "loss": 0.7811,
+      "step": 4593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3944850454503117,
+      "learning_rate": 4.756482238914285e-06,
+      "loss": 0.856,
+      "step": 4594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4009858805571114,
+      "learning_rate": 4.7523673567269095e-06,
+      "loss": 0.8674,
+      "step": 4595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.386660900386837,
+      "learning_rate": 4.7482537003870425e-06,
+      "loss": 0.8883,
+      "step": 4596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5430515381778944,
+      "learning_rate": 4.744141270855638e-06,
+      "loss": 0.8107,
+      "step": 4597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3471632988832563,
+      "learning_rate": 4.740030069093351e-06,
+      "loss": 0.8284,
+      "step": 4598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3261892829533413,
+      "learning_rate": 4.735920096060565e-06,
+      "loss": 0.8505,
+      "step": 4599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3722022600635457,
+      "learning_rate": 4.731811352717369e-06,
+      "loss": 0.8297,
+      "step": 4600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4497870477453527,
+      "learning_rate": 4.727703840023566e-06,
+      "loss": 0.7923,
+      "step": 4601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3970553224672155,
+      "learning_rate": 4.7235975589386715e-06,
+      "loss": 0.8887,
+      "step": 4602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8949589253544046,
+      "learning_rate": 4.719492510421913e-06,
+      "loss": 0.3419,
+      "step": 4603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2786837316951456,
+      "learning_rate": 4.715388695432232e-06,
+      "loss": 0.7902,
+      "step": 4604
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5366408350148864,
+      "learning_rate": 4.711286114928277e-06,
+      "loss": 0.7646,
+      "step": 4605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5567456309321708,
+      "learning_rate": 4.707184769868412e-06,
+      "loss": 0.8615,
+      "step": 4606
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.384110707474668,
+      "learning_rate": 4.7030846612107105e-06,
+      "loss": 0.7813,
+      "step": 4607
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4534471241967188,
+      "learning_rate": 4.6989857899129595e-06,
+      "loss": 0.8056,
+      "step": 4608
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2515046645075467,
+      "learning_rate": 4.694888156932657e-06,
+      "loss": 0.8657,
+      "step": 4609
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.185863418049476,
+      "learning_rate": 4.690791763227014e-06,
+      "loss": 0.826,
+      "step": 4610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.441683371713312,
+      "learning_rate": 4.68669660975294e-06,
+      "loss": 0.8537,
+      "step": 4611
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.412766075049699,
+      "learning_rate": 4.6826026974670665e-06,
+      "loss": 0.8745,
+      "step": 4612
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4241818680646237,
+      "learning_rate": 4.6785100273257335e-06,
+      "loss": 0.7919,
+      "step": 4613
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3799514787169287,
+      "learning_rate": 4.674418600284988e-06,
+      "loss": 0.7948,
+      "step": 4614
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.245829465995908,
+      "learning_rate": 4.670328417300588e-06,
+      "loss": 0.7695,
+      "step": 4615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5035562101700672,
+      "learning_rate": 4.666239479328004e-06,
+      "loss": 0.8887,
+      "step": 4616
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4349659938211743,
+      "learning_rate": 4.662151787322405e-06,
+      "loss": 0.7622,
+      "step": 4617
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4942491548054193,
+      "learning_rate": 4.658065342238681e-06,
+      "loss": 0.8055,
+      "step": 4618
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4261721754426413,
+      "learning_rate": 4.653980145031425e-06,
+      "loss": 0.8548,
+      "step": 4619
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.36462012444893,
+      "learning_rate": 4.64989619665494e-06,
+      "loss": 0.7615,
+      "step": 4620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2236799625284607,
+      "learning_rate": 4.645813498063235e-06,
+      "loss": 0.8346,
+      "step": 4621
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6040389836625868,
+      "learning_rate": 4.641732050210032e-06,
+      "loss": 0.8527,
+      "step": 4622
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.667602740115268,
+      "learning_rate": 4.637651854048759e-06,
+      "loss": 0.7332,
+      "step": 4623
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3652476404225031,
+      "learning_rate": 4.633572910532543e-06,
+      "loss": 0.7827,
+      "step": 4624
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2873324821688368,
+      "learning_rate": 4.62949522061423e-06,
+      "loss": 0.8192,
+      "step": 4625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.587948321834436,
+      "learning_rate": 4.6254187852463685e-06,
+      "loss": 0.8425,
+      "step": 4626
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0106800101685143,
+      "learning_rate": 4.621343605381215e-06,
+      "loss": 0.3402,
+      "step": 4627
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3883844177347062,
+      "learning_rate": 4.61726968197073e-06,
+      "loss": 0.8996,
+      "step": 4628
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4410600362744985,
+      "learning_rate": 4.613197015966587e-06,
+      "loss": 0.8247,
+      "step": 4629
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4033930885063715,
+      "learning_rate": 4.609125608320154e-06,
+      "loss": 0.8112,
+      "step": 4630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3584792142610111,
+      "learning_rate": 4.605055459982517e-06,
+      "loss": 0.9075,
+      "step": 4631
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5072101207536828,
+      "learning_rate": 4.600986571904461e-06,
+      "loss": 0.7514,
+      "step": 4632
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4723419099347086,
+      "learning_rate": 4.5969189450364804e-06,
+      "loss": 0.8128,
+      "step": 4633
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4686786946275787,
+      "learning_rate": 4.592852580328775e-06,
+      "loss": 0.8374,
+      "step": 4634
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4389864722866783,
+      "learning_rate": 4.588787478731242e-06,
+      "loss": 0.7918,
+      "step": 4635
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3863671528666628,
+      "learning_rate": 4.5847236411934945e-06,
+      "loss": 0.855,
+      "step": 4636
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4385407123388714,
+      "learning_rate": 4.580661068664844e-06,
+      "loss": 0.8695,
+      "step": 4637
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.33760559150418,
+      "learning_rate": 4.576599762094306e-06,
+      "loss": 0.9329,
+      "step": 4638
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.575854801348923,
+      "learning_rate": 4.5725397224306076e-06,
+      "loss": 0.8426,
+      "step": 4639
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5345019136284757,
+      "learning_rate": 4.568480950622169e-06,
+      "loss": 0.7612,
+      "step": 4640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.413658351733374,
+      "learning_rate": 4.56442344761712e-06,
+      "loss": 0.8526,
+      "step": 4641
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6060934881575104,
+      "learning_rate": 4.560367214363295e-06,
+      "loss": 0.8299,
+      "step": 4642
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.430545619058654,
+      "learning_rate": 4.556312251808232e-06,
+      "loss": 0.8646,
+      "step": 4643
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4653842653589146,
+      "learning_rate": 4.552258560899171e-06,
+      "loss": 0.808,
+      "step": 4644
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.428059233094383,
+      "learning_rate": 4.5482061425830504e-06,
+      "loss": 0.7432,
+      "step": 4645
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.65868272854902,
+      "learning_rate": 4.5441549978065166e-06,
+      "loss": 0.8384,
+      "step": 4646
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5749764364283159,
+      "learning_rate": 4.540105127515921e-06,
+      "loss": 0.8385,
+      "step": 4647
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4863594831349358,
+      "learning_rate": 4.53605653265731e-06,
+      "loss": 0.7843,
+      "step": 4648
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5667757379774738,
+      "learning_rate": 4.532009214176438e-06,
+      "loss": 0.712,
+      "step": 4649
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4829015263672782,
+      "learning_rate": 4.527963173018762e-06,
+      "loss": 0.7888,
+      "step": 4650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4465025339215238,
+      "learning_rate": 4.52391841012943e-06,
+      "loss": 0.7083,
+      "step": 4651
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.405534162716725,
+      "learning_rate": 4.519874926453303e-06,
+      "loss": 0.8239,
+      "step": 4652
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5563763462334577,
+      "learning_rate": 4.515832722934939e-06,
+      "loss": 0.8349,
+      "step": 4653
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5581598143871236,
+      "learning_rate": 4.511791800518596e-06,
+      "loss": 0.8495,
+      "step": 4654
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4292397488450292,
+      "learning_rate": 4.507752160148241e-06,
+      "loss": 0.7418,
+      "step": 4655
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8932544506107188,
+      "learning_rate": 4.503713802767523e-06,
+      "loss": 0.3297,
+      "step": 4656
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4965378261590467,
+      "learning_rate": 4.499676729319809e-06,
+      "loss": 0.8386,
+      "step": 4657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.515165323177369,
+      "learning_rate": 4.495640940748159e-06,
+      "loss": 0.8873,
+      "step": 4658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.439888403102316,
+      "learning_rate": 4.491606437995335e-06,
+      "loss": 0.8209,
+      "step": 4659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3864125397241902,
+      "learning_rate": 4.4875732220037935e-06,
+      "loss": 0.8238,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4600354363816694,
+      "learning_rate": 4.483541293715699e-06,
+      "loss": 0.8752,
+      "step": 4661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5157005313359344,
+      "learning_rate": 4.479510654072909e-06,
+      "loss": 0.8415,
+      "step": 4662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6642532675826205,
+      "learning_rate": 4.475481304016978e-06,
+      "loss": 0.8695,
+      "step": 4663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2668533571239595,
+      "learning_rate": 4.471453244489164e-06,
+      "loss": 0.8312,
+      "step": 4664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5066316284393733,
+      "learning_rate": 4.467426476430423e-06,
+      "loss": 0.9041,
+      "step": 4665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5060708000202778,
+      "learning_rate": 4.463401000781409e-06,
+      "loss": 0.7757,
+      "step": 4666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3390542785816126,
+      "learning_rate": 4.459376818482471e-06,
+      "loss": 0.7843,
+      "step": 4667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.436900641049538,
+      "learning_rate": 4.455353930473666e-06,
+      "loss": 0.8594,
+      "step": 4668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6462740588457703,
+      "learning_rate": 4.4513323376947304e-06,
+      "loss": 0.8773,
+      "step": 4669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4759910823819076,
+      "learning_rate": 4.447312041085113e-06,
+      "loss": 0.8712,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4827403981573728,
+      "learning_rate": 4.443293041583957e-06,
+      "loss": 0.8468,
+      "step": 4671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.396443301710576,
+      "learning_rate": 4.439275340130099e-06,
+      "loss": 0.8089,
+      "step": 4672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2549821331929478,
+      "learning_rate": 4.435258937662076e-06,
+      "loss": 0.7941,
+      "step": 4673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4418495124689839,
+      "learning_rate": 4.4312438351181246e-06,
+      "loss": 0.784,
+      "step": 4674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7972635512494446,
+      "learning_rate": 4.4272300334361616e-06,
+      "loss": 0.8095,
+      "step": 4675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4403862723580476,
+      "learning_rate": 4.42321753355382e-06,
+      "loss": 0.9377,
+      "step": 4676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6161400482673711,
+      "learning_rate": 4.419206336408418e-06,
+      "loss": 0.8809,
+      "step": 4677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2982877338535375,
+      "learning_rate": 4.415196442936971e-06,
+      "loss": 0.759,
+      "step": 4678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3484505620701703,
+      "learning_rate": 4.411187854076192e-06,
+      "loss": 0.8095,
+      "step": 4679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3972446506196152,
+      "learning_rate": 4.407180570762486e-06,
+      "loss": 0.8509,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9182674315182678,
+      "learning_rate": 4.40317459393196e-06,
+      "loss": 0.2946,
+      "step": 4681
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.45912375033366,
+      "learning_rate": 4.399169924520403e-06,
+      "loss": 0.8808,
+      "step": 4682
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3372329181206644,
+      "learning_rate": 4.39516656346331e-06,
+      "loss": 0.8189,
+      "step": 4683
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6867213837115818,
+      "learning_rate": 4.391164511695866e-06,
+      "loss": 0.8303,
+      "step": 4684
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3696883979157333,
+      "learning_rate": 4.38716377015295e-06,
+      "loss": 0.8647,
+      "step": 4685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3683280308809131,
+      "learning_rate": 4.383164339769137e-06,
+      "loss": 0.7426,
+      "step": 4686
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6572007765643664,
+      "learning_rate": 4.379166221478697e-06,
+      "loss": 0.9312,
+      "step": 4687
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3925392846339484,
+      "learning_rate": 4.375169416215584e-06,
+      "loss": 0.7916,
+      "step": 4688
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8757613192986392,
+      "learning_rate": 4.371173924913457e-06,
+      "loss": 0.3176,
+      "step": 4689
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4637444923799754,
+      "learning_rate": 4.36717974850566e-06,
+      "loss": 0.8318,
+      "step": 4690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7819713138015125,
+      "learning_rate": 4.363186887925236e-06,
+      "loss": 0.8466,
+      "step": 4691
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4104984652806851,
+      "learning_rate": 4.359195344104916e-06,
+      "loss": 0.7462,
+      "step": 4692
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5619620124179776,
+      "learning_rate": 4.355205117977126e-06,
+      "loss": 0.8214,
+      "step": 4693
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4586233537527196,
+      "learning_rate": 4.351216210473986e-06,
+      "loss": 0.7839,
+      "step": 4694
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6979397152850684,
+      "learning_rate": 4.3472286225272995e-06,
+      "loss": 0.829,
+      "step": 4695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3617905750744985,
+      "learning_rate": 4.343242355068569e-06,
+      "loss": 0.8012,
+      "step": 4696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4182092000368562,
+      "learning_rate": 4.339257409028987e-06,
+      "loss": 0.8156,
+      "step": 4697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.424496608891188,
+      "learning_rate": 4.33527378533944e-06,
+      "loss": 0.8184,
+      "step": 4698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.455989477998742,
+      "learning_rate": 4.3312914849305e-06,
+      "loss": 0.819,
+      "step": 4699
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5782519822918681,
+      "learning_rate": 4.3273105087324375e-06,
+      "loss": 0.8165,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3999424919648857,
+      "learning_rate": 4.323330857675202e-06,
+      "loss": 0.8229,
+      "step": 4701
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.534057270534777,
+      "learning_rate": 4.319352532688444e-06,
+      "loss": 0.8212,
+      "step": 4702
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4408446519193885,
+      "learning_rate": 4.315375534701499e-06,
+      "loss": 0.7798,
+      "step": 4703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2941510522769486,
+      "learning_rate": 4.311399864643396e-06,
+      "loss": 0.8171,
+      "step": 4704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4795992991888256,
+      "learning_rate": 4.307425523442852e-06,
+      "loss": 0.7836,
+      "step": 4705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6663074617930316,
+      "learning_rate": 4.303452512028272e-06,
+      "loss": 0.6779,
+      "step": 4706
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3665190506750673,
+      "learning_rate": 4.2994808313277565e-06,
+      "loss": 0.8058,
+      "step": 4707
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4618797234507428,
+      "learning_rate": 4.295510482269083e-06,
+      "loss": 0.8783,
+      "step": 4708
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5481642831369997,
+      "learning_rate": 4.29154146577973e-06,
+      "loss": 0.8546,
+      "step": 4709
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.43559915648515,
+      "learning_rate": 4.287573782786858e-06,
+      "loss": 0.8495,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6092210621960257,
+      "learning_rate": 4.2836074342173195e-06,
+      "loss": 0.8064,
+      "step": 4711
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5309822909220474,
+      "learning_rate": 4.279642420997655e-06,
+      "loss": 0.8266,
+      "step": 4712
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.362953973360087,
+      "learning_rate": 4.275678744054094e-06,
+      "loss": 0.8767,
+      "step": 4713
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4498389862743661,
+      "learning_rate": 4.271716404312545e-06,
+      "loss": 0.8809,
+      "step": 4714
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.532173708945647,
+      "learning_rate": 4.267755402698613e-06,
+      "loss": 0.7707,
+      "step": 4715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3243116948190425,
+      "learning_rate": 4.263795740137592e-06,
+      "loss": 0.8274,
+      "step": 4716
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.476025755020728,
+      "learning_rate": 4.259837417554457e-06,
+      "loss": 0.816,
+      "step": 4717
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4079421882295637,
+      "learning_rate": 4.2558804358738725e-06,
+      "loss": 0.7988,
+      "step": 4718
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4452387256093888,
+      "learning_rate": 4.251924796020191e-06,
+      "loss": 0.8519,
+      "step": 4719
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7772103385319755,
+      "learning_rate": 4.247970498917453e-06,
+      "loss": 0.9175,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4318473583385691,
+      "learning_rate": 4.244017545489376e-06,
+      "loss": 0.7419,
+      "step": 4721
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5810543046165035,
+      "learning_rate": 4.240065936659374e-06,
+      "loss": 0.7995,
+      "step": 4722
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5578898439366653,
+      "learning_rate": 4.236115673350544e-06,
+      "loss": 0.886,
+      "step": 4723
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4437783475875456,
+      "learning_rate": 4.232166756485665e-06,
+      "loss": 0.869,
+      "step": 4724
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4534665290887891,
+      "learning_rate": 4.228219186987207e-06,
+      "loss": 0.8314,
+      "step": 4725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5232528226128734,
+      "learning_rate": 4.224272965777326e-06,
+      "loss": 0.833,
+      "step": 4726
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3574944236633812,
+      "learning_rate": 4.220328093777851e-06,
+      "loss": 0.8731,
+      "step": 4727
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.363372775068747,
+      "learning_rate": 4.2163845719103106e-06,
+      "loss": 0.7636,
+      "step": 4728
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.419879949441352,
+      "learning_rate": 4.212442401095908e-06,
+      "loss": 0.8024,
+      "step": 4729
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5478196558988124,
+      "learning_rate": 4.20850158225554e-06,
+      "loss": 0.8649,
+      "step": 4730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6587733556465767,
+      "learning_rate": 4.2045621163097775e-06,
+      "loss": 0.859,
+      "step": 4731
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.413761817593304,
+      "learning_rate": 4.200624004178883e-06,
+      "loss": 0.8101,
+      "step": 4732
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.332535731383828,
+      "learning_rate": 4.196687246782801e-06,
+      "loss": 0.8221,
+      "step": 4733
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9546101248457464,
+      "learning_rate": 4.192751845041153e-06,
+      "loss": 0.3299,
+      "step": 4734
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5541536544415093,
+      "learning_rate": 4.188817799873254e-06,
+      "loss": 0.7875,
+      "step": 4735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.377765673467641,
+      "learning_rate": 4.184885112198094e-06,
+      "loss": 0.8193,
+      "step": 4736
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.361601929925436,
+      "learning_rate": 4.180953782934352e-06,
+      "loss": 0.7177,
+      "step": 4737
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5278395496840536,
+      "learning_rate": 4.177023813000386e-06,
+      "loss": 0.8795,
+      "step": 4738
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4926864272996119,
+      "learning_rate": 4.173095203314241e-06,
+      "loss": 0.8539,
+      "step": 4739
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.731921820523921,
+      "learning_rate": 4.169167954793633e-06,
+      "loss": 0.8566,
+      "step": 4740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8371866827097895,
+      "learning_rate": 4.1652420683559725e-06,
+      "loss": 0.334,
+      "step": 4741
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8822856624435376,
+      "learning_rate": 4.161317544918345e-06,
+      "loss": 0.3251,
+      "step": 4742
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3531579402367617,
+      "learning_rate": 4.157394385397521e-06,
+      "loss": 0.8454,
+      "step": 4743
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.585947068425145,
+      "learning_rate": 4.153472590709951e-06,
+      "loss": 0.8922,
+      "step": 4744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4739895838085741,
+      "learning_rate": 4.1495521617717695e-06,
+      "loss": 0.934,
+      "step": 4745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2765781203485074,
+      "learning_rate": 4.145633099498783e-06,
+      "loss": 0.8586,
+      "step": 4746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4175262312379457,
+      "learning_rate": 4.141715404806486e-06,
+      "loss": 0.7706,
+      "step": 4747
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3290769573347403,
+      "learning_rate": 4.137799078610055e-06,
+      "loss": 0.8767,
+      "step": 4748
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.446959698733549,
+      "learning_rate": 4.133884121824344e-06,
+      "loss": 0.8162,
+      "step": 4749
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7136313246230725,
+      "learning_rate": 4.129970535363885e-06,
+      "loss": 0.8187,
+      "step": 4750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9308276498009022,
+      "learning_rate": 4.126058320142895e-06,
+      "loss": 0.32,
+      "step": 4751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5418527357692045,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.8637,
+      "step": 4752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5596672032785333,
+      "learning_rate": 4.1182380070745755e-06,
+      "loss": 0.8271,
+      "step": 4753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.529606549684988,
+      "learning_rate": 4.114329911054069e-06,
+      "loss": 0.8066,
+      "step": 4754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5967319561630282,
+      "learning_rate": 4.110423189926682e-06,
+      "loss": 0.7884,
+      "step": 4755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5521734226059614,
+      "learning_rate": 4.106517844605023e-06,
+      "loss": 0.8523,
+      "step": 4756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4287801655380412,
+      "learning_rate": 4.1026138760013886e-06,
+      "loss": 0.8229,
+      "step": 4757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4106015522009068,
+      "learning_rate": 4.098711285027736e-06,
+      "loss": 0.7808,
+      "step": 4758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4666396817475729,
+      "learning_rate": 4.094810072595714e-06,
+      "loss": 0.8313,
+      "step": 4759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4927494148209086,
+      "learning_rate": 4.090910239616648e-06,
+      "loss": 0.8182,
+      "step": 4760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3403735295109507,
+      "learning_rate": 4.087011787001538e-06,
+      "loss": 0.7562,
+      "step": 4761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5008674785460179,
+      "learning_rate": 4.083114715661069e-06,
+      "loss": 0.7528,
+      "step": 4762
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5130862936844511,
+      "learning_rate": 4.079219026505586e-06,
+      "loss": 0.832,
+      "step": 4763
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5675065725552264,
+      "learning_rate": 4.07532472044513e-06,
+      "loss": 0.7786,
+      "step": 4764
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.454215364889777,
+      "learning_rate": 4.071431798389408e-06,
+      "loss": 0.8455,
+      "step": 4765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5021169353639472,
+      "learning_rate": 4.0675402612478095e-06,
+      "loss": 0.8277,
+      "step": 4766
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4966154832197276,
+      "learning_rate": 4.0636501099294e-06,
+      "loss": 0.7536,
+      "step": 4767
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6086826252898843,
+      "learning_rate": 4.059761345342913e-06,
+      "loss": 0.8441,
+      "step": 4768
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4273763467326928,
+      "learning_rate": 4.055873968396767e-06,
+      "loss": 0.9098,
+      "step": 4769
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4487811363521983,
+      "learning_rate": 4.051987979999056e-06,
+      "loss": 0.804,
+      "step": 4770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.674378791908379,
+      "learning_rate": 4.048103381057543e-06,
+      "loss": 0.8816,
+      "step": 4771
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3597268972714522,
+      "learning_rate": 4.044220172479675e-06,
+      "loss": 0.7997,
+      "step": 4772
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3826802885569884,
+      "learning_rate": 4.040338355172571e-06,
+      "loss": 0.8142,
+      "step": 4773
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9760999015720804,
+      "learning_rate": 4.0364579300430164e-06,
+      "loss": 0.8612,
+      "step": 4774
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5191566335247086,
+      "learning_rate": 4.032578897997485e-06,
+      "loss": 0.7454,
+      "step": 4775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.552452624447766,
+      "learning_rate": 4.028701259942116e-06,
+      "loss": 0.8588,
+      "step": 4776
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2612519758977432,
+      "learning_rate": 4.024825016782727e-06,
+      "loss": 0.7941,
+      "step": 4777
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3981196958767816,
+      "learning_rate": 4.020950169424815e-06,
+      "loss": 0.7622,
+      "step": 4778
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3804476999751516,
+      "learning_rate": 4.017076718773535e-06,
+      "loss": 0.832,
+      "step": 4779
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3309790715032808,
+      "learning_rate": 4.013204665733729e-06,
+      "loss": 0.7547,
+      "step": 4780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3757551771852794,
+      "learning_rate": 4.009334011209909e-06,
+      "loss": 0.8043,
+      "step": 4781
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3654193357223396,
+      "learning_rate": 4.0054647561062625e-06,
+      "loss": 0.7933,
+      "step": 4782
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4864574594156503,
+      "learning_rate": 4.001596901326644e-06,
+      "loss": 0.799,
+      "step": 4783
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.482259758464327,
+      "learning_rate": 3.997730447774591e-06,
+      "loss": 0.9302,
+      "step": 4784
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4225989636571883,
+      "learning_rate": 3.9938653963533e-06,
+      "loss": 0.8264,
+      "step": 4785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4878184307287057,
+      "learning_rate": 3.990001747965652e-06,
+      "loss": 0.8491,
+      "step": 4786
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3759775364793156,
+      "learning_rate": 3.9861395035141936e-06,
+      "loss": 0.8779,
+      "step": 4787
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5151635812092463,
+      "learning_rate": 3.982278663901146e-06,
+      "loss": 0.7878,
+      "step": 4788
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.194047763646673,
+      "learning_rate": 3.978419230028402e-06,
+      "loss": 0.7804,
+      "step": 4789
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4074109914243975,
+      "learning_rate": 3.974561202797525e-06,
+      "loss": 0.7491,
+      "step": 4790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.321865521333858,
+      "learning_rate": 3.970704583109755e-06,
+      "loss": 0.8734,
+      "step": 4791
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.466675656207691,
+      "learning_rate": 3.9668493718659924e-06,
+      "loss": 0.7656,
+      "step": 4792
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3905945966108024,
+      "learning_rate": 3.962995569966817e-06,
+      "loss": 0.8681,
+      "step": 4793
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8481108003508808,
+      "learning_rate": 3.9591431783124786e-06,
+      "loss": 0.8089,
+      "step": 4794
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5345186500111654,
+      "learning_rate": 3.955292197802895e-06,
+      "loss": 0.786,
+      "step": 4795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3616600576042024,
+      "learning_rate": 3.951442629337657e-06,
+      "loss": 0.8063,
+      "step": 4796
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4456912188901643,
+      "learning_rate": 3.947594473816026e-06,
+      "loss": 0.8064,
+      "step": 4797
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.471979114571703,
+      "learning_rate": 3.943747732136925e-06,
+      "loss": 0.7711,
+      "step": 4798
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4492812965756292,
+      "learning_rate": 3.939902405198959e-06,
+      "loss": 0.8662,
+      "step": 4799
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3353068567023287,
+      "learning_rate": 3.936058493900393e-06,
+      "loss": 0.8853,
+      "step": 4800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5290175325305821,
+      "learning_rate": 3.932215999139167e-06,
+      "loss": 0.7905,
+      "step": 4801
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4040433445236589,
+      "learning_rate": 3.9283749218128885e-06,
+      "loss": 0.8155,
+      "step": 4802
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.46126270881956,
+      "learning_rate": 3.924535262818836e-06,
+      "loss": 0.8694,
+      "step": 4803
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.632287188494207,
+      "learning_rate": 3.920697023053949e-06,
+      "loss": 0.8303,
+      "step": 4804
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3928589169948147,
+      "learning_rate": 3.916860203414843e-06,
+      "loss": 0.8813,
+      "step": 4805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.397834508641805,
+      "learning_rate": 3.913024804797798e-06,
+      "loss": 0.8743,
+      "step": 4806
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.512113429699455,
+      "learning_rate": 3.909190828098766e-06,
+      "loss": 0.784,
+      "step": 4807
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4448060836762557,
+      "learning_rate": 3.905358274213363e-06,
+      "loss": 0.7627,
+      "step": 4808
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3835596663047998,
+      "learning_rate": 3.901527144036875e-06,
+      "loss": 0.9225,
+      "step": 4809
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3675386495594217,
+      "learning_rate": 3.897697438464256e-06,
+      "loss": 0.8488,
+      "step": 4810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4752544122799878,
+      "learning_rate": 3.893869158390121e-06,
+      "loss": 0.8653,
+      "step": 4811
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5770190289200632,
+      "learning_rate": 3.890042304708758e-06,
+      "loss": 0.8394,
+      "step": 4812
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.874338005345734,
+      "learning_rate": 3.886216878314122e-06,
+      "loss": 0.9145,
+      "step": 4813
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3081891252765854,
+      "learning_rate": 3.882392880099832e-06,
+      "loss": 0.8998,
+      "step": 4814
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5232707030122634,
+      "learning_rate": 3.878570310959175e-06,
+      "loss": 0.8172,
+      "step": 4815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4112825499027488,
+      "learning_rate": 3.874749171785106e-06,
+      "loss": 0.8087,
+      "step": 4816
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3860203911991953,
+      "learning_rate": 3.8709294634702374e-06,
+      "loss": 0.8032,
+      "step": 4817
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4114405702309578,
+      "learning_rate": 3.867111186906857e-06,
+      "loss": 0.8686,
+      "step": 4818
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5032844688609845,
+      "learning_rate": 3.8632943429869145e-06,
+      "loss": 0.7949,
+      "step": 4819
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4174561611802248,
+      "learning_rate": 3.859478932602025e-06,
+      "loss": 0.8234,
+      "step": 4820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3569986795174422,
+      "learning_rate": 3.855664956643467e-06,
+      "loss": 0.8315,
+      "step": 4821
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4857589500337158,
+      "learning_rate": 3.8518524160021876e-06,
+      "loss": 0.8737,
+      "step": 4822
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.367728689573794,
+      "learning_rate": 3.848041311568801e-06,
+      "loss": 0.812,
+      "step": 4823
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8014481210062113,
+      "learning_rate": 3.844231644233572e-06,
+      "loss": 0.324,
+      "step": 4824
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2932778274280683,
+      "learning_rate": 3.840423414886445e-06,
+      "loss": 0.8386,
+      "step": 4825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4478635582400583,
+      "learning_rate": 3.836616624417022e-06,
+      "loss": 1.0061,
+      "step": 4826
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.417811438869292,
+      "learning_rate": 3.832811273714569e-06,
+      "loss": 0.9058,
+      "step": 4827
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8921443753417463,
+      "learning_rate": 3.829007363668018e-06,
+      "loss": 0.3239,
+      "step": 4828
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5137789436127937,
+      "learning_rate": 3.825204895165964e-06,
+      "loss": 0.8515,
+      "step": 4829
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.379460039861955,
+      "learning_rate": 3.821403869096658e-06,
+      "loss": 0.754,
+      "step": 4830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3290326957603285,
+      "learning_rate": 3.817604286348025e-06,
+      "loss": 0.8301,
+      "step": 4831
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4835734951551014,
+      "learning_rate": 3.813806147807645e-06,
+      "loss": 0.8526,
+      "step": 4832
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4383267507861999,
+      "learning_rate": 3.8100094543627666e-06,
+      "loss": 0.8591,
+      "step": 4833
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3725708866281356,
+      "learning_rate": 3.806214206900295e-06,
+      "loss": 0.8745,
+      "step": 4834
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4540064109992301,
+      "learning_rate": 3.8024204063068024e-06,
+      "loss": 0.8106,
+      "step": 4835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5596694932667077,
+      "learning_rate": 3.798628053468524e-06,
+      "loss": 0.8085,
+      "step": 4836
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3951547675296299,
+      "learning_rate": 3.7948371492713454e-06,
+      "loss": 0.8687,
+      "step": 4837
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9132438069304774,
+      "learning_rate": 3.791047694600828e-06,
+      "loss": 0.3558,
+      "step": 4838
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.420216989479524,
+      "learning_rate": 3.7872596903421876e-06,
+      "loss": 0.8197,
+      "step": 4839
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8687397487767968,
+      "learning_rate": 3.7834731373803023e-06,
+      "loss": 0.3534,
+      "step": 4840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.393845029363547,
+      "learning_rate": 3.7796880365997114e-06,
+      "loss": 0.7153,
+      "step": 4841
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4690922867111074,
+      "learning_rate": 3.775904388884618e-06,
+      "loss": 0.8483,
+      "step": 4842
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4726739719194737,
+      "learning_rate": 3.772122195118877e-06,
+      "loss": 0.7215,
+      "step": 4843
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5858698405316958,
+      "learning_rate": 3.7683414561860097e-06,
+      "loss": 0.9649,
+      "step": 4844
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3722663494991016,
+      "learning_rate": 3.7645621729692004e-06,
+      "loss": 0.8364,
+      "step": 4845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.55642083721182,
+      "learning_rate": 3.7607843463512894e-06,
+      "loss": 0.8412,
+      "step": 4846
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8704953899371346,
+      "learning_rate": 3.7570079772147748e-06,
+      "loss": 0.8075,
+      "step": 4847
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2505106330014943,
+      "learning_rate": 3.7532330664418202e-06,
+      "loss": 0.8618,
+      "step": 4848
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.458515907819654,
+      "learning_rate": 3.749459614914246e-06,
+      "loss": 0.8533,
+      "step": 4849
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4192929770699139,
+      "learning_rate": 3.7456876235135252e-06,
+      "loss": 0.7358,
+      "step": 4850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5055481196921887,
+      "learning_rate": 3.7419170931207994e-06,
+      "loss": 0.8199,
+      "step": 4851
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5142826478550857,
+      "learning_rate": 3.738148024616863e-06,
+      "loss": 0.8703,
+      "step": 4852
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5294559050168062,
+      "learning_rate": 3.7343804188821718e-06,
+      "loss": 0.7979,
+      "step": 4853
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5519530789845535,
+      "learning_rate": 3.73061427679684e-06,
+      "loss": 0.8235,
+      "step": 4854
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5608174014356124,
+      "learning_rate": 3.7268495992406404e-06,
+      "loss": 0.8163,
+      "step": 4855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5086427864597807,
+      "learning_rate": 3.723086387092997e-06,
+      "loss": 0.9018,
+      "step": 4856
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3719692274440831,
+      "learning_rate": 3.7193246412329976e-06,
+      "loss": 0.7802,
+      "step": 4857
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.456942407287454,
+      "learning_rate": 3.7155643625393878e-06,
+      "loss": 0.8336,
+      "step": 4858
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2060986746361193,
+      "learning_rate": 3.7118055518905693e-06,
+      "loss": 0.7113,
+      "step": 4859
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3452162934726444,
+      "learning_rate": 3.7080482101645998e-06,
+      "loss": 0.8275,
+      "step": 4860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4342690963752147,
+      "learning_rate": 3.7042923382391993e-06,
+      "loss": 0.7637,
+      "step": 4861
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3953516988278183,
+      "learning_rate": 3.700537936991733e-06,
+      "loss": 0.7598,
+      "step": 4862
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.585860891793868,
+      "learning_rate": 3.6967850072992307e-06,
+      "loss": 0.8729,
+      "step": 4863
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4687300449902814,
+      "learning_rate": 3.6930335500383795e-06,
+      "loss": 0.8169,
+      "step": 4864
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4885200011118405,
+      "learning_rate": 3.6892835660855184e-06,
+      "loss": 0.889,
+      "step": 4865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2646212306097024,
+      "learning_rate": 3.6855350563166457e-06,
+      "loss": 0.8271,
+      "step": 4866
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3582479975015693,
+      "learning_rate": 3.681788021607413e-06,
+      "loss": 0.7663,
+      "step": 4867
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5216727207530145,
+      "learning_rate": 3.6780424628331313e-06,
+      "loss": 0.8025,
+      "step": 4868
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5854804973142658,
+      "learning_rate": 3.674298380868756e-06,
+      "loss": 0.814,
+      "step": 4869
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5724932262994187,
+      "learning_rate": 3.67055577658891e-06,
+      "loss": 0.8147,
+      "step": 4870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5240502329583558,
+      "learning_rate": 3.6668146508678646e-06,
+      "loss": 0.8439,
+      "step": 4871
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6149185983653107,
+      "learning_rate": 3.6630750045795472e-06,
+      "loss": 0.8102,
+      "step": 4872
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6175554480042074,
+      "learning_rate": 3.65933683859754e-06,
+      "loss": 0.7363,
+      "step": 4873
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3917677897171943,
+      "learning_rate": 3.655600153795084e-06,
+      "loss": 0.8935,
+      "step": 4874
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4527951219889264,
+      "learning_rate": 3.6518649510450598e-06,
+      "loss": 0.8515,
+      "step": 4875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4030707992947349,
+      "learning_rate": 3.6481312312200144e-06,
+      "loss": 0.8567,
+      "step": 4876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5009433237588676,
+      "learning_rate": 3.6443989951921478e-06,
+      "loss": 0.835,
+      "step": 4877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.422809373076941,
+      "learning_rate": 3.6406682438333074e-06,
+      "loss": 0.8757,
+      "step": 4878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5316751167326643,
+      "learning_rate": 3.636938978015e-06,
+      "loss": 0.7215,
+      "step": 4879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.471397196863491,
+      "learning_rate": 3.633211198608384e-06,
+      "loss": 0.8552,
+      "step": 4880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5386594248149572,
+      "learning_rate": 3.6294849064842642e-06,
+      "loss": 0.8179,
+      "step": 4881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.605579674037567,
+      "learning_rate": 3.625760102513103e-06,
+      "loss": 0.7647,
+      "step": 4882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4130220583309183,
+      "learning_rate": 3.6220367875650175e-06,
+      "loss": 0.8248,
+      "step": 4883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3502545719118326,
+      "learning_rate": 3.6183149625097745e-06,
+      "loss": 0.7769,
+      "step": 4884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.563683135797119,
+      "learning_rate": 3.6145946282167944e-06,
+      "loss": 0.9104,
+      "step": 4885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4312655964724141,
+      "learning_rate": 3.6108757855551426e-06,
+      "loss": 0.7876,
+      "step": 4886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8904965521392547,
+      "learning_rate": 3.607158435393544e-06,
+      "loss": 0.3235,
+      "step": 4887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8261118634840814,
+      "learning_rate": 3.603442578600371e-06,
+      "loss": 0.3093,
+      "step": 4888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4719987889087596,
+      "learning_rate": 3.5997282160436488e-06,
+      "loss": 0.889,
+      "step": 4889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4901343120813264,
+      "learning_rate": 3.5960153485910564e-06,
+      "loss": 0.8714,
+      "step": 4890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5333902742885017,
+      "learning_rate": 3.592303977109914e-06,
+      "loss": 0.8535,
+      "step": 4891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4793218502024539,
+      "learning_rate": 3.5885941024672e-06,
+      "loss": 0.7675,
+      "step": 4892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.348049959777778,
+      "learning_rate": 3.5848857255295434e-06,
+      "loss": 0.8275,
+      "step": 4893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.54461516893757,
+      "learning_rate": 3.581178847163219e-06,
+      "loss": 0.8014,
+      "step": 4894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5353823118014385,
+      "learning_rate": 3.5774734682341563e-06,
+      "loss": 0.873,
+      "step": 4895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.439438940156898,
+      "learning_rate": 3.573769589607934e-06,
+      "loss": 0.8959,
+      "step": 4896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3809090171048923,
+      "learning_rate": 3.5700672121497728e-06,
+      "loss": 0.8152,
+      "step": 4897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3830247550446988,
+      "learning_rate": 3.5663663367245517e-06,
+      "loss": 0.8019,
+      "step": 4898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5182380493344791,
+      "learning_rate": 3.5626669641967948e-06,
+      "loss": 0.7764,
+      "step": 4899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3226309811070474,
+      "learning_rate": 3.5589690954306764e-06,
+      "loss": 0.7231,
+      "step": 4900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3586395222508696,
+      "learning_rate": 3.5552727312900228e-06,
+      "loss": 0.8259,
+      "step": 4901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3165007756817604,
+      "learning_rate": 3.5515778726382967e-06,
+      "loss": 0.8583,
+      "step": 4902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2576651442439926,
+      "learning_rate": 3.5478845203386227e-06,
+      "loss": 0.8827,
+      "step": 4903
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4480741034191904,
+      "learning_rate": 3.544192675253767e-06,
+      "loss": 0.806,
+      "step": 4904
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4602856622581222,
+      "learning_rate": 3.5405023382461457e-06,
+      "loss": 0.835,
+      "step": 4905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9027757157731147,
+      "learning_rate": 3.536813510177822e-06,
+      "loss": 0.2934,
+      "step": 4906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2897084712802276,
+      "learning_rate": 3.53312619191051e-06,
+      "loss": 0.7262,
+      "step": 4907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2227321683451722,
+      "learning_rate": 3.5294403843055604e-06,
+      "loss": 0.8158,
+      "step": 4908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.582154301579126,
+      "learning_rate": 3.5257560882239806e-06,
+      "loss": 0.8582,
+      "step": 4909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.453247587483461,
+      "learning_rate": 3.5220733045264245e-06,
+      "loss": 0.9205,
+      "step": 4910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5417394708801138,
+      "learning_rate": 3.5183920340731893e-06,
+      "loss": 0.8361,
+      "step": 4911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.415344627828099,
+      "learning_rate": 3.5147122777242203e-06,
+      "loss": 0.7334,
+      "step": 4912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3561795511934418,
+      "learning_rate": 3.5110340363391128e-06,
+      "loss": 0.9329,
+      "step": 4913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2808862931708658,
+      "learning_rate": 3.5073573107770977e-06,
+      "loss": 0.8246,
+      "step": 4914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4454705583647465,
+      "learning_rate": 3.5036821018970623e-06,
+      "loss": 0.7803,
+      "step": 4915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4250048247786875,
+      "learning_rate": 3.5000084105575338e-06,
+      "loss": 0.8314,
+      "step": 4916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3528221316158837,
+      "learning_rate": 3.4963362376166886e-06,
+      "loss": 0.8673,
+      "step": 4917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.500395836854606,
+      "learning_rate": 3.4926655839323463e-06,
+      "loss": 0.8191,
+      "step": 4918
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.527077861481613,
+      "learning_rate": 3.4889964503619723e-06,
+      "loss": 0.8824,
+      "step": 4919
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.469735482020893,
+      "learning_rate": 3.4853288377626793e-06,
+      "loss": 0.8314,
+      "step": 4920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3202190387252009,
+      "learning_rate": 3.4816627469912147e-06,
+      "loss": 0.793,
+      "step": 4921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4228220905820244,
+      "learning_rate": 3.477998178903982e-06,
+      "loss": 0.8116,
+      "step": 4922
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4091884588971832,
+      "learning_rate": 3.4743351343570253e-06,
+      "loss": 0.8918,
+      "step": 4923
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4314405410755784,
+      "learning_rate": 3.4706736142060305e-06,
+      "loss": 0.8379,
+      "step": 4924
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.509050342935842,
+      "learning_rate": 3.46701361930633e-06,
+      "loss": 0.8057,
+      "step": 4925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5487347150153037,
+      "learning_rate": 3.4633551505129025e-06,
+      "loss": 0.8935,
+      "step": 4926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.325865828913274,
+      "learning_rate": 3.459698208680359e-06,
+      "loss": 0.8681,
+      "step": 4927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4563853056717129,
+      "learning_rate": 3.456042794662966e-06,
+      "loss": 0.8923,
+      "step": 4928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3784967252019977,
+      "learning_rate": 3.452388909314628e-06,
+      "loss": 0.7922,
+      "step": 4929
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4515064541782943,
+      "learning_rate": 3.4487365534888926e-06,
+      "loss": 0.8274,
+      "step": 4930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4210344164301765,
+      "learning_rate": 3.4450857280389527e-06,
+      "loss": 0.7892,
+      "step": 4931
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6105679192603404,
+      "learning_rate": 3.441436433817641e-06,
+      "loss": 0.8734,
+      "step": 4932
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6438779495888407,
+      "learning_rate": 3.437788671677429e-06,
+      "loss": 0.746,
+      "step": 4933
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.388980935038465,
+      "learning_rate": 3.4341424424704373e-06,
+      "loss": 0.8235,
+      "step": 4934
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.540215415025226,
+      "learning_rate": 3.4304977470484257e-06,
+      "loss": 0.7908,
+      "step": 4935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6147232861518725,
+      "learning_rate": 3.4268545862627932e-06,
+      "loss": 0.8428,
+      "step": 4936
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5604165300879358,
+      "learning_rate": 3.423212960964586e-06,
+      "loss": 0.8757,
+      "step": 4937
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5906676546431284,
+      "learning_rate": 3.419572872004485e-06,
+      "loss": 0.8674,
+      "step": 4938
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3604494458081229,
+      "learning_rate": 3.415934320232821e-06,
+      "loss": 0.8945,
+      "step": 4939
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2812985347068435,
+      "learning_rate": 3.4122973064995513e-06,
+      "loss": 0.8319,
+      "step": 4940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5164318824952208,
+      "learning_rate": 3.4086618316542874e-06,
+      "loss": 0.8293,
+      "step": 4941
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4807960329669345,
+      "learning_rate": 3.405027896546277e-06,
+      "loss": 0.8723,
+      "step": 4942
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3964843623288026,
+      "learning_rate": 3.401395502024406e-06,
+      "loss": 0.8985,
+      "step": 4943
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3281162934859385,
+      "learning_rate": 3.3977646489372042e-06,
+      "loss": 0.8389,
+      "step": 4944
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3915344998446528,
+      "learning_rate": 3.39413533813284e-06,
+      "loss": 0.768,
+      "step": 4945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3920876554152817,
+      "learning_rate": 3.390507570459116e-06,
+      "loss": 0.875,
+      "step": 4946
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3177138846422447,
+      "learning_rate": 3.3868813467634833e-06,
+      "loss": 0.777,
+      "step": 4947
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5053367573710172,
+      "learning_rate": 3.3832566678930255e-06,
+      "loss": 0.8646,
+      "step": 4948
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3967249923098846,
+      "learning_rate": 3.3796335346944697e-06,
+      "loss": 0.8012,
+      "step": 4949
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5067589592446557,
+      "learning_rate": 3.3760119480141794e-06,
+      "loss": 0.8661,
+      "step": 4950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5924306672587207,
+      "learning_rate": 3.3723919086981584e-06,
+      "loss": 0.8267,
+      "step": 4951
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3137392575510314,
+      "learning_rate": 3.3687734175920505e-06,
+      "loss": 0.8414,
+      "step": 4952
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3798969731073352,
+      "learning_rate": 3.3651564755411293e-06,
+      "loss": 0.7691,
+      "step": 4953
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2526497970532375,
+      "learning_rate": 3.3615410833903174e-06,
+      "loss": 0.7858,
+      "step": 4954
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6640178804280839,
+      "learning_rate": 3.357927241984169e-06,
+      "loss": 0.815,
+      "step": 4955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4664663750115534,
+      "learning_rate": 3.3543149521668784e-06,
+      "loss": 0.947,
+      "step": 4956
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5390067054104086,
+      "learning_rate": 3.350704214782278e-06,
+      "loss": 0.8785,
+      "step": 4957
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2736941902371575,
+      "learning_rate": 3.347095030673838e-06,
+      "loss": 0.881,
+      "step": 4958
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8277592649853668,
+      "learning_rate": 3.3434874006846583e-06,
+      "loss": 0.3672,
+      "step": 4959
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2968850178504,
+      "learning_rate": 3.3398813256574847e-06,
+      "loss": 0.8256,
+      "step": 4960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2894751394057296,
+      "learning_rate": 3.3362768064346974e-06,
+      "loss": 0.8186,
+      "step": 4961
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.525085724019825,
+      "learning_rate": 3.3326738438583116e-06,
+      "loss": 0.9152,
+      "step": 4962
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7221369955523933,
+      "learning_rate": 3.3290724387699803e-06,
+      "loss": 0.812,
+      "step": 4963
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4134138223835346,
+      "learning_rate": 3.3254725920109922e-06,
+      "loss": 0.8234,
+      "step": 4964
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8848070614945303,
+      "learning_rate": 3.3218743044222746e-06,
+      "loss": 0.3164,
+      "step": 4965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5071622942571334,
+      "learning_rate": 3.318277576844381e-06,
+      "loss": 0.7947,
+      "step": 4966
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2562755819047962,
+      "learning_rate": 3.314682410117511e-06,
+      "loss": 0.7868,
+      "step": 4967
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6172549756573282,
+      "learning_rate": 3.311088805081497e-06,
+      "loss": 0.7891,
+      "step": 4968
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5749992253286484,
+      "learning_rate": 3.3074967625758037e-06,
+      "loss": 0.803,
+      "step": 4969
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.445036344951784,
+      "learning_rate": 3.303906283439534e-06,
+      "loss": 0.7489,
+      "step": 4970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6058600011632724,
+      "learning_rate": 3.3003173685114265e-06,
+      "loss": 0.7793,
+      "step": 4971
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5406243645877573,
+      "learning_rate": 3.2967300186298456e-06,
+      "loss": 0.8547,
+      "step": 4972
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3541374085701254,
+      "learning_rate": 3.2931442346328e-06,
+      "loss": 0.7588,
+      "step": 4973
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4130209886115168,
+      "learning_rate": 3.2895600173579302e-06,
+      "loss": 0.8479,
+      "step": 4974
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5548230799785667,
+      "learning_rate": 3.2859773676425077e-06,
+      "loss": 0.8801,
+      "step": 4975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6156861896843813,
+      "learning_rate": 3.2823962863234416e-06,
+      "loss": 0.9096,
+      "step": 4976
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.784636664226037,
+      "learning_rate": 3.2788167742372725e-06,
+      "loss": 0.3089,
+      "step": 4977
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3833322295907178,
+      "learning_rate": 3.275238832220178e-06,
+      "loss": 0.841,
+      "step": 4978
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2908992635542076,
+      "learning_rate": 3.271662461107958e-06,
+      "loss": 0.8154,
+      "step": 4979
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4818017759567386,
+      "learning_rate": 3.2680876617360578e-06,
+      "loss": 0.8424,
+      "step": 4980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.440983055632886,
+      "learning_rate": 3.264514434939551e-06,
+      "loss": 0.8194,
+      "step": 4981
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4442508400330965,
+      "learning_rate": 3.2609427815531426e-06,
+      "loss": 0.7527,
+      "step": 4982
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3619144099421232,
+      "learning_rate": 3.257372702411171e-06,
+      "loss": 0.8014,
+      "step": 4983
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4155440591046933,
+      "learning_rate": 3.253804198347612e-06,
+      "loss": 0.7911,
+      "step": 4984
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7960117277388261,
+      "learning_rate": 3.2502372701960603e-06,
+      "loss": 0.8031,
+      "step": 4985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4163487581840326,
+      "learning_rate": 3.2466719187897555e-06,
+      "loss": 0.7964,
+      "step": 4986
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.609020266395373,
+      "learning_rate": 3.243108144961563e-06,
+      "loss": 0.8117,
+      "step": 4987
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5806663577035889,
+      "learning_rate": 3.2395459495439817e-06,
+      "loss": 0.8302,
+      "step": 4988
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6151307336844092,
+      "learning_rate": 3.235985333369139e-06,
+      "loss": 0.8235,
+      "step": 4989
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4287413498046333,
+      "learning_rate": 3.2324262972688e-06,
+      "loss": 0.8531,
+      "step": 4990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4059978378234,
+      "learning_rate": 3.2288688420743487e-06,
+      "loss": 0.8024,
+      "step": 4991
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3787769062446489,
+      "learning_rate": 3.2253129686168105e-06,
+      "loss": 0.8391,
+      "step": 4992
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3963622259858146,
+      "learning_rate": 3.2217586777268383e-06,
+      "loss": 0.8302,
+      "step": 4993
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3595949123893245,
+      "learning_rate": 3.2182059702347135e-06,
+      "loss": 0.886,
+      "step": 4994
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4233739780108456,
+      "learning_rate": 3.2146548469703497e-06,
+      "loss": 0.7828,
+      "step": 4995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8106513386936745,
+      "learning_rate": 3.2111053087632904e-06,
+      "loss": 0.321,
+      "step": 4996
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4962990572622585,
+      "learning_rate": 3.2075573564427097e-06,
+      "loss": 0.8459,
+      "step": 4997
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3640774397370556,
+      "learning_rate": 3.204010990837404e-06,
+      "loss": 0.8962,
+      "step": 4998
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3858456300834159,
+      "learning_rate": 3.200466212775808e-06,
+      "loss": 0.8486,
+      "step": 4999
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4379594867977468,
+      "learning_rate": 3.1969230230859815e-06,
+      "loss": 0.7793,
+      "step": 5000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4347378066243224,
+      "learning_rate": 3.193381422595616e-06,
+      "loss": 0.8284,
+      "step": 5001
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5176999424925202,
+      "learning_rate": 3.1898414121320277e-06,
+      "loss": 0.7464,
+      "step": 5002
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4320486678996,
+      "learning_rate": 3.1863029925221667e-06,
+      "loss": 0.8051,
+      "step": 5003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5131113349621699,
+      "learning_rate": 3.1827661645926024e-06,
+      "loss": 0.8955,
+      "step": 5004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4197091477150343,
+      "learning_rate": 3.179230929169541e-06,
+      "loss": 0.7323,
+      "step": 5005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4592043909915928,
+      "learning_rate": 3.175697287078814e-06,
+      "loss": 0.8198,
+      "step": 5006
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4670316378769366,
+      "learning_rate": 3.1721652391458804e-06,
+      "loss": 0.8006,
+      "step": 5007
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3955671586110505,
+      "learning_rate": 3.1686347861958313e-06,
+      "loss": 0.7781,
+      "step": 5008
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.499166452003566,
+      "learning_rate": 3.1651059290533726e-06,
+      "loss": 0.8033,
+      "step": 5009
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7163630963892307,
+      "learning_rate": 3.1615786685428495e-06,
+      "loss": 0.7976,
+      "step": 5010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3711049794011652,
+      "learning_rate": 3.158053005488232e-06,
+      "loss": 0.7648,
+      "step": 5011
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4268633524707137,
+      "learning_rate": 3.1545289407131128e-06,
+      "loss": 0.8595,
+      "step": 5012
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4917406541529343,
+      "learning_rate": 3.151006475040719e-06,
+      "loss": 0.7695,
+      "step": 5013
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4830555811553476,
+      "learning_rate": 3.147485609293891e-06,
+      "loss": 0.8996,
+      "step": 5014
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.480986922176576,
+      "learning_rate": 3.1439663442951073e-06,
+      "loss": 0.7786,
+      "step": 5015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.479415921804418,
+      "learning_rate": 3.140448680866469e-06,
+      "loss": 0.8503,
+      "step": 5016
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.592533068876874,
+      "learning_rate": 3.1369326198297025e-06,
+      "loss": 0.8365,
+      "step": 5017
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.486062744249541,
+      "learning_rate": 3.13341816200616e-06,
+      "loss": 0.8036,
+      "step": 5018
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5557173332485443,
+      "learning_rate": 3.1299053082168206e-06,
+      "loss": 0.8483,
+      "step": 5019
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.493092759764533,
+      "learning_rate": 3.126394059282284e-06,
+      "loss": 0.7798,
+      "step": 5020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5227747649393084,
+      "learning_rate": 3.122884416022779e-06,
+      "loss": 0.8205,
+      "step": 5021
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4368630713763728,
+      "learning_rate": 3.11937637925816e-06,
+      "loss": 0.7469,
+      "step": 5022
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5487856744863249,
+      "learning_rate": 3.1158699498079037e-06,
+      "loss": 0.8654,
+      "step": 5023
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5658335899407716,
+      "learning_rate": 3.1123651284911162e-06,
+      "loss": 0.8348,
+      "step": 5024
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4453956373892873,
+      "learning_rate": 3.108861916126518e-06,
+      "loss": 0.8853,
+      "step": 5025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4394074957485403,
+      "learning_rate": 3.105360313532463e-06,
+      "loss": 0.8181,
+      "step": 5026
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5528121628859766,
+      "learning_rate": 3.101860321526924e-06,
+      "loss": 0.8283,
+      "step": 5027
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3932914351574135,
+      "learning_rate": 3.098361940927502e-06,
+      "loss": 0.8278,
+      "step": 5028
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.667759959310162,
+      "learning_rate": 3.0948651725514223e-06,
+      "loss": 0.8683,
+      "step": 5029
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3775178698087676,
+      "learning_rate": 3.0913700172155226e-06,
+      "loss": 0.7878,
+      "step": 5030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5710079754899224,
+      "learning_rate": 3.0878764757362745e-06,
+      "loss": 0.8395,
+      "step": 5031
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3338889322444747,
+      "learning_rate": 3.0843845489297698e-06,
+      "loss": 0.8098,
+      "step": 5032
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4173942261102668,
+      "learning_rate": 3.0808942376117244e-06,
+      "loss": 0.8592,
+      "step": 5033
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3592008638109572,
+      "learning_rate": 3.0774055425974737e-06,
+      "loss": 0.8769,
+      "step": 5034
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8606934636640169,
+      "learning_rate": 3.073918464701978e-06,
+      "loss": 0.3375,
+      "step": 5035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4968595300384728,
+      "learning_rate": 3.0704330047398233e-06,
+      "loss": 0.8324,
+      "step": 5036
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5703071753093727,
+      "learning_rate": 3.066949163525205e-06,
+      "loss": 0.7868,
+      "step": 5037
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6708134574393665,
+      "learning_rate": 3.063466941871952e-06,
+      "loss": 0.8006,
+      "step": 5038
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4595164971708143,
+      "learning_rate": 3.059986340593514e-06,
+      "loss": 0.8248,
+      "step": 5039
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8799428612374821,
+      "learning_rate": 3.056507360502957e-06,
+      "loss": 0.3184,
+      "step": 5040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.852615578048221,
+      "learning_rate": 3.053030002412972e-06,
+      "loss": 0.3233,
+      "step": 5041
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6089103963264473,
+      "learning_rate": 3.0495542671358745e-06,
+      "loss": 0.8355,
+      "step": 5042
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5634084444855452,
+      "learning_rate": 3.0460801554835894e-06,
+      "loss": 0.798,
+      "step": 5043
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6761728153835087,
+      "learning_rate": 3.042607668267672e-06,
+      "loss": 0.8607,
+      "step": 5044
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4439379044967995,
+      "learning_rate": 3.0391368062992976e-06,
+      "loss": 0.9213,
+      "step": 5045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4837084064000567,
+      "learning_rate": 3.035667570389258e-06,
+      "loss": 0.8277,
+      "step": 5046
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4727755847309505,
+      "learning_rate": 3.0321999613479668e-06,
+      "loss": 0.7724,
+      "step": 5047
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5219994898447782,
+      "learning_rate": 3.028733979985463e-06,
+      "loss": 0.8173,
+      "step": 5048
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4186030695155631,
+      "learning_rate": 3.0252696271113934e-06,
+      "loss": 0.8002,
+      "step": 5049
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3888898174130462,
+      "learning_rate": 3.0218069035350328e-06,
+      "loss": 0.8786,
+      "step": 5050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4197482543381839,
+      "learning_rate": 3.0183458100652752e-06,
+      "loss": 0.8537,
+      "step": 5051
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4045806897728466,
+      "learning_rate": 3.0148863475106315e-06,
+      "loss": 0.7819,
+      "step": 5052
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5951332104091749,
+      "learning_rate": 3.011428516679232e-06,
+      "loss": 0.8807,
+      "step": 5053
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4687694702988459,
+      "learning_rate": 3.007972318378829e-06,
+      "loss": 0.7757,
+      "step": 5054
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5007927235368792,
+      "learning_rate": 3.004517753416791e-06,
+      "loss": 0.8562,
+      "step": 5055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3634559360939043,
+      "learning_rate": 3.0010648226001004e-06,
+      "loss": 0.8298,
+      "step": 5056
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5942653075376565,
+      "learning_rate": 2.9976135267353636e-06,
+      "loss": 0.8789,
+      "step": 5057
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5196176653906448,
+      "learning_rate": 2.994163866628805e-06,
+      "loss": 0.7989,
+      "step": 5058
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.461773155903358,
+      "learning_rate": 2.990715843086265e-06,
+      "loss": 0.8552,
+      "step": 5059
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6170273312204086,
+      "learning_rate": 2.9872694569132022e-06,
+      "loss": 0.8673,
+      "step": 5060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5402932885493266,
+      "learning_rate": 2.9838247089146956e-06,
+      "loss": 0.9398,
+      "step": 5061
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3473375558661,
+      "learning_rate": 2.9803815998954334e-06,
+      "loss": 0.8014,
+      "step": 5062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.450757505960035,
+      "learning_rate": 2.976940130659729e-06,
+      "loss": 0.7577,
+      "step": 5063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.364956507899819,
+      "learning_rate": 2.9735003020115095e-06,
+      "loss": 0.7797,
+      "step": 5064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5904351482979222,
+      "learning_rate": 2.9700621147543194e-06,
+      "loss": 0.8334,
+      "step": 5065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.347287987412576,
+      "learning_rate": 2.9666255696913195e-06,
+      "loss": 0.7295,
+      "step": 5066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5721416993200201,
+      "learning_rate": 2.9631906676252865e-06,
+      "loss": 0.7833,
+      "step": 5067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.438067685640525,
+      "learning_rate": 2.959757409358619e-06,
+      "loss": 0.8828,
+      "step": 5068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4346065594564978,
+      "learning_rate": 2.9563257956933177e-06,
+      "loss": 0.8692,
+      "step": 5069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3165758742443843,
+      "learning_rate": 2.9528958274310126e-06,
+      "loss": 0.8155,
+      "step": 5070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4585942457879668,
+      "learning_rate": 2.949467505372945e-06,
+      "loss": 0.9053,
+      "step": 5071
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5773018422825356,
+      "learning_rate": 2.9460408303199696e-06,
+      "loss": 0.7672,
+      "step": 5072
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8638157887523645,
+      "learning_rate": 2.9426158030725592e-06,
+      "loss": 0.833,
+      "step": 5073
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4726185945609227,
+      "learning_rate": 2.9391924244308046e-06,
+      "loss": 0.7475,
+      "step": 5074
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4897454354613884,
+      "learning_rate": 2.9357706951943987e-06,
+      "loss": 0.8107,
+      "step": 5075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5411085993975648,
+      "learning_rate": 2.9323506161626613e-06,
+      "loss": 0.7344,
+      "step": 5076
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4441299536622616,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.925,
+      "step": 5077
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6700704831929407,
+      "learning_rate": 2.925515411908535e-06,
+      "loss": 0.808,
+      "step": 5078
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6239075805756247,
+      "learning_rate": 2.9221002882828486e-06,
+      "loss": 0.7671,
+      "step": 5079
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.631084893675169,
+      "learning_rate": 2.9186868180552396e-06,
+      "loss": 0.8497,
+      "step": 5080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4432021560009018,
+      "learning_rate": 2.915275002023099e-06,
+      "loss": 0.8866,
+      "step": 5081
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5917419792725953,
+      "learning_rate": 2.9118648409834205e-06,
+      "loss": 0.8394,
+      "step": 5082
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4742628738843244,
+      "learning_rate": 2.908456335732821e-06,
+      "loss": 0.8454,
+      "step": 5083
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.398670749818334,
+      "learning_rate": 2.905049487067528e-06,
+      "loss": 0.7637,
+      "step": 5084
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.539167197292273,
+      "learning_rate": 2.901644295783381e-06,
+      "loss": 0.7946,
+      "step": 5085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4893290093084923,
+      "learning_rate": 2.8982407626758348e-06,
+      "loss": 0.8733,
+      "step": 5086
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7945640541251401,
+      "learning_rate": 2.894838888539957e-06,
+      "loss": 0.3267,
+      "step": 5087
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4855221770122016,
+      "learning_rate": 2.891438674170419e-06,
+      "loss": 0.815,
+      "step": 5088
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3375040644456926,
+      "learning_rate": 2.888040120361515e-06,
+      "loss": 0.8673,
+      "step": 5089
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4070560953000169,
+      "learning_rate": 2.884643227907147e-06,
+      "loss": 0.8149,
+      "step": 5090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5364650165821996,
+      "learning_rate": 2.88124799760083e-06,
+      "loss": 0.8326,
+      "step": 5091
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5013267844391118,
+      "learning_rate": 2.8778544302356904e-06,
+      "loss": 0.8663,
+      "step": 5092
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.445383067599179,
+      "learning_rate": 2.8744625266044647e-06,
+      "loss": 0.8959,
+      "step": 5093
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.435227235254386,
+      "learning_rate": 2.8710722874995066e-06,
+      "loss": 0.7043,
+      "step": 5094
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4574933584856937,
+      "learning_rate": 2.8676837137127687e-06,
+      "loss": 0.828,
+      "step": 5095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.49241878055303,
+      "learning_rate": 2.8642968060358256e-06,
+      "loss": 0.7665,
+      "step": 5096
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4802101722200727,
+      "learning_rate": 2.8609115652598595e-06,
+      "loss": 0.7698,
+      "step": 5097
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4690246223836267,
+      "learning_rate": 2.8575279921756637e-06,
+      "loss": 0.8466,
+      "step": 5098
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5907910908075324,
+      "learning_rate": 2.8541460875736403e-06,
+      "loss": 0.8175,
+      "step": 5099
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3758410095306948,
+      "learning_rate": 2.8507658522438054e-06,
+      "loss": 0.8449,
+      "step": 5100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4552158911343096,
+      "learning_rate": 2.8473872869757768e-06,
+      "loss": 0.7662,
+      "step": 5101
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.396666607604129,
+      "learning_rate": 2.8440103925587904e-06,
+      "loss": 0.7347,
+      "step": 5102
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5318893799511526,
+      "learning_rate": 2.840635169781688e-06,
+      "loss": 0.7885,
+      "step": 5103
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4432453980308066,
+      "learning_rate": 2.837261619432925e-06,
+      "loss": 0.7513,
+      "step": 5104
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7019402709888787,
+      "learning_rate": 2.8338897423005607e-06,
+      "loss": 0.7712,
+      "step": 5105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3952607709665923,
+      "learning_rate": 2.8305195391722663e-06,
+      "loss": 0.835,
+      "step": 5106
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5190263120056213,
+      "learning_rate": 2.8271510108353237e-06,
+      "loss": 0.8781,
+      "step": 5107
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6806913303820863,
+      "learning_rate": 2.823784158076618e-06,
+      "loss": 0.7667,
+      "step": 5108
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4422855408966526,
+      "learning_rate": 2.820418981682648e-06,
+      "loss": 0.7727,
+      "step": 5109
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4882226191508583,
+      "learning_rate": 2.817055482439518e-06,
+      "loss": 0.816,
+      "step": 5110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4436284848058087,
+      "learning_rate": 2.8136936611329436e-06,
+      "loss": 0.7696,
+      "step": 5111
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3164923834991495,
+      "learning_rate": 2.810333518548246e-06,
+      "loss": 0.8996,
+      "step": 5112
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6070138765748827,
+      "learning_rate": 2.806975055470358e-06,
+      "loss": 0.8524,
+      "step": 5113
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4785965301469797,
+      "learning_rate": 2.803618272683809e-06,
+      "loss": 0.8403,
+      "step": 5114
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.499030225036364,
+      "learning_rate": 2.80026317097275e-06,
+      "loss": 0.8636,
+      "step": 5115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4707822651739537,
+      "learning_rate": 2.796909751120931e-06,
+      "loss": 0.7879,
+      "step": 5116
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4997675002731632,
+      "learning_rate": 2.7935580139117114e-06,
+      "loss": 0.8177,
+      "step": 5117
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.458039115421537,
+      "learning_rate": 2.7902079601280583e-06,
+      "loss": 0.9224,
+      "step": 5118
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2851376148078655,
+      "learning_rate": 2.7868595905525464e-06,
+      "loss": 0.837,
+      "step": 5119
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5736676747159453,
+      "learning_rate": 2.7835129059673496e-06,
+      "loss": 0.8696,
+      "step": 5120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5999291732597782,
+      "learning_rate": 2.7801679071542576e-06,
+      "loss": 0.8646,
+      "step": 5121
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4416532169220666,
+      "learning_rate": 2.7768245948946615e-06,
+      "loss": 0.8669,
+      "step": 5122
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9014195585070164,
+      "learning_rate": 2.7734829699695585e-06,
+      "loss": 0.3194,
+      "step": 5123
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5242080097123147,
+      "learning_rate": 2.770143033159555e-06,
+      "loss": 0.741,
+      "step": 5124
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4625193006207853,
+      "learning_rate": 2.766804785244859e-06,
+      "loss": 0.8942,
+      "step": 5125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5477927639018938,
+      "learning_rate": 2.763468227005288e-06,
+      "loss": 0.7551,
+      "step": 5126
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4738246408619002,
+      "learning_rate": 2.7601333592202583e-06,
+      "loss": 0.7635,
+      "step": 5127
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.866234631735783,
+      "learning_rate": 2.756800182668796e-06,
+      "loss": 0.374,
+      "step": 5128
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.559274426831016,
+      "learning_rate": 2.7534686981295335e-06,
+      "loss": 0.9176,
+      "step": 5129
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.408270645110137,
+      "learning_rate": 2.7501389063807037e-06,
+      "loss": 0.768,
+      "step": 5130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6788602843544644,
+      "learning_rate": 2.746810808200152e-06,
+      "loss": 0.7975,
+      "step": 5131
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.494269935148476,
+      "learning_rate": 2.743484404365314e-06,
+      "loss": 0.8175,
+      "step": 5132
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.497123785737367,
+      "learning_rate": 2.740159695653243e-06,
+      "loss": 0.8339,
+      "step": 5133
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.573388340919819,
+      "learning_rate": 2.7368366828405892e-06,
+      "loss": 0.8561,
+      "step": 5134
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3587759331226101,
+      "learning_rate": 2.7335153667036106e-06,
+      "loss": 0.866,
+      "step": 5135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8963988622800161,
+      "learning_rate": 2.7301957480181697e-06,
+      "loss": 0.3358,
+      "step": 5136
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4595957755480402,
+      "learning_rate": 2.7268778275597217e-06,
+      "loss": 0.8431,
+      "step": 5137
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.424294375124013,
+      "learning_rate": 2.7235616061033388e-06,
+      "loss": 0.7391,
+      "step": 5138
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3765513523843336,
+      "learning_rate": 2.7202470844236896e-06,
+      "loss": 0.7617,
+      "step": 5139
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.402381045740778,
+      "learning_rate": 2.7169342632950466e-06,
+      "loss": 0.8732,
+      "step": 5140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.651675469054545,
+      "learning_rate": 2.7136231434912854e-06,
+      "loss": 0.8727,
+      "step": 5141
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4181967333955,
+      "learning_rate": 2.7103137257858867e-06,
+      "loss": 0.8236,
+      "step": 5142
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4509007314845606,
+      "learning_rate": 2.707006010951925e-06,
+      "loss": 0.7604,
+      "step": 5143
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4976815508106767,
+      "learning_rate": 2.703699999762085e-06,
+      "loss": 0.8438,
+      "step": 5144
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3890570376023808,
+      "learning_rate": 2.7003956929886523e-06,
+      "loss": 0.8297,
+      "step": 5145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5252838903345383,
+      "learning_rate": 2.6970930914035134e-06,
+      "loss": 0.859,
+      "step": 5146
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3687283116847584,
+      "learning_rate": 2.6937921957781587e-06,
+      "loss": 0.8266,
+      "step": 5147
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4690525156816665,
+      "learning_rate": 2.690493006883671e-06,
+      "loss": 0.8691,
+      "step": 5148
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.428916280178367,
+      "learning_rate": 2.6871955254907455e-06,
+      "loss": 0.7763,
+      "step": 5149
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8549993759103848,
+      "learning_rate": 2.683899752369674e-06,
+      "loss": 0.3284,
+      "step": 5150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0003487786199141,
+      "learning_rate": 2.68060568829035e-06,
+      "loss": 0.3414,
+      "step": 5151
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.370832645926139,
+      "learning_rate": 2.6773133340222677e-06,
+      "loss": 0.8533,
+      "step": 5152
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.495586436241171,
+      "learning_rate": 2.674022690334518e-06,
+      "loss": 0.7907,
+      "step": 5153
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6276951697022042,
+      "learning_rate": 2.6707337579957983e-06,
+      "loss": 0.8283,
+      "step": 5154
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4199171059645121,
+      "learning_rate": 2.667446537774402e-06,
+      "loss": 0.8223,
+      "step": 5155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5410516943777657,
+      "learning_rate": 2.664161030438225e-06,
+      "loss": 0.8104,
+      "step": 5156
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4454646777894695,
+      "learning_rate": 2.660877236754762e-06,
+      "loss": 0.7811,
+      "step": 5157
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3515720656502974,
+      "learning_rate": 2.657595157491111e-06,
+      "loss": 0.8302,
+      "step": 5158
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4708979326894904,
+      "learning_rate": 2.654314793413959e-06,
+      "loss": 0.8698,
+      "step": 5159
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4599677975598686,
+      "learning_rate": 2.6510361452896038e-06,
+      "loss": 0.8004,
+      "step": 5160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.506253704189162,
+      "learning_rate": 2.6477592138839357e-06,
+      "loss": 0.8241,
+      "step": 5161
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.415136095699315,
+      "learning_rate": 2.6444839999624496e-06,
+      "loss": 0.7332,
+      "step": 5162
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4388816830176892,
+      "learning_rate": 2.6412105042902327e-06,
+      "loss": 0.8909,
+      "step": 5163
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6125252675046966,
+      "learning_rate": 2.6379387276319757e-06,
+      "loss": 0.7778,
+      "step": 5164
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.428628587989547,
+      "learning_rate": 2.6346686707519675e-06,
+      "loss": 0.7432,
+      "step": 5165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6225580323936093,
+      "learning_rate": 2.631400334414089e-06,
+      "loss": 0.8191,
+      "step": 5166
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.436347071481445,
+      "learning_rate": 2.6281337193818267e-06,
+      "loss": 0.7627,
+      "step": 5167
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3116512534546898,
+      "learning_rate": 2.624868826418262e-06,
+      "loss": 0.7553,
+      "step": 5168
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5899296275758623,
+      "learning_rate": 2.6216056562860746e-06,
+      "loss": 0.7602,
+      "step": 5169
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4261176809767078,
+      "learning_rate": 2.6183442097475398e-06,
+      "loss": 0.801,
+      "step": 5170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3800929706572926,
+      "learning_rate": 2.6150844875645364e-06,
+      "loss": 0.7783,
+      "step": 5171
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6495394684461355,
+      "learning_rate": 2.61182649049853e-06,
+      "loss": 0.8392,
+      "step": 5172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5925889519385894,
+      "learning_rate": 2.6085702193105913e-06,
+      "loss": 0.9093,
+      "step": 5173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.495658535440884,
+      "learning_rate": 2.605315674761385e-06,
+      "loss": 0.9176,
+      "step": 5174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4187953236982287,
+      "learning_rate": 2.602062857611174e-06,
+      "loss": 0.8022,
+      "step": 5175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.408316648671769,
+      "learning_rate": 2.5988117686198167e-06,
+      "loss": 0.7984,
+      "step": 5176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5680440434879845,
+      "learning_rate": 2.59556240854677e-06,
+      "loss": 0.8152,
+      "step": 5177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5053090457953842,
+      "learning_rate": 2.5923147781510795e-06,
+      "loss": 0.797,
+      "step": 5178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3465570887804141,
+      "learning_rate": 2.5890688781913954e-06,
+      "loss": 0.8021,
+      "step": 5179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5141151868089557,
+      "learning_rate": 2.585824709425958e-06,
+      "loss": 0.9689,
+      "step": 5180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6343181970856544,
+      "learning_rate": 2.5825822726126095e-06,
+      "loss": 0.741,
+      "step": 5181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.379115861408794,
+      "learning_rate": 2.5793415685087797e-06,
+      "loss": 0.8026,
+      "step": 5182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.446645431349882,
+      "learning_rate": 2.576102597871498e-06,
+      "loss": 0.762,
+      "step": 5183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4875574360555688,
+      "learning_rate": 2.5728653614573927e-06,
+      "loss": 0.7581,
+      "step": 5184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5517246123952102,
+      "learning_rate": 2.5696298600226766e-06,
+      "loss": 0.8235,
+      "step": 5185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4999064078537652,
+      "learning_rate": 2.566396094323165e-06,
+      "loss": 0.8537,
+      "step": 5186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4835738655312591,
+      "learning_rate": 2.5631640651142654e-06,
+      "loss": 0.7247,
+      "step": 5187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4463761398096484,
+      "learning_rate": 2.559933773150982e-06,
+      "loss": 0.8275,
+      "step": 5188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4602632828289412,
+      "learning_rate": 2.5567052191879104e-06,
+      "loss": 0.8509,
+      "step": 5189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4387524315018494,
+      "learning_rate": 2.5534784039792437e-06,
+      "loss": 0.7542,
+      "step": 5190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3635582970291584,
+      "learning_rate": 2.550253328278761e-06,
+      "loss": 0.804,
+      "step": 5191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2054497070192842,
+      "learning_rate": 2.5470299928398424e-06,
+      "loss": 0.8303,
+      "step": 5192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.285101037158337,
+      "learning_rate": 2.5438083984154606e-06,
+      "loss": 0.809,
+      "step": 5193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3858796677194265,
+      "learning_rate": 2.5405885457581793e-06,
+      "loss": 0.7714,
+      "step": 5194
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5169228055727215,
+      "learning_rate": 2.537370435620157e-06,
+      "loss": 0.7131,
+      "step": 5195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.538157340920337,
+      "learning_rate": 2.534154068753144e-06,
+      "loss": 0.8513,
+      "step": 5196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5409136558321017,
+      "learning_rate": 2.5309394459084878e-06,
+      "loss": 0.905,
+      "step": 5197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3537277369240417,
+      "learning_rate": 2.527726567837118e-06,
+      "loss": 0.8278,
+      "step": 5198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3521318008433085,
+      "learning_rate": 2.524515435289566e-06,
+      "loss": 0.8742,
+      "step": 5199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6118387465155175,
+      "learning_rate": 2.5213060490159536e-06,
+      "loss": 0.764,
+      "step": 5200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.350296708759324,
+      "learning_rate": 2.5180984097659924e-06,
+      "loss": 0.8,
+      "step": 5201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9074766211718266,
+      "learning_rate": 2.514892518288988e-06,
+      "loss": 0.318,
+      "step": 5202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2675993429738672,
+      "learning_rate": 2.511688375333842e-06,
+      "loss": 0.8022,
+      "step": 5203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6910893229222446,
+      "learning_rate": 2.5084859816490327e-06,
+      "loss": 0.8335,
+      "step": 5204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5454597820732972,
+      "learning_rate": 2.505285337982644e-06,
+      "loss": 0.7369,
+      "step": 5205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4088497899234877,
+      "learning_rate": 2.5020864450823477e-06,
+      "loss": 0.82,
+      "step": 5206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9735689390957943,
+      "learning_rate": 2.4988893036954045e-06,
+      "loss": 0.3349,
+      "step": 5207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.257075008397669,
+      "learning_rate": 2.4956939145686677e-06,
+      "loss": 0.8491,
+      "step": 5208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3366961353814741,
+      "learning_rate": 2.4925002784485796e-06,
+      "loss": 0.8911,
+      "step": 5209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.483889845568321,
+      "learning_rate": 2.489308396081178e-06,
+      "loss": 0.7483,
+      "step": 5210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.667345769867752,
+      "learning_rate": 2.48611826821208e-06,
+      "loss": 0.8605,
+      "step": 5211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6191098379376316,
+      "learning_rate": 2.4829298955865022e-06,
+      "loss": 0.8575,
+      "step": 5212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3941262429078851,
+      "learning_rate": 2.4797432789492506e-06,
+      "loss": 0.8482,
+      "step": 5213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5854755924570065,
+      "learning_rate": 2.476558419044718e-06,
+      "loss": 0.8346,
+      "step": 5214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3690406610453252,
+      "learning_rate": 2.4733753166168883e-06,
+      "loss": 0.7884,
+      "step": 5215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5323899439106292,
+      "learning_rate": 2.470193972409337e-06,
+      "loss": 0.8227,
+      "step": 5216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3838070664348467,
+      "learning_rate": 2.467014387165222e-06,
+      "loss": 0.7963,
+      "step": 5217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3533578391259928,
+      "learning_rate": 2.4638365616272952e-06,
+      "loss": 0.8548,
+      "step": 5218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4816083900048744,
+      "learning_rate": 2.460660496537899e-06,
+      "loss": 0.8006,
+      "step": 5219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9892379235304632,
+      "learning_rate": 2.4574861926389615e-06,
+      "loss": 0.3097,
+      "step": 5220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8787789305551243,
+      "learning_rate": 2.454313650672001e-06,
+      "loss": 0.3708,
+      "step": 5221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.380559118396946,
+      "learning_rate": 2.451142871378124e-06,
+      "loss": 0.7665,
+      "step": 5222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.631606092651518,
+      "learning_rate": 2.447973855498027e-06,
+      "loss": 0.8337,
+      "step": 5223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5255623267782115,
+      "learning_rate": 2.4448066037719865e-06,
+      "loss": 0.8143,
+      "step": 5224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2901481456682227,
+      "learning_rate": 2.4416411169398755e-06,
+      "loss": 0.8573,
+      "step": 5225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.371585847299984,
+      "learning_rate": 2.4384773957411533e-06,
+      "loss": 0.7782,
+      "step": 5226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4576708307762685,
+      "learning_rate": 2.4353154409148637e-06,
+      "loss": 0.8154,
+      "step": 5227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8625181709864146,
+      "learning_rate": 2.432155253199642e-06,
+      "loss": 0.3304,
+      "step": 5228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4268532491954804,
+      "learning_rate": 2.42899683333371e-06,
+      "loss": 0.7975,
+      "step": 5229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7498593891227354,
+      "learning_rate": 2.4258401820548682e-06,
+      "loss": 0.81,
+      "step": 5230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8353925158433698,
+      "learning_rate": 2.4226853001005146e-06,
+      "loss": 0.3,
+      "step": 5231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.472565166859187,
+      "learning_rate": 2.4195321882076295e-06,
+      "loss": 0.8038,
+      "step": 5232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6878072163181843,
+      "learning_rate": 2.4163808471127815e-06,
+      "loss": 0.7141,
+      "step": 5233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6696531085069557,
+      "learning_rate": 2.413231277552122e-06,
+      "loss": 0.7578,
+      "step": 5234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5066712475494022,
+      "learning_rate": 2.410083480261395e-06,
+      "loss": 0.7178,
+      "step": 5235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.476737938530119,
+      "learning_rate": 2.4069374559759207e-06,
+      "loss": 0.7588,
+      "step": 5236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5684129870884418,
+      "learning_rate": 2.4037932054306125e-06,
+      "loss": 0.839,
+      "step": 5237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3662739030944941,
+      "learning_rate": 2.400650729359969e-06,
+      "loss": 0.775,
+      "step": 5238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5272639783042223,
+      "learning_rate": 2.3975100284980713e-06,
+      "loss": 0.8904,
+      "step": 5239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4701769731117003,
+      "learning_rate": 2.394371103578589e-06,
+      "loss": 0.7936,
+      "step": 5240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8338514340417672,
+      "learning_rate": 2.3912339553347742e-06,
+      "loss": 0.311,
+      "step": 5241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3105285724265792,
+      "learning_rate": 2.3880985844994674e-06,
+      "loss": 0.8007,
+      "step": 5242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6320425498602253,
+      "learning_rate": 2.384964991805089e-06,
+      "loss": 0.8204,
+      "step": 5243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5809349853554442,
+      "learning_rate": 2.3818331779836447e-06,
+      "loss": 0.7954,
+      "step": 5244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4669171282749138,
+      "learning_rate": 2.3787031437667307e-06,
+      "loss": 0.9086,
+      "step": 5245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5352490520819126,
+      "learning_rate": 2.37557488988552e-06,
+      "loss": 0.8549,
+      "step": 5246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2499745634590351,
+      "learning_rate": 2.372448417070776e-06,
+      "loss": 0.7744,
+      "step": 5247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5954549075762638,
+      "learning_rate": 2.3693237260528436e-06,
+      "loss": 0.8664,
+      "step": 5248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3293878231609324,
+      "learning_rate": 2.366200817561647e-06,
+      "loss": 0.8828,
+      "step": 5249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4326423728616928,
+      "learning_rate": 2.3630796923266997e-06,
+      "loss": 0.8084,
+      "step": 5250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4560243550959084,
+      "learning_rate": 2.3599603510770962e-06,
+      "loss": 0.9566,
+      "step": 5251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.510083149119049,
+      "learning_rate": 2.3568427945415163e-06,
+      "loss": 0.7854,
+      "step": 5252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5400789252076748,
+      "learning_rate": 2.3537270234482213e-06,
+      "loss": 0.8988,
+      "step": 5253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5527720716686881,
+      "learning_rate": 2.350613038525058e-06,
+      "loss": 0.9005,
+      "step": 5254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2622735063177308,
+      "learning_rate": 2.347500840499447e-06,
+      "loss": 0.8947,
+      "step": 5255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5993420350708767,
+      "learning_rate": 2.3443904300984034e-06,
+      "loss": 0.879,
+      "step": 5256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.477986581250986,
+      "learning_rate": 2.3412818080485176e-06,
+      "loss": 0.7646,
+      "step": 5257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4935451605038887,
+      "learning_rate": 2.3381749750759643e-06,
+      "loss": 0.7298,
+      "step": 5258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5313736345048246,
+      "learning_rate": 2.335069931906503e-06,
+      "loss": 0.7812,
+      "step": 5259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5945991236264816,
+      "learning_rate": 2.331966679265467e-06,
+      "loss": 0.8085,
+      "step": 5260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3749900459882924,
+      "learning_rate": 2.3288652178777783e-06,
+      "loss": 0.8377,
+      "step": 5261
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.387328957902974,
+      "learning_rate": 2.3257655484679376e-06,
+      "loss": 0.8335,
+      "step": 5262
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2039454576295368,
+      "learning_rate": 2.3226676717600303e-06,
+      "loss": 0.7619,
+      "step": 5263
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3853506734165646,
+      "learning_rate": 2.3195715884777203e-06,
+      "loss": 0.8254,
+      "step": 5264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2965391899145648,
+      "learning_rate": 2.316477299344254e-06,
+      "loss": 0.8019,
+      "step": 5265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9064926046763753,
+      "learning_rate": 2.3133848050824536e-06,
+      "loss": 0.3247,
+      "step": 5266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4082301126663817,
+      "learning_rate": 2.3102941064147287e-06,
+      "loss": 0.7995,
+      "step": 5267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6248746087681787,
+      "learning_rate": 2.3072052040630666e-06,
+      "loss": 0.8671,
+      "step": 5268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3629218163259866,
+      "learning_rate": 2.3041180987490354e-06,
+      "loss": 0.7719,
+      "step": 5269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4157957129881158,
+      "learning_rate": 2.3010327911937856e-06,
+      "loss": 0.8142,
+      "step": 5270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5878835480873652,
+      "learning_rate": 2.2979492821180394e-06,
+      "loss": 0.7817,
+      "step": 5271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5204970072152255,
+      "learning_rate": 2.2948675722421086e-06,
+      "loss": 0.893,
+      "step": 5272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.505893533717531,
+      "learning_rate": 2.2917876622858814e-06,
+      "loss": 0.7169,
+      "step": 5273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4025162488470824,
+      "learning_rate": 2.288709552968823e-06,
+      "loss": 0.8427,
+      "step": 5274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.37039198332275,
+      "learning_rate": 2.285633245009984e-06,
+      "loss": 0.8525,
+      "step": 5275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6165986290472205,
+      "learning_rate": 2.2825587391279857e-06,
+      "loss": 0.8273,
+      "step": 5276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4875401293626862,
+      "learning_rate": 2.279486036041034e-06,
+      "loss": 0.8569,
+      "step": 5277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4496691362536043,
+      "learning_rate": 2.276415136466913e-06,
+      "loss": 0.7318,
+      "step": 5278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2522938185915398,
+      "learning_rate": 2.273346041122987e-06,
+      "loss": 0.7418,
+      "step": 5279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3592608817749643,
+      "learning_rate": 2.270278750726194e-06,
+      "loss": 0.8044,
+      "step": 5280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3956000475771186,
+      "learning_rate": 2.267213265993058e-06,
+      "loss": 0.9067,
+      "step": 5281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6103565656677958,
+      "learning_rate": 2.2641495876396713e-06,
+      "loss": 0.8752,
+      "step": 5282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4045429133576957,
+      "learning_rate": 2.2610877163817113e-06,
+      "loss": 0.822,
+      "step": 5283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2697186999216175,
+      "learning_rate": 2.2580276529344312e-06,
+      "loss": 0.8148,
+      "step": 5284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4721477791839408,
+      "learning_rate": 2.254969398012663e-06,
+      "loss": 0.8856,
+      "step": 5285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.358093066277862,
+      "learning_rate": 2.2519129523308146e-06,
+      "loss": 0.8148,
+      "step": 5286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3097182267848255,
+      "learning_rate": 2.2488583166028754e-06,
+      "loss": 0.85,
+      "step": 5287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6008923320524924,
+      "learning_rate": 2.2458054915424033e-06,
+      "loss": 0.8212,
+      "step": 5288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5377344741108048,
+      "learning_rate": 2.242754477862541e-06,
+      "loss": 0.9197,
+      "step": 5289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3935879968673492,
+      "learning_rate": 2.239705276276004e-06,
+      "loss": 0.6686,
+      "step": 5290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4539789976955533,
+      "learning_rate": 2.2366578874950894e-06,
+      "loss": 0.8263,
+      "step": 5291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4132237100301375,
+      "learning_rate": 2.2336123122316642e-06,
+      "loss": 0.8171,
+      "step": 5292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4586722526510985,
+      "learning_rate": 2.2305685511971777e-06,
+      "loss": 0.7955,
+      "step": 5293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3740479406789656,
+      "learning_rate": 2.2275266051026533e-06,
+      "loss": 0.8507,
+      "step": 5294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3185802465432919,
+      "learning_rate": 2.2244864746586858e-06,
+      "loss": 0.7254,
+      "step": 5295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5739235116017158,
+      "learning_rate": 2.2214481605754532e-06,
+      "loss": 0.8483,
+      "step": 5296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4799101405989241,
+      "learning_rate": 2.218411663562704e-06,
+      "loss": 0.8058,
+      "step": 5297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4417786188295285,
+      "learning_rate": 2.215376984329767e-06,
+      "loss": 0.7945,
+      "step": 5298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4823076407706,
+      "learning_rate": 2.2123441235855404e-06,
+      "loss": 0.7885,
+      "step": 5299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.409585830784883,
+      "learning_rate": 2.2093130820385057e-06,
+      "loss": 0.9087,
+      "step": 5300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3536009300620115,
+      "learning_rate": 2.2062838603967086e-06,
+      "loss": 0.7439,
+      "step": 5301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6376922723404126,
+      "learning_rate": 2.2032564593677773e-06,
+      "loss": 0.8306,
+      "step": 5302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6286567421425713,
+      "learning_rate": 2.2002308796589146e-06,
+      "loss": 0.8224,
+      "step": 5303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3952200274512352,
+      "learning_rate": 2.197207121976895e-06,
+      "loss": 0.8228,
+      "step": 5304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5100682857093606,
+      "learning_rate": 2.1941851870280697e-06,
+      "loss": 0.8633,
+      "step": 5305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4221243770650307,
+      "learning_rate": 2.1911650755183646e-06,
+      "loss": 0.8321,
+      "step": 5306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.462208790142082,
+      "learning_rate": 2.1881467881532737e-06,
+      "loss": 0.7677,
+      "step": 5307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4424789373436055,
+      "learning_rate": 2.18513032563787e-06,
+      "loss": 0.7192,
+      "step": 5308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4481557118018122,
+      "learning_rate": 2.1821156886768013e-06,
+      "loss": 0.8455,
+      "step": 5309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3664316424469038,
+      "learning_rate": 2.179102877974287e-06,
+      "loss": 0.7968,
+      "step": 5310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6382194725462393,
+      "learning_rate": 2.1760918942341193e-06,
+      "loss": 0.8453,
+      "step": 5311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3278957112921814,
+      "learning_rate": 2.1730827381596643e-06,
+      "loss": 0.791,
+      "step": 5312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.549786987636284,
+      "learning_rate": 2.1700754104538645e-06,
+      "loss": 0.8028,
+      "step": 5313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1473408579159785,
+      "learning_rate": 2.167069911819225e-06,
+      "loss": 0.8213,
+      "step": 5314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.382246412771008,
+      "learning_rate": 2.164066242957836e-06,
+      "loss": 0.8631,
+      "step": 5315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8344783487008602,
+      "learning_rate": 2.1610644045713525e-06,
+      "loss": 0.3076,
+      "step": 5316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5441011492544239,
+      "learning_rate": 2.158064397361005e-06,
+      "loss": 0.8532,
+      "step": 5317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.950255354839666,
+      "learning_rate": 2.1550662220275955e-06,
+      "loss": 0.3057,
+      "step": 5318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4853759784936431,
+      "learning_rate": 2.152069879271501e-06,
+      "loss": 0.8486,
+      "step": 5319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4454438136993746,
+      "learning_rate": 2.1490753697926613e-06,
+      "loss": 0.8341,
+      "step": 5320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4622404209473119,
+      "learning_rate": 2.1460826942905977e-06,
+      "loss": 0.8743,
+      "step": 5321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3797156454716968,
+      "learning_rate": 2.1430918534643996e-06,
+      "loss": 0.7598,
+      "step": 5322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5968018872923193,
+      "learning_rate": 2.140102848012726e-06,
+      "loss": 0.8278,
+      "step": 5323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7060098238102301,
+      "learning_rate": 2.1371156786338108e-06,
+      "loss": 0.3128,
+      "step": 5324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3758143227136712,
+      "learning_rate": 2.134130346025457e-06,
+      "loss": 0.8858,
+      "step": 5325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8866841618266074,
+      "learning_rate": 2.13114685088504e-06,
+      "loss": 0.3158,
+      "step": 5326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.618890873061229,
+      "learning_rate": 2.1281651939094996e-06,
+      "loss": 0.7638,
+      "step": 5327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8191641502751412,
+      "learning_rate": 2.1251853757953546e-06,
+      "loss": 0.8254,
+      "step": 5328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.406772042438374,
+      "learning_rate": 2.1222073972386903e-06,
+      "loss": 0.7824,
+      "step": 5329
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9188725734587385,
+      "learning_rate": 2.1192312589351626e-06,
+      "loss": 0.3072,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.550843961826565,
+      "learning_rate": 2.1162569615799978e-06,
+      "loss": 0.8175,
+      "step": 5331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5359053052232876,
+      "learning_rate": 2.1132845058679942e-06,
+      "loss": 0.8831,
+      "step": 5332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3691604942044595,
+      "learning_rate": 2.110313892493514e-06,
+      "loss": 0.7478,
+      "step": 5333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.879156354710742,
+      "learning_rate": 2.107345122150495e-06,
+      "loss": 0.3026,
+      "step": 5334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5930274302444687,
+      "learning_rate": 2.1043781955324406e-06,
+      "loss": 0.8315,
+      "step": 5335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5767135817808149,
+      "learning_rate": 2.101413113332427e-06,
+      "loss": 0.8328,
+      "step": 5336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4387097891525165,
+      "learning_rate": 2.098449876243096e-06,
+      "loss": 0.8215,
+      "step": 5337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.567902221655809,
+      "learning_rate": 2.0954884849566613e-06,
+      "loss": 0.8905,
+      "step": 5338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.312496117177397,
+      "learning_rate": 2.0925289401649074e-06,
+      "loss": 0.8174,
+      "step": 5339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3971126283604443,
+      "learning_rate": 2.0895712425591776e-06,
+      "loss": 0.7667,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5511720378763254,
+      "learning_rate": 2.0866153928303947e-06,
+      "loss": 0.8443,
+      "step": 5341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.326341440515166,
+      "learning_rate": 2.083661391669043e-06,
+      "loss": 0.8235,
+      "step": 5342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3307545513156853,
+      "learning_rate": 2.0807092397651795e-06,
+      "loss": 0.8682,
+      "step": 5343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4253670642799852,
+      "learning_rate": 2.0777589378084263e-06,
+      "loss": 0.8324,
+      "step": 5344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4893130020171332,
+      "learning_rate": 2.074810486487977e-06,
+      "loss": 0.8671,
+      "step": 5345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4683105719418144,
+      "learning_rate": 2.0718638864925845e-06,
+      "loss": 0.8913,
+      "step": 5346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.485124944420155,
+      "learning_rate": 2.0689191385105787e-06,
+      "loss": 0.7797,
+      "step": 5347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4233495198308606,
+      "learning_rate": 2.0659762432298527e-06,
+      "loss": 0.7999,
+      "step": 5348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5214002510094353,
+      "learning_rate": 2.063035201337865e-06,
+      "loss": 0.8696,
+      "step": 5349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.432322488077478,
+      "learning_rate": 2.0600960135216463e-06,
+      "loss": 0.8727,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3551453168787113,
+      "learning_rate": 2.057158680467789e-06,
+      "loss": 0.8017,
+      "step": 5351
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4328855823486144,
+      "learning_rate": 2.0542232028624585e-06,
+      "loss": 0.8454,
+      "step": 5352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3750459164842304,
+      "learning_rate": 2.051289581391377e-06,
+      "loss": 0.7989,
+      "step": 5353
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4856494596060186,
+      "learning_rate": 2.04835781673984e-06,
+      "loss": 0.791,
+      "step": 5354
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3117024561544246,
+      "learning_rate": 2.04542790959271e-06,
+      "loss": 0.851,
+      "step": 5355
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4377549971515386,
+      "learning_rate": 2.0424998606344127e-06,
+      "loss": 0.8665,
+      "step": 5356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9241354861484622,
+      "learning_rate": 2.03957367054894e-06,
+      "loss": 0.2912,
+      "step": 5357
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4316737260934016,
+      "learning_rate": 2.036649340019854e-06,
+      "loss": 0.8389,
+      "step": 5358
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.443895065695472,
+      "learning_rate": 2.0337268697302738e-06,
+      "loss": 0.7649,
+      "step": 5359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4539548730739582,
+      "learning_rate": 2.0308062603628888e-06,
+      "loss": 0.8618,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5010686975171228,
+      "learning_rate": 2.0278875125999565e-06,
+      "loss": 0.7263,
+      "step": 5361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8260585406193655,
+      "learning_rate": 2.024970627123295e-06,
+      "loss": 0.3464,
+      "step": 5362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4445796919489027,
+      "learning_rate": 2.022055604614289e-06,
+      "loss": 0.7995,
+      "step": 5363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.368011901802301,
+      "learning_rate": 2.0191424457538923e-06,
+      "loss": 0.8362,
+      "step": 5364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4680967679116654,
+      "learning_rate": 2.016231151222612e-06,
+      "loss": 0.8054,
+      "step": 5365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5570133174700023,
+      "learning_rate": 2.0133217217005295e-06,
+      "loss": 0.7711,
+      "step": 5366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5520615277476053,
+      "learning_rate": 2.0104141578672887e-06,
+      "loss": 0.8724,
+      "step": 5367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5891897668906345,
+      "learning_rate": 2.0075084604020967e-06,
+      "loss": 0.8117,
+      "step": 5368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3704469913285864,
+      "learning_rate": 2.004604629983725e-06,
+      "loss": 0.8463,
+      "step": 5369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8332186817945408,
+      "learning_rate": 2.001702667290508e-06,
+      "loss": 0.8072,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.719631561783302,
+      "learning_rate": 1.998802573000348e-06,
+      "loss": 0.8008,
+      "step": 5371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4088136034876957,
+      "learning_rate": 1.9959043477907e-06,
+      "loss": 0.8846,
+      "step": 5372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4511524817812553,
+      "learning_rate": 1.9930079923385946e-06,
+      "loss": 0.8108,
+      "step": 5373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4416291799769558,
+      "learning_rate": 1.99011350732062e-06,
+      "loss": 0.8025,
+      "step": 5374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2962009627312474,
+      "learning_rate": 1.987220893412928e-06,
+      "loss": 0.7932,
+      "step": 5375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4065173464534155,
+      "learning_rate": 1.984330151291233e-06,
+      "loss": 0.8619,
+      "step": 5376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4356291974396183,
+      "learning_rate": 1.981441281630816e-06,
+      "loss": 0.716,
+      "step": 5377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.420530063030565,
+      "learning_rate": 1.978554285106512e-06,
+      "loss": 0.8435,
+      "step": 5378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.386981267661283,
+      "learning_rate": 1.975669162392726e-06,
+      "loss": 0.8503,
+      "step": 5379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3077479661038947,
+      "learning_rate": 1.972785914163423e-06,
+      "loss": 0.7825,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4674984921255925,
+      "learning_rate": 1.9699045410921303e-06,
+      "loss": 0.803,
+      "step": 5381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4783606189721945,
+      "learning_rate": 1.967025043851939e-06,
+      "loss": 0.8165,
+      "step": 5382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.471136624506106,
+      "learning_rate": 1.9641474231154956e-06,
+      "loss": 0.7785,
+      "step": 5383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4972478962810782,
+      "learning_rate": 1.9612716795550146e-06,
+      "loss": 0.7779,
+      "step": 5384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.33398224633579,
+      "learning_rate": 1.95839781384227e-06,
+      "loss": 0.8951,
+      "step": 5385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4877879045961764,
+      "learning_rate": 1.9555258266485965e-06,
+      "loss": 0.9326,
+      "step": 5386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2889626811133876,
+      "learning_rate": 1.9526557186448924e-06,
+      "loss": 0.7704,
+      "step": 5387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4819842542004975,
+      "learning_rate": 1.9497874905016156e-06,
+      "loss": 0.7892,
+      "step": 5388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.566893118420538,
+      "learning_rate": 1.9469211428887813e-06,
+      "loss": 0.8179,
+      "step": 5389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4152449845672597,
+      "learning_rate": 1.9440566764759704e-06,
+      "loss": 0.9246,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4861657396053622,
+      "learning_rate": 1.9411940919323215e-06,
+      "loss": 0.8003,
+      "step": 5391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.514813741250979,
+      "learning_rate": 1.9383333899265368e-06,
+      "loss": 0.8184,
+      "step": 5392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3701586672865653,
+      "learning_rate": 1.9354745711268763e-06,
+      "loss": 0.8368,
+      "step": 5393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3485465888843264,
+      "learning_rate": 1.9326176362011584e-06,
+      "loss": 0.8578,
+      "step": 5394
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4747462436491496,
+      "learning_rate": 1.9297625858167636e-06,
+      "loss": 0.7842,
+      "step": 5395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5091024054912812,
+      "learning_rate": 1.9269094206406326e-06,
+      "loss": 0.8427,
+      "step": 5396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5720568014752216,
+      "learning_rate": 1.9240581413392647e-06,
+      "loss": 0.8886,
+      "step": 5397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.297846175226786,
+      "learning_rate": 1.9212087485787233e-06,
+      "loss": 0.7749,
+      "step": 5398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4141657288423555,
+      "learning_rate": 1.918361243024619e-06,
+      "loss": 0.8324,
+      "step": 5399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4273951645429928,
+      "learning_rate": 1.9155156253421348e-06,
+      "loss": 0.8196,
+      "step": 5400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4874439812979077,
+      "learning_rate": 1.9126718961960056e-06,
+      "loss": 0.776,
+      "step": 5401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4351381245446067,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.7787,
+      "step": 5402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5731438725047595,
+      "learning_rate": 1.906990106169555e-06,
+      "loss": 0.8282,
+      "step": 5403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3985393622930269,
+      "learning_rate": 1.9041520466164988e-06,
+      "loss": 0.8458,
+      "step": 5404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.368068726462965,
+      "learning_rate": 1.9013158782543307e-06,
+      "loss": 0.8215,
+      "step": 5405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9142689359705787,
+      "learning_rate": 1.8984816017455798e-06,
+      "loss": 0.3243,
+      "step": 5406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5347651366817776,
+      "learning_rate": 1.8956492177523345e-06,
+      "loss": 0.803,
+      "step": 5407
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4512161072296008,
+      "learning_rate": 1.8928187269362398e-06,
+      "loss": 0.7971,
+      "step": 5408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.635393545659041,
+      "learning_rate": 1.8899901299584965e-06,
+      "loss": 0.8271,
+      "step": 5409
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.461825113896843,
+      "learning_rate": 1.8871634274798701e-06,
+      "loss": 0.7527,
+      "step": 5410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5204580589195327,
+      "learning_rate": 1.8843386201606729e-06,
+      "loss": 0.8204,
+      "step": 5411
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3681925437789058,
+      "learning_rate": 1.8815157086607826e-06,
+      "loss": 0.7905,
+      "step": 5412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3815746363605288,
+      "learning_rate": 1.878694693639631e-06,
+      "loss": 0.8228,
+      "step": 5413
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4192743005331077,
+      "learning_rate": 1.8758755757562087e-06,
+      "loss": 0.8051,
+      "step": 5414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.637184468861064,
+      "learning_rate": 1.8730583556690607e-06,
+      "loss": 0.8609,
+      "step": 5415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5111275078575743,
+      "learning_rate": 1.8702430340362932e-06,
+      "loss": 0.8047,
+      "step": 5416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.60423389568482,
+      "learning_rate": 1.86742961151556e-06,
+      "loss": 0.7251,
+      "step": 5417
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3892790376448796,
+      "learning_rate": 1.8646180887640808e-06,
+      "loss": 0.8207,
+      "step": 5418
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4942206841256545,
+      "learning_rate": 1.861808466438625e-06,
+      "loss": 0.8745,
+      "step": 5419
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5195657191743155,
+      "learning_rate": 1.8590007451955227e-06,
+      "loss": 0.855,
+      "step": 5420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3182416460665347,
+      "learning_rate": 1.8561949256906576e-06,
+      "loss": 0.7961,
+      "step": 5421
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4030934540663471,
+      "learning_rate": 1.8533910085794714e-06,
+      "loss": 0.6997,
+      "step": 5422
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3457299328086845,
+      "learning_rate": 1.8505889945169552e-06,
+      "loss": 0.7676,
+      "step": 5423
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7200951162355735,
+      "learning_rate": 1.8477888841576619e-06,
+      "loss": 0.833,
+      "step": 5424
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5666898663294038,
+      "learning_rate": 1.8449906781556959e-06,
+      "loss": 0.7709,
+      "step": 5425
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4403661668640364,
+      "learning_rate": 1.842194377164721e-06,
+      "loss": 0.8095,
+      "step": 5426
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3491385214056537,
+      "learning_rate": 1.8393999818379527e-06,
+      "loss": 0.7634,
+      "step": 5427
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3301694743398509,
+      "learning_rate": 1.8366074928281608e-06,
+      "loss": 0.8202,
+      "step": 5428
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8879331586424297,
+      "learning_rate": 1.8338169107876746e-06,
+      "loss": 0.3175,
+      "step": 5429
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3519710070880397,
+      "learning_rate": 1.8310282363683686e-06,
+      "loss": 0.8608,
+      "step": 5430
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5633425384109751,
+      "learning_rate": 1.828241470221681e-06,
+      "loss": 0.8305,
+      "step": 5431
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4968910744526684,
+      "learning_rate": 1.8254566129985996e-06,
+      "loss": 0.8011,
+      "step": 5432
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4919415715662103,
+      "learning_rate": 1.822673665349668e-06,
+      "loss": 0.8718,
+      "step": 5433
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9065972620641687,
+      "learning_rate": 1.8198926279249828e-06,
+      "loss": 0.3274,
+      "step": 5434
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3408161261896738,
+      "learning_rate": 1.817113501374197e-06,
+      "loss": 0.8617,
+      "step": 5435
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4891788486746218,
+      "learning_rate": 1.8143362863465098e-06,
+      "loss": 0.7444,
+      "step": 5436
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.608909984761927,
+      "learning_rate": 1.8115609834906821e-06,
+      "loss": 0.8493,
+      "step": 5437
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9057507054155469,
+      "learning_rate": 1.8087875934550237e-06,
+      "loss": 0.3092,
+      "step": 5438
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.480485169966335,
+      "learning_rate": 1.8060161168874002e-06,
+      "loss": 0.8688,
+      "step": 5439
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3153707806850248,
+      "learning_rate": 1.8032465544352274e-06,
+      "loss": 0.8455,
+      "step": 5440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6466245964860347,
+      "learning_rate": 1.8004789067454763e-06,
+      "loss": 0.7466,
+      "step": 5441
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5009255769765546,
+      "learning_rate": 1.7977131744646724e-06,
+      "loss": 0.7905,
+      "step": 5442
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5511504641793747,
+      "learning_rate": 1.7949493582388856e-06,
+      "loss": 0.7825,
+      "step": 5443
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.551146739245963,
+      "learning_rate": 1.7921874587137455e-06,
+      "loss": 0.7932,
+      "step": 5444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4813303748840447,
+      "learning_rate": 1.7894274765344322e-06,
+      "loss": 0.8227,
+      "step": 5445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.495632531310112,
+      "learning_rate": 1.7866694123456796e-06,
+      "loss": 0.9659,
+      "step": 5446
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.42540853410187,
+      "learning_rate": 1.7839132667917692e-06,
+      "loss": 0.7582,
+      "step": 5447
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.407862157718545,
+      "learning_rate": 1.7811590405165413e-06,
+      "loss": 0.7592,
+      "step": 5448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3616127417179036,
+      "learning_rate": 1.7784067341633781e-06,
+      "loss": 0.8078,
+      "step": 5449
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.837409960876165,
+      "learning_rate": 1.775656348375221e-06,
+      "loss": 0.3608,
+      "step": 5450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7862977420501167,
+      "learning_rate": 1.7729078837945602e-06,
+      "loss": 0.8365,
+      "step": 5451
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7679983407166945,
+      "learning_rate": 1.7701613410634367e-06,
+      "loss": 0.3726,
+      "step": 5452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.824373552736497,
+      "learning_rate": 1.7674167208234438e-06,
+      "loss": 0.3395,
+      "step": 5453
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4208820342070756,
+      "learning_rate": 1.7646740237157256e-06,
+      "loss": 0.8493,
+      "step": 5454
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7536404929779834,
+      "learning_rate": 1.7619332503809771e-06,
+      "loss": 0.8722,
+      "step": 5455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8505266871876735,
+      "learning_rate": 1.7591944014594408e-06,
+      "loss": 0.3603,
+      "step": 5456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5725560251819182,
+      "learning_rate": 1.7564574775909127e-06,
+      "loss": 0.8108,
+      "step": 5457
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4459537436782959,
+      "learning_rate": 1.75372247941474e-06,
+      "loss": 0.7431,
+      "step": 5458
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4518695940688044,
+      "learning_rate": 1.7509894075698165e-06,
+      "loss": 0.7779,
+      "step": 5459
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2579137060960675,
+      "learning_rate": 1.74825826269459e-06,
+      "loss": 0.7821,
+      "step": 5460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7987288411006441,
+      "learning_rate": 1.7455290454270567e-06,
+      "loss": 0.3288,
+      "step": 5461
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4483407244918518,
+      "learning_rate": 1.7428017564047594e-06,
+      "loss": 0.8205,
+      "step": 5462
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3118202103777108,
+      "learning_rate": 1.7400763962647937e-06,
+      "loss": 0.8173,
+      "step": 5463
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5396675993508369,
+      "learning_rate": 1.7373529656438048e-06,
+      "loss": 0.8424,
+      "step": 5464
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.389698906931674,
+      "learning_rate": 1.7346314651779872e-06,
+      "loss": 0.8204,
+      "step": 5465
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.494668317622261,
+      "learning_rate": 1.7319118955030812e-06,
+      "loss": 0.8112,
+      "step": 5466
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.297368352188807,
+      "learning_rate": 1.7291942572543806e-06,
+      "loss": 0.7652,
+      "step": 5467
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.633014915094076,
+      "learning_rate": 1.7264785510667281e-06,
+      "loss": 0.7763,
+      "step": 5468
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3795988886870105,
+      "learning_rate": 1.723764777574508e-06,
+      "loss": 0.829,
+      "step": 5469
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5189054188364544,
+      "learning_rate": 1.7210529374116603e-06,
+      "loss": 0.7594,
+      "step": 5470
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.480291764895458,
+      "learning_rate": 1.718343031211671e-06,
+      "loss": 0.814,
+      "step": 5471
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5080355296449017,
+      "learning_rate": 1.7156350596075743e-06,
+      "loss": 0.7721,
+      "step": 5472
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5572793959024667,
+      "learning_rate": 1.712929023231954e-06,
+      "loss": 0.9408,
+      "step": 5473
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5542340264241221,
+      "learning_rate": 1.710224922716941e-06,
+      "loss": 0.793,
+      "step": 5474
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5082668289507613,
+      "learning_rate": 1.7075227586942101e-06,
+      "loss": 0.8061,
+      "step": 5475
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4416454906202698,
+      "learning_rate": 1.7048225317949873e-06,
+      "loss": 0.7459,
+      "step": 5476
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.490440949287603,
+      "learning_rate": 1.7021242426500495e-06,
+      "loss": 0.8798,
+      "step": 5477
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5788130841710728,
+      "learning_rate": 1.6994278918897146e-06,
+      "loss": 0.7776,
+      "step": 5478
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4502246295007821,
+      "learning_rate": 1.6967334801438507e-06,
+      "loss": 0.8834,
+      "step": 5479
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4164442249021554,
+      "learning_rate": 1.6940410080418723e-06,
+      "loss": 0.7219,
+      "step": 5480
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4303587284987394,
+      "learning_rate": 1.6913504762127452e-06,
+      "loss": 0.8324,
+      "step": 5481
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4338681350810047,
+      "learning_rate": 1.6886618852849723e-06,
+      "loss": 0.8403,
+      "step": 5482
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.566794449013292,
+      "learning_rate": 1.685975235886611e-06,
+      "loss": 0.8762,
+      "step": 5483
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5650936153858075,
+      "learning_rate": 1.6832905286452616e-06,
+      "loss": 0.8565,
+      "step": 5484
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5441902866285886,
+      "learning_rate": 1.6806077641880746e-06,
+      "loss": 0.77,
+      "step": 5485
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4211188115955646,
+      "learning_rate": 1.6779269431417421e-06,
+      "loss": 0.8214,
+      "step": 5486
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.469878819332048,
+      "learning_rate": 1.6752480661325077e-06,
+      "loss": 0.8167,
+      "step": 5487
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5274607084327922,
+      "learning_rate": 1.6725711337861517e-06,
+      "loss": 0.7898,
+      "step": 5488
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.724208090636435,
+      "learning_rate": 1.6698961467280073e-06,
+      "loss": 0.8674,
+      "step": 5489
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.323917594496611,
+      "learning_rate": 1.6672231055829547e-06,
+      "loss": 0.8017,
+      "step": 5490
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.450098527703443,
+      "learning_rate": 1.6645520109754132e-06,
+      "loss": 0.8718,
+      "step": 5491
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4373245611629832,
+      "learning_rate": 1.6618828635293538e-06,
+      "loss": 0.8192,
+      "step": 5492
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4608870774252785,
+      "learning_rate": 1.6592156638682887e-06,
+      "loss": 0.782,
+      "step": 5493
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8681398448419817,
+      "learning_rate": 1.6565504126152742e-06,
+      "loss": 0.3591,
+      "step": 5494
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.468811983084041,
+      "learning_rate": 1.6538871103929144e-06,
+      "loss": 0.7336,
+      "step": 5495
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4895131208110255,
+      "learning_rate": 1.651225757823357e-06,
+      "loss": 0.7878,
+      "step": 5496
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.481807201846488,
+      "learning_rate": 1.6485663555282949e-06,
+      "loss": 0.8492,
+      "step": 5497
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5016800181957053,
+      "learning_rate": 1.6459089041289634e-06,
+      "loss": 0.8835,
+      "step": 5498
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.447151729797157,
+      "learning_rate": 1.6432534042461446e-06,
+      "loss": 0.8699,
+      "step": 5499
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6934858085748248,
+      "learning_rate": 1.6405998565001646e-06,
+      "loss": 0.904,
+      "step": 5500
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3007438313483457,
+      "learning_rate": 1.6379482615108888e-06,
+      "loss": 0.8035,
+      "step": 5501
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4078265296991646,
+      "learning_rate": 1.6352986198977327e-06,
+      "loss": 0.813,
+      "step": 5502
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4457536552501893,
+      "learning_rate": 1.6326509322796513e-06,
+      "loss": 0.7866,
+      "step": 5503
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5888623704702443,
+      "learning_rate": 1.6300051992751463e-06,
+      "loss": 0.8373,
+      "step": 5504
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5273461323546347,
+      "learning_rate": 1.6273614215022636e-06,
+      "loss": 0.7989,
+      "step": 5505
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5966160367825208,
+      "learning_rate": 1.6247195995785836e-06,
+      "loss": 0.877,
+      "step": 5506
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4706451707290387,
+      "learning_rate": 1.6220797341212401e-06,
+      "loss": 0.7165,
+      "step": 5507
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.409521968747777,
+      "learning_rate": 1.6194418257469047e-06,
+      "loss": 0.8675,
+      "step": 5508
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5564335717663023,
+      "learning_rate": 1.616805875071794e-06,
+      "loss": 0.8843,
+      "step": 5509
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2637774601711118,
+      "learning_rate": 1.6141718827116659e-06,
+      "loss": 0.7899,
+      "step": 5510
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5686378730543915,
+      "learning_rate": 1.6115398492818235e-06,
+      "loss": 0.7816,
+      "step": 5511
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.566122470511675,
+      "learning_rate": 1.6089097753971061e-06,
+      "loss": 0.8672,
+      "step": 5512
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.697413481427657,
+      "learning_rate": 1.6062816616719013e-06,
+      "loss": 0.7988,
+      "step": 5513
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5111844365468987,
+      "learning_rate": 1.6036555087201366e-06,
+      "loss": 0.8115,
+      "step": 5514
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4064719084651394,
+      "learning_rate": 1.6010313171552804e-06,
+      "loss": 0.8659,
+      "step": 5515
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4776909845260506,
+      "learning_rate": 1.598409087590349e-06,
+      "loss": 0.8006,
+      "step": 5516
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2557623367432647,
+      "learning_rate": 1.595788820637888e-06,
+      "loss": 0.8216,
+      "step": 5517
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9164182432906981,
+      "learning_rate": 1.5931705169099965e-06,
+      "loss": 0.3066,
+      "step": 5518
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5187019706027092,
+      "learning_rate": 1.5905541770183096e-06,
+      "loss": 0.8389,
+      "step": 5519
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4841139363050238,
+      "learning_rate": 1.5879398015740043e-06,
+      "loss": 0.7987,
+      "step": 5520
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8975412075519973,
+      "learning_rate": 1.5853273911878031e-06,
+      "loss": 0.3308,
+      "step": 5521
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4671420708327252,
+      "learning_rate": 1.5827169464699576e-06,
+      "loss": 0.7869,
+      "step": 5522
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6414938703587976,
+      "learning_rate": 1.5801084680302725e-06,
+      "loss": 0.7815,
+      "step": 5523
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3630373252883512,
+      "learning_rate": 1.577501956478088e-06,
+      "loss": 0.9018,
+      "step": 5524
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.391208658132168,
+      "learning_rate": 1.574897412422286e-06,
+      "loss": 0.8026,
+      "step": 5525
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4961974752938094,
+      "learning_rate": 1.5722948364712898e-06,
+      "loss": 0.8504,
+      "step": 5526
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3970538658742992,
+      "learning_rate": 1.5696942292330574e-06,
+      "loss": 0.8241,
+      "step": 5527
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3240620518979163,
+      "learning_rate": 1.567095591315092e-06,
+      "loss": 0.85,
+      "step": 5528
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.496021946934474,
+      "learning_rate": 1.5644989233244367e-06,
+      "loss": 0.8403,
+      "step": 5529
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7780968941032695,
+      "learning_rate": 1.5619042258676743e-06,
+      "loss": 0.8073,
+      "step": 5530
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2583673301924199,
+      "learning_rate": 1.559311499550924e-06,
+      "loss": 0.7656,
+      "step": 5531
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5084843361784621,
+      "learning_rate": 1.5567207449798517e-06,
+      "loss": 0.8093,
+      "step": 5532
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6460516067871112,
+      "learning_rate": 1.5541319627596517e-06,
+      "loss": 0.776,
+      "step": 5533
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5494695221710313,
+      "learning_rate": 1.5515451534950665e-06,
+      "loss": 0.8247,
+      "step": 5534
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4142065481981463,
+      "learning_rate": 1.5489603177903756e-06,
+      "loss": 0.8355,
+      "step": 5535
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.444022779838901,
+      "learning_rate": 1.546377456249396e-06,
+      "loss": 0.8107,
+      "step": 5536
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3935138660325066,
+      "learning_rate": 1.5437965694754842e-06,
+      "loss": 0.8107,
+      "step": 5537
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4773841885156629,
+      "learning_rate": 1.5412176580715367e-06,
+      "loss": 0.813,
+      "step": 5538
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.350258353742779,
+      "learning_rate": 1.5386407226399892e-06,
+      "loss": 0.9231,
+      "step": 5539
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4658333125290868,
+      "learning_rate": 1.5360657637828103e-06,
+      "loss": 0.8656,
+      "step": 5540
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5984906393784741,
+      "learning_rate": 1.533492782101511e-06,
+      "loss": 0.8379,
+      "step": 5541
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9607959445338982,
+      "learning_rate": 1.5309217781971419e-06,
+      "loss": 0.3203,
+      "step": 5542
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5163276766486915,
+      "learning_rate": 1.5283527526702891e-06,
+      "loss": 0.8196,
+      "step": 5543
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4747659600090726,
+      "learning_rate": 1.525785706121077e-06,
+      "loss": 0.7726,
+      "step": 5544
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.433930879055901,
+      "learning_rate": 1.52322063914917e-06,
+      "loss": 0.9098,
+      "step": 5545
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4701327234782449,
+      "learning_rate": 1.5206575523537649e-06,
+      "loss": 0.8306,
+      "step": 5546
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6304313189272786,
+      "learning_rate": 1.518096446333599e-06,
+      "loss": 0.7637,
+      "step": 5547
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.65617738308151,
+      "learning_rate": 1.5155373216869485e-06,
+      "loss": 0.7441,
+      "step": 5548
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5414975316928146,
+      "learning_rate": 1.5129801790116261e-06,
+      "loss": 0.8168,
+      "step": 5549
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3932105857881663,
+      "learning_rate": 1.5104250189049786e-06,
+      "loss": 0.8248,
+      "step": 5550
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6404144594793981,
+      "learning_rate": 1.5078718419638939e-06,
+      "loss": 0.8879,
+      "step": 5551
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5051287190843852,
+      "learning_rate": 1.5053206487847916e-06,
+      "loss": 0.8251,
+      "step": 5552
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5778456490253587,
+      "learning_rate": 1.5027714399636318e-06,
+      "loss": 0.7344,
+      "step": 5553
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4540634957295167,
+      "learning_rate": 1.5002242160959102e-06,
+      "loss": 0.876,
+      "step": 5554
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3944013702726914,
+      "learning_rate": 1.4976789777766576e-06,
+      "loss": 0.8261,
+      "step": 5555
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.459976009941761,
+      "learning_rate": 1.4951357256004439e-06,
+      "loss": 0.8651,
+      "step": 5556
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4746105623556245,
+      "learning_rate": 1.4925944601613718e-06,
+      "loss": 0.858,
+      "step": 5557
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.644449789245113,
+      "learning_rate": 1.490055182053083e-06,
+      "loss": 0.7707,
+      "step": 5558
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7752689156187316,
+      "learning_rate": 1.4875178918687493e-06,
+      "loss": 0.8024,
+      "step": 5559
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6857290766204718,
+      "learning_rate": 1.4849825902010851e-06,
+      "loss": 0.8048,
+      "step": 5560
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8098440466767464,
+      "learning_rate": 1.4824492776423349e-06,
+      "loss": 0.3148,
+      "step": 5561
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7196549666587329,
+      "learning_rate": 1.4799179547842823e-06,
+      "loss": 0.81,
+      "step": 5562
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4645435413452035,
+      "learning_rate": 1.4773886222182442e-06,
+      "loss": 0.8788,
+      "step": 5563
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2583547143180198,
+      "learning_rate": 1.474861280535076e-06,
+      "loss": 0.8497,
+      "step": 5564
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3721015946972879,
+      "learning_rate": 1.4723359303251594e-06,
+      "loss": 0.7628,
+      "step": 5565
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.58826013842811,
+      "learning_rate": 1.4698125721784183e-06,
+      "loss": 0.7902,
+      "step": 5566
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.829573092816346,
+      "learning_rate": 1.4672912066843103e-06,
+      "loss": 0.3181,
+      "step": 5567
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5418678877428342,
+      "learning_rate": 1.4647718344318263e-06,
+      "loss": 0.8186,
+      "step": 5568
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4416269105037618,
+      "learning_rate": 1.462254456009493e-06,
+      "loss": 0.7823,
+      "step": 5569
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.35936507887344,
+      "learning_rate": 1.4597390720053683e-06,
+      "loss": 0.7987,
+      "step": 5570
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5937556355507636,
+      "learning_rate": 1.4572256830070497e-06,
+      "loss": 0.8052,
+      "step": 5571
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.602847615303778,
+      "learning_rate": 1.454714289601661e-06,
+      "loss": 0.8419,
+      "step": 5572
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.425390778596702,
+      "learning_rate": 1.4522048923758647e-06,
+      "loss": 0.7829,
+      "step": 5573
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.352883551895712,
+      "learning_rate": 1.4496974919158569e-06,
+      "loss": 0.7635,
+      "step": 5574
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6195130523840662,
+      "learning_rate": 1.4471920888073676e-06,
+      "loss": 0.8647,
+      "step": 5575
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3764801105736542,
+      "learning_rate": 1.4446886836356578e-06,
+      "loss": 0.8518,
+      "step": 5576
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.530840901786308,
+      "learning_rate": 1.4421872769855262e-06,
+      "loss": 0.823,
+      "step": 5577
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4800179519865388,
+      "learning_rate": 1.4396878694412975e-06,
+      "loss": 0.9144,
+      "step": 5578
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3808580072118306,
+      "learning_rate": 1.4371904615868348e-06,
+      "loss": 0.8762,
+      "step": 5579
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.535561323716555,
+      "learning_rate": 1.4346950540055327e-06,
+      "loss": 0.8773,
+      "step": 5580
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3278806482656107,
+      "learning_rate": 1.4322016472803202e-06,
+      "loss": 0.864,
+      "step": 5581
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.495017309359749,
+      "learning_rate": 1.4297102419936559e-06,
+      "loss": 0.8616,
+      "step": 5582
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5488413392858595,
+      "learning_rate": 1.4272208387275332e-06,
+      "loss": 0.805,
+      "step": 5583
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.589270863427278,
+      "learning_rate": 1.4247334380634792e-06,
+      "loss": 0.8344,
+      "step": 5584
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5553199198233327,
+      "learning_rate": 1.4222480405825455e-06,
+      "loss": 0.8337,
+      "step": 5585
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5185311128558847,
+      "learning_rate": 1.4197646468653236e-06,
+      "loss": 0.9153,
+      "step": 5586
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3820352329727235,
+      "learning_rate": 1.4172832574919359e-06,
+      "loss": 0.7893,
+      "step": 5587
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4654419643879775,
+      "learning_rate": 1.4148038730420333e-06,
+      "loss": 0.8081,
+      "step": 5588
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4031886859981557,
+      "learning_rate": 1.4123264940948022e-06,
+      "loss": 0.8936,
+      "step": 5589
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5201834490088904,
+      "learning_rate": 1.409851121228959e-06,
+      "loss": 0.8423,
+      "step": 5590
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3663162568502987,
+      "learning_rate": 1.4073777550227485e-06,
+      "loss": 0.7801,
+      "step": 5591
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8827130379812916,
+      "learning_rate": 1.4049063960539488e-06,
+      "loss": 0.3037,
+      "step": 5592
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.394609849782385,
+      "learning_rate": 1.4024370448998726e-06,
+      "loss": 0.7811,
+      "step": 5593
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4127957323526816,
+      "learning_rate": 1.3999697021373582e-06,
+      "loss": 0.7325,
+      "step": 5594
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7082815452586944,
+      "learning_rate": 1.3975043683427791e-06,
+      "loss": 0.8212,
+      "step": 5595
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5110768777820272,
+      "learning_rate": 1.3950410440920359e-06,
+      "loss": 0.8165,
+      "step": 5596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4289574355055963,
+      "learning_rate": 1.3925797299605649e-06,
+      "loss": 0.8385,
+      "step": 5597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3500965380534669,
+      "learning_rate": 1.3901204265233237e-06,
+      "loss": 0.7188,
+      "step": 5598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5445498321140911,
+      "learning_rate": 1.3876631343548085e-06,
+      "loss": 0.8636,
+      "step": 5599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4966427428235587,
+      "learning_rate": 1.3852078540290437e-06,
+      "loss": 0.7899,
+      "step": 5600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5137203048833996,
+      "learning_rate": 1.3827545861195813e-06,
+      "loss": 0.9025,
+      "step": 5601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6069025415128586,
+      "learning_rate": 1.3803033311995072e-06,
+      "loss": 0.7619,
+      "step": 5602
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.481638423049099,
+      "learning_rate": 1.3778540898414349e-06,
+      "loss": 0.7315,
+      "step": 5603
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3281603114258382,
+      "learning_rate": 1.3754068626175043e-06,
+      "loss": 0.844,
+      "step": 5604
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.803280401038341,
+      "learning_rate": 1.3729616500993902e-06,
+      "loss": 0.3147,
+      "step": 5605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4297202906470103,
+      "learning_rate": 1.3705184528582937e-06,
+      "loss": 0.8571,
+      "step": 5606
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4129285382000398,
+      "learning_rate": 1.368077271464946e-06,
+      "loss": 0.8243,
+      "step": 5607
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5345366552564756,
+      "learning_rate": 1.3656381064896084e-06,
+      "loss": 0.8709,
+      "step": 5608
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4250483314365596,
+      "learning_rate": 1.3632009585020712e-06,
+      "loss": 0.7837,
+      "step": 5609
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.36223177424182,
+      "learning_rate": 1.3607658280716474e-06,
+      "loss": 0.8424,
+      "step": 5610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4879863076304276,
+      "learning_rate": 1.3583327157671878e-06,
+      "loss": 0.84,
+      "step": 5611
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.352572915104578,
+      "learning_rate": 1.3559016221570663e-06,
+      "loss": 0.7236,
+      "step": 5612
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2885910385786041,
+      "learning_rate": 1.353472547809187e-06,
+      "loss": 0.8153,
+      "step": 5613
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3929198883641805,
+      "learning_rate": 1.3510454932909823e-06,
+      "loss": 0.7998,
+      "step": 5614
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4842694094395823,
+      "learning_rate": 1.3486204591694118e-06,
+      "loss": 0.861,
+      "step": 5615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6298579173532424,
+      "learning_rate": 1.346197446010965e-06,
+      "loss": 0.7964,
+      "step": 5616
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5871884598816404,
+      "learning_rate": 1.3437764543816556e-06,
+      "loss": 0.8198,
+      "step": 5617
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5068796933954247,
+      "learning_rate": 1.341357484847029e-06,
+      "loss": 0.7854,
+      "step": 5618
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4252774628818057,
+      "learning_rate": 1.3389405379721564e-06,
+      "loss": 0.8259,
+      "step": 5619
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4353888677715951,
+      "learning_rate": 1.3365256143216377e-06,
+      "loss": 0.827,
+      "step": 5620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3669370939639536,
+      "learning_rate": 1.3341127144595978e-06,
+      "loss": 0.8322,
+      "step": 5621
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3917847201630018,
+      "learning_rate": 1.3317018389496927e-06,
+      "loss": 0.7853,
+      "step": 5622
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.317118778025492,
+      "learning_rate": 1.3292929883550998e-06,
+      "loss": 0.8805,
+      "step": 5623
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.360222807635599,
+      "learning_rate": 1.3268861632385288e-06,
+      "loss": 0.8551,
+      "step": 5624
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3977123370252795,
+      "learning_rate": 1.3244813641622146e-06,
+      "loss": 0.7454,
+      "step": 5625
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.595640924845315,
+      "learning_rate": 1.3220785916879165e-06,
+      "loss": 0.8409,
+      "step": 5626
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.411216738623978,
+      "learning_rate": 1.3196778463769256e-06,
+      "loss": 0.816,
+      "step": 5627
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5395991120887047,
+      "learning_rate": 1.3172791287900555e-06,
+      "loss": 0.8501,
+      "step": 5628
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.543553886041695,
+      "learning_rate": 1.3148824394876437e-06,
+      "loss": 0.8512,
+      "step": 5629
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5065576280259123,
+      "learning_rate": 1.3124877790295597e-06,
+      "loss": 0.739,
+      "step": 5630
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.31284031117535,
+      "learning_rate": 1.3100951479751967e-06,
+      "loss": 0.8386,
+      "step": 5631
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3337472241278043,
+      "learning_rate": 1.3077045468834714e-06,
+      "loss": 0.826,
+      "step": 5632
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4845560434009752,
+      "learning_rate": 1.3053159763128308e-06,
+      "loss": 0.8674,
+      "step": 5633
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5253173197521832,
+      "learning_rate": 1.3029294368212464e-06,
+      "loss": 0.7816,
+      "step": 5634
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4925019509595872,
+      "learning_rate": 1.3005449289662099e-06,
+      "loss": 0.7995,
+      "step": 5635
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4176195538307048,
+      "learning_rate": 1.2981624533047432e-06,
+      "loss": 0.7658,
+      "step": 5636
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.500556523428137,
+      "learning_rate": 1.295782010393396e-06,
+      "loss": 0.8436,
+      "step": 5637
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5447101538367003,
+      "learning_rate": 1.2934036007882378e-06,
+      "loss": 0.8523,
+      "step": 5638
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.444016724390116,
+      "learning_rate": 1.2910272250448675e-06,
+      "loss": 0.7755,
+      "step": 5639
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6557898883228868,
+      "learning_rate": 1.288652883718403e-06,
+      "loss": 0.8196,
+      "step": 5640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2910261932840776,
+      "learning_rate": 1.2862805773634934e-06,
+      "loss": 0.8632,
+      "step": 5641
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4065579474552827,
+      "learning_rate": 1.2839103065343084e-06,
+      "loss": 0.8273,
+      "step": 5642
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.383671166641272,
+      "learning_rate": 1.281542071784544e-06,
+      "loss": 0.7879,
+      "step": 5643
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4156518225465577,
+      "learning_rate": 1.2791758736674232e-06,
+      "loss": 0.8987,
+      "step": 5644
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.479803108051562,
+      "learning_rate": 1.2768117127356838e-06,
+      "loss": 0.7887,
+      "step": 5645
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6259523166532743,
+      "learning_rate": 1.2744495895415975e-06,
+      "loss": 0.8528,
+      "step": 5646
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4272012497979003,
+      "learning_rate": 1.2720895046369564e-06,
+      "loss": 0.7828,
+      "step": 5647
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5802915709514433,
+      "learning_rate": 1.269731458573077e-06,
+      "loss": 0.8399,
+      "step": 5648
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.576318908295224,
+      "learning_rate": 1.2673754519008008e-06,
+      "loss": 0.8486,
+      "step": 5649
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.724227901424948,
+      "learning_rate": 1.2650214851704866e-06,
+      "loss": 0.822,
+      "step": 5650
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1321056578034416,
+      "learning_rate": 1.2626695589320226e-06,
+      "loss": 0.8174,
+      "step": 5651
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.454999180386745,
+      "learning_rate": 1.2603196737348211e-06,
+      "loss": 0.7969,
+      "step": 5652
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6057886190190596,
+      "learning_rate": 1.2579718301278143e-06,
+      "loss": 0.7726,
+      "step": 5653
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4388468907285137,
+      "learning_rate": 1.255626028659459e-06,
+      "loss": 0.7955,
+      "step": 5654
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3757111656895906,
+      "learning_rate": 1.253282269877737e-06,
+      "loss": 0.8094,
+      "step": 5655
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5699015145975979,
+      "learning_rate": 1.2509405543301456e-06,
+      "loss": 0.8156,
+      "step": 5656
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4009171025629052,
+      "learning_rate": 1.2486008825637119e-06,
+      "loss": 0.7991,
+      "step": 5657
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.628825473269102,
+      "learning_rate": 1.2462632551249842e-06,
+      "loss": 0.7587,
+      "step": 5658
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.529396740197073,
+      "learning_rate": 1.2439276725600324e-06,
+      "loss": 0.7718,
+      "step": 5659
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.535025012252462,
+      "learning_rate": 1.2415941354144478e-06,
+      "loss": 0.8602,
+      "step": 5660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6595181776048544,
+      "learning_rate": 1.2392626442333488e-06,
+      "loss": 0.73,
+      "step": 5661
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3737998874084563,
+      "learning_rate": 1.2369331995613664e-06,
+      "loss": 0.8746,
+      "step": 5662
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3606653146067516,
+      "learning_rate": 1.234605801942661e-06,
+      "loss": 0.7773,
+      "step": 5663
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4311778828617514,
+      "learning_rate": 1.232280451920914e-06,
+      "loss": 0.8644,
+      "step": 5664
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3928005654786797,
+      "learning_rate": 1.229957150039327e-06,
+      "loss": 0.7957,
+      "step": 5665
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4631977140958095,
+      "learning_rate": 1.2276358968406233e-06,
+      "loss": 0.8095,
+      "step": 5666
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5743956596606044,
+      "learning_rate": 1.2253166928670478e-06,
+      "loss": 0.8352,
+      "step": 5667
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3886411116206447,
+      "learning_rate": 1.222999538660369e-06,
+      "loss": 0.7899,
+      "step": 5668
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2069141491625306,
+      "learning_rate": 1.2206844347618707e-06,
+      "loss": 0.8168,
+      "step": 5669
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4132497016375687,
+      "learning_rate": 1.2183713817123622e-06,
+      "loss": 0.7725,
+      "step": 5670
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5733510174513445,
+      "learning_rate": 1.2160603800521742e-06,
+      "loss": 0.8459,
+      "step": 5671
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4265737408302719,
+      "learning_rate": 1.213751430321156e-06,
+      "loss": 0.8258,
+      "step": 5672
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.437067274470808,
+      "learning_rate": 1.21144453305868e-06,
+      "loss": 0.8088,
+      "step": 5673
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8530367117609204,
+      "learning_rate": 1.2091396888036388e-06,
+      "loss": 0.3131,
+      "step": 5674
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3021184310156435,
+      "learning_rate": 1.206836898094439e-06,
+      "loss": 0.8166,
+      "step": 5675
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5845480138318666,
+      "learning_rate": 1.2045361614690166e-06,
+      "loss": 0.8347,
+      "step": 5676
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5298449452560832,
+      "learning_rate": 1.2022374794648229e-06,
+      "loss": 0.8639,
+      "step": 5677
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5372755598222279,
+      "learning_rate": 1.1999408526188295e-06,
+      "loss": 0.7902,
+      "step": 5678
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5181375212651633,
+      "learning_rate": 1.1976462814675305e-06,
+      "loss": 0.7438,
+      "step": 5679
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.426182147061947,
+      "learning_rate": 1.1953537665469383e-06,
+      "loss": 0.795,
+      "step": 5680
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.301857246241473,
+      "learning_rate": 1.1930633083925824e-06,
+      "loss": 0.7949,
+      "step": 5681
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5857344312979722,
+      "learning_rate": 1.1907749075395147e-06,
+      "loss": 0.8395,
+      "step": 5682
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.595842514113353,
+      "learning_rate": 1.1884885645223055e-06,
+      "loss": 0.8312,
+      "step": 5683
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4108327685210615,
+      "learning_rate": 1.1862042798750462e-06,
+      "loss": 0.8865,
+      "step": 5684
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2966530700459167,
+      "learning_rate": 1.1839220541313445e-06,
+      "loss": 0.7852,
+      "step": 5685
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.384203636723357,
+      "learning_rate": 1.1816418878243296e-06,
+      "loss": 0.8297,
+      "step": 5686
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6024638913253402,
+      "learning_rate": 1.179363781486651e-06,
+      "loss": 0.8701,
+      "step": 5687
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3896996353969324,
+      "learning_rate": 1.1770877356504684e-06,
+      "loss": 0.7804,
+      "step": 5688
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8686173337085714,
+      "learning_rate": 1.1748137508474699e-06,
+      "loss": 0.314,
+      "step": 5689
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3516349118465154,
+      "learning_rate": 1.1725418276088596e-06,
+      "loss": 0.7975,
+      "step": 5690
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5003963644647405,
+      "learning_rate": 1.170271966465356e-06,
+      "loss": 0.8027,
+      "step": 5691
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5614163777746215,
+      "learning_rate": 1.168004167947202e-06,
+      "loss": 0.8543,
+      "step": 5692
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4618083199562741,
+      "learning_rate": 1.1657384325841558e-06,
+      "loss": 0.8432,
+      "step": 5693
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5359316892455774,
+      "learning_rate": 1.1634747609054897e-06,
+      "loss": 0.8702,
+      "step": 5694
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4256592737565792,
+      "learning_rate": 1.1612131534399995e-06,
+      "loss": 0.86,
+      "step": 5695
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7752052562870355,
+      "learning_rate": 1.1589536107159981e-06,
+      "loss": 0.7744,
+      "step": 5696
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.351909281502772,
+      "learning_rate": 1.1566961332613136e-06,
+      "loss": 0.7192,
+      "step": 5697
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4024335665085583,
+      "learning_rate": 1.1544407216032928e-06,
+      "loss": 0.7249,
+      "step": 5698
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4207985079256367,
+      "learning_rate": 1.1521873762688007e-06,
+      "loss": 0.8791,
+      "step": 5699
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4849432616691427,
+      "learning_rate": 1.1499360977842212e-06,
+      "loss": 0.7577,
+      "step": 5700
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4694271207164782,
+      "learning_rate": 1.1476868866754488e-06,
+      "loss": 0.8661,
+      "step": 5701
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2823296074815393,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 0.8279,
+      "step": 5702
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5509524587197696,
+      "learning_rate": 1.1431946686865124e-06,
+      "loss": 0.7952,
+      "step": 5703
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5756226567134055,
+      "learning_rate": 1.1409516628557315e-06,
+      "loss": 0.8093,
+      "step": 5704
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5036382708895017,
+      "learning_rate": 1.1387107264995234e-06,
+      "loss": 0.8505,
+      "step": 5705
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5551125242582047,
+      "learning_rate": 1.136471860141376e-06,
+      "loss": 0.8046,
+      "step": 5706
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5147034194234463,
+      "learning_rate": 1.1342350643042822e-06,
+      "loss": 0.8299,
+      "step": 5707
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2882249632765759,
+      "learning_rate": 1.1320003395107604e-06,
+      "loss": 0.6946,
+      "step": 5708
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.527704240862158,
+      "learning_rate": 1.1297676862828421e-06,
+      "loss": 0.8621,
+      "step": 5709
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7554174081058926,
+      "learning_rate": 1.1275371051420769e-06,
+      "loss": 0.3287,
+      "step": 5710
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4371161712937863,
+      "learning_rate": 1.1253085966095278e-06,
+      "loss": 0.8085,
+      "step": 5711
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4776900788489071,
+      "learning_rate": 1.123082161205775e-06,
+      "loss": 0.8682,
+      "step": 5712
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4585296095734914,
+      "learning_rate": 1.120857799450915e-06,
+      "loss": 0.8439,
+      "step": 5713
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3931721948995697,
+      "learning_rate": 1.1186355118645552e-06,
+      "loss": 0.8381,
+      "step": 5714
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4748332302123681,
+      "learning_rate": 1.1164152989658251e-06,
+      "loss": 0.8336,
+      "step": 5715
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5041398498631433,
+      "learning_rate": 1.114197161273367e-06,
+      "loss": 0.7713,
+      "step": 5716
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.560401465715829,
+      "learning_rate": 1.111981099305336e-06,
+      "loss": 0.837,
+      "step": 5717
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3731257523173033,
+      "learning_rate": 1.1097671135794063e-06,
+      "loss": 0.8013,
+      "step": 5718
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6030186220943285,
+      "learning_rate": 1.1075552046127658e-06,
+      "loss": 0.7719,
+      "step": 5719
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5143900533446337,
+      "learning_rate": 1.1053453729221142e-06,
+      "loss": 0.7533,
+      "step": 5720
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.491812052103907,
+      "learning_rate": 1.1031376190236687e-06,
+      "loss": 0.8135,
+      "step": 5721
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5813760118728517,
+      "learning_rate": 1.1009319434331623e-06,
+      "loss": 0.7529,
+      "step": 5722
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4710801643072446,
+      "learning_rate": 1.0987283466658404e-06,
+      "loss": 0.823,
+      "step": 5723
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6725489837678664,
+      "learning_rate": 1.0965268292364639e-06,
+      "loss": 0.7201,
+      "step": 5724
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.36205055368535,
+      "learning_rate": 1.0943273916593067e-06,
+      "loss": 0.8893,
+      "step": 5725
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.544847948401028,
+      "learning_rate": 1.092130034448159e-06,
+      "loss": 0.8332,
+      "step": 5726
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.345572158658101,
+      "learning_rate": 1.0899347581163222e-06,
+      "loss": 0.7924,
+      "step": 5727
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3997858742149203,
+      "learning_rate": 1.087741563176613e-06,
+      "loss": 0.919,
+      "step": 5728
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5506090866150255,
+      "learning_rate": 1.0855504501413616e-06,
+      "loss": 0.8058,
+      "step": 5729
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4596590706039845,
+      "learning_rate": 1.0833614195224141e-06,
+      "loss": 0.8247,
+      "step": 5730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.574355821699966,
+      "learning_rate": 1.0811744718311267e-06,
+      "loss": 0.8068,
+      "step": 5731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5252985114444997,
+      "learning_rate": 1.0789896075783734e-06,
+      "loss": 0.8754,
+      "step": 5732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.498463974222597,
+      "learning_rate": 1.0768068272745347e-06,
+      "loss": 0.8187,
+      "step": 5733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.644244947792092,
+      "learning_rate": 1.0746261314295104e-06,
+      "loss": 0.893,
+      "step": 5734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6551908140463691,
+      "learning_rate": 1.0724475205527107e-06,
+      "loss": 0.8681,
+      "step": 5735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4688769555769485,
+      "learning_rate": 1.0702709951530587e-06,
+      "loss": 0.7887,
+      "step": 5736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4819465661777282,
+      "learning_rate": 1.0680965557389934e-06,
+      "loss": 0.7924,
+      "step": 5737
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3203300977132322,
+      "learning_rate": 1.0659242028184635e-06,
+      "loss": 0.7207,
+      "step": 5738
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4262053701510407,
+      "learning_rate": 1.063753936898928e-06,
+      "loss": 0.8571,
+      "step": 5739
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3777035501274875,
+      "learning_rate": 1.0615857584873624e-06,
+      "loss": 0.7935,
+      "step": 5740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2632930599066385,
+      "learning_rate": 1.0594196680902547e-06,
+      "loss": 0.7612,
+      "step": 5741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4726409468627246,
+      "learning_rate": 1.0572556662136036e-06,
+      "loss": 0.7954,
+      "step": 5742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.622198734792019,
+      "learning_rate": 1.055093753362919e-06,
+      "loss": 0.7772,
+      "step": 5743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3993958460706282,
+      "learning_rate": 1.052933930043225e-06,
+      "loss": 0.7605,
+      "step": 5744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6258069519841674,
+      "learning_rate": 1.050776196759058e-06,
+      "loss": 0.79,
+      "step": 5745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4824049600633822,
+      "learning_rate": 1.0486205540144612e-06,
+      "loss": 0.8563,
+      "step": 5746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.501868208380895,
+      "learning_rate": 1.0464670023129952e-06,
+      "loss": 0.8536,
+      "step": 5747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9202954531970664,
+      "learning_rate": 1.044315542157729e-06,
+      "loss": 0.3553,
+      "step": 5748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6241696521866726,
+      "learning_rate": 1.0421661740512445e-06,
+      "loss": 0.7516,
+      "step": 5749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6776822673801417,
+      "learning_rate": 1.0400188984956339e-06,
+      "loss": 0.8252,
+      "step": 5750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.439800028491611,
+      "learning_rate": 1.0378737159925023e-06,
+      "loss": 0.8404,
+      "step": 5751
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.377604308260554,
+      "learning_rate": 1.0357306270429623e-06,
+      "loss": 0.7838,
+      "step": 5752
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4277866023904988,
+      "learning_rate": 1.0335896321476413e-06,
+      "loss": 0.8116,
+      "step": 5753
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3260175821665998,
+      "learning_rate": 1.0314507318066757e-06,
+      "loss": 0.8293,
+      "step": 5754
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1277469173166015,
+      "learning_rate": 1.0293139265197116e-06,
+      "loss": 0.7013,
+      "step": 5755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3449864326868755,
+      "learning_rate": 1.0271792167859084e-06,
+      "loss": 0.8138,
+      "step": 5756
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4059085424770785,
+      "learning_rate": 1.0250466031039353e-06,
+      "loss": 0.8375,
+      "step": 5757
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6031715927733514,
+      "learning_rate": 1.0229160859719688e-06,
+      "loss": 0.81,
+      "step": 5758
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4429648535097033,
+      "learning_rate": 1.020787665887699e-06,
+      "loss": 0.778,
+      "step": 5759
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6734038285911446,
+      "learning_rate": 1.0186613433483238e-06,
+      "loss": 0.7931,
+      "step": 5760
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4500883473979898,
+      "learning_rate": 1.0165371188505546e-06,
+      "loss": 0.9094,
+      "step": 5761
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5036194763234356,
+      "learning_rate": 1.014414992890611e-06,
+      "loss": 0.8898,
+      "step": 5762
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.464445203605549,
+      "learning_rate": 1.012294965964218e-06,
+      "loss": 0.7474,
+      "step": 5763
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7599123365443723,
+      "learning_rate": 1.0101770385666166e-06,
+      "loss": 0.7693,
+      "step": 5764
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5430975135964056,
+      "learning_rate": 1.0080612111925547e-06,
+      "loss": 0.8473,
+      "step": 5765
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3302683323151863,
+      "learning_rate": 1.0059474843362893e-06,
+      "loss": 0.9066,
+      "step": 5766
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5519377350254149,
+      "learning_rate": 1.0038358584915896e-06,
+      "loss": 0.8193,
+      "step": 5767
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4810117214026877,
+      "learning_rate": 1.001726334151728e-06,
+      "loss": 0.8091,
+      "step": 5768
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4257675464180606,
+      "learning_rate": 9.996189118094901e-07,
+      "loss": 0.7591,
+      "step": 5769
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5802017076351729,
+      "learning_rate": 9.97513591957172e-07,
+      "loss": 0.8658,
+      "step": 5770
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.360368185633235,
+      "learning_rate": 9.954103750865762e-07,
+      "loss": 0.8092,
+      "step": 5771
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4468781870298115,
+      "learning_rate": 9.93309261689015e-07,
+      "loss": 0.8227,
+      "step": 5772
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.454892611358579,
+      "learning_rate": 9.912102522553047e-07,
+      "loss": 0.8374,
+      "step": 5773
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4704066227167842,
+      "learning_rate": 9.891133472757774e-07,
+      "loss": 0.8417,
+      "step": 5774
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2723028149877522,
+      "learning_rate": 9.870185472402705e-07,
+      "loss": 0.7401,
+      "step": 5775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4894490628871158,
+      "learning_rate": 9.849258526381288e-07,
+      "loss": 0.8302,
+      "step": 5776
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4345225306084823,
+      "learning_rate": 9.828352639582073e-07,
+      "loss": 0.8345,
+      "step": 5777
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4535279131992453,
+      "learning_rate": 9.80746781688865e-07,
+      "loss": 0.8294,
+      "step": 5778
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9183404251213945,
+      "learning_rate": 9.786604063179728e-07,
+      "loss": 0.3042,
+      "step": 5779
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3514829003041369,
+      "learning_rate": 9.765761383329087e-07,
+      "loss": 0.8808,
+      "step": 5780
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5428358829608562,
+      "learning_rate": 9.744939782205575e-07,
+      "loss": 0.8282,
+      "step": 5781
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8707008379114233,
+      "learning_rate": 9.724139264673116e-07,
+      "loss": 0.3138,
+      "step": 5782
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9674334588089352,
+      "learning_rate": 9.703359835590731e-07,
+      "loss": 0.3321,
+      "step": 5783
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3459840365070956,
+      "learning_rate": 9.682601499812494e-07,
+      "loss": 0.8321,
+      "step": 5784
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4689395863345713,
+      "learning_rate": 9.661864262187527e-07,
+      "loss": 0.8771,
+      "step": 5785
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4383817416169693,
+      "learning_rate": 9.641148127560063e-07,
+      "loss": 0.8344,
+      "step": 5786
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7361414243647821,
+      "learning_rate": 9.6204531007694e-07,
+      "loss": 0.9285,
+      "step": 5787
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.283474659749838,
+      "learning_rate": 9.599779186649893e-07,
+      "loss": 0.8295,
+      "step": 5788
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.394214252832118,
+      "learning_rate": 9.57912639003098e-07,
+      "loss": 0.8236,
+      "step": 5789
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3760682712367116,
+      "learning_rate": 9.558494715737166e-07,
+      "loss": 0.8192,
+      "step": 5790
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5853118143788905,
+      "learning_rate": 9.537884168587974e-07,
+      "loss": 0.7848,
+      "step": 5791
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5994496003727312,
+      "learning_rate": 9.517294753398066e-07,
+      "loss": 0.8238,
+      "step": 5792
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5367172933551225,
+      "learning_rate": 9.496726474977103e-07,
+      "loss": 0.8078,
+      "step": 5793
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5255289576737288,
+      "learning_rate": 9.476179338129854e-07,
+      "loss": 0.8875,
+      "step": 5794
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6827260799673724,
+      "learning_rate": 9.455653347656135e-07,
+      "loss": 0.8962,
+      "step": 5795
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8724072774454232,
+      "learning_rate": 9.435148508350822e-07,
+      "loss": 0.8302,
+      "step": 5796
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3734770002098882,
+      "learning_rate": 9.414664825003838e-07,
+      "loss": 0.7681,
+      "step": 5797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.492997322976864,
+      "learning_rate": 9.394202302400158e-07,
+      "loss": 0.7453,
+      "step": 5798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4009183158131107,
+      "learning_rate": 9.373760945319854e-07,
+      "loss": 0.8687,
+      "step": 5799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8318579391405551,
+      "learning_rate": 9.353340758538021e-07,
+      "loss": 0.3236,
+      "step": 5800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6650743508724601,
+      "learning_rate": 9.332941746824819e-07,
+      "loss": 0.8434,
+      "step": 5801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5911997301075265,
+      "learning_rate": 9.312563914945461e-07,
+      "loss": 0.831,
+      "step": 5802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3763131703916958,
+      "learning_rate": 9.292207267660214e-07,
+      "loss": 0.8581,
+      "step": 5803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.863228281691854,
+      "learning_rate": 9.271871809724375e-07,
+      "loss": 0.306,
+      "step": 5804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3637326430519523,
+      "learning_rate": 9.251557545888312e-07,
+      "loss": 0.8285,
+      "step": 5805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5170808686874004,
+      "learning_rate": 9.231264480897461e-07,
+      "loss": 0.9109,
+      "step": 5806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4544174089568205,
+      "learning_rate": 9.210992619492254e-07,
+      "loss": 0.8129,
+      "step": 5807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.395209327233474,
+      "learning_rate": 9.190741966408224e-07,
+      "loss": 0.8445,
+      "step": 5808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8813065463805019,
+      "learning_rate": 9.17051252637593e-07,
+      "loss": 0.2995,
+      "step": 5809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3091140773028045,
+      "learning_rate": 9.15030430412095e-07,
+      "loss": 0.8705,
+      "step": 5810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4537442000700627,
+      "learning_rate": 9.130117304363928e-07,
+      "loss": 0.7422,
+      "step": 5811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3806688434908208,
+      "learning_rate": 9.10995153182056e-07,
+      "loss": 0.764,
+      "step": 5812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5180552185569036,
+      "learning_rate": 9.089806991201567e-07,
+      "loss": 0.8315,
+      "step": 5813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3785202064124553,
+      "learning_rate": 9.069683687212716e-07,
+      "loss": 0.8407,
+      "step": 5814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4673116583184187,
+      "learning_rate": 9.049581624554816e-07,
+      "loss": 0.8306,
+      "step": 5815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6167578207010136,
+      "learning_rate": 9.029500807923719e-07,
+      "loss": 0.8216,
+      "step": 5816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3910951536208571,
+      "learning_rate": 9.009441242010287e-07,
+      "loss": 0.8002,
+      "step": 5817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4898255176058681,
+      "learning_rate": 8.989402931500434e-07,
+      "loss": 0.8257,
+      "step": 5818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4987953863726593,
+      "learning_rate": 8.969385881075132e-07,
+      "loss": 0.8308,
+      "step": 5819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4010799990932457,
+      "learning_rate": 8.94939009541036e-07,
+      "loss": 0.8521,
+      "step": 5820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5226489006714568,
+      "learning_rate": 8.929415579177125e-07,
+      "loss": 0.8066,
+      "step": 5821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4828013504853244,
+      "learning_rate": 8.909462337041508e-07,
+      "loss": 0.798,
+      "step": 5822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5389632105871618,
+      "learning_rate": 8.889530373664546e-07,
+      "loss": 0.7639,
+      "step": 5823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4121760204886484,
+      "learning_rate": 8.869619693702358e-07,
+      "loss": 0.7407,
+      "step": 5824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4112065074909304,
+      "learning_rate": 8.849730301806092e-07,
+      "loss": 0.8603,
+      "step": 5825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3596342212335895,
+      "learning_rate": 8.829862202621908e-07,
+      "loss": 0.8309,
+      "step": 5826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3652803559261828,
+      "learning_rate": 8.810015400790994e-07,
+      "loss": 0.7371,
+      "step": 5827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5847949900690912,
+      "learning_rate": 8.790189900949563e-07,
+      "loss": 0.8003,
+      "step": 5828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.531503888656653,
+      "learning_rate": 8.770385707728879e-07,
+      "loss": 0.7864,
+      "step": 5829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2646239525033796,
+      "learning_rate": 8.750602825755172e-07,
+      "loss": 0.8302,
+      "step": 5830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5924891575432214,
+      "learning_rate": 8.730841259649725e-07,
+      "loss": 0.7938,
+      "step": 5831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.509769617349298,
+      "learning_rate": 8.711101014028855e-07,
+      "loss": 0.8567,
+      "step": 5832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6722493966858294,
+      "learning_rate": 8.691382093503886e-07,
+      "loss": 0.7371,
+      "step": 5833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4924078586405232,
+      "learning_rate": 8.671684502681155e-07,
+      "loss": 0.8129,
+      "step": 5834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4985604344656946,
+      "learning_rate": 8.652008246162036e-07,
+      "loss": 0.8161,
+      "step": 5835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4942446217465666,
+      "learning_rate": 8.632353328542875e-07,
+      "loss": 0.8597,
+      "step": 5836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3692937423860445,
+      "learning_rate": 8.612719754415078e-07,
+      "loss": 0.8034,
+      "step": 5837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.486290822206405,
+      "learning_rate": 8.593107528365052e-07,
+      "loss": 0.8376,
+      "step": 5838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2882952480295942,
+      "learning_rate": 8.573516654974212e-07,
+      "loss": 0.8052,
+      "step": 5839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6290616846366113,
+      "learning_rate": 8.553947138818985e-07,
+      "loss": 0.8781,
+      "step": 5840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4831647990290548,
+      "learning_rate": 8.534398984470827e-07,
+      "loss": 0.7592,
+      "step": 5841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8115019858562805,
+      "learning_rate": 8.514872196496182e-07,
+      "loss": 0.3453,
+      "step": 5842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.491721846413568,
+      "learning_rate": 8.495366779456493e-07,
+      "loss": 0.7949,
+      "step": 5843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4194444512265951,
+      "learning_rate": 8.475882737908248e-07,
+      "loss": 0.8349,
+      "step": 5844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4785707355140427,
+      "learning_rate": 8.456420076402904e-07,
+      "loss": 0.7385,
+      "step": 5845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6845582104676677,
+      "learning_rate": 8.436978799486962e-07,
+      "loss": 0.795,
+      "step": 5846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6073742096973014,
+      "learning_rate": 8.417558911701884e-07,
+      "loss": 0.8416,
+      "step": 5847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.65362204205427,
+      "learning_rate": 8.398160417584178e-07,
+      "loss": 0.8626,
+      "step": 5848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5279714545727174,
+      "learning_rate": 8.378783321665318e-07,
+      "loss": 0.8275,
+      "step": 5849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4573769696162062,
+      "learning_rate": 8.359427628471806e-07,
+      "loss": 0.7494,
+      "step": 5850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.347179425928135,
+      "learning_rate": 8.340093342525113e-07,
+      "loss": 0.8045,
+      "step": 5851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8442951887023357,
+      "learning_rate": 8.320780468341761e-07,
+      "loss": 0.3437,
+      "step": 5852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5100270423733437,
+      "learning_rate": 8.301489010433216e-07,
+      "loss": 0.8728,
+      "step": 5853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4718822670157852,
+      "learning_rate": 8.282218973305978e-07,
+      "loss": 0.8375,
+      "step": 5854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4244416488806975,
+      "learning_rate": 8.262970361461542e-07,
+      "loss": 0.8348,
+      "step": 5855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.578791154349969,
+      "learning_rate": 8.243743179396346e-07,
+      "loss": 0.8614,
+      "step": 5856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.446664990566375,
+      "learning_rate": 8.224537431601886e-07,
+      "loss": 0.821,
+      "step": 5857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8475259068692577,
+      "learning_rate": 8.205353122564629e-07,
+      "loss": 0.3364,
+      "step": 5858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.408534325183491,
+      "learning_rate": 8.186190256766025e-07,
+      "loss": 0.7731,
+      "step": 5859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.649701800863386,
+      "learning_rate": 8.167048838682523e-07,
+      "loss": 0.8906,
+      "step": 5860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7061926562678056,
+      "learning_rate": 8.147928872785571e-07,
+      "loss": 0.7741,
+      "step": 5861
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6046694926335212,
+      "learning_rate": 8.128830363541574e-07,
+      "loss": 0.8543,
+      "step": 5862
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5276635711978692,
+      "learning_rate": 8.109753315411962e-07,
+      "loss": 0.7865,
+      "step": 5863
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5439641940502322,
+      "learning_rate": 8.090697732853125e-07,
+      "loss": 0.8035,
+      "step": 5864
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4847064037299842,
+      "learning_rate": 8.07166362031645e-07,
+      "loss": 0.7861,
+      "step": 5865
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4189935053801672,
+      "learning_rate": 8.052650982248311e-07,
+      "loss": 0.8629,
+      "step": 5866
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4735022740471926,
+      "learning_rate": 8.03365982309009e-07,
+      "loss": 0.7917,
+      "step": 5867
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4740504544527968,
+      "learning_rate": 8.01469014727807e-07,
+      "loss": 0.7894,
+      "step": 5868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3676577263498024,
+      "learning_rate": 7.995741959243597e-07,
+      "loss": 0.8685,
+      "step": 5869
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4609799060174877,
+      "learning_rate": 7.976815263412963e-07,
+      "loss": 0.8569,
+      "step": 5870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.545162037581966,
+      "learning_rate": 7.957910064207453e-07,
+      "loss": 0.8005,
+      "step": 5871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4012629184411343,
+      "learning_rate": 7.939026366043323e-07,
+      "loss": 0.7933,
+      "step": 5872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3771260484372194,
+      "learning_rate": 7.920164173331812e-07,
+      "loss": 0.8361,
+      "step": 5873
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4917728849161653,
+      "learning_rate": 7.901323490479129e-07,
+      "loss": 0.8597,
+      "step": 5874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3083003163505948,
+      "learning_rate": 7.882504321886442e-07,
+      "loss": 0.7648,
+      "step": 5875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2868280819729914,
+      "learning_rate": 7.863706671949922e-07,
+      "loss": 0.7772,
+      "step": 5876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5135939799539269,
+      "learning_rate": 7.844930545060703e-07,
+      "loss": 0.807,
+      "step": 5877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7674986110105817,
+      "learning_rate": 7.826175945604886e-07,
+      "loss": 0.8494,
+      "step": 5878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4639300510378377,
+      "learning_rate": 7.807442877963556e-07,
+      "loss": 0.8043,
+      "step": 5879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.406425716097815,
+      "learning_rate": 7.788731346512768e-07,
+      "loss": 0.703,
+      "step": 5880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4434702686094867,
+      "learning_rate": 7.770041355623504e-07,
+      "loss": 0.8474,
+      "step": 5881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1867874381627033,
+      "learning_rate": 7.75137290966177e-07,
+      "loss": 0.7855,
+      "step": 5882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4689276592876295,
+      "learning_rate": 7.732726012988512e-07,
+      "loss": 0.8909,
+      "step": 5883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.47282270608883,
+      "learning_rate": 7.714100669959656e-07,
+      "loss": 0.831,
+      "step": 5884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.428883323777507,
+      "learning_rate": 7.695496884926079e-07,
+      "loss": 0.8295,
+      "step": 5885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3342588241499156,
+      "learning_rate": 7.676914662233625e-07,
+      "loss": 0.787,
+      "step": 5886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3721511898436154,
+      "learning_rate": 7.65835400622309e-07,
+      "loss": 0.7669,
+      "step": 5887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.433935255307242,
+      "learning_rate": 7.639814921230271e-07,
+      "loss": 0.8445,
+      "step": 5888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.491144372419929,
+      "learning_rate": 7.621297411585881e-07,
+      "loss": 0.7324,
+      "step": 5889
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6064407313783018,
+      "learning_rate": 7.602801481615629e-07,
+      "loss": 0.8837,
+      "step": 5890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7871585651569413,
+      "learning_rate": 7.584327135640146e-07,
+      "loss": 0.7751,
+      "step": 5891
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4574514849192333,
+      "learning_rate": 7.565874377975046e-07,
+      "loss": 0.8443,
+      "step": 5892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.459437021769356,
+      "learning_rate": 7.547443212930906e-07,
+      "loss": 0.7697,
+      "step": 5893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.440790127415659,
+      "learning_rate": 7.529033644813232e-07,
+      "loss": 0.8447,
+      "step": 5894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.256512669246677,
+      "learning_rate": 7.510645677922534e-07,
+      "loss": 0.7913,
+      "step": 5895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3945551991743672,
+      "learning_rate": 7.492279316554207e-07,
+      "loss": 0.8086,
+      "step": 5896
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5442572826822383,
+      "learning_rate": 7.473934564998641e-07,
+      "loss": 0.8898,
+      "step": 5897
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6020052053752276,
+      "learning_rate": 7.455611427541176e-07,
+      "loss": 0.7849,
+      "step": 5898
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.853193944264774,
+      "learning_rate": 7.43730990846211e-07,
+      "loss": 0.2979,
+      "step": 5899
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6672546736511864,
+      "learning_rate": 7.419030012036676e-07,
+      "loss": 0.7982,
+      "step": 5900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4526160346089265,
+      "learning_rate": 7.400771742535051e-07,
+      "loss": 0.8094,
+      "step": 5901
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5524239079764113,
+      "learning_rate": 7.382535104222366e-07,
+      "loss": 0.8098,
+      "step": 5902
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4528487927438605,
+      "learning_rate": 7.364320101358701e-07,
+      "loss": 0.7176,
+      "step": 5903
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4708249667282491,
+      "learning_rate": 7.346126738199089e-07,
+      "loss": 0.7953,
+      "step": 5904
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.350785066291444,
+      "learning_rate": 7.327955018993504e-07,
+      "loss": 0.8541,
+      "step": 5905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4043682820359351,
+      "learning_rate": 7.309804947986876e-07,
+      "loss": 0.8801,
+      "step": 5906
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3474453816398573,
+      "learning_rate": 7.291676529419034e-07,
+      "loss": 0.7746,
+      "step": 5907
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8010749208623257,
+      "learning_rate": 7.273569767524791e-07,
+      "loss": 0.3604,
+      "step": 5908
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.399684426677342,
+      "learning_rate": 7.255484666533874e-07,
+      "loss": 0.8385,
+      "step": 5909
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5450925660392787,
+      "learning_rate": 7.237421230670994e-07,
+      "loss": 0.8281,
+      "step": 5910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6584249517702134,
+      "learning_rate": 7.21937946415574e-07,
+      "loss": 0.8738,
+      "step": 5911
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5627101947546354,
+      "learning_rate": 7.201359371202698e-07,
+      "loss": 0.8822,
+      "step": 5912
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3450334997884406,
+      "learning_rate": 7.183360956021368e-07,
+      "loss": 0.802,
+      "step": 5913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.390558199671098,
+      "learning_rate": 7.165384222816141e-07,
+      "loss": 0.9109,
+      "step": 5914
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.295718793987034,
+      "learning_rate": 7.147429175786413e-07,
+      "loss": 0.8183,
+      "step": 5915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4065951062967756,
+      "learning_rate": 7.129495819126476e-07,
+      "loss": 0.8235,
+      "step": 5916
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4245071015388078,
+      "learning_rate": 7.111584157025575e-07,
+      "loss": 0.8289,
+      "step": 5917
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3064255238005393,
+      "learning_rate": 7.093694193667866e-07,
+      "loss": 0.7531,
+      "step": 5918
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3899770121723252,
+      "learning_rate": 7.075825933232461e-07,
+      "loss": 0.8687,
+      "step": 5919
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5346226407869021,
+      "learning_rate": 7.057979379893353e-07,
+      "loss": 0.8846,
+      "step": 5920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4240059362195934,
+      "learning_rate": 7.040154537819533e-07,
+      "loss": 0.8722,
+      "step": 5921
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4310095282065138,
+      "learning_rate": 7.022351411174866e-07,
+      "loss": 0.7597,
+      "step": 5922
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.603588444105267,
+      "learning_rate": 7.004570004118172e-07,
+      "loss": 0.8156,
+      "step": 5923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4914003700292304,
+      "learning_rate": 6.986810320803195e-07,
+      "loss": 0.7692,
+      "step": 5924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6335567804036393,
+      "learning_rate": 6.969072365378605e-07,
+      "loss": 0.8702,
+      "step": 5925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5112571294894828,
+      "learning_rate": 6.951356141987963e-07,
+      "loss": 0.7707,
+      "step": 5926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2762290949935435,
+      "learning_rate": 6.933661654769797e-07,
+      "loss": 0.8104,
+      "step": 5927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.53685108084369,
+      "learning_rate": 6.91598890785754e-07,
+      "loss": 0.8344,
+      "step": 5928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.429970967874016,
+      "learning_rate": 6.898337905379549e-07,
+      "loss": 0.8886,
+      "step": 5929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4579141475590403,
+      "learning_rate": 6.88070865145909e-07,
+      "loss": 0.7821,
+      "step": 5930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9416034346551806,
+      "learning_rate": 6.863101150214369e-07,
+      "loss": 0.3091,
+      "step": 5931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3314424385195347,
+      "learning_rate": 6.845515405758518e-07,
+      "loss": 0.8465,
+      "step": 5932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5400321902240883,
+      "learning_rate": 6.827951422199531e-07,
+      "loss": 0.787,
+      "step": 5933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.527361783118756,
+      "learning_rate": 6.810409203640378e-07,
+      "loss": 0.7712,
+      "step": 5934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.414291817686033,
+      "learning_rate": 6.792888754178906e-07,
+      "loss": 0.7771,
+      "step": 5935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3719478328516528,
+      "learning_rate": 6.775390077907918e-07,
+      "loss": 0.7731,
+      "step": 5936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4162112814461316,
+      "learning_rate": 6.757913178915087e-07,
+      "loss": 0.7218,
+      "step": 5937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8770896406052634,
+      "learning_rate": 6.74045806128305e-07,
+      "loss": 0.3018,
+      "step": 5938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.543155433733017,
+      "learning_rate": 6.723024729089278e-07,
+      "loss": 0.795,
+      "step": 5939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3829758091038407,
+      "learning_rate": 6.705613186406223e-07,
+      "loss": 0.7688,
+      "step": 5940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5224843203869813,
+      "learning_rate": 6.688223437301222e-07,
+      "loss": 0.7841,
+      "step": 5941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3913430204648534,
+      "learning_rate": 6.670855485836525e-07,
+      "loss": 0.8429,
+      "step": 5942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4823748495662123,
+      "learning_rate": 6.653509336069285e-07,
+      "loss": 0.8302,
+      "step": 5943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.364959787415724,
+      "learning_rate": 6.636184992051553e-07,
+      "loss": 0.8156,
+      "step": 5944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5895254674139034,
+      "learning_rate": 6.618882457830334e-07,
+      "loss": 0.7907,
+      "step": 5945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3317540836872253,
+      "learning_rate": 6.601601737447461e-07,
+      "loss": 0.7529,
+      "step": 5946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.380769605969531,
+      "learning_rate": 6.584342834939717e-07,
+      "loss": 0.7617,
+      "step": 5947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.486678832756588,
+      "learning_rate": 6.567105754338798e-07,
+      "loss": 0.7975,
+      "step": 5948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5142665894518585,
+      "learning_rate": 6.549890499671285e-07,
+      "loss": 0.798,
+      "step": 5949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5457823288642993,
+      "learning_rate": 6.532697074958661e-07,
+      "loss": 0.8038,
+      "step": 5950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4400660699205168,
+      "learning_rate": 6.515525484217323e-07,
+      "loss": 0.8534,
+      "step": 5951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5195908792166468,
+      "learning_rate": 6.498375731458529e-07,
+      "loss": 0.7921,
+      "step": 5952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4485575627746758,
+      "learning_rate": 6.481247820688475e-07,
+      "loss": 0.7321,
+      "step": 5953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3029254282814693,
+      "learning_rate": 6.46414175590826e-07,
+      "loss": 0.8233,
+      "step": 5954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5290624918802223,
+      "learning_rate": 6.447057541113832e-07,
+      "loss": 0.8634,
+      "step": 5955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5153888138614064,
+      "learning_rate": 6.42999518029609e-07,
+      "loss": 0.7795,
+      "step": 5956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5781135086547653,
+      "learning_rate": 6.412954677440797e-07,
+      "loss": 0.78,
+      "step": 5957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.431357275022252,
+      "learning_rate": 6.395936036528627e-07,
+      "loss": 0.7856,
+      "step": 5958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4299018300540858,
+      "learning_rate": 6.378939261535111e-07,
+      "loss": 0.8316,
+      "step": 5959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3564976379320886,
+      "learning_rate": 6.361964356430717e-07,
+      "loss": 0.9327,
+      "step": 5960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4516939954152928,
+      "learning_rate": 6.345011325180772e-07,
+      "loss": 0.768,
+      "step": 5961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4536822824499733,
+      "learning_rate": 6.32808017174551e-07,
+      "loss": 0.8605,
+      "step": 5962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.54892466016969,
+      "learning_rate": 6.311170900080044e-07,
+      "loss": 0.8204,
+      "step": 5963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.378243499157131,
+      "learning_rate": 6.294283514134414e-07,
+      "loss": 0.8382,
+      "step": 5964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5842245204461887,
+      "learning_rate": 6.277418017853476e-07,
+      "loss": 0.7621,
+      "step": 5965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2991322863957695,
+      "learning_rate": 6.260574415177012e-07,
+      "loss": 0.8059,
+      "step": 5966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7342637752404075,
+      "learning_rate": 6.243752710039719e-07,
+      "loss": 0.8289,
+      "step": 5967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5609923563588473,
+      "learning_rate": 6.226952906371131e-07,
+      "loss": 0.9273,
+      "step": 5968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5331418134494013,
+      "learning_rate": 6.210175008095675e-07,
+      "loss": 0.7992,
+      "step": 5969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.261713053776074,
+      "learning_rate": 6.193419019132685e-07,
+      "loss": 0.8286,
+      "step": 5970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5298776163358483,
+      "learning_rate": 6.176684943396383e-07,
+      "loss": 0.8028,
+      "step": 5971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4465507995565334,
+      "learning_rate": 6.159972784795798e-07,
+      "loss": 0.8406,
+      "step": 5972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.465692630216025,
+      "learning_rate": 6.143282547234919e-07,
+      "loss": 0.7757,
+      "step": 5973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.542709106618553,
+      "learning_rate": 6.126614234612593e-07,
+      "loss": 0.7998,
+      "step": 5974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4419882214597715,
+      "learning_rate": 6.109967850822529e-07,
+      "loss": 0.9108,
+      "step": 5975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4451114978722102,
+      "learning_rate": 6.093343399753327e-07,
+      "loss": 0.7805,
+      "step": 5976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3334239554683949,
+      "learning_rate": 6.076740885288479e-07,
+      "loss": 0.8371,
+      "step": 5977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8213178071696831,
+      "learning_rate": 6.060160311306307e-07,
+      "loss": 0.3207,
+      "step": 5978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.414291328414453,
+      "learning_rate": 6.043601681680045e-07,
+      "loss": 0.7822,
+      "step": 5979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5378556300109016,
+      "learning_rate": 6.027065000277776e-07,
+      "loss": 0.8647,
+      "step": 5980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5324473042727753,
+      "learning_rate": 6.010550270962501e-07,
+      "loss": 0.8052,
+      "step": 5981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5700500752958246,
+      "learning_rate": 5.994057497592032e-07,
+      "loss": 0.7401,
+      "step": 5982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4364867805404387,
+      "learning_rate": 5.97758668401911e-07,
+      "loss": 0.8555,
+      "step": 5983
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6340851820368967,
+      "learning_rate": 5.961137834091313e-07,
+      "loss": 0.7655,
+      "step": 5984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.805879616656503,
+      "learning_rate": 5.944710951651067e-07,
+      "loss": 0.287,
+      "step": 5985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4135264110222188,
+      "learning_rate": 5.928306040535725e-07,
+      "loss": 0.8703,
+      "step": 5986
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3645196402647846,
+      "learning_rate": 5.911923104577455e-07,
+      "loss": 0.8089,
+      "step": 5987
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3706676426923676,
+      "learning_rate": 5.895562147603317e-07,
+      "loss": 0.8147,
+      "step": 5988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5334474551148063,
+      "learning_rate": 5.879223173435245e-07,
+      "loss": 0.8056,
+      "step": 5989
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5710583796662052,
+      "learning_rate": 5.862906185890027e-07,
+      "loss": 0.823,
+      "step": 5990
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3071111258208394,
+      "learning_rate": 5.846611188779283e-07,
+      "loss": 0.8242,
+      "step": 5991
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5030716301688591,
+      "learning_rate": 5.830338185909545e-07,
+      "loss": 0.7478,
+      "step": 5992
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4878308855063345,
+      "learning_rate": 5.814087181082195e-07,
+      "loss": 0.7888,
+      "step": 5993
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.430201095560194,
+      "learning_rate": 5.797858178093463e-07,
+      "loss": 0.7913,
+      "step": 5994
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.395487065090944,
+      "learning_rate": 5.781651180734438e-07,
+      "loss": 0.8612,
+      "step": 5995
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.536681800421341,
+      "learning_rate": 5.765466192791103e-07,
+      "loss": 0.845,
+      "step": 5996
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.668963353532185,
+      "learning_rate": 5.749303218044234e-07,
+      "loss": 0.8468,
+      "step": 5997
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5744757785586105,
+      "learning_rate": 5.73316226026952e-07,
+      "loss": 0.8692,
+      "step": 5998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.775092832268406,
+      "learning_rate": 5.717043323237503e-07,
+      "loss": 0.8236,
+      "step": 5999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4651047245247877,
+      "learning_rate": 5.700946410713548e-07,
+      "loss": 0.9029,
+      "step": 6000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6702,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 2.000184256211242e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}