diff --git "a/ajaymin28/vl-sg/checkpoint-2000/trainer_state.json" "b/ajaymin28/vl-sg/checkpoint-2000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/ajaymin28/vl-sg/checkpoint-2000/trainer_state.json"
@@ -0,0 +1,14021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9762845849802373,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.631305851388058,
+      "learning_rate": 1.639344262295082e-07,
+      "loss": 0.7943,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.346641941057879,
+      "learning_rate": 3.278688524590164e-07,
+      "loss": 0.8013,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.925720270742593,
+      "learning_rate": 4.918032786885246e-07,
+      "loss": 0.8416,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.580924168656784,
+      "learning_rate": 6.557377049180328e-07,
+      "loss": 0.8441,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.551913967770986,
+      "learning_rate": 8.196721311475409e-07,
+      "loss": 0.735,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.620840204280576,
+      "learning_rate": 9.836065573770493e-07,
+      "loss": 0.8406,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.245707797327382,
+      "learning_rate": 1.1475409836065575e-06,
+      "loss": 0.6569,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.19931522169994,
+      "learning_rate": 1.3114754098360657e-06,
+      "loss": 0.6519,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.80206123230587,
+      "learning_rate": 1.4754098360655739e-06,
+      "loss": 0.652,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.075520262991187,
+      "learning_rate": 1.6393442622950819e-06,
+      "loss": 0.5741,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.673104969696313,
+      "learning_rate": 1.8032786885245903e-06,
+      "loss": 0.7034,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.465018340927124,
+      "learning_rate": 1.9672131147540985e-06,
+      "loss": 0.5484,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.595314473550937,
+      "learning_rate": 2.1311475409836067e-06,
+      "loss": 0.4478,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.3040612954225494,
+      "learning_rate": 2.295081967213115e-06,
+      "loss": 0.3268,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.722360257746658,
+      "learning_rate": 2.459016393442623e-06,
+      "loss": 0.3072,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.0490329481332985,
+      "learning_rate": 2.6229508196721314e-06,
+      "loss": 0.2465,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.131500605280447,
+      "learning_rate": 2.786885245901639e-06,
+      "loss": 0.2873,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.0745203091956474,
+      "learning_rate": 2.9508196721311478e-06,
+      "loss": 0.2323,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.161054943272229,
+      "learning_rate": 3.114754098360656e-06,
+      "loss": 0.2304,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.211697637623634,
+      "learning_rate": 3.2786885245901638e-06,
+      "loss": 0.2446,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.081749210877428,
+      "learning_rate": 3.4426229508196724e-06,
+      "loss": 0.1992,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9576919724324435,
+      "learning_rate": 3.6065573770491806e-06,
+      "loss": 0.1976,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.919484126413122,
+      "learning_rate": 3.7704918032786884e-06,
+      "loss": 0.1694,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.990436329009411,
+      "learning_rate": 3.934426229508197e-06,
+      "loss": 0.1704,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.904975657739626,
+      "learning_rate": 4.098360655737705e-06,
+      "loss": 0.2249,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.400257571284763,
+      "learning_rate": 4.2622950819672135e-06,
+      "loss": 0.1817,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6630000285111053,
+      "learning_rate": 4.426229508196722e-06,
+      "loss": 0.1724,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.8660635864126,
+      "learning_rate": 4.59016393442623e-06,
+      "loss": 0.2231,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.041616731907119,
+      "learning_rate": 4.754098360655738e-06,
+      "loss": 0.2331,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.699798847116446,
+      "learning_rate": 4.918032786885246e-06,
+      "loss": 0.1944,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.051559842542771,
+      "learning_rate": 5.0819672131147545e-06,
+      "loss": 0.1742,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.0126589941382855,
+      "learning_rate": 5.245901639344263e-06,
+      "loss": 0.1737,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.962155734254877,
+      "learning_rate": 5.409836065573772e-06,
+      "loss": 0.1951,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.367924200321454,
+      "learning_rate": 5.573770491803278e-06,
+      "loss": 0.2053,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.581007083931671,
+      "learning_rate": 5.737704918032787e-06,
+      "loss": 0.1344,
+      "step": 35
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.814175913297936,
+      "learning_rate": 5.9016393442622956e-06,
+      "loss": 0.185,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.923834811559088,
+      "learning_rate": 6.065573770491804e-06,
+      "loss": 0.1955,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8227363541701314,
+      "learning_rate": 6.229508196721312e-06,
+      "loss": 0.1639,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6972292584990414,
+      "learning_rate": 6.393442622950821e-06,
+      "loss": 0.1703,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6507311826944577,
+      "learning_rate": 6.5573770491803276e-06,
+      "loss": 0.178,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9993573884654845,
+      "learning_rate": 6.721311475409837e-06,
+      "loss": 0.1552,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.629217820811896,
+      "learning_rate": 6.885245901639345e-06,
+      "loss": 0.1642,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2303769162789946,
+      "learning_rate": 7.049180327868853e-06,
+      "loss": 0.1433,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9022939442067437,
+      "learning_rate": 7.213114754098361e-06,
+      "loss": 0.1671,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.518448605554198,
+      "learning_rate": 7.3770491803278695e-06,
+      "loss": 0.1336,
+      "step": 45
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.034618971631891,
+      "learning_rate": 7.540983606557377e-06,
+      "loss": 0.1501,
+      "step": 46
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.26390119636968,
+      "learning_rate": 7.704918032786886e-06,
+      "loss": 0.1275,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5565200003009316,
+      "learning_rate": 7.868852459016394e-06,
+      "loss": 0.1572,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4801405609392058,
+      "learning_rate": 8.032786885245902e-06,
+      "loss": 0.1354,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.152281452474444,
+      "learning_rate": 8.19672131147541e-06,
+      "loss": 0.1358,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8697633273885104,
+      "learning_rate": 8.360655737704919e-06,
+      "loss": 0.1512,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.5821217782901185,
+      "learning_rate": 8.524590163934427e-06,
+      "loss": 0.1327,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.915857828990885,
+      "learning_rate": 8.688524590163935e-06,
+      "loss": 0.1177,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.643567686766952,
+      "learning_rate": 8.852459016393443e-06,
+      "loss": 0.1116,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.827720168787582,
+      "learning_rate": 9.016393442622952e-06,
+      "loss": 0.1399,
+      "step": 55
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3325522278488298,
+      "learning_rate": 9.18032786885246e-06,
+      "loss": 0.1297,
+      "step": 56
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.214346732745731,
+      "learning_rate": 9.344262295081968e-06,
+      "loss": 0.1792,
+      "step": 57
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.828284793640583,
+      "learning_rate": 9.508196721311476e-06,
+      "loss": 0.1461,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9596409765258485,
+      "learning_rate": 9.672131147540984e-06,
+      "loss": 0.1513,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2036564881308625,
+      "learning_rate": 9.836065573770493e-06,
+      "loss": 0.1485,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.456609953730327,
+      "learning_rate": 1e-05,
+      "loss": 0.1056,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.707216229085872,
+      "learning_rate": 9.999993596770583e-06,
+      "loss": 0.1925,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7136348282084795,
+      "learning_rate": 9.999974387098732e-06,
+      "loss": 0.1534,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8902574550035913,
+      "learning_rate": 9.999942371033649e-06,
+      "loss": 0.1443,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4920991371319463,
+      "learning_rate": 9.999897548657335e-06,
+      "loss": 0.1736,
+      "step": 65
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.910381719634011,
+      "learning_rate": 9.999839920084596e-06,
+      "loss": 0.1346,
+      "step": 66
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8276514936410475,
+      "learning_rate": 9.999769485463033e-06,
+      "loss": 0.1439,
+      "step": 67
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8112591500388016,
+      "learning_rate": 9.999686244973048e-06,
+      "loss": 0.1058,
+      "step": 68
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3201219851502324,
+      "learning_rate": 9.999590198827851e-06,
+      "loss": 0.1096,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.675174239613632,
+      "learning_rate": 9.999481347273437e-06,
+      "loss": 0.1616,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4734469072988774,
+      "learning_rate": 9.99935969058861e-06,
+      "loss": 0.1475,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.029244154732283,
+      "learning_rate": 9.999225229084967e-06,
+      "loss": 0.1441,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.74263319793326,
+      "learning_rate": 9.999077963106906e-06,
+      "loss": 0.1166,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1161867356606834,
+      "learning_rate": 9.998917893031615e-06,
+      "loss": 0.1099,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.184783373365182,
+      "learning_rate": 9.99874501926908e-06,
+      "loss": 0.1147,
+      "step": 75
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.477787258132454,
+      "learning_rate": 9.998559342262086e-06,
+      "loss": 0.1247,
+      "step": 76
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.334018069534599,
+      "learning_rate": 9.998360862486199e-06,
+      "loss": 0.1204,
+      "step": 77
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8005941630072586,
+      "learning_rate": 9.998149580449787e-06,
+      "loss": 0.142,
+      "step": 78
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5403091682911296,
+      "learning_rate": 9.997925496694006e-06,
+      "loss": 0.0946,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9804191935413014,
+      "learning_rate": 9.9976886117928e-06,
+      "loss": 0.1286,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.923557713497968,
+      "learning_rate": 9.997438926352896e-06,
+      "loss": 0.1837,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.670219851094306,
+      "learning_rate": 9.997176441013817e-06,
+      "loss": 0.1241,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.898925596705727,
+      "learning_rate": 9.99690115644786e-06,
+      "loss": 0.103,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3909556938573298,
+      "learning_rate": 9.996613073360111e-06,
+      "loss": 0.1261,
+      "step": 84
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.56655736296895,
+      "learning_rate": 9.996312192488436e-06,
+      "loss": 0.1625,
+      "step": 85
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7522296492231564,
+      "learning_rate": 9.995998514603478e-06,
+      "loss": 0.126,
+      "step": 86
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0743647967016217,
+      "learning_rate": 9.995672040508656e-06,
+      "loss": 0.127,
+      "step": 87
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3293214870202488,
+      "learning_rate": 9.995332771040167e-06,
+      "loss": 0.1302,
+      "step": 88
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5399329041251377,
+      "learning_rate": 9.994980707066979e-06,
+      "loss": 0.1361,
+      "step": 89
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.628114945814919,
+      "learning_rate": 9.99461584949083e-06,
+      "loss": 0.1685,
+      "step": 90
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.105678014096815,
+      "learning_rate": 9.994238199246226e-06,
+      "loss": 0.1358,
+      "step": 91
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.807630136340985,
+      "learning_rate": 9.993847757300441e-06,
+      "loss": 0.1251,
+      "step": 92
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0778212654654142,
+      "learning_rate": 9.993444524653509e-06,
+      "loss": 0.1315,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7611031344727106,
+      "learning_rate": 9.99302850233823e-06,
+      "loss": 0.1507,
+      "step": 94
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3385040063959,
+      "learning_rate": 9.992599691420154e-06,
+      "loss": 0.1321,
+      "step": 95
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.032463630698966,
+      "learning_rate": 9.992158092997593e-06,
+      "loss": 0.1338,
+      "step": 96
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.121891250986305,
+      "learning_rate": 9.99170370820161e-06,
+      "loss": 0.1069,
+      "step": 97
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7807561340523663,
+      "learning_rate": 9.991236538196016e-06,
+      "loss": 0.0973,
+      "step": 98
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2095866667891007,
+      "learning_rate": 9.99075658417737e-06,
+      "loss": 0.1206,
+      "step": 99
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.005155338304001,
+      "learning_rate": 9.990263847374976e-06,
+      "loss": 0.1251,
+      "step": 100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7454326241934206,
+      "learning_rate": 9.989758329050874e-06,
+      "loss": 0.1259,
+      "step": 101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4997644178125036,
+      "learning_rate": 9.989240030499846e-06,
+      "loss": 0.1137,
+      "step": 102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7860502564975003,
+      "learning_rate": 9.988708953049402e-06,
+      "loss": 0.0806,
+      "step": 103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.709783249623465,
+      "learning_rate": 9.988165098059791e-06,
+      "loss": 0.0849,
+      "step": 104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.329249225305103,
+      "learning_rate": 9.987608466923982e-06,
+      "loss": 0.1124,
+      "step": 105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7750026523349485,
+      "learning_rate": 9.987039061067672e-06,
+      "loss": 0.1364,
+      "step": 106
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.747140189403959,
+      "learning_rate": 9.986456881949271e-06,
+      "loss": 0.1191,
+      "step": 107
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.165112231084418,
+      "learning_rate": 9.985861931059914e-06,
+      "loss": 0.1024,
+      "step": 108
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6540364928011821,
+      "learning_rate": 9.985254209923441e-06,
+      "loss": 0.0785,
+      "step": 109
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6773329820403036,
+      "learning_rate": 9.984633720096405e-06,
+      "loss": 0.1381,
+      "step": 110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.342199690015637,
+      "learning_rate": 9.984000463168059e-06,
+      "loss": 0.1246,
+      "step": 111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5558883663248728,
+      "learning_rate": 9.983354440760363e-06,
+      "loss": 0.0819,
+      "step": 112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.170180142906283,
+      "learning_rate": 9.982695654527966e-06,
+      "loss": 0.1205,
+      "step": 113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0678134875414313,
+      "learning_rate": 9.98202410615821e-06,
+      "loss": 0.1162,
+      "step": 114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.364069460723156,
+      "learning_rate": 9.98133979737113e-06,
+      "loss": 0.1225,
+      "step": 115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3581386805595628,
+      "learning_rate": 9.98064272991944e-06,
+      "loss": 0.0847,
+      "step": 116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.200954920692463,
+      "learning_rate": 9.97993290558853e-06,
+      "loss": 0.1115,
+      "step": 117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.471145135562813,
+      "learning_rate": 9.97921032619647e-06,
+      "loss": 0.1026,
+      "step": 118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.810870720141375,
+      "learning_rate": 9.978474993593996e-06,
+      "loss": 0.0932,
+      "step": 119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1161776267633927,
+      "learning_rate": 9.977726909664509e-06,
+      "loss": 0.0858,
+      "step": 120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3126812719954857,
+      "learning_rate": 9.97696607632407e-06,
+      "loss": 0.1193,
+      "step": 121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.38494203181323,
+      "learning_rate": 9.976192495521397e-06,
+      "loss": 0.1434,
+      "step": 122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8894160681506684,
+      "learning_rate": 9.975406169237853e-06,
+      "loss": 0.119,
+      "step": 123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5708693258091966,
+      "learning_rate": 9.974607099487452e-06,
+      "loss": 0.1178,
+      "step": 124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3884674963007506,
+      "learning_rate": 9.973795288316843e-06,
+      "loss": 0.1452,
+      "step": 125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1641350610587784,
+      "learning_rate": 9.972970737805312e-06,
+      "loss": 0.1058,
+      "step": 126
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4935395649774468,
+      "learning_rate": 9.972133450064774e-06,
+      "loss": 0.1261,
+      "step": 127
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.529944326619432,
+      "learning_rate": 9.971283427239764e-06,
+      "loss": 0.1574,
+      "step": 128
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.152117071068294,
+      "learning_rate": 9.970420671507444e-06,
+      "loss": 0.1287,
+      "step": 129
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7931505522057154,
+      "learning_rate": 9.969545185077578e-06,
+      "loss": 0.0834,
+      "step": 130
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5714230930757864,
+      "learning_rate": 9.968656970192545e-06,
+      "loss": 0.1086,
+      "step": 131
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4226134781975723,
+      "learning_rate": 9.967756029127321e-06,
+      "loss": 0.0746,
+      "step": 132
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1304563124483606,
+      "learning_rate": 9.96684236418948e-06,
+      "loss": 0.081,
+      "step": 133
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.997037756466334,
+      "learning_rate": 9.965915977719186e-06,
+      "loss": 0.0828,
+      "step": 134
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8748016550446636,
+      "learning_rate": 9.964976872089181e-06,
+      "loss": 0.1029,
+      "step": 135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1324337837984615,
+      "learning_rate": 9.964025049704791e-06,
+      "loss": 0.0912,
+      "step": 136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.729065553381391,
+      "learning_rate": 9.96306051300391e-06,
+      "loss": 0.1081,
+      "step": 137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0390277690579706,
+      "learning_rate": 9.962083264456999e-06,
+      "loss": 0.0983,
+      "step": 138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.865830909717086,
+      "learning_rate": 9.961093306567076e-06,
+      "loss": 0.114,
+      "step": 139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3533721656364404,
+      "learning_rate": 9.960090641869713e-06,
+      "loss": 0.1084,
+      "step": 140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5551902725664026,
+      "learning_rate": 9.959075272933026e-06,
+      "loss": 0.1113,
+      "step": 141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.071888359696449,
+      "learning_rate": 9.958047202357669e-06,
+      "loss": 0.1174,
+      "step": 142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8366168158210243,
+      "learning_rate": 9.957006432776835e-06,
+      "loss": 0.1237,
+      "step": 143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2814774928629244,
+      "learning_rate": 9.955952966856235e-06,
+      "loss": 0.0816,
+      "step": 144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6787467519373642,
+      "learning_rate": 9.954886807294104e-06,
+      "loss": 0.1389,
+      "step": 145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.776350872381563,
+      "learning_rate": 9.953807956821186e-06,
+      "loss": 0.0973,
+      "step": 146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.743539597608596,
+      "learning_rate": 9.952716418200736e-06,
+      "loss": 0.0925,
+      "step": 147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4252300401619833,
+      "learning_rate": 9.9516121942285e-06,
+      "loss": 0.0931,
+      "step": 148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.263822153457368,
+      "learning_rate": 9.950495287732718e-06,
+      "loss": 0.0909,
+      "step": 149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.140538985744693,
+      "learning_rate": 9.949365701574111e-06,
+      "loss": 0.0912,
+      "step": 150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.138218262804208,
+      "learning_rate": 9.948223438645885e-06,
+      "loss": 0.1138,
+      "step": 151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8680837459099324,
+      "learning_rate": 9.947068501873702e-06,
+      "loss": 0.1038,
+      "step": 152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4482792126490553,
+      "learning_rate": 9.945900894215697e-06,
+      "loss": 0.1034,
+      "step": 153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.585416283038379,
+      "learning_rate": 9.94472061866245e-06,
+      "loss": 0.0943,
+      "step": 154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.242948767982364,
+      "learning_rate": 9.943527678236993e-06,
+      "loss": 0.1127,
+      "step": 155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7899257398914066,
+      "learning_rate": 9.942322075994795e-06,
+      "loss": 0.105,
+      "step": 156
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.796907562819366,
+      "learning_rate": 9.941103815023754e-06,
+      "loss": 0.0979,
+      "step": 157
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.992648855587167,
+      "learning_rate": 9.939872898444193e-06,
+      "loss": 0.0841,
+      "step": 158
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4007767172605858,
+      "learning_rate": 9.938629329408846e-06,
+      "loss": 0.0873,
+      "step": 159
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5723243881307907,
+      "learning_rate": 9.937373111102859e-06,
+      "loss": 0.0946,
+      "step": 160
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.816342977635474,
+      "learning_rate": 9.936104246743773e-06,
+      "loss": 0.0912,
+      "step": 161
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2290152164656765,
+      "learning_rate": 9.934822739581518e-06,
+      "loss": 0.1103,
+      "step": 162
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.487123775788363,
+      "learning_rate": 9.93352859289841e-06,
+      "loss": 0.1247,
+      "step": 163
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3289229263800526,
+      "learning_rate": 9.932221810009136e-06,
+      "loss": 0.0765,
+      "step": 164
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.441273715142216,
+      "learning_rate": 9.930902394260746e-06,
+      "loss": 0.0795,
+      "step": 165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9404987464382477,
+      "learning_rate": 9.92957034903265e-06,
+      "loss": 0.1015,
+      "step": 166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3410400098803896,
+      "learning_rate": 9.928225677736605e-06,
+      "loss": 0.073,
+      "step": 167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0502444988691617,
+      "learning_rate": 9.926868383816707e-06,
+      "loss": 0.0969,
+      "step": 168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0877152198641795,
+      "learning_rate": 9.92549847074938e-06,
+      "loss": 0.1356,
+      "step": 169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9152870160820563,
+      "learning_rate": 9.924115942043372e-06,
+      "loss": 0.1136,
+      "step": 170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0743638774511117,
+      "learning_rate": 9.922720801239744e-06,
+      "loss": 0.1299,
+      "step": 171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7953929763715504,
+      "learning_rate": 9.921313051911856e-06,
+      "loss": 0.0908,
+      "step": 172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9157646131826696,
+      "learning_rate": 9.919892697665366e-06,
+      "loss": 0.0934,
+      "step": 173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0178472764702429,
+      "learning_rate": 9.918459742138216e-06,
+      "loss": 0.0663,
+      "step": 174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.943463027055093,
+      "learning_rate": 9.917014189000622e-06,
+      "loss": 0.1212,
+      "step": 175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0231520392386524,
+      "learning_rate": 9.915556041955067e-06,
+      "loss": 0.1125,
+      "step": 176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0028604714913203,
+      "learning_rate": 9.914085304736292e-06,
+      "loss": 0.0851,
+      "step": 177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3013421574166353,
+      "learning_rate": 9.912601981111287e-06,
+      "loss": 0.0693,
+      "step": 178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.039237793918344,
+      "learning_rate": 9.911106074879272e-06,
+      "loss": 0.1213,
+      "step": 179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4853225489196868,
+      "learning_rate": 9.9095975898717e-06,
+      "loss": 0.0717,
+      "step": 180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5225237739191446,
+      "learning_rate": 9.908076529952244e-06,
+      "loss": 0.0734,
+      "step": 181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8510887662937323,
+      "learning_rate": 9.90654289901678e-06,
+      "loss": 0.0916,
+      "step": 182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9812919920187775,
+      "learning_rate": 9.904996700993383e-06,
+      "loss": 0.1082,
+      "step": 183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6600233528402641,
+      "learning_rate": 9.90343793984232e-06,
+      "loss": 0.1138,
+      "step": 184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.383212353529967,
+      "learning_rate": 9.901866619556033e-06,
+      "loss": 0.0827,
+      "step": 185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.925331137805128,
+      "learning_rate": 9.90028274415913e-06,
+      "loss": 0.1037,
+      "step": 186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.37916222656882,
+      "learning_rate": 9.898686317708381e-06,
+      "loss": 0.0784,
+      "step": 187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2716641949041647,
+      "learning_rate": 9.897077344292695e-06,
+      "loss": 0.0951,
+      "step": 188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9242092279254224,
+      "learning_rate": 9.89545582803313e-06,
+      "loss": 0.0993,
+      "step": 189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.12901206008378,
+      "learning_rate": 9.893821773082852e-06,
+      "loss": 0.1273,
+      "step": 190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.105626108417437,
+      "learning_rate": 9.892175183627161e-06,
+      "loss": 0.1047,
+      "step": 191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.3441589498848545,
+      "learning_rate": 9.890516063883449e-06,
+      "loss": 0.0932,
+      "step": 192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8098066422942596,
+      "learning_rate": 9.888844418101207e-06,
+      "loss": 0.1016,
+      "step": 193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0529364761864364,
+      "learning_rate": 9.887160250562007e-06,
+      "loss": 0.1207,
+      "step": 194
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9596501504369412,
+      "learning_rate": 9.885463565579494e-06,
+      "loss": 0.0956,
+      "step": 195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1140435352484783,
+      "learning_rate": 9.883754367499372e-06,
+      "loss": 0.0734,
+      "step": 196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7698226079660913,
+      "learning_rate": 9.8820326606994e-06,
+      "loss": 0.1388,
+      "step": 197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6701850513217271,
+      "learning_rate": 9.880298449589365e-06,
+      "loss": 0.0958,
+      "step": 198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2916206595956448,
+      "learning_rate": 9.878551738611092e-06,
+      "loss": 0.1151,
+      "step": 199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6787179895305941,
+      "learning_rate": 9.876792532238417e-06,
+      "loss": 0.0862,
+      "step": 200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4125068562773224,
+      "learning_rate": 9.87502083497718e-06,
+      "loss": 0.1087,
+      "step": 201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3258740208971436,
+      "learning_rate": 9.873236651365215e-06,
+      "loss": 0.0681,
+      "step": 202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1361575934052066,
+      "learning_rate": 9.871439985972338e-06,
+      "loss": 0.0855,
+      "step": 203
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.417534590578083,
+      "learning_rate": 9.869630843400331e-06,
+      "loss": 0.064,
+      "step": 204
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0072193710609736,
+      "learning_rate": 9.867809228282937e-06,
+      "loss": 0.0842,
+      "step": 205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.084886745710113,
+      "learning_rate": 9.865975145285842e-06,
+      "loss": 0.0837,
+      "step": 206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.2565455054149535,
+      "learning_rate": 9.864128599106672e-06,
+      "loss": 0.093,
+      "step": 207
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4752672341475757,
+      "learning_rate": 9.862269594474967e-06,
+      "loss": 0.1319,
+      "step": 208
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2901130251571824,
+      "learning_rate": 9.86039813615218e-06,
+      "loss": 0.0852,
+      "step": 209
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4976044292493496,
+      "learning_rate": 9.858514228931664e-06,
+      "loss": 0.0856,
+      "step": 210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8407522923248512,
+      "learning_rate": 9.856617877638656e-06,
+      "loss": 0.0928,
+      "step": 211
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.476856421690967,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.0785,
+      "step": 212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6556235800226011,
+      "learning_rate": 9.85278786229545e-06,
+      "loss": 0.076,
+      "step": 213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.534209397246005,
+      "learning_rate": 9.850854208055043e-06,
+      "loss": 0.0913,
+      "step": 214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.409274827048923,
+      "learning_rate": 9.848908129361689e-06,
+      "loss": 0.0856,
+      "step": 215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6845946861454264,
+      "learning_rate": 9.846949631199865e-06,
+      "loss": 0.0765,
+      "step": 216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0629627391913066,
+      "learning_rate": 9.844978718585855e-06,
+      "loss": 0.0907,
+      "step": 217
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0848081390621207,
+      "learning_rate": 9.842995396567744e-06,
+      "loss": 0.0755,
+      "step": 218
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.937175272950418,
+      "learning_rate": 9.840999670225396e-06,
+      "loss": 0.0825,
+      "step": 219
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3382983974983316,
+      "learning_rate": 9.83899154467045e-06,
+      "loss": 0.1139,
+      "step": 220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8713367237587217,
+      "learning_rate": 9.8369710250463e-06,
+      "loss": 0.0955,
+      "step": 221
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.04053344806577,
+      "learning_rate": 9.834938116528086e-06,
+      "loss": 0.0851,
+      "step": 222
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8318815354282614,
+      "learning_rate": 9.832892824322681e-06,
+      "loss": 0.1112,
+      "step": 223
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.8606361693646125,
+      "learning_rate": 9.830835153668675e-06,
+      "loss": 0.1507,
+      "step": 224
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7538317703216597,
+      "learning_rate": 9.828765109836362e-06,
+      "loss": 0.0914,
+      "step": 225
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9703680557018663,
+      "learning_rate": 9.826682698127731e-06,
+      "loss": 0.1028,
+      "step": 226
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1041065295008856,
+      "learning_rate": 9.824587923876442e-06,
+      "loss": 0.1184,
+      "step": 227
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.39277051465413,
+      "learning_rate": 9.822480792447826e-06,
+      "loss": 0.1179,
+      "step": 228
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7491779318374767,
+      "learning_rate": 9.82036130923886e-06,
+      "loss": 0.0969,
+      "step": 229
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.726332596955443,
+      "learning_rate": 9.81822947967816e-06,
+      "loss": 0.0698,
+      "step": 230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0927820863569315,
+      "learning_rate": 9.81608530922596e-06,
+      "loss": 0.0586,
+      "step": 231
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3063159183295077,
+      "learning_rate": 9.81392880337411e-06,
+      "loss": 0.0859,
+      "step": 232
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9256645726866466,
+      "learning_rate": 9.81175996764605e-06,
+      "loss": 0.1002,
+      "step": 233
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.631213101565748,
+      "learning_rate": 9.8095788075968e-06,
+      "loss": 0.0905,
+      "step": 234
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.758854636637926,
+      "learning_rate": 9.807385328812948e-06,
+      "loss": 0.076,
+      "step": 235
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.051432984550155,
+      "learning_rate": 9.805179536912635e-06,
+      "loss": 0.0792,
+      "step": 236
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1924640722920685,
+      "learning_rate": 9.802961437545532e-06,
+      "loss": 0.0997,
+      "step": 237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.414190947211591,
+      "learning_rate": 9.800731036392846e-06,
+      "loss": 0.0715,
+      "step": 238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.039154923532322,
+      "learning_rate": 9.79848833916728e-06,
+      "loss": 0.112,
+      "step": 239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.277383152775661,
+      "learning_rate": 9.796233351613036e-06,
+      "loss": 0.1544,
+      "step": 240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.823021976103906,
+      "learning_rate": 9.793966079505797e-06,
+      "loss": 0.0842,
+      "step": 241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.912512017914575,
+      "learning_rate": 9.791686528652708e-06,
+      "loss": 0.0974,
+      "step": 242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9153410476183683,
+      "learning_rate": 9.789394704892364e-06,
+      "loss": 0.1026,
+      "step": 243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.807843325639007,
+      "learning_rate": 9.787090614094795e-06,
+      "loss": 0.0807,
+      "step": 244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4132064235794966,
+      "learning_rate": 9.784774262161448e-06,
+      "loss": 0.0944,
+      "step": 245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5293791979763536,
+      "learning_rate": 9.782445655025175e-06,
+      "loss": 0.115,
+      "step": 246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.428020426458093,
+      "learning_rate": 9.780104798650221e-06,
+      "loss": 0.0797,
+      "step": 247
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.016570138879204,
+      "learning_rate": 9.777751699032203e-06,
+      "loss": 0.0894,
+      "step": 248
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0703814315935216,
+      "learning_rate": 9.775386362198093e-06,
+      "loss": 0.0686,
+      "step": 249
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.246554230901814,
+      "learning_rate": 9.773008794206209e-06,
+      "loss": 0.0863,
+      "step": 250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2872072039177085,
+      "learning_rate": 9.770619001146198e-06,
+      "loss": 0.0961,
+      "step": 251
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4319299004238022,
+      "learning_rate": 9.768216989139017e-06,
+      "loss": 0.0741,
+      "step": 252
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.665388204235485,
+      "learning_rate": 9.765802764336917e-06,
+      "loss": 0.0858,
+      "step": 253
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.083220714768147,
+      "learning_rate": 9.763376332923435e-06,
+      "loss": 0.1004,
+      "step": 254
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8288065714795982,
+      "learning_rate": 9.760937701113368e-06,
+      "loss": 0.082,
+      "step": 255
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.42391892235683,
+      "learning_rate": 9.758486875152766e-06,
+      "loss": 0.064,
+      "step": 256
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2151765663560514,
+      "learning_rate": 9.756023861318907e-06,
+      "loss": 0.1338,
+      "step": 257
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.149962952777695,
+      "learning_rate": 9.75354866592029e-06,
+      "loss": 0.0832,
+      "step": 258
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7962412801494425,
+      "learning_rate": 9.75106129529661e-06,
+      "loss": 0.0816,
+      "step": 259
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3866114809470351,
+      "learning_rate": 9.748561755818751e-06,
+      "loss": 0.0848,
+      "step": 260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2395857876895726,
+      "learning_rate": 9.746050053888761e-06,
+      "loss": 0.1089,
+      "step": 261
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1697229926909465,
+      "learning_rate": 9.743526195939844e-06,
+      "loss": 0.1046,
+      "step": 262
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8188765207908812,
+      "learning_rate": 9.740990188436336e-06,
+      "loss": 0.0854,
+      "step": 263
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6284142753747066,
+      "learning_rate": 9.73844203787369e-06,
+      "loss": 0.0848,
+      "step": 264
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5105954281748584,
+      "learning_rate": 9.735881750778463e-06,
+      "loss": 0.0816,
+      "step": 265
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4226387847216695,
+      "learning_rate": 9.733309333708303e-06,
+      "loss": 0.1207,
+      "step": 266
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2394786486778886,
+      "learning_rate": 9.730724793251914e-06,
+      "loss": 0.0801,
+      "step": 267
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1909334570172754,
+      "learning_rate": 9.72812813602906e-06,
+      "loss": 0.1096,
+      "step": 268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8150706378609962,
+      "learning_rate": 9.725519368690539e-06,
+      "loss": 0.0795,
+      "step": 269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8365063330132647,
+      "learning_rate": 9.722898497918165e-06,
+      "loss": 0.0919,
+      "step": 270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.664954614981809,
+      "learning_rate": 9.720265530424752e-06,
+      "loss": 0.1014,
+      "step": 271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7255068806392266,
+      "learning_rate": 9.717620472954098e-06,
+      "loss": 0.0803,
+      "step": 272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.77913692453882,
+      "learning_rate": 9.714963332280967e-06,
+      "loss": 0.0962,
+      "step": 273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4814102439238948,
+      "learning_rate": 9.712294115211074e-06,
+      "loss": 0.0913,
+      "step": 274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.253001053387503,
+      "learning_rate": 9.709612828581058e-06,
+      "loss": 0.0763,
+      "step": 275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7833683606000394,
+      "learning_rate": 9.70691947925848e-06,
+      "loss": 0.0775,
+      "step": 276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4058700337148404,
+      "learning_rate": 9.704214074141793e-06,
+      "loss": 0.0874,
+      "step": 277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3466667907354781,
+      "learning_rate": 9.701496620160327e-06,
+      "loss": 0.0704,
+      "step": 278
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1309554442052114,
+      "learning_rate": 9.698767124274276e-06,
+      "loss": 0.1082,
+      "step": 279
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5829938262087564,
+      "learning_rate": 9.696025593474676e-06,
+      "loss": 0.0826,
+      "step": 280
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5487626505131384,
+      "learning_rate": 9.693272034783385e-06,
+      "loss": 0.0889,
+      "step": 281
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6596439121539222,
+      "learning_rate": 9.690506455253073e-06,
+      "loss": 0.099,
+      "step": 282
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.69312107368622,
+      "learning_rate": 9.687728861967192e-06,
+      "loss": 0.0737,
+      "step": 283
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.355971373182455,
+      "learning_rate": 9.684939262039974e-06,
+      "loss": 0.0688,
+      "step": 284
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2553069216995882,
+      "learning_rate": 9.682137662616393e-06,
+      "loss": 0.0604,
+      "step": 285
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4072716167061843,
+      "learning_rate": 9.679324070872168e-06,
+      "loss": 0.061,
+      "step": 286
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3801539197582835,
+      "learning_rate": 9.676498494013724e-06,
+      "loss": 0.068,
+      "step": 287
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0356080364751667,
+      "learning_rate": 9.67366093927819e-06,
+      "loss": 0.0797,
+      "step": 288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5713160560356783,
+      "learning_rate": 9.670811413933372e-06,
+      "loss": 0.0581,
+      "step": 289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1345366422930265,
+      "learning_rate": 9.667949925277733e-06,
+      "loss": 0.0583,
+      "step": 290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6352674907213263,
+      "learning_rate": 9.665076480640383e-06,
+      "loss": 0.074,
+      "step": 291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4038554071150915,
+      "learning_rate": 9.662191087381051e-06,
+      "loss": 0.0543,
+      "step": 292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1170162395093253,
+      "learning_rate": 9.659293752890072e-06,
+      "loss": 0.1129,
+      "step": 293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.708030080537653,
+      "learning_rate": 9.656384484588363e-06,
+      "loss": 0.0907,
+      "step": 294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7196125677896754,
+      "learning_rate": 9.65346328992741e-06,
+      "loss": 0.0874,
+      "step": 295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.160266089759155,
+      "learning_rate": 9.650530176389245e-06,
+      "loss": 0.0771,
+      "step": 296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2038797082738868,
+      "learning_rate": 9.64758515148643e-06,
+      "loss": 0.0809,
+      "step": 297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7119010674264274,
+      "learning_rate": 9.644628222762027e-06,
+      "loss": 0.0707,
+      "step": 298
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.087952105670146,
+      "learning_rate": 9.641659397789599e-06,
+      "loss": 0.0469,
+      "step": 299
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.956435229689038,
+      "learning_rate": 9.63867868417317e-06,
+      "loss": 0.1137,
+      "step": 300
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9933370820366156,
+      "learning_rate": 9.635686089547219e-06,
+      "loss": 0.069,
+      "step": 301
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.081689772329396,
+      "learning_rate": 9.632681621576651e-06,
+      "loss": 0.1044,
+      "step": 302
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2096559086670005,
+      "learning_rate": 9.629665287956788e-06,
+      "loss": 0.0821,
+      "step": 303
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.484080717855697,
+      "learning_rate": 9.626637096413339e-06,
+      "loss": 0.0693,
+      "step": 304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7412247084697006,
+      "learning_rate": 9.623597054702388e-06,
+      "loss": 0.0699,
+      "step": 305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9095850402493102,
+      "learning_rate": 9.620545170610364e-06,
+      "loss": 0.096,
+      "step": 306
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9340127820809814,
+      "learning_rate": 9.61748145195404e-06,
+      "loss": 0.0881,
+      "step": 307
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.063160163996744,
+      "learning_rate": 9.614405906580486e-06,
+      "loss": 0.0947,
+      "step": 308
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4516427837059722,
+      "learning_rate": 9.611318542367076e-06,
+      "loss": 0.0718,
+      "step": 309
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6438745401278574,
+      "learning_rate": 9.60821936722145e-06,
+      "loss": 0.0708,
+      "step": 310
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7053858895529537,
+      "learning_rate": 9.605108389081497e-06,
+      "loss": 0.0815,
+      "step": 311
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.495838987672316,
+      "learning_rate": 9.601985615915342e-06,
+      "loss": 0.088,
+      "step": 312
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.182730205772672,
+      "learning_rate": 9.598851055721317e-06,
+      "loss": 0.0956,
+      "step": 313
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.355337420631783,
+      "learning_rate": 9.595704716527946e-06,
+      "loss": 0.0618,
+      "step": 314
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9182810450895473,
+      "learning_rate": 9.592546606393922e-06,
+      "loss": 0.1141,
+      "step": 315
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3996443318689997,
+      "learning_rate": 9.589376733408086e-06,
+      "loss": 0.068,
+      "step": 316
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0459197241210223,
+      "learning_rate": 9.586195105689406e-06,
+      "loss": 0.0631,
+      "step": 317
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0378716872013047,
+      "learning_rate": 9.583001731386962e-06,
+      "loss": 0.1295,
+      "step": 318
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6754804382652202,
+      "learning_rate": 9.579796618679917e-06,
+      "loss": 0.0889,
+      "step": 319
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.216230794120504,
+      "learning_rate": 9.576579775777497e-06,
+      "loss": 0.095,
+      "step": 320
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4318931227506908,
+      "learning_rate": 9.573351210918976e-06,
+      "loss": 0.0766,
+      "step": 321
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6404177698368994,
+      "learning_rate": 9.570110932373652e-06,
+      "loss": 0.0901,
+      "step": 322
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3109869919883583,
+      "learning_rate": 9.566858948440823e-06,
+      "loss": 0.063,
+      "step": 323
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8968886949685448,
+      "learning_rate": 9.56359526744977e-06,
+      "loss": 0.0965,
+      "step": 324
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5650335354640215,
+      "learning_rate": 9.560319897759729e-06,
+      "loss": 0.099,
+      "step": 325
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.435820283477607,
+      "learning_rate": 9.557032847759879e-06,
+      "loss": 0.075,
+      "step": 326
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3933170068610747,
+      "learning_rate": 9.553734125869316e-06,
+      "loss": 0.0721,
+      "step": 327
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0619677922888786,
+      "learning_rate": 9.550423740537028e-06,
+      "loss": 0.0868,
+      "step": 328
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2967042911447932,
+      "learning_rate": 9.547101700241878e-06,
+      "loss": 0.1052,
+      "step": 329
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.131619635803148,
+      "learning_rate": 9.543768013492578e-06,
+      "loss": 0.1245,
+      "step": 330
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5230807628829646,
+      "learning_rate": 9.540422688827677e-06,
+      "loss": 0.0829,
+      "step": 331
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9367377983487091,
+      "learning_rate": 9.537065734815523e-06,
+      "loss": 0.0662,
+      "step": 332
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.066131384955803,
+      "learning_rate": 9.533697160054256e-06,
+      "loss": 0.0801,
+      "step": 333
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.597529173903688,
+      "learning_rate": 9.53031697317178e-06,
+      "loss": 0.0916,
+      "step": 334
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2965629730886212,
+      "learning_rate": 9.52692518282574e-06,
+      "loss": 0.0799,
+      "step": 335
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2843598864453172,
+      "learning_rate": 9.5235217977035e-06,
+      "loss": 0.0658,
+      "step": 336
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.624715892770218,
+      "learning_rate": 9.520106826522123e-06,
+      "loss": 0.0933,
+      "step": 337
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1720007241534482,
+      "learning_rate": 9.516680278028343e-06,
+      "loss": 0.0527,
+      "step": 338
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3607490046691666,
+      "learning_rate": 9.513242160998555e-06,
+      "loss": 0.0742,
+      "step": 339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3952909438425838,
+      "learning_rate": 9.509792484238776e-06,
+      "loss": 0.0715,
+      "step": 340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7424553551612125,
+      "learning_rate": 9.506331256584637e-06,
+      "loss": 0.0963,
+      "step": 341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.578495355923729,
+      "learning_rate": 9.502858486901354e-06,
+      "loss": 0.113,
+      "step": 342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2366093718995437,
+      "learning_rate": 9.4993741840837e-06,
+      "loss": 0.0901,
+      "step": 343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.983175168502808,
+      "learning_rate": 9.495878357055992e-06,
+      "loss": 0.1064,
+      "step": 344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9500124640313773,
+      "learning_rate": 9.492371014772063e-06,
+      "loss": 0.0427,
+      "step": 345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8851187458756449,
+      "learning_rate": 9.48885216621524e-06,
+      "loss": 0.0579,
+      "step": 346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.622075510857947,
+      "learning_rate": 9.485321820398321e-06,
+      "loss": 0.0668,
+      "step": 347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6580856557518695,
+      "learning_rate": 9.481779986363552e-06,
+      "loss": 0.0789,
+      "step": 348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.512095570610664,
+      "learning_rate": 9.478226673182602e-06,
+      "loss": 0.069,
+      "step": 349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6702707971480697,
+      "learning_rate": 9.474661889956544e-06,
+      "loss": 0.0994,
+      "step": 350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1155812559129124,
+      "learning_rate": 9.471085645815828e-06,
+      "loss": 0.0884,
+      "step": 351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0091628967886317,
+      "learning_rate": 9.467497949920256e-06,
+      "loss": 0.0753,
+      "step": 352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9773778602091453,
+      "learning_rate": 9.463898811458968e-06,
+      "loss": 0.0535,
+      "step": 353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3011646246801316,
+      "learning_rate": 9.460288239650409e-06,
+      "loss": 0.096,
+      "step": 354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6098500437903527,
+      "learning_rate": 9.4566662437423e-06,
+      "loss": 0.0645,
+      "step": 355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8545640076569774,
+      "learning_rate": 9.453032833011633e-06,
+      "loss": 0.0535,
+      "step": 356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.13656418875037,
+      "learning_rate": 9.449388016764634e-06,
+      "loss": 0.0536,
+      "step": 357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1924165470684622,
+      "learning_rate": 9.44573180433674e-06,
+      "loss": 0.11,
+      "step": 358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8340819553384473,
+      "learning_rate": 9.442064205092578e-06,
+      "loss": 0.0812,
+      "step": 359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5107648883572784,
+      "learning_rate": 9.43838522842594e-06,
+      "loss": 0.0834,
+      "step": 360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9803211912182362,
+      "learning_rate": 9.434694883759757e-06,
+      "loss": 0.0527,
+      "step": 361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.113097249223458,
+      "learning_rate": 9.43099318054608e-06,
+      "loss": 0.1023,
+      "step": 362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7775339990135852,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.0714,
+      "step": 363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6899887297282574,
+      "learning_rate": 9.42355573642988e-06,
+      "loss": 0.0728,
+      "step": 364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7074309015837548,
+      "learning_rate": 9.419820014576818e-06,
+      "loss": 0.0687,
+      "step": 365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8657976560603164,
+      "learning_rate": 9.416072972275144e-06,
+      "loss": 0.1106,
+      "step": 366
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2617207451485761,
+      "learning_rate": 9.412314619122124e-06,
+      "loss": 0.0524,
+      "step": 367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2819250350418456,
+      "learning_rate": 9.408544964743996e-06,
+      "loss": 0.0807,
+      "step": 368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1937701721663347,
+      "learning_rate": 9.404764018795947e-06,
+      "loss": 0.0518,
+      "step": 369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2777976949870833,
+      "learning_rate": 9.400971790962081e-06,
+      "loss": 0.0563,
+      "step": 370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4766786972271955,
+      "learning_rate": 9.3971682909554e-06,
+      "loss": 0.0751,
+      "step": 371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5716995031811691,
+      "learning_rate": 9.393353528517778e-06,
+      "loss": 0.0532,
+      "step": 372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8565525773799597,
+      "learning_rate": 9.389527513419935e-06,
+      "loss": 0.0546,
+      "step": 373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2523781687323514,
+      "learning_rate": 9.38569025546141e-06,
+      "loss": 0.0663,
+      "step": 374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5941548292464838,
+      "learning_rate": 9.381841764470543e-06,
+      "loss": 0.0849,
+      "step": 375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9991597120815358,
+      "learning_rate": 9.377982050304441e-06,
+      "loss": 0.0468,
+      "step": 376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2320914466684454,
+      "learning_rate": 9.374111122848957e-06,
+      "loss": 0.0932,
+      "step": 377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5056596004131972,
+      "learning_rate": 9.370228992018666e-06,
+      "loss": 0.0737,
+      "step": 378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.229304673858704,
+      "learning_rate": 9.36633566775684e-06,
+      "loss": 0.0602,
+      "step": 379
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5220192209727423,
+      "learning_rate": 9.362431160035415e-06,
+      "loss": 0.0698,
+      "step": 380
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7867448206398417,
+      "learning_rate": 9.358515478854977e-06,
+      "loss": 0.0849,
+      "step": 381
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0935869033882721,
+      "learning_rate": 9.354588634244728e-06,
+      "loss": 0.0748,
+      "step": 382
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.532490450509265,
+      "learning_rate": 9.35065063626246e-06,
+      "loss": 0.0703,
+      "step": 383
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4938766268199541,
+      "learning_rate": 9.346701494994537e-06,
+      "loss": 0.0704,
+      "step": 384
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3175682655861047,
+      "learning_rate": 9.342741220555862e-06,
+      "loss": 0.0658,
+      "step": 385
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8799920557327616,
+      "learning_rate": 9.338769823089853e-06,
+      "loss": 0.072,
+      "step": 386
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.091485141339395,
+      "learning_rate": 9.334787312768419e-06,
+      "loss": 0.1154,
+      "step": 387
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1795003198750242,
+      "learning_rate": 9.330793699791929e-06,
+      "loss": 0.0644,
+      "step": 388
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.29652939495859,
+      "learning_rate": 9.326788994389191e-06,
+      "loss": 0.1329,
+      "step": 389
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.755267217555506,
+      "learning_rate": 9.322773206817425e-06,
+      "loss": 0.0632,
+      "step": 390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5860689233573197,
+      "learning_rate": 9.318746347362234e-06,
+      "loss": 0.0723,
+      "step": 391
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9815352814144065,
+      "learning_rate": 9.31470842633758e-06,
+      "loss": 0.0838,
+      "step": 392
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4741872607662527,
+      "learning_rate": 9.310659454085757e-06,
+      "loss": 0.0663,
+      "step": 393
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.330523711040219,
+      "learning_rate": 9.306599440977364e-06,
+      "loss": 0.0671,
+      "step": 394
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.538380618597983,
+      "learning_rate": 9.302528397411282e-06,
+      "loss": 0.0898,
+      "step": 395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4259119324898288,
+      "learning_rate": 9.298446333814634e-06,
+      "loss": 0.077,
+      "step": 396
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4798744530874832,
+      "learning_rate": 9.294353260642785e-06,
+      "loss": 0.0697,
+      "step": 397
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.546206268272879,
+      "learning_rate": 9.290249188379284e-06,
+      "loss": 0.0627,
+      "step": 398
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9537128285464975,
+      "learning_rate": 9.286134127535859e-06,
+      "loss": 0.0867,
+      "step": 399
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3995970063745107,
+      "learning_rate": 9.282008088652381e-06,
+      "loss": 0.0795,
+      "step": 400
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5437480493944655,
+      "learning_rate": 9.27787108229684e-06,
+      "loss": 0.0605,
+      "step": 401
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3231852490977054,
+      "learning_rate": 9.273723119065316e-06,
+      "loss": 0.0642,
+      "step": 402
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0983511223510702,
+      "learning_rate": 9.269564209581953e-06,
+      "loss": 0.0698,
+      "step": 403
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9638250329004523,
+      "learning_rate": 9.265394364498933e-06,
+      "loss": 0.0613,
+      "step": 404
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1237185184057257,
+      "learning_rate": 9.261213594496443e-06,
+      "loss": 0.0605,
+      "step": 405
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.111228491358394,
+      "learning_rate": 9.257021910282658e-06,
+      "loss": 0.0594,
+      "step": 406
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.412987219933755,
+      "learning_rate": 9.252819322593702e-06,
+      "loss": 0.0632,
+      "step": 407
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.238843236884171,
+      "learning_rate": 9.248605842193628e-06,
+      "loss": 0.0562,
+      "step": 408
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5281808809312805,
+      "learning_rate": 9.24438147987439e-06,
+      "loss": 0.0645,
+      "step": 409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.360991380105523,
+      "learning_rate": 9.240146246455814e-06,
+      "loss": 0.0753,
+      "step": 410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2922523007020124,
+      "learning_rate": 9.235900152785564e-06,
+      "loss": 0.0434,
+      "step": 411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4373970506401919,
+      "learning_rate": 9.231643209739128e-06,
+      "loss": 0.0927,
+      "step": 412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.034620905490409,
+      "learning_rate": 9.227375428219777e-06,
+      "loss": 0.0797,
+      "step": 413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4667858443208852,
+      "learning_rate": 9.22309681915855e-06,
+      "loss": 0.0542,
+      "step": 414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1848260113622535,
+      "learning_rate": 9.218807393514207e-06,
+      "loss": 0.0788,
+      "step": 415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9929252392688195,
+      "learning_rate": 9.21450716227322e-06,
+      "loss": 0.0633,
+      "step": 416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4775624542456316,
+      "learning_rate": 9.210196136449738e-06,
+      "loss": 0.0664,
+      "step": 417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.235653428937132,
+      "learning_rate": 9.205874327085554e-06,
+      "loss": 0.0664,
+      "step": 418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4916279720222827,
+      "learning_rate": 9.201541745250084e-06,
+      "loss": 0.0651,
+      "step": 419
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.346805901014103,
+      "learning_rate": 9.197198402040335e-06,
+      "loss": 0.0668,
+      "step": 420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7380719887472837,
+      "learning_rate": 9.192844308580872e-06,
+      "loss": 0.1061,
+      "step": 421
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1977746654700658,
+      "learning_rate": 9.188479476023803e-06,
+      "loss": 0.0513,
+      "step": 422
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4928135949497092,
+      "learning_rate": 9.184103915548737e-06,
+      "loss": 0.0527,
+      "step": 423
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0948400388091197,
+      "learning_rate": 9.179717638362756e-06,
+      "loss": 0.0609,
+      "step": 424
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.367623112561712,
+      "learning_rate": 9.175320655700407e-06,
+      "loss": 0.0636,
+      "step": 425
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9122244810169786,
+      "learning_rate": 9.170912978823634e-06,
+      "loss": 0.0924,
+      "step": 426
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7194683882049424,
+      "learning_rate": 9.166494619021788e-06,
+      "loss": 0.0869,
+      "step": 427
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8058781739102134,
+      "learning_rate": 9.162065587611578e-06,
+      "loss": 0.0498,
+      "step": 428
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9174740077703543,
+      "learning_rate": 9.157625895937045e-06,
+      "loss": 0.049,
+      "step": 429
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.538856642201719,
+      "learning_rate": 9.153175555369536e-06,
+      "loss": 0.0762,
+      "step": 430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1280379258945326,
+      "learning_rate": 9.14871457730767e-06,
+      "loss": 0.0493,
+      "step": 431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.546891200724216,
+      "learning_rate": 9.144242973177311e-06,
+      "loss": 0.0579,
+      "step": 432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1259850013209407,
+      "learning_rate": 9.139760754431549e-06,
+      "loss": 0.0561,
+      "step": 433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.390978156286863,
+      "learning_rate": 9.135267932550647e-06,
+      "loss": 0.0758,
+      "step": 434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.314675628407316,
+      "learning_rate": 9.130764519042036e-06,
+      "loss": 0.0511,
+      "step": 435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0433867643994965,
+      "learning_rate": 9.126250525440273e-06,
+      "loss": 0.0421,
+      "step": 436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.955888645256769,
+      "learning_rate": 9.12172596330701e-06,
+      "loss": 0.0874,
+      "step": 437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5882258901883926,
+      "learning_rate": 9.117190844230971e-06,
+      "loss": 0.0507,
+      "step": 438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.900326904241668,
+      "learning_rate": 9.112645179827922e-06,
+      "loss": 0.0686,
+      "step": 439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.011111441922898,
+      "learning_rate": 9.108088981740633e-06,
+      "loss": 0.0862,
+      "step": 440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4966326512252073,
+      "learning_rate": 9.103522261638857e-06,
+      "loss": 0.074,
+      "step": 441
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4385678568699105,
+      "learning_rate": 9.098945031219297e-06,
+      "loss": 0.0734,
+      "step": 442
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5204707684735033,
+      "learning_rate": 9.094357302205575e-06,
+      "loss": 0.0841,
+      "step": 443
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2937317248210154,
+      "learning_rate": 9.089759086348204e-06,
+      "loss": 0.0587,
+      "step": 444
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3861131844899162,
+      "learning_rate": 9.085150395424557e-06,
+      "loss": 0.0526,
+      "step": 445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.749413906607524,
+      "learning_rate": 9.080531241238836e-06,
+      "loss": 0.0681,
+      "step": 446
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1655901911911675,
+      "learning_rate": 9.075901635622041e-06,
+      "loss": 0.0651,
+      "step": 447
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.563849142939697,
+      "learning_rate": 9.071261590431945e-06,
+      "loss": 0.0615,
+      "step": 448
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9666664856605189,
+      "learning_rate": 9.066611117553058e-06,
+      "loss": 0.0547,
+      "step": 449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6304500518473257,
+      "learning_rate": 9.061950228896594e-06,
+      "loss": 0.0719,
+      "step": 450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.208068296719404,
+      "learning_rate": 9.057278936400453e-06,
+      "loss": 0.0499,
+      "step": 451
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2910850367163424,
+      "learning_rate": 9.052597252029175e-06,
+      "loss": 0.0506,
+      "step": 452
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0096660024374238,
+      "learning_rate": 9.047905187773922e-06,
+      "loss": 0.0598,
+      "step": 453
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3017894716148746,
+      "learning_rate": 9.043202755652436e-06,
+      "loss": 0.093,
+      "step": 454
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1809136704456458,
+      "learning_rate": 9.038489967709023e-06,
+      "loss": 0.0717,
+      "step": 455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4118385240601037,
+      "learning_rate": 9.033766836014504e-06,
+      "loss": 0.0607,
+      "step": 456
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.190671617119357,
+      "learning_rate": 9.029033372666199e-06,
+      "loss": 0.0676,
+      "step": 457
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4608113102052052,
+      "learning_rate": 9.024289589787885e-06,
+      "loss": 0.0912,
+      "step": 458
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0716327370456122,
+      "learning_rate": 9.019535499529781e-06,
+      "loss": 0.0704,
+      "step": 459
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9547421435045815,
+      "learning_rate": 9.014771114068492e-06,
+      "loss": 0.0786,
+      "step": 460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5264171632599215,
+      "learning_rate": 9.009996445607004e-06,
+      "loss": 0.063,
+      "step": 461
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3271932708271126,
+      "learning_rate": 9.005211506374636e-06,
+      "loss": 0.045,
+      "step": 462
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.455795720502849,
+      "learning_rate": 9.00041630862701e-06,
+      "loss": 0.0755,
+      "step": 463
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5728504984245728,
+      "learning_rate": 8.99561086464603e-06,
+      "loss": 0.0689,
+      "step": 464
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0184694441905213,
+      "learning_rate": 8.990795186739838e-06,
+      "loss": 0.0465,
+      "step": 465
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.711545145839596,
+      "learning_rate": 8.985969287242791e-06,
+      "loss": 0.0713,
+      "step": 466
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6749239768622135,
+      "learning_rate": 8.981133178515424e-06,
+      "loss": 0.0877,
+      "step": 467
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4534332873265494,
+      "learning_rate": 8.976286872944425e-06,
+      "loss": 0.0534,
+      "step": 468
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7939842165258273,
+      "learning_rate": 8.971430382942597e-06,
+      "loss": 0.0655,
+      "step": 469
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.45853725629627,
+      "learning_rate": 8.966563720948824e-06,
+      "loss": 0.1031,
+      "step": 470
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.177368183337995,
+      "learning_rate": 8.96168689942805e-06,
+      "loss": 0.0606,
+      "step": 471
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.235444390625768,
+      "learning_rate": 8.956799930871238e-06,
+      "loss": 0.0688,
+      "step": 472
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.452038429310052,
+      "learning_rate": 8.95190282779534e-06,
+      "loss": 0.0655,
+      "step": 473
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4724243348741484,
+      "learning_rate": 8.946995602743267e-06,
+      "loss": 0.0603,
+      "step": 474
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8899006566216181,
+      "learning_rate": 8.94207826828385e-06,
+      "loss": 0.0533,
+      "step": 475
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4619154759372903,
+      "learning_rate": 8.937150837011822e-06,
+      "loss": 0.0754,
+      "step": 476
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9738974265046594,
+      "learning_rate": 8.932213321547769e-06,
+      "loss": 0.0507,
+      "step": 477
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5501236401082463,
+      "learning_rate": 8.927265734538111e-06,
+      "loss": 0.0565,
+      "step": 478
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1332840368532724,
+      "learning_rate": 8.92230808865506e-06,
+      "loss": 0.051,
+      "step": 479
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2256642469778365,
+      "learning_rate": 8.917340396596594e-06,
+      "loss": 0.0835,
+      "step": 480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7683836409006217,
+      "learning_rate": 8.912362671086424e-06,
+      "loss": 0.0805,
+      "step": 481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1807223874785262,
+      "learning_rate": 8.907374924873953e-06,
+      "loss": 0.0606,
+      "step": 482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3547721846092768,
+      "learning_rate": 8.902377170734258e-06,
+      "loss": 0.0615,
+      "step": 483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5945685777575744,
+      "learning_rate": 8.897369421468042e-06,
+      "loss": 0.0934,
+      "step": 484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9304262314397095,
+      "learning_rate": 8.892351689901618e-06,
+      "loss": 0.0498,
+      "step": 485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5883917689539437,
+      "learning_rate": 8.887323988886853e-06,
+      "loss": 0.063,
+      "step": 486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.322868869946993,
+      "learning_rate": 8.882286331301162e-06,
+      "loss": 0.0786,
+      "step": 487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0519135254924814,
+      "learning_rate": 8.877238730047454e-06,
+      "loss": 0.0596,
+      "step": 488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.624611435100832,
+      "learning_rate": 8.872181198054109e-06,
+      "loss": 0.0605,
+      "step": 489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7411620989235403,
+      "learning_rate": 8.86711374827494e-06,
+      "loss": 0.0644,
+      "step": 490
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2561485952196334,
+      "learning_rate": 8.862036393689167e-06,
+      "loss": 0.0609,
+      "step": 491
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.565086561625876,
+      "learning_rate": 8.856949147301375e-06,
+      "loss": 0.0802,
+      "step": 492
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.118910169047855,
+      "learning_rate": 8.851852022141486e-06,
+      "loss": 0.0738,
+      "step": 493
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8276607532145808,
+      "learning_rate": 8.846745031264727e-06,
+      "loss": 0.0836,
+      "step": 494
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.573960942810051,
+      "learning_rate": 8.84162818775159e-06,
+      "loss": 0.1106,
+      "step": 495
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.338345683797441,
+      "learning_rate": 8.836501504707802e-06,
+      "loss": 0.0633,
+      "step": 496
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.357804125150053,
+      "learning_rate": 8.831364995264297e-06,
+      "loss": 0.0683,
+      "step": 497
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1612434605900008,
+      "learning_rate": 8.826218672577175e-06,
+      "loss": 0.056,
+      "step": 498
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7127111262285413,
+      "learning_rate": 8.821062549827669e-06,
+      "loss": 0.047,
+      "step": 499
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.413765018338368,
+      "learning_rate": 8.815896640222112e-06,
+      "loss": 0.0639,
+      "step": 500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7809132641091092,
+      "learning_rate": 8.810720956991906e-06,
+      "loss": 0.0761,
+      "step": 501
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.7508580300066034,
+      "learning_rate": 8.805535513393488e-06,
+      "loss": 0.1202,
+      "step": 502
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1850119164950286,
+      "learning_rate": 8.800340322708291e-06,
+      "loss": 0.0711,
+      "step": 503
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9803941992462488,
+      "learning_rate": 8.795135398242715e-06,
+      "loss": 0.0494,
+      "step": 504
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4729217571131208,
+      "learning_rate": 8.78992075332809e-06,
+      "loss": 0.0698,
+      "step": 505
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2395336577118885,
+      "learning_rate": 8.78469640132064e-06,
+      "loss": 0.0525,
+      "step": 506
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1422345460944723,
+      "learning_rate": 8.779462355601457e-06,
+      "loss": 0.0527,
+      "step": 507
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3434980922499284,
+      "learning_rate": 8.774218629576461e-06,
+      "loss": 0.0568,
+      "step": 508
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.902815923319694,
+      "learning_rate": 8.768965236676362e-06,
+      "loss": 0.0619,
+      "step": 509
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5712968770389895,
+      "learning_rate": 8.763702190356632e-06,
+      "loss": 0.0977,
+      "step": 510
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6657851322761215,
+      "learning_rate": 8.758429504097469e-06,
+      "loss": 0.0716,
+      "step": 511
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3756418955263106,
+      "learning_rate": 8.753147191403762e-06,
+      "loss": 0.0573,
+      "step": 512
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.066335150336722,
+      "learning_rate": 8.747855265805053e-06,
+      "loss": 0.0571,
+      "step": 513
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0604816767014786,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.0745,
+      "step": 514
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5747177810123547,
+      "learning_rate": 8.737242630133878e-06,
+      "loss": 0.0585,
+      "step": 515
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.265484983287345,
+      "learning_rate": 8.73192194724347e-06,
+      "loss": 0.0841,
+      "step": 516
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5874226214432121,
+      "learning_rate": 8.726591705812102e-06,
+      "loss": 0.0676,
+      "step": 517
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.428005055418884,
+      "learning_rate": 8.721251919492083e-06,
+      "loss": 0.099,
+      "step": 518
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6128730700351466,
+      "learning_rate": 8.715902601960157e-06,
+      "loss": 0.0909,
+      "step": 519
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2540385526514841,
+      "learning_rate": 8.71054376691749e-06,
+      "loss": 0.0598,
+      "step": 520
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8172924798142867,
+      "learning_rate": 8.705175428089622e-06,
+      "loss": 0.085,
+      "step": 521
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3520076068544826,
+      "learning_rate": 8.699797599226438e-06,
+      "loss": 0.0716,
+      "step": 522
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3037215232354775,
+      "learning_rate": 8.694410294102121e-06,
+      "loss": 0.0749,
+      "step": 523
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.947508559116718,
+      "learning_rate": 8.689013526515136e-06,
+      "loss": 0.0517,
+      "step": 524
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.504232673619683,
+      "learning_rate": 8.683607310288176e-06,
+      "loss": 0.0827,
+      "step": 525
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1415599175520932,
+      "learning_rate": 8.67819165926814e-06,
+      "loss": 0.0673,
+      "step": 526
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.134782275091129,
+      "learning_rate": 8.672766587326091e-06,
+      "loss": 0.074,
+      "step": 527
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2286611260487121,
+      "learning_rate": 8.66733210835722e-06,
+      "loss": 0.0701,
+      "step": 528
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7430230560995867,
+      "learning_rate": 8.661888236280813e-06,
+      "loss": 0.084,
+      "step": 529
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1726076433801744,
+      "learning_rate": 8.656434985040215e-06,
+      "loss": 0.0446,
+      "step": 530
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.416331100895462,
+      "learning_rate": 8.650972368602793e-06,
+      "loss": 0.0641,
+      "step": 531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.943146778515395,
+      "learning_rate": 8.645500400959904e-06,
+      "loss": 0.0417,
+      "step": 532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0610796045934787,
+      "learning_rate": 8.640019096126851e-06,
+      "loss": 0.0586,
+      "step": 533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1955873477768537,
+      "learning_rate": 8.634528468142855e-06,
+      "loss": 0.082,
+      "step": 534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3787963183214098,
+      "learning_rate": 8.629028531071018e-06,
+      "loss": 0.0577,
+      "step": 535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2226100702680849,
+      "learning_rate": 8.623519298998282e-06,
+      "loss": 0.0488,
+      "step": 536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4620543475274264,
+      "learning_rate": 8.618000786035399e-06,
+      "loss": 0.0608,
+      "step": 537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6117757963010761,
+      "learning_rate": 8.612473006316888e-06,
+      "loss": 0.0669,
+      "step": 538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4385278132607686,
+      "learning_rate": 8.606935974001009e-06,
+      "loss": 0.0532,
+      "step": 539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9070996845688652,
+      "learning_rate": 8.601389703269716e-06,
+      "loss": 0.0749,
+      "step": 540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8584314450629578,
+      "learning_rate": 8.595834208328629e-06,
+      "loss": 0.0693,
+      "step": 541
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8975535225824105,
+      "learning_rate": 8.590269503406986e-06,
+      "loss": 0.0447,
+      "step": 542
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6838852869871483,
+      "learning_rate": 8.584695602757624e-06,
+      "loss": 0.0722,
+      "step": 543
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4605905386561235,
+      "learning_rate": 8.579112520656928e-06,
+      "loss": 0.072,
+      "step": 544
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1773973628095062,
+      "learning_rate": 8.5735202714048e-06,
+      "loss": 0.0511,
+      "step": 545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5662400268745837,
+      "learning_rate": 8.56791886932462e-06,
+      "loss": 0.0646,
+      "step": 546
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1584626449271131,
+      "learning_rate": 8.562308328763217e-06,
+      "loss": 0.0663,
+      "step": 547
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2096222560437233,
+      "learning_rate": 8.556688664090822e-06,
+      "loss": 0.0541,
+      "step": 548
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1044021637797712,
+      "learning_rate": 8.551059889701032e-06,
+      "loss": 0.0462,
+      "step": 549
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.275190179335817,
+      "learning_rate": 8.545422020010782e-06,
+      "loss": 0.0436,
+      "step": 550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.758686455350391,
+      "learning_rate": 8.539775069460306e-06,
+      "loss": 0.0646,
+      "step": 551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5896830704748524,
+      "learning_rate": 8.534119052513084e-06,
+      "loss": 0.0652,
+      "step": 552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.94046553261661,
+      "learning_rate": 8.52845398365583e-06,
+      "loss": 0.0767,
+      "step": 553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.538412265816462,
+      "learning_rate": 8.52277987739844e-06,
+      "loss": 0.0815,
+      "step": 554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1662847269102934,
+      "learning_rate": 8.517096748273951e-06,
+      "loss": 0.0665,
+      "step": 555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6022494338744144,
+      "learning_rate": 8.511404610838519e-06,
+      "loss": 0.0516,
+      "step": 556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.939781419965344,
+      "learning_rate": 8.505703479671365e-06,
+      "loss": 0.0448,
+      "step": 557
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1018914419618033,
+      "learning_rate": 8.499993369374754e-06,
+      "loss": 0.0521,
+      "step": 558
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0558610033685443,
+      "learning_rate": 8.49427429457394e-06,
+      "loss": 0.0513,
+      "step": 559
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2787207221715446,
+      "learning_rate": 8.488546269917145e-06,
+      "loss": 0.0554,
+      "step": 560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2983076996891234,
+      "learning_rate": 8.48280931007551e-06,
+      "loss": 0.0526,
+      "step": 561
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5470353895885407,
+      "learning_rate": 8.477063429743061e-06,
+      "loss": 0.0652,
+      "step": 562
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.643121959436738,
+      "learning_rate": 8.471308643636682e-06,
+      "loss": 0.0878,
+      "step": 563
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2378141085176515,
+      "learning_rate": 8.465544966496049e-06,
+      "loss": 0.069,
+      "step": 564
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3515805346107896,
+      "learning_rate": 8.459772413083628e-06,
+      "loss": 0.0583,
+      "step": 565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6101441673355676,
+      "learning_rate": 8.45399099818461e-06,
+      "loss": 0.0592,
+      "step": 566
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2878452013970008,
+      "learning_rate": 8.448200736606883e-06,
+      "loss": 0.0674,
+      "step": 567
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.07020821775226,
+      "learning_rate": 8.442401643181e-06,
+      "loss": 0.0557,
+      "step": 568
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.522275745146963,
+      "learning_rate": 8.43659373276013e-06,
+      "loss": 0.0647,
+      "step": 569
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9372445462464527,
+      "learning_rate": 8.430777020220026e-06,
+      "loss": 0.0505,
+      "step": 570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0766527001955493,
+      "learning_rate": 8.424951520458987e-06,
+      "loss": 0.0497,
+      "step": 571
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3515114497395717,
+      "learning_rate": 8.419117248397816e-06,
+      "loss": 0.0592,
+      "step": 572
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2860874473607398,
+      "learning_rate": 8.413274218979787e-06,
+      "loss": 0.0602,
+      "step": 573
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.060996016732545,
+      "learning_rate": 8.407422447170603e-06,
+      "loss": 0.0731,
+      "step": 574
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.142925453362851,
+      "learning_rate": 8.401561947958357e-06,
+      "loss": 0.0392,
+      "step": 575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7919299238414284,
+      "learning_rate": 8.3956927363535e-06,
+      "loss": 0.0728,
+      "step": 576
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6619277286864411,
+      "learning_rate": 8.389814827388795e-06,
+      "loss": 0.0583,
+      "step": 577
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1741375011364992,
+      "learning_rate": 8.383928236119282e-06,
+      "loss": 0.0442,
+      "step": 578
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.019745040704028,
+      "learning_rate": 8.378032977622235e-06,
+      "loss": 0.0633,
+      "step": 579
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5237063624620941,
+      "learning_rate": 8.372129066997137e-06,
+      "loss": 0.0689,
+      "step": 580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3155095691531686,
+      "learning_rate": 8.366216519365623e-06,
+      "loss": 0.0484,
+      "step": 581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0155526815115254,
+      "learning_rate": 8.36029534987145e-06,
+      "loss": 0.0625,
+      "step": 582
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0948218257424078,
+      "learning_rate": 8.354365573680465e-06,
+      "loss": 0.0538,
+      "step": 583
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5326474826791807,
+      "learning_rate": 8.348427205980552e-06,
+      "loss": 0.0663,
+      "step": 584
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4121187656089296,
+      "learning_rate": 8.342480261981604e-06,
+      "loss": 0.066,
+      "step": 585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.961029784661072,
+      "learning_rate": 8.336524756915482e-06,
+      "loss": 0.0456,
+      "step": 586
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7902506613018505,
+      "learning_rate": 8.330560706035968e-06,
+      "loss": 0.0755,
+      "step": 587
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8123904283951255,
+      "learning_rate": 8.32458812461874e-06,
+      "loss": 0.0576,
+      "step": 588
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1387807867413784,
+      "learning_rate": 8.318607027961318e-06,
+      "loss": 0.0769,
+      "step": 589
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5212598171396146,
+      "learning_rate": 8.312617431383039e-06,
+      "loss": 0.0678,
+      "step": 590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1337647098881944,
+      "learning_rate": 8.306619350225006e-06,
+      "loss": 0.0475,
+      "step": 591
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6095612646846799,
+      "learning_rate": 8.300612799850055e-06,
+      "loss": 0.0288,
+      "step": 592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2650131261757638,
+      "learning_rate": 8.294597795642715e-06,
+      "loss": 0.0547,
+      "step": 593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7591819855712723,
+      "learning_rate": 8.288574353009164e-06,
+      "loss": 0.0939,
+      "step": 594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3474754038703902,
+      "learning_rate": 8.282542487377199e-06,
+      "loss": 0.0664,
+      "step": 595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9668815112287411,
+      "learning_rate": 8.276502214196187e-06,
+      "loss": 0.0472,
+      "step": 596
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.637641184362882,
+      "learning_rate": 8.270453548937028e-06,
+      "loss": 0.0716,
+      "step": 597
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8896896880386068,
+      "learning_rate": 8.264396507092121e-06,
+      "loss": 0.0494,
+      "step": 598
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8835890673936688,
+      "learning_rate": 8.258331104175317e-06,
+      "loss": 0.0623,
+      "step": 599
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9496660157175251,
+      "learning_rate": 8.25225735572188e-06,
+      "loss": 0.0524,
+      "step": 600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5322847657196466,
+      "learning_rate": 8.246175277288457e-06,
+      "loss": 0.0498,
+      "step": 601
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0174572189523983,
+      "learning_rate": 8.240084884453022e-06,
+      "loss": 0.0614,
+      "step": 602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.91746310590934,
+      "learning_rate": 8.233986192814846e-06,
+      "loss": 0.0513,
+      "step": 603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7684397279630049,
+      "learning_rate": 8.227879217994461e-06,
+      "loss": 0.0748,
+      "step": 604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9785240601515683,
+      "learning_rate": 8.221763975633611e-06,
+      "loss": 0.0733,
+      "step": 605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1291467469735303,
+      "learning_rate": 8.215640481395215e-06,
+      "loss": 0.06,
+      "step": 606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4847914237132795,
+      "learning_rate": 8.20950875096333e-06,
+      "loss": 0.0557,
+      "step": 607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9325607449405862,
+      "learning_rate": 8.203368800043103e-06,
+      "loss": 0.0436,
+      "step": 608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7643923412720106,
+      "learning_rate": 8.197220644360745e-06,
+      "loss": 0.0643,
+      "step": 609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3674698089019297,
+      "learning_rate": 8.19106429966347e-06,
+      "loss": 0.0569,
+      "step": 610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0667370558475722,
+      "learning_rate": 8.18489978171948e-06,
+      "loss": 0.0592,
+      "step": 611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7980358087158415,
+      "learning_rate": 8.178727106317899e-06,
+      "loss": 0.08,
+      "step": 612
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5273761560502266,
+      "learning_rate": 8.17254628926875e-06,
+      "loss": 0.0546,
+      "step": 613
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.116897956255585,
+      "learning_rate": 8.166357346402914e-06,
+      "loss": 0.0628,
+      "step": 614
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7103264990604705,
+      "learning_rate": 8.160160293572073e-06,
+      "loss": 0.0488,
+      "step": 615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1284051204409844,
+      "learning_rate": 8.15395514664869e-06,
+      "loss": 0.048,
+      "step": 616
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2783445700303377,
+      "learning_rate": 8.147741921525956e-06,
+      "loss": 0.0666,
+      "step": 617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4795816320080954,
+      "learning_rate": 8.141520634117756e-06,
+      "loss": 0.0564,
+      "step": 618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.181586224479883,
+      "learning_rate": 8.13529130035862e-06,
+      "loss": 0.0677,
+      "step": 619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1124677628660073,
+      "learning_rate": 8.129053936203688e-06,
+      "loss": 0.0522,
+      "step": 620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7114645633433125,
+      "learning_rate": 8.122808557628674e-06,
+      "loss": 0.0646,
+      "step": 621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.531141332248485,
+      "learning_rate": 8.116555180629808e-06,
+      "loss": 0.0524,
+      "step": 622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0519226390878296,
+      "learning_rate": 8.11029382122382e-06,
+      "loss": 0.0904,
+      "step": 623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.411218515136369,
+      "learning_rate": 8.104024495447873e-06,
+      "loss": 0.0436,
+      "step": 624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7136185797858745,
+      "learning_rate": 8.097747219359543e-06,
+      "loss": 0.0513,
+      "step": 625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1230390583540037,
+      "learning_rate": 8.091462009036764e-06,
+      "loss": 0.0358,
+      "step": 626
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3037179652892361,
+      "learning_rate": 8.085168880577792e-06,
+      "loss": 0.0471,
+      "step": 627
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9244134810312545,
+      "learning_rate": 8.078867850101168e-06,
+      "loss": 0.0395,
+      "step": 628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.219173315497727,
+      "learning_rate": 8.072558933745668e-06,
+      "loss": 0.0499,
+      "step": 629
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4030210979804876,
+      "learning_rate": 8.066242147670268e-06,
+      "loss": 0.0418,
+      "step": 630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3156097733110124,
+      "learning_rate": 8.059917508054101e-06,
+      "loss": 0.0488,
+      "step": 631
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6817087679339362,
+      "learning_rate": 8.05358503109641e-06,
+      "loss": 0.0656,
+      "step": 632
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7541039302247473,
+      "learning_rate": 8.04724473301652e-06,
+      "loss": 0.0353,
+      "step": 633
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7192355457284463,
+      "learning_rate": 8.040896630053786e-06,
+      "loss": 0.0645,
+      "step": 634
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5761084488991353,
+      "learning_rate": 8.034540738467549e-06,
+      "loss": 0.0644,
+      "step": 635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0289767918987414,
+      "learning_rate": 8.0281770745371e-06,
+      "loss": 0.0504,
+      "step": 636
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.845359052407356,
+      "learning_rate": 8.021805654561643e-06,
+      "loss": 0.051,
+      "step": 637
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3253111334584065,
+      "learning_rate": 8.015426494860242e-06,
+      "loss": 0.0589,
+      "step": 638
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0579407657699587,
+      "learning_rate": 8.009039611771785e-06,
+      "loss": 0.0632,
+      "step": 639
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1426778255786547,
+      "learning_rate": 8.002645021654944e-06,
+      "loss": 0.0625,
+      "step": 640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4755077125900704,
+      "learning_rate": 7.996242740888131e-06,
+      "loss": 0.0867,
+      "step": 641
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.469653970091983,
+      "learning_rate": 7.989832785869454e-06,
+      "loss": 0.1006,
+      "step": 642
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.418240485230458,
+      "learning_rate": 7.98341517301668e-06,
+      "loss": 0.0957,
+      "step": 643
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1863463247272923,
+      "learning_rate": 7.976989918767183e-06,
+      "loss": 0.0476,
+      "step": 644
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9320047106615985,
+      "learning_rate": 7.970557039577918e-06,
+      "loss": 0.1027,
+      "step": 645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8756315880718288,
+      "learning_rate": 7.964116551925365e-06,
+      "loss": 0.0512,
+      "step": 646
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2079349714663918,
+      "learning_rate": 7.957668472305491e-06,
+      "loss": 0.0608,
+      "step": 647
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.067062794690031,
+      "learning_rate": 7.951212817233709e-06,
+      "loss": 0.0498,
+      "step": 648
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4353105762665406,
+      "learning_rate": 7.944749603244837e-06,
+      "loss": 0.0693,
+      "step": 649
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.120606097195914,
+      "learning_rate": 7.938278846893051e-06,
+      "loss": 0.0471,
+      "step": 650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.669764308329259,
+      "learning_rate": 7.931800564751844e-06,
+      "loss": 0.0426,
+      "step": 651
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4490252401737391,
+      "learning_rate": 7.925314773413988e-06,
+      "loss": 0.0769,
+      "step": 652
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0443343523013158,
+      "learning_rate": 7.918821489491489e-06,
+      "loss": 0.0449,
+      "step": 653
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9141229109009441,
+      "learning_rate": 7.91232072961554e-06,
+      "loss": 0.0572,
+      "step": 654
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.972278870552116,
+      "learning_rate": 7.905812510436483e-06,
+      "loss": 0.0588,
+      "step": 655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0982712477608652,
+      "learning_rate": 7.899296848623766e-06,
+      "loss": 0.0727,
+      "step": 656
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3069436042751426,
+      "learning_rate": 7.892773760865901e-06,
+      "loss": 0.0616,
+      "step": 657
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.036432390647132,
+      "learning_rate": 7.886243263870421e-06,
+      "loss": 0.0398,
+      "step": 658
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3003358641954792,
+      "learning_rate": 7.879705374363831e-06,
+      "loss": 0.0517,
+      "step": 659
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9881779327346152,
+      "learning_rate": 7.873160109091574e-06,
+      "loss": 0.0466,
+      "step": 660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2844285680285321,
+      "learning_rate": 7.866607484817984e-06,
+      "loss": 0.048,
+      "step": 661
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1217634195547732,
+      "learning_rate": 7.860047518326243e-06,
+      "loss": 0.0469,
+      "step": 662
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7765193310118434,
+      "learning_rate": 7.853480226418345e-06,
+      "loss": 0.0727,
+      "step": 663
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.351110060170768,
+      "learning_rate": 7.846905625915032e-06,
+      "loss": 0.0602,
+      "step": 664
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2480431610205707,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.0668,
+      "step": 665
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.279228731526593,
+      "learning_rate": 7.833734566498732e-06,
+      "loss": 0.0619,
+      "step": 666
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3069130651640115,
+      "learning_rate": 7.827138141320672e-06,
+      "loss": 0.0487,
+      "step": 667
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6649550741739837,
+      "learning_rate": 7.820534475016965e-06,
+      "loss": 0.0357,
+      "step": 668
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1780747788613053,
+      "learning_rate": 7.813923584501529e-06,
+      "loss": 0.0507,
+      "step": 669
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3976292130290968,
+      "learning_rate": 7.807305486706782e-06,
+      "loss": 0.0455,
+      "step": 670
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0135391390071877,
+      "learning_rate": 7.800680198583608e-06,
+      "loss": 0.0352,
+      "step": 671
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8518127433708236,
+      "learning_rate": 7.794047737101298e-06,
+      "loss": 0.0768,
+      "step": 672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1646871241715808,
+      "learning_rate": 7.787408119247522e-06,
+      "loss": 0.0368,
+      "step": 673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0054665399271567,
+      "learning_rate": 7.78076136202828e-06,
+      "loss": 0.0444,
+      "step": 674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0060984550389096,
+      "learning_rate": 7.774107482467856e-06,
+      "loss": 0.058,
+      "step": 675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8843245325848659,
+      "learning_rate": 7.767446497608775e-06,
+      "loss": 0.0485,
+      "step": 676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.007540144322301,
+      "learning_rate": 7.760778424511766e-06,
+      "loss": 0.0589,
+      "step": 677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8352098796640809,
+      "learning_rate": 7.754103280255707e-06,
+      "loss": 0.0405,
+      "step": 678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.47213607467956525,
+      "learning_rate": 7.747421081937592e-06,
+      "loss": 0.026,
+      "step": 679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2589822578996794,
+      "learning_rate": 7.740731846672479e-06,
+      "loss": 0.0397,
+      "step": 680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2489792445554098,
+      "learning_rate": 7.734035591593452e-06,
+      "loss": 0.0629,
+      "step": 681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2130156604486495,
+      "learning_rate": 7.727332333851574e-06,
+      "loss": 0.0623,
+      "step": 682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0670881828943366,
+      "learning_rate": 7.720622090615844e-06,
+      "loss": 0.0444,
+      "step": 683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5910472088638483,
+      "learning_rate": 7.713904879073153e-06,
+      "loss": 0.063,
+      "step": 684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0945352572471834,
+      "learning_rate": 7.707180716428237e-06,
+      "loss": 0.0543,
+      "step": 685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1888354792478368,
+      "learning_rate": 7.700449619903643e-06,
+      "loss": 0.0494,
+      "step": 686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234737397503048,
+      "learning_rate": 7.693711606739667e-06,
+      "loss": 0.0654,
+      "step": 687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5921467818053312,
+      "learning_rate": 7.686966694194334e-06,
+      "loss": 0.0473,
+      "step": 688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.068287322661853,
+      "learning_rate": 7.680214899543328e-06,
+      "loss": 0.0498,
+      "step": 689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.090304711427149,
+      "learning_rate": 7.673456240079966e-06,
+      "loss": 0.0537,
+      "step": 690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2623514077463773,
+      "learning_rate": 7.666690733115147e-06,
+      "loss": 0.0323,
+      "step": 691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0821798809113374,
+      "learning_rate": 7.65991839597731e-06,
+      "loss": 0.0541,
+      "step": 692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3982738877088425,
+      "learning_rate": 7.653139246012383e-06,
+      "loss": 0.0995,
+      "step": 693
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0333414715948794,
+      "learning_rate": 7.64635330058375e-06,
+      "loss": 0.0534,
+      "step": 694
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5282306413091515,
+      "learning_rate": 7.639560577072194e-06,
+      "loss": 0.0507,
+      "step": 695
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4155931930782537,
+      "learning_rate": 7.632761092875864e-06,
+      "loss": 0.0767,
+      "step": 696
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.421296001948551,
+      "learning_rate": 7.625954865410224e-06,
+      "loss": 0.0688,
+      "step": 697
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5818062433164756,
+      "learning_rate": 7.619141912108008e-06,
+      "loss": 0.0368,
+      "step": 698
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8119759623455076,
+      "learning_rate": 7.612322250419173e-06,
+      "loss": 0.0357,
+      "step": 699
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4030844108026284,
+      "learning_rate": 7.605495897810868e-06,
+      "loss": 0.0652,
+      "step": 700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7435621952165166,
+      "learning_rate": 7.598662871767371e-06,
+      "loss": 0.0595,
+      "step": 701
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0199936017098497,
+      "learning_rate": 7.591823189790054e-06,
+      "loss": 0.0564,
+      "step": 702
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.013801793151265,
+      "learning_rate": 7.5849768693973415e-06,
+      "loss": 0.0506,
+      "step": 703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1994253630174772,
+      "learning_rate": 7.578123928124654e-06,
+      "loss": 0.0514,
+      "step": 704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.695659698644204,
+      "learning_rate": 7.571264383524377e-06,
+      "loss": 0.0868,
+      "step": 705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.080323902841376,
+      "learning_rate": 7.564398253165804e-06,
+      "loss": 0.0635,
+      "step": 706
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0873572448586641,
+      "learning_rate": 7.557525554635097e-06,
+      "loss": 0.0488,
+      "step": 707
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1921038081951376,
+      "learning_rate": 7.550646305535243e-06,
+      "loss": 0.0537,
+      "step": 708
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3417929684907905,
+      "learning_rate": 7.543760523486008e-06,
+      "loss": 0.0705,
+      "step": 709
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.770501025309428,
+      "learning_rate": 7.5368682261238865e-06,
+      "loss": 0.0348,
+      "step": 710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7913968514488297,
+      "learning_rate": 7.529969431102063e-06,
+      "loss": 0.0469,
+      "step": 711
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.372638464172713,
+      "learning_rate": 7.523064156090367e-06,
+      "loss": 0.0647,
+      "step": 712
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4168525758105495,
+      "learning_rate": 7.51615241877522e-06,
+      "loss": 0.0684,
+      "step": 713
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5952202718121256,
+      "learning_rate": 7.509234236859597e-06,
+      "loss": 0.0285,
+      "step": 714
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1132503184302827,
+      "learning_rate": 7.502309628062984e-06,
+      "loss": 0.055,
+      "step": 715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2940171549098398,
+      "learning_rate": 7.49537861012132e-06,
+      "loss": 0.0568,
+      "step": 716
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.432400009253595,
+      "learning_rate": 7.488441200786969e-06,
+      "loss": 0.0568,
+      "step": 717
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7357982264667429,
+      "learning_rate": 7.481497417828657e-06,
+      "loss": 0.0584,
+      "step": 718
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.152811480884973,
+      "learning_rate": 7.474547279031439e-06,
+      "loss": 0.0356,
+      "step": 719
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9457070217244251,
+      "learning_rate": 7.467590802196648e-06,
+      "loss": 0.0421,
+      "step": 720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3703852382166088,
+      "learning_rate": 7.460628005141853e-06,
+      "loss": 0.0603,
+      "step": 721
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8000228910289057,
+      "learning_rate": 7.453658905700804e-06,
+      "loss": 0.0428,
+      "step": 722
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6863165441999962,
+      "learning_rate": 7.4466835217234034e-06,
+      "loss": 0.0393,
+      "step": 723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4274426214090807,
+      "learning_rate": 7.4397018710756415e-06,
+      "loss": 0.0666,
+      "step": 724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.615013887200379,
+      "learning_rate": 7.432713971639565e-06,
+      "loss": 0.0351,
+      "step": 725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3432841177404575,
+      "learning_rate": 7.42571984131322e-06,
+      "loss": 0.0576,
+      "step": 726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1694400922031816,
+      "learning_rate": 7.418719498010618e-06,
+      "loss": 0.0495,
+      "step": 727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7107862218528511,
+      "learning_rate": 7.411712959661677e-06,
+      "loss": 0.065,
+      "step": 728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.114796883559968,
+      "learning_rate": 7.40470024421219e-06,
+      "loss": 0.0451,
+      "step": 729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.510226449257016,
+      "learning_rate": 7.397681369623766e-06,
+      "loss": 0.049,
+      "step": 730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9849390108162892,
+      "learning_rate": 7.39065635387379e-06,
+      "loss": 0.0487,
+      "step": 731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.938148015960924,
+      "learning_rate": 7.383625214955377e-06,
+      "loss": 0.0524,
+      "step": 732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.344406418148889,
+      "learning_rate": 7.376587970877327e-06,
+      "loss": 0.0574,
+      "step": 733
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.43575449802591,
+      "learning_rate": 7.369544639664072e-06,
+      "loss": 0.0559,
+      "step": 734
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.800158124092139,
+      "learning_rate": 7.362495239355642e-06,
+      "loss": 0.0402,
+      "step": 735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.162183628676278,
+      "learning_rate": 7.355439788007604e-06,
+      "loss": 0.0504,
+      "step": 736
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7051824963315141,
+      "learning_rate": 7.34837830369103e-06,
+      "loss": 0.04,
+      "step": 737
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9660305349076502,
+      "learning_rate": 7.341310804492443e-06,
+      "loss": 0.041,
+      "step": 738
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2103272636528846,
+      "learning_rate": 7.334237308513766e-06,
+      "loss": 0.0588,
+      "step": 739
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6247813317536415,
+      "learning_rate": 7.327157833872291e-06,
+      "loss": 0.0388,
+      "step": 740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.950242229839862,
+      "learning_rate": 7.320072398700615e-06,
+      "loss": 0.0415,
+      "step": 741
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5224829709523,
+      "learning_rate": 7.312981021146606e-06,
+      "loss": 0.0654,
+      "step": 742
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.029686304470058,
+      "learning_rate": 7.30588371937335e-06,
+      "loss": 0.0532,
+      "step": 743
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1599326200613038,
+      "learning_rate": 7.298780511559108e-06,
+      "loss": 0.0569,
+      "step": 744
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.997871435156531,
+      "learning_rate": 7.291671415897268e-06,
+      "loss": 0.0597,
+      "step": 745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.276453119868504,
+      "learning_rate": 7.284556450596298e-06,
+      "loss": 0.0502,
+      "step": 746
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3737928110013409,
+      "learning_rate": 7.2774356338797004e-06,
+      "loss": 0.0417,
+      "step": 747
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9285847428586894,
+      "learning_rate": 7.270308983985963e-06,
+      "loss": 0.0469,
+      "step": 748
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8504652175843128,
+      "learning_rate": 7.2631765191685174e-06,
+      "loss": 0.0373,
+      "step": 749
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8153321056774139,
+      "learning_rate": 7.2560382576956875e-06,
+      "loss": 0.0408,
+      "step": 750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1454444295449129,
+      "learning_rate": 7.24889421785064e-06,
+      "loss": 0.0512,
+      "step": 751
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8552009648324275,
+      "learning_rate": 7.24174441793135e-06,
+      "loss": 0.0566,
+      "step": 752
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8568618511392941,
+      "learning_rate": 7.234588876250538e-06,
+      "loss": 0.0408,
+      "step": 753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.395789889097727,
+      "learning_rate": 7.227427611135634e-06,
+      "loss": 0.0487,
+      "step": 754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1473068375872773,
+      "learning_rate": 7.22026064092873e-06,
+      "loss": 0.0453,
+      "step": 755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9107552740533715,
+      "learning_rate": 7.2130879839865255e-06,
+      "loss": 0.0484,
+      "step": 756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.479569569975316,
+      "learning_rate": 7.205909658680287e-06,
+      "loss": 0.0704,
+      "step": 757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.614669912893105,
+      "learning_rate": 7.198725683395802e-06,
+      "loss": 0.0476,
+      "step": 758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8291391376525924,
+      "learning_rate": 7.191536076533326e-06,
+      "loss": 0.0449,
+      "step": 759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6819517582924615,
+      "learning_rate": 7.184340856507541e-06,
+      "loss": 0.0439,
+      "step": 760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7986954604883367,
+      "learning_rate": 7.177140041747503e-06,
+      "loss": 0.0522,
+      "step": 761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0191355693655744,
+      "learning_rate": 7.169933650696602e-06,
+      "loss": 0.0493,
+      "step": 762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8276853219246487,
+      "learning_rate": 7.162721701812506e-06,
+      "loss": 0.04,
+      "step": 763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.273271861109226,
+      "learning_rate": 7.155504213567122e-06,
+      "loss": 0.064,
+      "step": 764
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8211027414183176,
+      "learning_rate": 7.148281204446541e-06,
+      "loss": 0.0375,
+      "step": 765
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.324686842180071,
+      "learning_rate": 7.141052692950999e-06,
+      "loss": 0.0617,
+      "step": 766
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3516945141218741,
+      "learning_rate": 7.133818697594821e-06,
+      "loss": 0.0495,
+      "step": 767
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.546142032224951,
+      "learning_rate": 7.126579236906382e-06,
+      "loss": 0.0664,
+      "step": 768
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6871920840680835,
+      "learning_rate": 7.1193343294280516e-06,
+      "loss": 0.0413,
+      "step": 769
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6318268345167921,
+      "learning_rate": 7.112083993716151e-06,
+      "loss": 0.0638,
+      "step": 770
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9005360464139678,
+      "learning_rate": 7.104828248340907e-06,
+      "loss": 0.042,
+      "step": 771
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7820265958769346,
+      "learning_rate": 7.0975671118864e-06,
+      "loss": 0.0428,
+      "step": 772
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8437539128584735,
+      "learning_rate": 7.090300602950518e-06,
+      "loss": 0.0471,
+      "step": 773
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8279688633282087,
+      "learning_rate": 7.083028740144912e-06,
+      "loss": 0.0299,
+      "step": 774
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.302893161500799,
+      "learning_rate": 7.075751542094944e-06,
+      "loss": 0.0499,
+      "step": 775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0146660381609616,
+      "learning_rate": 7.068469027439642e-06,
+      "loss": 0.0466,
+      "step": 776
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9711265895426145,
+      "learning_rate": 7.0611812148316475e-06,
+      "loss": 0.0426,
+      "step": 777
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.43048903542425127,
+      "learning_rate": 7.0538881229371795e-06,
+      "loss": 0.0262,
+      "step": 778
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1457379520309836,
+      "learning_rate": 7.046589770435971e-06,
+      "loss": 0.0457,
+      "step": 779
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2453652954639685,
+      "learning_rate": 7.039286176021233e-06,
+      "loss": 0.0538,
+      "step": 780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9963928129972024,
+      "learning_rate": 7.031977358399602e-06,
+      "loss": 0.0382,
+      "step": 781
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2900988865570673,
+      "learning_rate": 7.024663336291093e-06,
+      "loss": 0.0509,
+      "step": 782
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4027443850614405,
+      "learning_rate": 7.017344128429049e-06,
+      "loss": 0.07,
+      "step": 783
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0371750020391997,
+      "learning_rate": 7.0100197535601e-06,
+      "loss": 0.052,
+      "step": 784
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7843862300889355,
+      "learning_rate": 7.002690230444104e-06,
+      "loss": 0.0543,
+      "step": 785
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8603855097473864,
+      "learning_rate": 6.9953555778541074e-06,
+      "loss": 0.0342,
+      "step": 786
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0511727640363182,
+      "learning_rate": 6.988015814576299e-06,
+      "loss": 0.0574,
+      "step": 787
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.439368303862243,
+      "learning_rate": 6.980670959409952e-06,
+      "loss": 0.0504,
+      "step": 788
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2912175634676242,
+      "learning_rate": 6.9733210311673826e-06,
+      "loss": 0.0512,
+      "step": 789
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7768806755603833,
+      "learning_rate": 6.965966048673905e-06,
+      "loss": 0.0344,
+      "step": 790
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9256238998707683,
+      "learning_rate": 6.958606030767771e-06,
+      "loss": 0.0298,
+      "step": 791
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.87432810891287,
+      "learning_rate": 6.951240996300135e-06,
+      "loss": 0.0548,
+      "step": 792
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2413364634857271,
+      "learning_rate": 6.9438709641350025e-06,
+      "loss": 0.0632,
+      "step": 793
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5436192413434815,
+      "learning_rate": 6.9364959531491714e-06,
+      "loss": 0.0276,
+      "step": 794
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1211151078772463,
+      "learning_rate": 6.9291159822322e-06,
+      "loss": 0.0431,
+      "step": 795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9611971763181043,
+      "learning_rate": 6.921731070286346e-06,
+      "loss": 0.0405,
+      "step": 796
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0122635098927624,
+      "learning_rate": 6.914341236226524e-06,
+      "loss": 0.0402,
+      "step": 797
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.136260979127934,
+      "learning_rate": 6.906946498980254e-06,
+      "loss": 0.0461,
+      "step": 798
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5458396979143085,
+      "learning_rate": 6.899546877487616e-06,
+      "loss": 0.0456,
+      "step": 799
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.913407988966103,
+      "learning_rate": 6.8921423907012e-06,
+      "loss": 0.0355,
+      "step": 800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8881887819273989,
+      "learning_rate": 6.884733057586057e-06,
+      "loss": 0.0333,
+      "step": 801
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0538766237007566,
+      "learning_rate": 6.8773188971196515e-06,
+      "loss": 0.0637,
+      "step": 802
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8453030323277484,
+      "learning_rate": 6.869899928291809e-06,
+      "loss": 0.0356,
+      "step": 803
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4774705218128774,
+      "learning_rate": 6.862476170104677e-06,
+      "loss": 0.047,
+      "step": 804
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3921989694543493,
+      "learning_rate": 6.8550476415726645e-06,
+      "loss": 0.0651,
+      "step": 805
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1488627023685396,
+      "learning_rate": 6.847614361722399e-06,
+      "loss": 0.0478,
+      "step": 806
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6811740696636209,
+      "learning_rate": 6.840176349592681e-06,
+      "loss": 0.0296,
+      "step": 807
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4898953199819551,
+      "learning_rate": 6.83273362423443e-06,
+      "loss": 0.0609,
+      "step": 808
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3565601020815854,
+      "learning_rate": 6.825286204710635e-06,
+      "loss": 0.0597,
+      "step": 809
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1415886329244214,
+      "learning_rate": 6.817834110096313e-06,
+      "loss": 0.0368,
+      "step": 810
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.77611927624329,
+      "learning_rate": 6.8103773594784506e-06,
+      "loss": 0.0449,
+      "step": 811
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4555453825233526,
+      "learning_rate": 6.8029159719559615e-06,
+      "loss": 0.0394,
+      "step": 812
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8409635536876771,
+      "learning_rate": 6.795449966639638e-06,
+      "loss": 0.0367,
+      "step": 813
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2610559094175553,
+      "learning_rate": 6.787979362652097e-06,
+      "loss": 0.0379,
+      "step": 814
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6529011944025385,
+      "learning_rate": 6.780504179127735e-06,
+      "loss": 0.0672,
+      "step": 815
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8307567736703948,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.024,
+      "step": 816
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9432272440319054,
+      "learning_rate": 6.765540150064734e-06,
+      "loss": 0.0345,
+      "step": 817
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7703612656478014,
+      "learning_rate": 6.758051342853337e-06,
+      "loss": 0.0462,
+      "step": 818
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.314910616474518,
+      "learning_rate": 6.750558032759512e-06,
+      "loss": 0.0548,
+      "step": 819
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4226314354369014,
+      "learning_rate": 6.743060238975811e-06,
+      "loss": 0.0417,
+      "step": 820
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.919616358303647,
+      "learning_rate": 6.735557980706268e-06,
+      "loss": 0.0464,
+      "step": 821
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2640004664395663,
+      "learning_rate": 6.728051277166358e-06,
+      "loss": 0.0485,
+      "step": 822
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1906980417732456,
+      "learning_rate": 6.720540147582941e-06,
+      "loss": 0.0487,
+      "step": 823
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.241031549174688,
+      "learning_rate": 6.713024611194208e-06,
+      "loss": 0.0509,
+      "step": 824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.988899982913446,
+      "learning_rate": 6.705504687249642e-06,
+      "loss": 0.0452,
+      "step": 825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.969809384808336,
+      "learning_rate": 6.697980395009963e-06,
+      "loss": 0.0388,
+      "step": 826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2267200517217354,
+      "learning_rate": 6.690451753747076e-06,
+      "loss": 0.0332,
+      "step": 827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3868305312881741,
+      "learning_rate": 6.682918782744033e-06,
+      "loss": 0.0405,
+      "step": 828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7644839921848126,
+      "learning_rate": 6.675381501294965e-06,
+      "loss": 0.0396,
+      "step": 829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.221670982443101,
+      "learning_rate": 6.667839928705051e-06,
+      "loss": 0.0503,
+      "step": 830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.824025036996117,
+      "learning_rate": 6.660294084290461e-06,
+      "loss": 0.0703,
+      "step": 831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.299267850755918,
+      "learning_rate": 6.6527439873783004e-06,
+      "loss": 0.0507,
+      "step": 832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9081222896678484,
+      "learning_rate": 6.645189657306572e-06,
+      "loss": 0.0422,
+      "step": 833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8290197497835838,
+      "learning_rate": 6.6376311134241216e-06,
+      "loss": 0.039,
+      "step": 834
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.289255419574439,
+      "learning_rate": 6.630068375090581e-06,
+      "loss": 0.0462,
+      "step": 835
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9270669463180224,
+      "learning_rate": 6.622501461676333e-06,
+      "loss": 0.033,
+      "step": 836
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1795154469114397,
+      "learning_rate": 6.6149303925624495e-06,
+      "loss": 0.0391,
+      "step": 837
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.250248808799025,
+      "learning_rate": 6.607355187140647e-06,
+      "loss": 0.0436,
+      "step": 838
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8718051208128388,
+      "learning_rate": 6.599775864813237e-06,
+      "loss": 0.0347,
+      "step": 839
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0027766395054658,
+      "learning_rate": 6.592192444993078e-06,
+      "loss": 0.0429,
+      "step": 840
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3182757491833073,
+      "learning_rate": 6.584604947103515e-06,
+      "loss": 0.0701,
+      "step": 841
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.812740598782862,
+      "learning_rate": 6.5770133905783485e-06,
+      "loss": 0.0387,
+      "step": 842
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.082109636231357,
+      "learning_rate": 6.569417794861768e-06,
+      "loss": 0.0457,
+      "step": 843
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6750257856342989,
+      "learning_rate": 6.56181817940831e-06,
+      "loss": 0.0386,
+      "step": 844
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.037029132615747,
+      "learning_rate": 6.554214563682809e-06,
+      "loss": 0.0589,
+      "step": 845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2770572481644458,
+      "learning_rate": 6.546606967160343e-06,
+      "loss": 0.0449,
+      "step": 846
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0837849954557062,
+      "learning_rate": 6.538995409326183e-06,
+      "loss": 0.0417,
+      "step": 847
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.036021742726003,
+      "learning_rate": 6.531379909675753e-06,
+      "loss": 0.038,
+      "step": 848
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.403469464122417,
+      "learning_rate": 6.5237604877145685e-06,
+      "loss": 0.0495,
+      "step": 849
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8107811814670984,
+      "learning_rate": 6.516137162958192e-06,
+      "loss": 0.0323,
+      "step": 850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.174091980809628,
+      "learning_rate": 6.5085099549321826e-06,
+      "loss": 0.0532,
+      "step": 851
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5788132792682261,
+      "learning_rate": 6.500878883172045e-06,
+      "loss": 0.0348,
+      "step": 852
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9128356981769381,
+      "learning_rate": 6.49324396722318e-06,
+      "loss": 0.0408,
+      "step": 853
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.251124315603189,
+      "learning_rate": 6.4856052266408375e-06,
+      "loss": 0.0506,
+      "step": 854
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0456200167180727,
+      "learning_rate": 6.4779626809900585e-06,
+      "loss": 0.0267,
+      "step": 855
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6946818933509344,
+      "learning_rate": 6.470316349845632e-06,
+      "loss": 0.0451,
+      "step": 856
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.807980051547909,
+      "learning_rate": 6.462666252792044e-06,
+      "loss": 0.0557,
+      "step": 857
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4075680826831167,
+      "learning_rate": 6.455012409423427e-06,
+      "loss": 0.0549,
+      "step": 858
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.370907384848696,
+      "learning_rate": 6.447354839343504e-06,
+      "loss": 0.0547,
+      "step": 859
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5469420948622832,
+      "learning_rate": 6.439693562165546e-06,
+      "loss": 0.0617,
+      "step": 860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6537681598841625,
+      "learning_rate": 6.432028597512322e-06,
+      "loss": 0.04,
+      "step": 861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3509749104300668,
+      "learning_rate": 6.424359965016041e-06,
+      "loss": 0.0471,
+      "step": 862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9152872083214765,
+      "learning_rate": 6.416687684318309e-06,
+      "loss": 0.0447,
+      "step": 863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4295335018939017,
+      "learning_rate": 6.409011775070073e-06,
+      "loss": 0.0585,
+      "step": 864
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7808982276220853,
+      "learning_rate": 6.40133225693158e-06,
+      "loss": 0.0393,
+      "step": 865
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8901219228184742,
+      "learning_rate": 6.393649149572313e-06,
+      "loss": 0.0394,
+      "step": 866
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4096812791604068,
+      "learning_rate": 6.385962472670953e-06,
+      "loss": 0.0492,
+      "step": 867
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6751166911982542,
+      "learning_rate": 6.3782722459153246e-06,
+      "loss": 0.0469,
+      "step": 868
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1596847414679285,
+      "learning_rate": 6.370578489002339e-06,
+      "loss": 0.042,
+      "step": 869
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0083992591231499,
+      "learning_rate": 6.362881221637953e-06,
+      "loss": 0.0428,
+      "step": 870
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8285729394403849,
+      "learning_rate": 6.355180463537116e-06,
+      "loss": 0.0412,
+      "step": 871
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4255200540690596,
+      "learning_rate": 6.347476234423715e-06,
+      "loss": 0.0441,
+      "step": 872
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0790904056026727,
+      "learning_rate": 6.339768554030528e-06,
+      "loss": 0.0322,
+      "step": 873
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0712417336999227,
+      "learning_rate": 6.3320574420991735e-06,
+      "loss": 0.0397,
+      "step": 874
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0713829508974753,
+      "learning_rate": 6.324342918380061e-06,
+      "loss": 0.0395,
+      "step": 875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8121714693017822,
+      "learning_rate": 6.316625002632333e-06,
+      "loss": 0.0295,
+      "step": 876
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4494665701652905,
+      "learning_rate": 6.3089037146238275e-06,
+      "loss": 0.0508,
+      "step": 877
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.753263308113668,
+      "learning_rate": 6.301179074131014e-06,
+      "loss": 0.0366,
+      "step": 878
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4386877756964491,
+      "learning_rate": 6.29345110093895e-06,
+      "loss": 0.053,
+      "step": 879
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0394724656689378,
+      "learning_rate": 6.28571981484123e-06,
+      "loss": 0.0389,
+      "step": 880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9400400857708836,
+      "learning_rate": 6.277985235639935e-06,
+      "loss": 0.0367,
+      "step": 881
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9516717348503178,
+      "learning_rate": 6.2702473831455755e-06,
+      "loss": 0.0374,
+      "step": 882
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8182276197656334,
+      "learning_rate": 6.262506277177055e-06,
+      "loss": 0.0468,
+      "step": 883
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.35296221468063,
+      "learning_rate": 6.254761937561599e-06,
+      "loss": 0.0449,
+      "step": 884
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9797346235633131,
+      "learning_rate": 6.247014384134723e-06,
+      "loss": 0.0328,
+      "step": 885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0396717421424964,
+      "learning_rate": 6.2392636367401725e-06,
+      "loss": 0.0452,
+      "step": 886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.515308122456777,
+      "learning_rate": 6.2315097152298705e-06,
+      "loss": 0.0348,
+      "step": 887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6496972916065884,
+      "learning_rate": 6.223752639463876e-06,
+      "loss": 0.0692,
+      "step": 888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8274307431826953,
+      "learning_rate": 6.2159924293103205e-06,
+      "loss": 0.027,
+      "step": 889
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6250106328575086,
+      "learning_rate": 6.208229104645366e-06,
+      "loss": 0.0302,
+      "step": 890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9417114215197753,
+      "learning_rate": 6.2004626853531545e-06,
+      "loss": 0.0454,
+      "step": 891
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2897997496298013,
+      "learning_rate": 6.1926931913257515e-06,
+      "loss": 0.0305,
+      "step": 892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1784748626118067,
+      "learning_rate": 6.184920642463095e-06,
+      "loss": 0.0327,
+      "step": 893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8166216570833613,
+      "learning_rate": 6.177145058672954e-06,
+      "loss": 0.0315,
+      "step": 894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.798575884334304,
+      "learning_rate": 6.169366459870866e-06,
+      "loss": 0.0323,
+      "step": 895
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2946830918663645,
+      "learning_rate": 6.16158486598009e-06,
+      "loss": 0.044,
+      "step": 896
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.0979367059633836,
+      "learning_rate": 6.153800296931561e-06,
+      "loss": 0.0538,
+      "step": 897
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.005181417036001,
+      "learning_rate": 6.146012772663832e-06,
+      "loss": 0.04,
+      "step": 898
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2335761673269008,
+      "learning_rate": 6.138222313123022e-06,
+      "loss": 0.046,
+      "step": 899
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.8252291418130575,
+      "learning_rate": 6.130428938262774e-06,
+      "loss": 0.0411,
+      "step": 900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2349381048462125,
+      "learning_rate": 6.122632668044193e-06,
+      "loss": 0.0365,
+      "step": 901
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7469611077478491,
+      "learning_rate": 6.1148335224358016e-06,
+      "loss": 0.0312,
+      "step": 902
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6468983595440548,
+      "learning_rate": 6.107031521413488e-06,
+      "loss": 0.0289,
+      "step": 903
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2638411098954696,
+      "learning_rate": 6.099226684960454e-06,
+      "loss": 0.0537,
+      "step": 904
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6411174182139878,
+      "learning_rate": 6.091419033067159e-06,
+      "loss": 0.0473,
+      "step": 905
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1672958610885626,
+      "learning_rate": 6.083608585731283e-06,
+      "loss": 0.0348,
+      "step": 906
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3478950565393775,
+      "learning_rate": 6.075795362957657e-06,
+      "loss": 0.0422,
+      "step": 907
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1232483592786926,
+      "learning_rate": 6.067979384758225e-06,
+      "loss": 0.0441,
+      "step": 908
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7482395181364767,
+      "learning_rate": 6.0601606711519875e-06,
+      "loss": 0.0333,
+      "step": 909
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1293888728654633,
+      "learning_rate": 6.0523392421649515e-06,
+      "loss": 0.0606,
+      "step": 910
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8673764096236666,
+      "learning_rate": 6.044515117830078e-06,
+      "loss": 0.0323,
+      "step": 911
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3722075174223107,
+      "learning_rate": 6.036688318187234e-06,
+      "loss": 0.0357,
+      "step": 912
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4110866208689847,
+      "learning_rate": 6.028858863283135e-06,
+      "loss": 0.0369,
+      "step": 913
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.067372279350801,
+      "learning_rate": 6.021026773171299e-06,
+      "loss": 0.0346,
+      "step": 914
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.430328387623402,
+      "learning_rate": 6.013192067911997e-06,
+      "loss": 0.0516,
+      "step": 915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0292400791593719,
+      "learning_rate": 6.005354767572194e-06,
+      "loss": 0.0401,
+      "step": 916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1345574370447715,
+      "learning_rate": 5.997514892225499e-06,
+      "loss": 0.0345,
+      "step": 917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9843319026076962,
+      "learning_rate": 5.9896724619521256e-06,
+      "loss": 0.0416,
+      "step": 918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.022037359513478,
+      "learning_rate": 5.9818274968388225e-06,
+      "loss": 0.0422,
+      "step": 919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0909811445781186,
+      "learning_rate": 5.973980016978834e-06,
+      "loss": 0.0375,
+      "step": 920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5586504018752385,
+      "learning_rate": 5.966130042471848e-06,
+      "loss": 0.0493,
+      "step": 921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0696048332400143,
+      "learning_rate": 5.95827759342394e-06,
+      "loss": 0.0463,
+      "step": 922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7850891902664793,
+      "learning_rate": 5.950422689947519e-06,
+      "loss": 0.0368,
+      "step": 923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6268144662037067,
+      "learning_rate": 5.942565352161289e-06,
+      "loss": 0.0685,
+      "step": 924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6995902180149687,
+      "learning_rate": 5.934705600190183e-06,
+      "loss": 0.0283,
+      "step": 925
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8735482858574848,
+      "learning_rate": 5.926843454165317e-06,
+      "loss": 0.0388,
+      "step": 926
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9215046522914323,
+      "learning_rate": 5.918978934223945e-06,
+      "loss": 0.0451,
+      "step": 927
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8746979507415403,
+      "learning_rate": 5.911112060509392e-06,
+      "loss": 0.0362,
+      "step": 928
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8621716940708184,
+      "learning_rate": 5.903242853171022e-06,
+      "loss": 0.0363,
+      "step": 929
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.136759596361992,
+      "learning_rate": 5.895371332364167e-06,
+      "loss": 0.0479,
+      "step": 930
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2815977401133114,
+      "learning_rate": 5.88749751825009e-06,
+      "loss": 0.0435,
+      "step": 931
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0950033959553411,
+      "learning_rate": 5.879621430995927e-06,
+      "loss": 0.038,
+      "step": 932
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2293706227587744,
+      "learning_rate": 5.8717430907746365e-06,
+      "loss": 0.0497,
+      "step": 933
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9493611725216664,
+      "learning_rate": 5.863862517764942e-06,
+      "loss": 0.0407,
+      "step": 934
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6413160972295684,
+      "learning_rate": 5.8559797321512946e-06,
+      "loss": 0.0482,
+      "step": 935
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7828006455912568,
+      "learning_rate": 5.848094754123808e-06,
+      "loss": 0.0265,
+      "step": 936
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0709946055942174,
+      "learning_rate": 5.8402076038782065e-06,
+      "loss": 0.0466,
+      "step": 937
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1460909174115927,
+      "learning_rate": 5.832318301615789e-06,
+      "loss": 0.0456,
+      "step": 938
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3080987894141531,
+      "learning_rate": 5.824426867543358e-06,
+      "loss": 0.0563,
+      "step": 939
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8941656475776596,
+      "learning_rate": 5.816533321873178e-06,
+      "loss": 0.0394,
+      "step": 940
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.39318900492243813,
+      "learning_rate": 5.808637684822924e-06,
+      "loss": 0.0259,
+      "step": 941
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.34674884621565,
+      "learning_rate": 5.800739976615626e-06,
+      "loss": 0.0431,
+      "step": 942
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7805739408684246,
+      "learning_rate": 5.792840217479616e-06,
+      "loss": 0.0341,
+      "step": 943
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7497753410737589,
+      "learning_rate": 5.784938427648488e-06,
+      "loss": 0.031,
+      "step": 944
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0072527798290345,
+      "learning_rate": 5.777034627361025e-06,
+      "loss": 0.038,
+      "step": 945
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8811001342598082,
+      "learning_rate": 5.769128836861169e-06,
+      "loss": 0.0463,
+      "step": 946
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6974121452637448,
+      "learning_rate": 5.761221076397957e-06,
+      "loss": 0.0255,
+      "step": 947
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.476097592972458,
+      "learning_rate": 5.753311366225467e-06,
+      "loss": 0.0501,
+      "step": 948
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8008420028609479,
+      "learning_rate": 5.7453997266027784e-06,
+      "loss": 0.0331,
+      "step": 949
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5582987049019141,
+      "learning_rate": 5.737486177793907e-06,
+      "loss": 0.0239,
+      "step": 950
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2081504866571806,
+      "learning_rate": 5.729570740067758e-06,
+      "loss": 0.053,
+      "step": 951
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6927334913056521,
+      "learning_rate": 5.721653433698082e-06,
+      "loss": 0.0322,
+      "step": 952
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0192490892327637,
+      "learning_rate": 5.713734278963407e-06,
+      "loss": 0.0478,
+      "step": 953
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4021546573898167,
+      "learning_rate": 5.705813296146998e-06,
+      "loss": 0.0281,
+      "step": 954
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.283203718705625,
+      "learning_rate": 5.697890505536805e-06,
+      "loss": 0.0494,
+      "step": 955
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8723678181245383,
+      "learning_rate": 5.689965927425407e-06,
+      "loss": 0.039,
+      "step": 956
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9154126878492291,
+      "learning_rate": 5.682039582109959e-06,
+      "loss": 0.039,
+      "step": 957
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0000734940623763,
+      "learning_rate": 5.674111489892144e-06,
+      "loss": 0.0477,
+      "step": 958
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.797034006036536,
+      "learning_rate": 5.666181671078123e-06,
+      "loss": 0.043,
+      "step": 959
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0236055666082888,
+      "learning_rate": 5.658250145978469e-06,
+      "loss": 0.0273,
+      "step": 960
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9664406384627875,
+      "learning_rate": 5.650316934908138e-06,
+      "loss": 0.0336,
+      "step": 961
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.487854141369925,
+      "learning_rate": 5.642382058186394e-06,
+      "loss": 0.0265,
+      "step": 962
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9545374665607708,
+      "learning_rate": 5.634445536136774e-06,
+      "loss": 0.0446,
+      "step": 963
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8919658607697852,
+      "learning_rate": 5.626507389087026e-06,
+      "loss": 0.0366,
+      "step": 964
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7391178186132781,
+      "learning_rate": 5.61856763736906e-06,
+      "loss": 0.0311,
+      "step": 965
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9268738865398216,
+      "learning_rate": 5.610626301318897e-06,
+      "loss": 0.0498,
+      "step": 966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9120637509821803,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.0359,
+      "step": 967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0265774489726658,
+      "learning_rate": 5.594738957586302e-06,
+      "loss": 0.025,
+      "step": 968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5550363052752717,
+      "learning_rate": 5.586792990595992e-06,
+      "loss": 0.0299,
+      "step": 969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6178986655486775,
+      "learning_rate": 5.578845520657625e-06,
+      "loss": 0.0249,
+      "step": 970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6321963136394885,
+      "learning_rate": 5.570896568126994e-06,
+      "loss": 0.027,
+      "step": 971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8370481748494285,
+      "learning_rate": 5.562946153363681e-06,
+      "loss": 0.0403,
+      "step": 972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0627524764235494,
+      "learning_rate": 5.5549942967310214e-06,
+      "loss": 0.0336,
+      "step": 973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3932288711882305,
+      "learning_rate": 5.547041018596039e-06,
+      "loss": 0.0198,
+      "step": 974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8964752757758289,
+      "learning_rate": 5.539086339329398e-06,
+      "loss": 0.0444,
+      "step": 975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49690998582614554,
+      "learning_rate": 5.531130279305357e-06,
+      "loss": 0.0273,
+      "step": 976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9808838263146912,
+      "learning_rate": 5.523172858901703e-06,
+      "loss": 0.0477,
+      "step": 977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8666016629499839,
+      "learning_rate": 5.515214098499712e-06,
+      "loss": 0.0361,
+      "step": 978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.028073378457674,
+      "learning_rate": 5.507254018484094e-06,
+      "loss": 0.0392,
+      "step": 979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2707931952783673,
+      "learning_rate": 5.499292639242933e-06,
+      "loss": 0.0294,
+      "step": 980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0376209731413644,
+      "learning_rate": 5.491329981167647e-06,
+      "loss": 0.0436,
+      "step": 981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8080034651931363,
+      "learning_rate": 5.483366064652925e-06,
+      "loss": 0.035,
+      "step": 982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5472502566509414,
+      "learning_rate": 5.475400910096682e-06,
+      "loss": 0.0371,
+      "step": 983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7416379930088657,
+      "learning_rate": 5.4674345379e-06,
+      "loss": 0.0339,
+      "step": 984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3090824691105016,
+      "learning_rate": 5.4594669684670874e-06,
+      "loss": 0.0523,
+      "step": 985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3517769265087425,
+      "learning_rate": 5.451498222205211e-06,
+      "loss": 0.0509,
+      "step": 986
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5490271439552697,
+      "learning_rate": 5.443528319524653e-06,
+      "loss": 0.0248,
+      "step": 987
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8278875192150178,
+      "learning_rate": 5.435557280838663e-06,
+      "loss": 0.0359,
+      "step": 988
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4484530755745902,
+      "learning_rate": 5.427585126563395e-06,
+      "loss": 0.0236,
+      "step": 989
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.370692615590824,
+      "learning_rate": 5.419611877117864e-06,
+      "loss": 0.0392,
+      "step": 990
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0708257474470337,
+      "learning_rate": 5.411637552923887e-06,
+      "loss": 0.0496,
+      "step": 991
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5272878667231355,
+      "learning_rate": 5.403662174406033e-06,
+      "loss": 0.027,
+      "step": 992
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.799116176048434,
+      "learning_rate": 5.395685761991576e-06,
+      "loss": 0.0306,
+      "step": 993
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5795420159041744,
+      "learning_rate": 5.387708336110434e-06,
+      "loss": 0.0514,
+      "step": 994
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3530404342180486,
+      "learning_rate": 5.379729917195124e-06,
+      "loss": 0.0495,
+      "step": 995
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1542699573921147,
+      "learning_rate": 5.371750525680703e-06,
+      "loss": 0.0414,
+      "step": 996
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6465003437919752,
+      "learning_rate": 5.36377018200472e-06,
+      "loss": 0.0344,
+      "step": 997
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0695478253107236,
+      "learning_rate": 5.355788906607164e-06,
+      "loss": 0.024,
+      "step": 998
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.819386420050746,
+      "learning_rate": 5.347806719930414e-06,
+      "loss": 0.0206,
+      "step": 999
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6586229658287968,
+      "learning_rate": 5.339823642419174e-06,
+      "loss": 0.0542,
+      "step": 1000
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8943229287072868,
+      "learning_rate": 5.331839694520436e-06,
+      "loss": 0.0306,
+      "step": 1001
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0061767635442918,
+      "learning_rate": 5.323854896683422e-06,
+      "loss": 0.0305,
+      "step": 1002
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3117131384625127,
+      "learning_rate": 5.315869269359527e-06,
+      "loss": 0.0483,
+      "step": 1003
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.483692159272879,
+      "learning_rate": 5.307882833002271e-06,
+      "loss": 0.0306,
+      "step": 1004
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4928013172540946,
+      "learning_rate": 5.299895608067253e-06,
+      "loss": 0.0272,
+      "step": 1005
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9350981952793932,
+      "learning_rate": 5.291907615012081e-06,
+      "loss": 0.0257,
+      "step": 1006
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1880304932342811,
+      "learning_rate": 5.283918874296338e-06,
+      "loss": 0.0478,
+      "step": 1007
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.567658791017144,
+      "learning_rate": 5.2759294063815205e-06,
+      "loss": 0.0207,
+      "step": 1008
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5878617256713423,
+      "learning_rate": 5.267939231730985e-06,
+      "loss": 0.0259,
+      "step": 1009
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6296036095600541,
+      "learning_rate": 5.259948370809902e-06,
+      "loss": 0.03,
+      "step": 1010
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.086582723137742,
+      "learning_rate": 5.251956844085198e-06,
+      "loss": 0.0315,
+      "step": 1011
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7493884416834921,
+      "learning_rate": 5.243964672025502e-06,
+      "loss": 0.027,
+      "step": 1012
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5755763721876781,
+      "learning_rate": 5.235971875101101e-06,
+      "loss": 0.024,
+      "step": 1013
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0336876559829822,
+      "learning_rate": 5.22797847378388e-06,
+      "loss": 0.0324,
+      "step": 1014
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7680003907100028,
+      "learning_rate": 5.219984488547269e-06,
+      "loss": 0.0341,
+      "step": 1015
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8226456680704546,
+      "learning_rate": 5.2119899398662e-06,
+      "loss": 0.0324,
+      "step": 1016
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6982266202238494,
+      "learning_rate": 5.203994848217043e-06,
+      "loss": 0.0286,
+      "step": 1017
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7139453038989596,
+      "learning_rate": 5.195999234077561e-06,
+      "loss": 0.0263,
+      "step": 1018
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7446444911541764,
+      "learning_rate": 5.188003117926854e-06,
+      "loss": 0.0177,
+      "step": 1019
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.3514532648365165,
+      "learning_rate": 5.1800065202453095e-06,
+      "loss": 0.0361,
+      "step": 1020
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9721865980606375,
+      "learning_rate": 5.1720094615145455e-06,
+      "loss": 0.0389,
+      "step": 1021
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3703407080177544,
+      "learning_rate": 5.164011962217366e-06,
+      "loss": 0.019,
+      "step": 1022
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9599534838359112,
+      "learning_rate": 5.156014042837696e-06,
+      "loss": 0.0285,
+      "step": 1023
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6398162968662446,
+      "learning_rate": 5.148015723860543e-06,
+      "loss": 0.0251,
+      "step": 1024
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6225810158973047,
+      "learning_rate": 5.140017025771936e-06,
+      "loss": 0.0332,
+      "step": 1025
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7141483456674433,
+      "learning_rate": 5.1320179690588735e-06,
+      "loss": 0.0286,
+      "step": 1026
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9907387551938683,
+      "learning_rate": 5.124018574209272e-06,
+      "loss": 0.0432,
+      "step": 1027
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7904414734004622,
+      "learning_rate": 5.116018861711919e-06,
+      "loss": 0.0287,
+      "step": 1028
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.3076855298975811,
+      "learning_rate": 5.108018852056411e-06,
+      "loss": 0.0428,
+      "step": 1029
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.47849407068874555,
+      "learning_rate": 5.100018565733106e-06,
+      "loss": 0.0245,
+      "step": 1030
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.2447035809415234,
+      "learning_rate": 5.092018023233072e-06,
+      "loss": 0.0332,
+      "step": 1031
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8381997558957843,
+      "learning_rate": 5.084017245048034e-06,
+      "loss": 0.0321,
+      "step": 1032
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7945475267321247,
+      "learning_rate": 5.0760162516703156e-06,
+      "loss": 0.0382,
+      "step": 1033
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8671733382239873,
+      "learning_rate": 5.068015063592799e-06,
+      "loss": 0.0372,
+      "step": 1034
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8604675588518298,
+      "learning_rate": 5.06001370130886e-06,
+      "loss": 0.0178,
+      "step": 1035
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.0987612417558899,
+      "learning_rate": 5.052012185312322e-06,
+      "loss": 0.0402,
+      "step": 1036
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.107296934615035,
+      "learning_rate": 5.044010536097402e-06,
+      "loss": 0.03,
+      "step": 1037
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8614135705236582,
+      "learning_rate": 5.036008774158658e-06,
+      "loss": 0.034,
+      "step": 1038
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7233307057992016,
+      "learning_rate": 5.028006919990936e-06,
+      "loss": 0.027,
+      "step": 1039
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6366263075262523,
+      "learning_rate": 5.0200049940893225e-06,
+      "loss": 0.03,
+      "step": 1040
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5525916393083419,
+      "learning_rate": 5.012003016949082e-06,
+      "loss": 0.0271,
+      "step": 1041
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8220356610097129,
+      "learning_rate": 5.004001009065611e-06,
+      "loss": 0.031,
+      "step": 1042
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8725303219826496,
+      "learning_rate": 4.99599899093439e-06,
+      "loss": 0.0317,
+      "step": 1043
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.969863129435251,
+      "learning_rate": 4.98799698305092e-06,
+      "loss": 0.0257,
+      "step": 1044
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7438173977745536,
+      "learning_rate": 4.979995005910679e-06,
+      "loss": 0.0189,
+      "step": 1045
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.3051006180712026,
+      "learning_rate": 4.971993080009065e-06,
+      "loss": 0.0304,
+      "step": 1046
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5562679937676483,
+      "learning_rate": 4.9639912258413435e-06,
+      "loss": 0.0237,
+      "step": 1047
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8810611430495988,
+      "learning_rate": 4.955989463902599e-06,
+      "loss": 0.0297,
+      "step": 1048
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7041819608996103,
+      "learning_rate": 4.94798781468768e-06,
+      "loss": 0.0306,
+      "step": 1049
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.850027647311031,
+      "learning_rate": 4.939986298691141e-06,
+      "loss": 0.0218,
+      "step": 1050
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.8282525428235044,
+      "learning_rate": 4.931984936407202e-06,
+      "loss": 0.0393,
+      "step": 1051
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.018508686889688,
+      "learning_rate": 4.923983748329685e-06,
+      "loss": 0.0312,
+      "step": 1052
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.6184868153222927,
+      "learning_rate": 4.9159827549519676e-06,
+      "loss": 0.0345,
+      "step": 1053
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.964887841469877,
+      "learning_rate": 4.907981976766928e-06,
+      "loss": 0.0469,
+      "step": 1054
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7869754204093123,
+      "learning_rate": 4.899981434266895e-06,
+      "loss": 0.0314,
+      "step": 1055
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.514845941566159,
+      "learning_rate": 4.891981147943589e-06,
+      "loss": 0.0255,
+      "step": 1056
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8788715161807736,
+      "learning_rate": 4.883981138288081e-06,
+      "loss": 0.0244,
+      "step": 1057
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.046570067083794,
+      "learning_rate": 4.8759814257907275e-06,
+      "loss": 0.0316,
+      "step": 1058
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.0154582399674494,
+      "learning_rate": 4.867982030941127e-06,
+      "loss": 0.0345,
+      "step": 1059
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.3103756035719953,
+      "learning_rate": 4.859982974228065e-06,
+      "loss": 0.0248,
+      "step": 1060
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.23368046234692,
+      "learning_rate": 4.851984276139458e-06,
+      "loss": 0.0395,
+      "step": 1061
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1098419076483146,
+      "learning_rate": 4.843985957162304e-06,
+      "loss": 0.0276,
+      "step": 1062
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.942607118035288,
+      "learning_rate": 4.835988037782635e-06,
+      "loss": 0.0533,
+      "step": 1063
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6421131635296107,
+      "learning_rate": 4.827990538485456e-06,
+      "loss": 0.0239,
+      "step": 1064
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.3053339694382833,
+      "learning_rate": 4.819993479754693e-06,
+      "loss": 0.0353,
+      "step": 1065
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9986906730806447,
+      "learning_rate": 4.811996882073148e-06,
+      "loss": 0.0265,
+      "step": 1066
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8181143387578583,
+      "learning_rate": 4.804000765922441e-06,
+      "loss": 0.0216,
+      "step": 1067
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.502546600151344,
+      "learning_rate": 4.796005151782958e-06,
+      "loss": 0.0445,
+      "step": 1068
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7044947654375584,
+      "learning_rate": 4.788010060133802e-06,
+      "loss": 0.0251,
+      "step": 1069
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6013429646716092,
+      "learning_rate": 4.780015511452732e-06,
+      "loss": 0.0256,
+      "step": 1070
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5427439673220715,
+      "learning_rate": 4.772021526216123e-06,
+      "loss": 0.0263,
+      "step": 1071
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.769644548462023,
+      "learning_rate": 4.764028124898901e-06,
+      "loss": 0.0335,
+      "step": 1072
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7447494663693941,
+      "learning_rate": 4.756035327974499e-06,
+      "loss": 0.0223,
+      "step": 1073
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7204796166780826,
+      "learning_rate": 4.748043155914804e-06,
+      "loss": 0.03,
+      "step": 1074
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.0177867742746178,
+      "learning_rate": 4.740051629190099e-06,
+      "loss": 0.0341,
+      "step": 1075
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7421615617814211,
+      "learning_rate": 4.732060768269016e-06,
+      "loss": 0.0273,
+      "step": 1076
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.8620381131051364,
+      "learning_rate": 4.724070593618482e-06,
+      "loss": 0.0263,
+      "step": 1077
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.20839082094382,
+      "learning_rate": 4.716081125703665e-06,
+      "loss": 0.0272,
+      "step": 1078
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.2494484460524444,
+      "learning_rate": 4.708092384987921e-06,
+      "loss": 0.0414,
+      "step": 1079
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9688768305749882,
+      "learning_rate": 4.70010439193275e-06,
+      "loss": 0.0307,
+      "step": 1080
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9176018744878512,
+      "learning_rate": 4.6921171669977304e-06,
+      "loss": 0.0333,
+      "step": 1081
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.4328823254667193,
+      "learning_rate": 4.684130730640475e-06,
+      "loss": 0.037,
+      "step": 1082
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0825269206875192,
+      "learning_rate": 4.676145103316579e-06,
+      "loss": 0.0403,
+      "step": 1083
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.159048100788253,
+      "learning_rate": 4.6681603054795654e-06,
+      "loss": 0.0327,
+      "step": 1084
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8399124550508421,
+      "learning_rate": 4.660176357580827e-06,
+      "loss": 0.0295,
+      "step": 1085
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.42909450654776243,
+      "learning_rate": 4.652193280069588e-06,
+      "loss": 0.0223,
+      "step": 1086
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5824237081174857,
+      "learning_rate": 4.644211093392837e-06,
+      "loss": 0.0277,
+      "step": 1087
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7172546484982484,
+      "learning_rate": 4.636229817995281e-06,
+      "loss": 0.027,
+      "step": 1088
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.4487630330036912,
+      "learning_rate": 4.6282494743193e-06,
+      "loss": 0.0156,
+      "step": 1089
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.9578673232337767,
+      "learning_rate": 4.620270082804879e-06,
+      "loss": 0.035,
+      "step": 1090
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8541729174877927,
+      "learning_rate": 4.612291663889567e-06,
+      "loss": 0.0333,
+      "step": 1091
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.413518111201729,
+      "learning_rate": 4.604314238008426e-06,
+      "loss": 0.0293,
+      "step": 1092
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.013397355900455,
+      "learning_rate": 4.596337825593969e-06,
+      "loss": 0.0198,
+      "step": 1093
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8488149353686695,
+      "learning_rate": 4.588362447076115e-06,
+      "loss": 0.0308,
+      "step": 1094
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.808337943132434,
+      "learning_rate": 4.5803881228821375e-06,
+      "loss": 0.0277,
+      "step": 1095
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.9732316718127773,
+      "learning_rate": 4.572414873436606e-06,
+      "loss": 0.0268,
+      "step": 1096
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6459252848175363,
+      "learning_rate": 4.564442719161338e-06,
+      "loss": 0.0279,
+      "step": 1097
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6935117069628923,
+      "learning_rate": 4.556471680475348e-06,
+      "loss": 0.0239,
+      "step": 1098
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.607783773980419,
+      "learning_rate": 4.548501777794792e-06,
+      "loss": 0.0186,
+      "step": 1099
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.3877976209094554,
+      "learning_rate": 4.540533031532913e-06,
+      "loss": 0.0331,
+      "step": 1100
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.6712739416461726,
+      "learning_rate": 4.532565462099999e-06,
+      "loss": 0.0432,
+      "step": 1101
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7879664908382655,
+      "learning_rate": 4.524599089903319e-06,
+      "loss": 0.028,
+      "step": 1102
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7875123164738963,
+      "learning_rate": 4.516633935347075e-06,
+      "loss": 0.0301,
+      "step": 1103
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.9561176864842041,
+      "learning_rate": 4.508670018832353e-06,
+      "loss": 0.0316,
+      "step": 1104
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5519096705716723,
+      "learning_rate": 4.5007073607570674e-06,
+      "loss": 0.024,
+      "step": 1105
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.958646382749322,
+      "learning_rate": 4.492745981515907e-06,
+      "loss": 0.036,
+      "step": 1106
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.924537960903203,
+      "learning_rate": 4.48478590150029e-06,
+      "loss": 0.0306,
+      "step": 1107
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.852576984550701,
+      "learning_rate": 4.4768271410983e-06,
+      "loss": 0.0354,
+      "step": 1108
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8059026291970833,
+      "learning_rate": 4.468869720694647e-06,
+      "loss": 0.022,
+      "step": 1109
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8622128063886197,
+      "learning_rate": 4.460913660670604e-06,
+      "loss": 0.0302,
+      "step": 1110
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.0005211058131525,
+      "learning_rate": 4.452958981403963e-06,
+      "loss": 0.0244,
+      "step": 1111
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.49953654433337896,
+      "learning_rate": 4.445005703268981e-06,
+      "loss": 0.0246,
+      "step": 1112
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5211757187596917,
+      "learning_rate": 4.4370538466363216e-06,
+      "loss": 0.0282,
+      "step": 1113
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7049283173553994,
+      "learning_rate": 4.429103431873009e-06,
+      "loss": 0.0275,
+      "step": 1114
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8760836408069194,
+      "learning_rate": 4.421154479342377e-06,
+      "loss": 0.0258,
+      "step": 1115
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5497290975156301,
+      "learning_rate": 4.413207009404012e-06,
+      "loss": 0.0234,
+      "step": 1116
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.43160638727821743,
+      "learning_rate": 4.4052610424137e-06,
+      "loss": 0.0204,
+      "step": 1117
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5232450640935504,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.0224,
+      "step": 1118
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.47825044777409836,
+      "learning_rate": 4.389373698681105e-06,
+      "loss": 0.0206,
+      "step": 1119
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.991085904207292,
+      "learning_rate": 4.381432362630942e-06,
+      "loss": 0.0321,
+      "step": 1120
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.987193721368169,
+      "learning_rate": 4.373492610912976e-06,
+      "loss": 0.0312,
+      "step": 1121
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6093671176495833,
+      "learning_rate": 4.365554463863228e-06,
+      "loss": 0.0219,
+      "step": 1122
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.496517302197376,
+      "learning_rate": 4.3576179418136075e-06,
+      "loss": 0.0203,
+      "step": 1123
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6130892595308474,
+      "learning_rate": 4.349683065091864e-06,
+      "loss": 0.0306,
+      "step": 1124
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7776082447658934,
+      "learning_rate": 4.3417498540215325e-06,
+      "loss": 0.0398,
+      "step": 1125
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.3171830830330598,
+      "learning_rate": 4.33381832892188e-06,
+      "loss": 0.038,
+      "step": 1126
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6685509894900467,
+      "learning_rate": 4.3258885101078565e-06,
+      "loss": 0.0302,
+      "step": 1127
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.1144124292702757,
+      "learning_rate": 4.317960417890043e-06,
+      "loss": 0.0275,
+      "step": 1128
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.501368764729424,
+      "learning_rate": 4.3100340725745934e-06,
+      "loss": 0.0374,
+      "step": 1129
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.2663880435731152,
+      "learning_rate": 4.3021094944631955e-06,
+      "loss": 0.0341,
+      "step": 1130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9252347942434317,
+      "learning_rate": 4.294186703853004e-06,
+      "loss": 0.0277,
+      "step": 1131
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6717103063934182,
+      "learning_rate": 4.286265721036595e-06,
+      "loss": 0.027,
+      "step": 1132
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.1171647120173334,
+      "learning_rate": 4.27834656630192e-06,
+      "loss": 0.0428,
+      "step": 1133
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7077621527686202,
+      "learning_rate": 4.270429259932243e-06,
+      "loss": 0.0268,
+      "step": 1134
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.1261178104190204,
+      "learning_rate": 4.262513822206095e-06,
+      "loss": 0.0346,
+      "step": 1135
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5321911431043505,
+      "learning_rate": 4.254600273397223e-06,
+      "loss": 0.0205,
+      "step": 1136
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6395519526652998,
+      "learning_rate": 4.246688633774534e-06,
+      "loss": 0.0236,
+      "step": 1137
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5594885034804874,
+      "learning_rate": 4.238778923602045e-06,
+      "loss": 0.014,
+      "step": 1138
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8895582256492076,
+      "learning_rate": 4.230871163138831e-06,
+      "loss": 0.0257,
+      "step": 1139
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8120185835677661,
+      "learning_rate": 4.2229653726389765e-06,
+      "loss": 0.0261,
+      "step": 1140
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7699512160395745,
+      "learning_rate": 4.215061572351513e-06,
+      "loss": 0.0371,
+      "step": 1141
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7904391026632073,
+      "learning_rate": 4.207159782520383e-06,
+      "loss": 0.0143,
+      "step": 1142
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.04855099373956,
+      "learning_rate": 4.199260023384376e-06,
+      "loss": 0.0431,
+      "step": 1143
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.2533318783486824,
+      "learning_rate": 4.1913623151770765e-06,
+      "loss": 0.0318,
+      "step": 1144
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.9220419806025165,
+      "learning_rate": 4.183466678126822e-06,
+      "loss": 0.0317,
+      "step": 1145
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.647797635933002,
+      "learning_rate": 4.175573132456644e-06,
+      "loss": 0.0221,
+      "step": 1146
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5143353606003064,
+      "learning_rate": 4.167681698384211e-06,
+      "loss": 0.026,
+      "step": 1147
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.9659467748981236,
+      "learning_rate": 4.1597923961217935e-06,
+      "loss": 0.028,
+      "step": 1148
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8694130191332563,
+      "learning_rate": 4.151905245876194e-06,
+      "loss": 0.0302,
+      "step": 1149
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.587958711576844,
+      "learning_rate": 4.144020267848707e-06,
+      "loss": 0.02,
+      "step": 1150
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7156546748797903,
+      "learning_rate": 4.13613748223506e-06,
+      "loss": 0.0158,
+      "step": 1151
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0851432279340858,
+      "learning_rate": 4.128256909225366e-06,
+      "loss": 0.0286,
+      "step": 1152
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7096178232213832,
+      "learning_rate": 4.120378569004074e-06,
+      "loss": 0.0316,
+      "step": 1153
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5717606844829977,
+      "learning_rate": 4.112502481749911e-06,
+      "loss": 0.0194,
+      "step": 1154
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0662768483868281,
+      "learning_rate": 4.104628667635835e-06,
+      "loss": 0.0272,
+      "step": 1155
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0802239084531717,
+      "learning_rate": 4.0967571468289815e-06,
+      "loss": 0.0419,
+      "step": 1156
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.075912303145742,
+      "learning_rate": 4.0888879394906094e-06,
+      "loss": 0.0311,
+      "step": 1157
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.2670503288679043,
+      "learning_rate": 4.081021065776058e-06,
+      "loss": 0.0345,
+      "step": 1158
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5912311253811383,
+      "learning_rate": 4.073156545834685e-06,
+      "loss": 0.0306,
+      "step": 1159
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7740612434514832,
+      "learning_rate": 4.065294399809819e-06,
+      "loss": 0.0248,
+      "step": 1160
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6660766846869574,
+      "learning_rate": 4.057434647838713e-06,
+      "loss": 0.0272,
+      "step": 1161
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.4464175518732789,
+      "learning_rate": 4.049577310052482e-06,
+      "loss": 0.0187,
+      "step": 1162
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6420961971799071,
+      "learning_rate": 4.041722406576062e-06,
+      "loss": 0.0178,
+      "step": 1163
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5933792075717889,
+      "learning_rate": 4.033869957528153e-06,
+      "loss": 0.023,
+      "step": 1164
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7545983123520259,
+      "learning_rate": 4.026019983021168e-06,
+      "loss": 0.0279,
+      "step": 1165
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.0413887638237633,
+      "learning_rate": 4.018172503161179e-06,
+      "loss": 0.0371,
+      "step": 1166
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.0934504551905797,
+      "learning_rate": 4.010327538047877e-06,
+      "loss": 0.0328,
+      "step": 1167
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.36078229020799923,
+      "learning_rate": 4.002485107774503e-06,
+      "loss": 0.016,
+      "step": 1168
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6444553615010122,
+      "learning_rate": 3.994645232427809e-06,
+      "loss": 0.0291,
+      "step": 1169
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6594877072308999,
+      "learning_rate": 3.986807932088004e-06,
+      "loss": 0.0232,
+      "step": 1170
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9328266110281278,
+      "learning_rate": 3.978973226828702e-06,
+      "loss": 0.0312,
+      "step": 1171
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5447087731555137,
+      "learning_rate": 3.971141136716866e-06,
+      "loss": 0.0189,
+      "step": 1172
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9056993750867337,
+      "learning_rate": 3.963311681812768e-06,
+      "loss": 0.0228,
+      "step": 1173
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7617244908266496,
+      "learning_rate": 3.955484882169923e-06,
+      "loss": 0.0207,
+      "step": 1174
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.053093794917535,
+      "learning_rate": 3.947660757835049e-06,
+      "loss": 0.0368,
+      "step": 1175
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7773155500120877,
+      "learning_rate": 3.939839328848014e-06,
+      "loss": 0.0275,
+      "step": 1176
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.1309111614765357,
+      "learning_rate": 3.932020615241777e-06,
+      "loss": 0.0391,
+      "step": 1177
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7772381331788334,
+      "learning_rate": 3.9242046370423434e-06,
+      "loss": 0.0228,
+      "step": 1178
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0497661692021192,
+      "learning_rate": 3.9163914142687185e-06,
+      "loss": 0.0197,
+      "step": 1179
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.802443273282112,
+      "learning_rate": 3.9085809669328415e-06,
+      "loss": 0.0287,
+      "step": 1180
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.6422605212693915,
+      "learning_rate": 3.900773315039548e-06,
+      "loss": 0.0263,
+      "step": 1181
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.4827986025988822,
+      "learning_rate": 3.892968478586513e-06,
+      "loss": 0.0385,
+      "step": 1182
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0726420346174323,
+      "learning_rate": 3.885166477564199e-06,
+      "loss": 0.0315,
+      "step": 1183
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.4868137126849361,
+      "learning_rate": 3.877367331955808e-06,
+      "loss": 0.0301,
+      "step": 1184
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7793921575940534,
+      "learning_rate": 3.869571061737226e-06,
+      "loss": 0.0278,
+      "step": 1185
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7423673491553581,
+      "learning_rate": 3.861777686876978e-06,
+      "loss": 0.0172,
+      "step": 1186
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8445082531782259,
+      "learning_rate": 3.853987227336168e-06,
+      "loss": 0.0234,
+      "step": 1187
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.4866791378338005,
+      "learning_rate": 3.8461997030684386e-06,
+      "loss": 0.0413,
+      "step": 1188
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.696171863459194,
+      "learning_rate": 3.838415134019911e-06,
+      "loss": 0.0278,
+      "step": 1189
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9645765538518006,
+      "learning_rate": 3.830633540129135e-06,
+      "loss": 0.0333,
+      "step": 1190
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.1588316179391511,
+      "learning_rate": 3.822854941327046e-06,
+      "loss": 0.0247,
+      "step": 1191
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8964545735956162,
+      "learning_rate": 3.815079357536907e-06,
+      "loss": 0.0244,
+      "step": 1192
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9118777787298978,
+      "learning_rate": 3.8073068086742514e-06,
+      "loss": 0.0275,
+      "step": 1193
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.777626923225908,
+      "learning_rate": 3.799537314646848e-06,
+      "loss": 0.0205,
+      "step": 1194
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6685174415093665,
+      "learning_rate": 3.791770895354635e-06,
+      "loss": 0.0206,
+      "step": 1195
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.48232058435281616,
+      "learning_rate": 3.7840075706896824e-06,
+      "loss": 0.0207,
+      "step": 1196
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.0217841454148604,
+      "learning_rate": 3.776247360536127e-06,
+      "loss": 0.0393,
+      "step": 1197
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6119942879107466,
+      "learning_rate": 3.768490284770131e-06,
+      "loss": 0.0203,
+      "step": 1198
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5206581259536639,
+      "learning_rate": 3.7607363632598305e-06,
+      "loss": 0.0196,
+      "step": 1199
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.9881570998597756,
+      "learning_rate": 3.7529856158652792e-06,
+      "loss": 0.0163,
+      "step": 1200
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6382121125866328,
+      "learning_rate": 3.7452380624384026e-06,
+      "loss": 0.0231,
+      "step": 1201
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.4224758933081467,
+      "learning_rate": 3.7374937228229472e-06,
+      "loss": 0.0139,
+      "step": 1202
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.38564278891273984,
+      "learning_rate": 3.7297526168544253e-06,
+      "loss": 0.0173,
+      "step": 1203
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0962566818321713,
+      "learning_rate": 3.722014764360067e-06,
+      "loss": 0.025,
+      "step": 1204
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.2246173963024596,
+      "learning_rate": 3.714280185158771e-06,
+      "loss": 0.0406,
+      "step": 1205
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8603176514449102,
+      "learning_rate": 3.706548899061052e-06,
+      "loss": 0.026,
+      "step": 1206
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.2822870337099297,
+      "learning_rate": 3.6988209258689877e-06,
+      "loss": 0.0313,
+      "step": 1207
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.2715509865915087,
+      "learning_rate": 3.6910962853761738e-06,
+      "loss": 0.0355,
+      "step": 1208
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.1853601452520386,
+      "learning_rate": 3.683374997367668e-06,
+      "loss": 0.0321,
+      "step": 1209
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.902921247787625,
+      "learning_rate": 3.675657081619941e-06,
+      "loss": 0.0191,
+      "step": 1210
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7821409311223014,
+      "learning_rate": 3.6679425579008278e-06,
+      "loss": 0.0326,
+      "step": 1211
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7903619927425193,
+      "learning_rate": 3.6602314459694743e-06,
+      "loss": 0.0301,
+      "step": 1212
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.46174044405418857,
+      "learning_rate": 3.652523765576287e-06,
+      "loss": 0.022,
+      "step": 1213
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6323491372648098,
+      "learning_rate": 3.6448195364628857e-06,
+      "loss": 0.0191,
+      "step": 1214
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9099434352944868,
+      "learning_rate": 3.6371187783620486e-06,
+      "loss": 0.0241,
+      "step": 1215
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9207530906993158,
+      "learning_rate": 3.6294215109976628e-06,
+      "loss": 0.0222,
+      "step": 1216
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5852047866123646,
+      "learning_rate": 3.6217277540846775e-06,
+      "loss": 0.0268,
+      "step": 1217
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7561125846165206,
+      "learning_rate": 3.614037527329048e-06,
+      "loss": 0.0286,
+      "step": 1218
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.7062712342887891,
+      "learning_rate": 3.606350850427688e-06,
+      "loss": 0.0356,
+      "step": 1219
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.681472660161719,
+      "learning_rate": 3.5986677430684224e-06,
+      "loss": 0.0225,
+      "step": 1220
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8267818175730947,
+      "learning_rate": 3.5909882249299287e-06,
+      "loss": 0.0318,
+      "step": 1221
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7355695025915546,
+      "learning_rate": 3.583312315681693e-06,
+      "loss": 0.0223,
+      "step": 1222
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.833987497480372,
+      "learning_rate": 3.5756400349839603e-06,
+      "loss": 0.0337,
+      "step": 1223
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.9794172004233616,
+      "learning_rate": 3.567971402487679e-06,
+      "loss": 0.0218,
+      "step": 1224
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.6501686738425505,
+      "learning_rate": 3.5603064378344536e-06,
+      "loss": 0.0262,
+      "step": 1225
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5674790693617395,
+      "learning_rate": 3.552645160656497e-06,
+      "loss": 0.0265,
+      "step": 1226
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0102812039517735,
+      "learning_rate": 3.544987590576574e-06,
+      "loss": 0.0226,
+      "step": 1227
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5226619099252662,
+      "learning_rate": 3.537333747207955e-06,
+      "loss": 0.0183,
+      "step": 1228
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0690310624262982,
+      "learning_rate": 3.529683650154368e-06,
+      "loss": 0.0367,
+      "step": 1229
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.019580208227227,
+      "learning_rate": 3.5220373190099428e-06,
+      "loss": 0.0379,
+      "step": 1230
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6233956991684912,
+      "learning_rate": 3.5143947733591633e-06,
+      "loss": 0.0231,
+      "step": 1231
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6281650494544383,
+      "learning_rate": 3.50675603277682e-06,
+      "loss": 0.024,
+      "step": 1232
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5383027091207303,
+      "learning_rate": 3.499121116827956e-06,
+      "loss": 0.0167,
+      "step": 1233
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7215097341859587,
+      "learning_rate": 3.491490045067818e-06,
+      "loss": 0.0258,
+      "step": 1234
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5910503130056679,
+      "learning_rate": 3.48386283704181e-06,
+      "loss": 0.0231,
+      "step": 1235
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.4693134066224136,
+      "learning_rate": 3.4762395122854336e-06,
+      "loss": 0.0363,
+      "step": 1236
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.166411864946178,
+      "learning_rate": 3.4686200903242497e-06,
+      "loss": 0.0259,
+      "step": 1237
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.4817375402839268,
+      "learning_rate": 3.461004590673819e-06,
+      "loss": 0.0344,
+      "step": 1238
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.046182158547702,
+      "learning_rate": 3.45339303283966e-06,
+      "loss": 0.0242,
+      "step": 1239
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.292224774416532,
+      "learning_rate": 3.445785436317193e-06,
+      "loss": 0.0375,
+      "step": 1240
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.861521851112122,
+      "learning_rate": 3.4381818205916907e-06,
+      "loss": 0.0218,
+      "step": 1241
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5716383717347113,
+      "learning_rate": 3.4305822051382345e-06,
+      "loss": 0.0254,
+      "step": 1242
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8052011902200021,
+      "learning_rate": 3.422986609421655e-06,
+      "loss": 0.0266,
+      "step": 1243
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9384565798258717,
+      "learning_rate": 3.4153950528964867e-06,
+      "loss": 0.0399,
+      "step": 1244
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9711098718625382,
+      "learning_rate": 3.4078075550069255e-06,
+      "loss": 0.0241,
+      "step": 1245
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6655061872931324,
+      "learning_rate": 3.400224135186765e-06,
+      "loss": 0.0364,
+      "step": 1246
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.5502248667043002,
+      "learning_rate": 3.392644812859354e-06,
+      "loss": 0.0416,
+      "step": 1247
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.844181133418411,
+      "learning_rate": 3.385069607437552e-06,
+      "loss": 0.0354,
+      "step": 1248
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.1257355441450307,
+      "learning_rate": 3.3774985383236685e-06,
+      "loss": 0.0232,
+      "step": 1249
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9739750889298195,
+      "learning_rate": 3.3699316249094195e-06,
+      "loss": 0.0395,
+      "step": 1250
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7690071951479069,
+      "learning_rate": 3.36236888657588e-06,
+      "loss": 0.0344,
+      "step": 1251
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.6660183823158623,
+      "learning_rate": 3.3548103426934287e-06,
+      "loss": 0.0213,
+      "step": 1252
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6154954723041898,
+      "learning_rate": 3.3472560126217004e-06,
+      "loss": 0.0192,
+      "step": 1253
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6999616838858199,
+      "learning_rate": 3.3397059157095412e-06,
+      "loss": 0.0222,
+      "step": 1254
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.625168930364848,
+      "learning_rate": 3.33216007129495e-06,
+      "loss": 0.0281,
+      "step": 1255
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7423624969087769,
+      "learning_rate": 3.3246184987050366e-06,
+      "loss": 0.0285,
+      "step": 1256
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5785260967877592,
+      "learning_rate": 3.3170812172559695e-06,
+      "loss": 0.0258,
+      "step": 1257
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.053003162298107,
+      "learning_rate": 3.309548246252925e-06,
+      "loss": 0.0259,
+      "step": 1258
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9626579997361799,
+      "learning_rate": 3.3020196049900386e-06,
+      "loss": 0.0341,
+      "step": 1259
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.5378411567759094,
+      "learning_rate": 3.2944953127503593e-06,
+      "loss": 0.0229,
+      "step": 1260
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.40918902129839757,
+      "learning_rate": 3.2869753888057936e-06,
+      "loss": 0.0174,
+      "step": 1261
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.948390041474773,
+      "learning_rate": 3.2794598524170606e-06,
+      "loss": 0.021,
+      "step": 1262
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9761585653303926,
+      "learning_rate": 3.271948722833643e-06,
+      "loss": 0.0245,
+      "step": 1263
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.1402260326532885,
+      "learning_rate": 3.264442019293734e-06,
+      "loss": 0.0289,
+      "step": 1264
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8272407776944543,
+      "learning_rate": 3.2569397610241915e-06,
+      "loss": 0.022,
+      "step": 1265
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.1065240923106912,
+      "learning_rate": 3.249441967240489e-06,
+      "loss": 0.0387,
+      "step": 1266
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7964507556545418,
+      "learning_rate": 3.241948657146663e-06,
+      "loss": 0.0258,
+      "step": 1267
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9722127655862852,
+      "learning_rate": 3.2344598499352663e-06,
+      "loss": 0.0233,
+      "step": 1268
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6702923565192391,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.02,
+      "step": 1269
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6781473563830941,
+      "learning_rate": 3.2194958208722656e-06,
+      "loss": 0.027,
+      "step": 1270
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8764587765192495,
+      "learning_rate": 3.2120206373479024e-06,
+      "loss": 0.0259,
+      "step": 1271
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.204359727983454,
+      "learning_rate": 3.204550033360362e-06,
+      "loss": 0.0261,
+      "step": 1272
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.3643673401914205,
+      "learning_rate": 3.1970840280440384e-06,
+      "loss": 0.0189,
+      "step": 1273
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7324647980496716,
+      "learning_rate": 3.1896226405215503e-06,
+      "loss": 0.0221,
+      "step": 1274
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8575165229725605,
+      "learning_rate": 3.1821658899036876e-06,
+      "loss": 0.0276,
+      "step": 1275
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7541136711033276,
+      "learning_rate": 3.174713795289366e-06,
+      "loss": 0.0322,
+      "step": 1276
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.9834424977627805,
+      "learning_rate": 3.1672663757655707e-06,
+      "loss": 0.0441,
+      "step": 1277
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.3702494488264347,
+      "learning_rate": 3.1598236504073215e-06,
+      "loss": 0.0329,
+      "step": 1278
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.02349427189603,
+      "learning_rate": 3.152385638277603e-06,
+      "loss": 0.0442,
+      "step": 1279
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.0267724041865374,
+      "learning_rate": 3.144952358427339e-06,
+      "loss": 0.0265,
+      "step": 1280
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.4656151906951841,
+      "learning_rate": 3.137523829895326e-06,
+      "loss": 0.0178,
+      "step": 1281
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9467462721261859,
+      "learning_rate": 3.1301000717081926e-06,
+      "loss": 0.0325,
+      "step": 1282
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.1709202769320373,
+      "learning_rate": 3.1226811028803514e-06,
+      "loss": 0.0333,
+      "step": 1283
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.3576000887150794,
+      "learning_rate": 3.115266942413946e-06,
+      "loss": 0.0301,
+      "step": 1284
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7941095758553105,
+      "learning_rate": 3.107857609298802e-06,
+      "loss": 0.021,
+      "step": 1285
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7929011356573532,
+      "learning_rate": 3.1004531225123857e-06,
+      "loss": 0.0256,
+      "step": 1286
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8786043128992731,
+      "learning_rate": 3.093053501019748e-06,
+      "loss": 0.0258,
+      "step": 1287
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.260719000880376,
+      "learning_rate": 3.0856587637734783e-06,
+      "loss": 0.0282,
+      "step": 1288
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.0540374948932114,
+      "learning_rate": 3.0782689297136554e-06,
+      "loss": 0.0309,
+      "step": 1289
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5725262960615425,
+      "learning_rate": 3.070884017767801e-06,
+      "loss": 0.022,
+      "step": 1290
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.959101589090888,
+      "learning_rate": 3.0635040468508294e-06,
+      "loss": 0.028,
+      "step": 1291
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5386872857085101,
+      "learning_rate": 3.0561290358649996e-06,
+      "loss": 0.0151,
+      "step": 1292
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.569838897864246,
+      "learning_rate": 3.048759003699866e-06,
+      "loss": 0.0207,
+      "step": 1293
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.581532645283473,
+      "learning_rate": 3.0413939692322304e-06,
+      "loss": 0.0225,
+      "step": 1294
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5028362927727279,
+      "learning_rate": 3.0340339513260976e-06,
+      "loss": 0.0235,
+      "step": 1295
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8007831617356519,
+      "learning_rate": 3.0266789688326187e-06,
+      "loss": 0.0227,
+      "step": 1296
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7695969527513412,
+      "learning_rate": 3.0193290405900494e-06,
+      "loss": 0.0196,
+      "step": 1297
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.0289670306604963,
+      "learning_rate": 3.0119841854237027e-06,
+      "loss": 0.0307,
+      "step": 1298
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6835324484005986,
+      "learning_rate": 3.004644422145894e-06,
+      "loss": 0.0292,
+      "step": 1299
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.581528963254537,
+      "learning_rate": 2.9973097695558982e-06,
+      "loss": 0.0259,
+      "step": 1300
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8906463659166122,
+      "learning_rate": 2.9899802464399018e-06,
+      "loss": 0.0441,
+      "step": 1301
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8464019379319843,
+      "learning_rate": 2.982655871570952e-06,
+      "loss": 0.0156,
+      "step": 1302
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.3655786669874588,
+      "learning_rate": 2.9753366637089075e-06,
+      "loss": 0.0263,
+      "step": 1303
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.4302998086198895,
+      "learning_rate": 2.9680226416003986e-06,
+      "loss": 0.0331,
+      "step": 1304
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6007442648446595,
+      "learning_rate": 2.960713823978769e-06,
+      "loss": 0.0188,
+      "step": 1305
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5396731652610417,
+      "learning_rate": 2.9534102295640305e-06,
+      "loss": 0.0216,
+      "step": 1306
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.2810585899025055,
+      "learning_rate": 2.9461118770628226e-06,
+      "loss": 0.0248,
+      "step": 1307
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8106101777855437,
+      "learning_rate": 2.9388187851683537e-06,
+      "loss": 0.0223,
+      "step": 1308
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6780089328890525,
+      "learning_rate": 2.9315309725603596e-06,
+      "loss": 0.0251,
+      "step": 1309
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5254825720095616,
+      "learning_rate": 2.9242484579050566e-06,
+      "loss": 0.0233,
+      "step": 1310
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5856391286139735,
+      "learning_rate": 2.9169712598550885e-06,
+      "loss": 0.0183,
+      "step": 1311
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6144133333430777,
+      "learning_rate": 2.9096993970494825e-06,
+      "loss": 0.0207,
+      "step": 1312
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7793656594295922,
+      "learning_rate": 2.9024328881136e-06,
+      "loss": 0.0275,
+      "step": 1313
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8547383245915084,
+      "learning_rate": 2.895171751659093e-06,
+      "loss": 0.024,
+      "step": 1314
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.136725575523068,
+      "learning_rate": 2.887916006283849e-06,
+      "loss": 0.0296,
+      "step": 1315
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.000339457408409,
+      "learning_rate": 2.8806656705719492e-06,
+      "loss": 0.0216,
+      "step": 1316
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5570093686866413,
+      "learning_rate": 2.8734207630936195e-06,
+      "loss": 0.0209,
+      "step": 1317
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.4454859134820076,
+      "learning_rate": 2.86618130240518e-06,
+      "loss": 0.0147,
+      "step": 1318
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.563113334058305,
+      "learning_rate": 2.858947307049001e-06,
+      "loss": 0.0187,
+      "step": 1319
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8038186022459026,
+      "learning_rate": 2.851718795553461e-06,
+      "loss": 0.0214,
+      "step": 1320
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8962721319951323,
+      "learning_rate": 2.84449578643288e-06,
+      "loss": 0.0332,
+      "step": 1321
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.0943982653356987,
+      "learning_rate": 2.8372782981874964e-06,
+      "loss": 0.0203,
+      "step": 1322
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6210789822485363,
+      "learning_rate": 2.830066349303401e-06,
+      "loss": 0.0207,
+      "step": 1323
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7558588675119113,
+      "learning_rate": 2.8228599582524975e-06,
+      "loss": 0.0279,
+      "step": 1324
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.2117255873190556,
+      "learning_rate": 2.815659143492461e-06,
+      "loss": 0.0197,
+      "step": 1325
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8905208919281724,
+      "learning_rate": 2.8084639234666753e-06,
+      "loss": 0.0254,
+      "step": 1326
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5673813607386939,
+      "learning_rate": 2.8012743166042002e-06,
+      "loss": 0.0195,
+      "step": 1327
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7629642074483579,
+      "learning_rate": 2.794090341319715e-06,
+      "loss": 0.0234,
+      "step": 1328
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.0756254635065226,
+      "learning_rate": 2.786912016013478e-06,
+      "loss": 0.0197,
+      "step": 1329
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7397629259597135,
+      "learning_rate": 2.7797393590712706e-06,
+      "loss": 0.0286,
+      "step": 1330
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.3606802714238877,
+      "learning_rate": 2.7725723888643663e-06,
+      "loss": 0.0195,
+      "step": 1331
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.1951944803879382,
+      "learning_rate": 2.765411123749463e-06,
+      "loss": 0.0227,
+      "step": 1332
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.572616845327841,
+      "learning_rate": 2.758255582068651e-06,
+      "loss": 0.0214,
+      "step": 1333
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6254891645360283,
+      "learning_rate": 2.75110578214936e-06,
+      "loss": 0.0226,
+      "step": 1334
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8258197500603863,
+      "learning_rate": 2.7439617423043146e-06,
+      "loss": 0.0253,
+      "step": 1335
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5567925238200621,
+      "learning_rate": 2.736823480831482e-06,
+      "loss": 0.0103,
+      "step": 1336
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9486102645398342,
+      "learning_rate": 2.7296910160140365e-06,
+      "loss": 0.036,
+      "step": 1337
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.9259932643996045,
+      "learning_rate": 2.7225643661203004e-06,
+      "loss": 0.0322,
+      "step": 1338
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.691222736049963,
+      "learning_rate": 2.7154435494037033e-06,
+      "loss": 0.0148,
+      "step": 1339
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8575355131535669,
+      "learning_rate": 2.708328584102734e-06,
+      "loss": 0.0181,
+      "step": 1340
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.1963430964304536,
+      "learning_rate": 2.7012194884408942e-06,
+      "loss": 0.0282,
+      "step": 1341
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9228574782155844,
+      "learning_rate": 2.6941162806266506e-06,
+      "loss": 0.0259,
+      "step": 1342
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6683360136989352,
+      "learning_rate": 2.6870189788533953e-06,
+      "loss": 0.0233,
+      "step": 1343
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6158335243969865,
+      "learning_rate": 2.679927601299386e-06,
+      "loss": 0.0224,
+      "step": 1344
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.4342317868225625,
+      "learning_rate": 2.6728421661277105e-06,
+      "loss": 0.0187,
+      "step": 1345
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8702462153937798,
+      "learning_rate": 2.665762691486235e-06,
+      "loss": 0.0279,
+      "step": 1346
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9076288865381044,
+      "learning_rate": 2.65868919550756e-06,
+      "loss": 0.0256,
+      "step": 1347
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7546622654755015,
+      "learning_rate": 2.6516216963089698e-06,
+      "loss": 0.0239,
+      "step": 1348
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5596958413844281,
+      "learning_rate": 2.6445602119923963e-06,
+      "loss": 0.0159,
+      "step": 1349
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.817812674260711,
+      "learning_rate": 2.637504760644359e-06,
+      "loss": 0.0173,
+      "step": 1350
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9082917695452071,
+      "learning_rate": 2.630455360335929e-06,
+      "loss": 0.0284,
+      "step": 1351
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0181782420057977,
+      "learning_rate": 2.623412029122675e-06,
+      "loss": 0.0235,
+      "step": 1352
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.4308595058667782,
+      "learning_rate": 2.616374785044624e-06,
+      "loss": 0.0195,
+      "step": 1353
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0010421051313705,
+      "learning_rate": 2.60934364612621e-06,
+      "loss": 0.0405,
+      "step": 1354
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.1072856388623673,
+      "learning_rate": 2.602318630376234e-06,
+      "loss": 0.0346,
+      "step": 1355
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.570554903732537,
+      "learning_rate": 2.59529975578781e-06,
+      "loss": 0.02,
+      "step": 1356
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.763034168480365,
+      "learning_rate": 2.588287040338323e-06,
+      "loss": 0.028,
+      "step": 1357
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.9818915105741374,
+      "learning_rate": 2.5812805019893837e-06,
+      "loss": 0.0255,
+      "step": 1358
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.37647009133737225,
+      "learning_rate": 2.574280158686782e-06,
+      "loss": 0.0128,
+      "step": 1359
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8141530396717688,
+      "learning_rate": 2.567286028360435e-06,
+      "loss": 0.0279,
+      "step": 1360
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7619954885706852,
+      "learning_rate": 2.560298128924358e-06,
+      "loss": 0.0212,
+      "step": 1361
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7786618256826614,
+      "learning_rate": 2.5533164782765974e-06,
+      "loss": 0.0244,
+      "step": 1362
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8083725721324221,
+      "learning_rate": 2.5463410942991986e-06,
+      "loss": 0.0298,
+      "step": 1363
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8305456421254139,
+      "learning_rate": 2.5393719948581507e-06,
+      "loss": 0.029,
+      "step": 1364
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.4458021272805112,
+      "learning_rate": 2.532409197803353e-06,
+      "loss": 0.0178,
+      "step": 1365
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6645665194259693,
+      "learning_rate": 2.525452720968563e-06,
+      "loss": 0.0207,
+      "step": 1366
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6310746208910438,
+      "learning_rate": 2.5185025821713454e-06,
+      "loss": 0.0175,
+      "step": 1367
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.0364074720417116,
+      "learning_rate": 2.5115587992130342e-06,
+      "loss": 0.0275,
+      "step": 1368
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6205551171830646,
+      "learning_rate": 2.504621389878682e-06,
+      "loss": 0.0301,
+      "step": 1369
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.49694932253221913,
+      "learning_rate": 2.4976903719370193e-06,
+      "loss": 0.0142,
+      "step": 1370
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5103829516719648,
+      "learning_rate": 2.4907657631404037e-06,
+      "loss": 0.023,
+      "step": 1371
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6578952733958501,
+      "learning_rate": 2.483847581224782e-06,
+      "loss": 0.0266,
+      "step": 1372
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.124638769126836,
+      "learning_rate": 2.4769358439096347e-06,
+      "loss": 0.0318,
+      "step": 1373
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8613357003477377,
+      "learning_rate": 2.470030568897938e-06,
+      "loss": 0.0329,
+      "step": 1374
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7481385202102532,
+      "learning_rate": 2.4631317738761155e-06,
+      "loss": 0.0261,
+      "step": 1375
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5082467310959725,
+      "learning_rate": 2.456239476513994e-06,
+      "loss": 0.0202,
+      "step": 1376
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.34702336652624877,
+      "learning_rate": 2.4493536944647566e-06,
+      "loss": 0.0202,
+      "step": 1377
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6745601361393228,
+      "learning_rate": 2.442474445364904e-06,
+      "loss": 0.0223,
+      "step": 1378
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.5011297804780708,
+      "learning_rate": 2.4356017468341977e-06,
+      "loss": 0.0338,
+      "step": 1379
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.934845789299032,
+      "learning_rate": 2.4287356164756244e-06,
+      "loss": 0.0204,
+      "step": 1380
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5768390021313147,
+      "learning_rate": 2.421876071875347e-06,
+      "loss": 0.0189,
+      "step": 1381
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.4666835633560529,
+      "learning_rate": 2.415023130602661e-06,
+      "loss": 0.0217,
+      "step": 1382
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7251505144838705,
+      "learning_rate": 2.408176810209946e-06,
+      "loss": 0.0245,
+      "step": 1383
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7030905046001982,
+      "learning_rate": 2.401337128232631e-06,
+      "loss": 0.0238,
+      "step": 1384
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7104189569845816,
+      "learning_rate": 2.3945041021891335e-06,
+      "loss": 0.0212,
+      "step": 1385
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.2295751661808245,
+      "learning_rate": 2.387677749580828e-06,
+      "loss": 0.0355,
+      "step": 1386
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6330065194593735,
+      "learning_rate": 2.3808580878919948e-06,
+      "loss": 0.0348,
+      "step": 1387
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.918874073383736,
+      "learning_rate": 2.3740451345897773e-06,
+      "loss": 0.0164,
+      "step": 1388
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.582892515629951,
+      "learning_rate": 2.3672389071241354e-06,
+      "loss": 0.0195,
+      "step": 1389
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.0182000574709098,
+      "learning_rate": 2.3604394229278064e-06,
+      "loss": 0.0366,
+      "step": 1390
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9912460164229965,
+      "learning_rate": 2.3536466994162522e-06,
+      "loss": 0.0272,
+      "step": 1391
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9809821583130328,
+      "learning_rate": 2.3468607539876186e-06,
+      "loss": 0.0209,
+      "step": 1392
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6201084241577994,
+      "learning_rate": 2.3400816040226925e-06,
+      "loss": 0.0199,
+      "step": 1393
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8996803265244651,
+      "learning_rate": 2.3333092668848544e-06,
+      "loss": 0.0333,
+      "step": 1394
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.4792146683688987,
+      "learning_rate": 2.326543759920034e-06,
+      "loss": 0.0203,
+      "step": 1395
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.45049080136019715,
+      "learning_rate": 2.3197851004566723e-06,
+      "loss": 0.0215,
+      "step": 1396
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.2387095657348046,
+      "learning_rate": 2.313033305805667e-06,
+      "loss": 0.0269,
+      "step": 1397
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5598441840881433,
+      "learning_rate": 2.3062883932603326e-06,
+      "loss": 0.0252,
+      "step": 1398
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7873601630358676,
+      "learning_rate": 2.2995503800963593e-06,
+      "loss": 0.0215,
+      "step": 1399
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.907763784643883,
+      "learning_rate": 2.2928192835717642e-06,
+      "loss": 0.0264,
+      "step": 1400
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8360874450536064,
+      "learning_rate": 2.2860951209268485e-06,
+      "loss": 0.0254,
+      "step": 1401
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.1729734710689084,
+      "learning_rate": 2.2793779093841565e-06,
+      "loss": 0.0172,
+      "step": 1402
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.894340993070462,
+      "learning_rate": 2.2726676661484265e-06,
+      "loss": 0.0138,
+      "step": 1403
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.609965868879223,
+      "learning_rate": 2.2659644084065483e-06,
+      "loss": 0.0214,
+      "step": 1404
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.4737416535129027,
+      "learning_rate": 2.2592681533275223e-06,
+      "loss": 0.0187,
+      "step": 1405
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.0187050959181063,
+      "learning_rate": 2.2525789180624096e-06,
+      "loss": 0.0157,
+      "step": 1406
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.9494581574836511,
+      "learning_rate": 2.2458967197442942e-06,
+      "loss": 0.0264,
+      "step": 1407
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7222904042380601,
+      "learning_rate": 2.2392215754882363e-06,
+      "loss": 0.0316,
+      "step": 1408
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5612517935965755,
+      "learning_rate": 2.232553502391227e-06,
+      "loss": 0.0208,
+      "step": 1409
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.564448457114273,
+      "learning_rate": 2.2258925175321476e-06,
+      "loss": 0.0358,
+      "step": 1410
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.9895281557974522,
+      "learning_rate": 2.2192386379717214e-06,
+      "loss": 0.0322,
+      "step": 1411
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5522292721758334,
+      "learning_rate": 2.2125918807524795e-06,
+      "loss": 0.0238,
+      "step": 1412
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0731253951864272,
+      "learning_rate": 2.205952262898704e-06,
+      "loss": 0.0333,
+      "step": 1413
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.48693201879768505,
+      "learning_rate": 2.199319801416394e-06,
+      "loss": 0.0185,
+      "step": 1414
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8312161420257296,
+      "learning_rate": 2.1926945132932188e-06,
+      "loss": 0.0296,
+      "step": 1415
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6626417816503829,
+      "learning_rate": 2.1860764154984736e-06,
+      "loss": 0.0207,
+      "step": 1416
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6049530155376529,
+      "learning_rate": 2.179465524983036e-06,
+      "loss": 0.0172,
+      "step": 1417
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8572149751190988,
+      "learning_rate": 2.1728618586793292e-06,
+      "loss": 0.0263,
+      "step": 1418
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.553519448666031,
+      "learning_rate": 2.1662654335012684e-06,
+      "loss": 0.0133,
+      "step": 1419
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7689498757925647,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.0232,
+      "step": 1420
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5344206051884095,
+      "learning_rate": 2.1530943740849696e-06,
+      "loss": 0.0204,
+      "step": 1421
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7641828859197944,
+      "learning_rate": 2.1465197735816585e-06,
+      "loss": 0.0246,
+      "step": 1422
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5707472485620008,
+      "learning_rate": 2.139952481673756e-06,
+      "loss": 0.0209,
+      "step": 1423
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5335231672331546,
+      "learning_rate": 2.1333925151820175e-06,
+      "loss": 0.0184,
+      "step": 1424
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.37509628954758883,
+      "learning_rate": 2.126839890908428e-06,
+      "loss": 0.0132,
+      "step": 1425
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.4228335969477914,
+      "learning_rate": 2.120294625636171e-06,
+      "loss": 0.0198,
+      "step": 1426
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5467505342742434,
+      "learning_rate": 2.113756736129581e-06,
+      "loss": 0.0232,
+      "step": 1427
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.6241443244997975,
+      "learning_rate": 2.1072262391340996e-06,
+      "loss": 0.0226,
+      "step": 1428
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.914095950785102,
+      "learning_rate": 2.100703151376234e-06,
+      "loss": 0.037,
+      "step": 1429
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5764144696089817,
+      "learning_rate": 2.0941874895635184e-06,
+      "loss": 0.0216,
+      "step": 1430
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5566137561193788,
+      "learning_rate": 2.087679270384461e-06,
+      "loss": 0.0139,
+      "step": 1431
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9897236771294801,
+      "learning_rate": 2.081178510508512e-06,
+      "loss": 0.0279,
+      "step": 1432
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9771936407490407,
+      "learning_rate": 2.0746852265860123e-06,
+      "loss": 0.0223,
+      "step": 1433
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.291363849454442,
+      "learning_rate": 2.0681994352481575e-06,
+      "loss": 0.015,
+      "step": 1434
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.3268806800493107,
+      "learning_rate": 2.0617211531069503e-06,
+      "loss": 0.0294,
+      "step": 1435
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7891869090943812,
+      "learning_rate": 2.0552503967551634e-06,
+      "loss": 0.0247,
+      "step": 1436
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.0008723260609602,
+      "learning_rate": 2.0487871827662913e-06,
+      "loss": 0.0326,
+      "step": 1437
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.43601598237058325,
+      "learning_rate": 2.0423315276945103e-06,
+      "loss": 0.02,
+      "step": 1438
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6944295941030163,
+      "learning_rate": 2.0358834480746363e-06,
+      "loss": 0.0269,
+      "step": 1439
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6707427363028934,
+      "learning_rate": 2.0294429604220833e-06,
+      "loss": 0.0203,
+      "step": 1440
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.685925821741964,
+      "learning_rate": 2.0230100812328167e-06,
+      "loss": 0.023,
+      "step": 1441
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5147866054111775,
+      "learning_rate": 2.0165848269833215e-06,
+      "loss": 0.0203,
+      "step": 1442
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.776632642956283,
+      "learning_rate": 2.010167214130546e-06,
+      "loss": 0.0246,
+      "step": 1443
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.0411426767089536,
+      "learning_rate": 2.003757259111869e-06,
+      "loss": 0.0282,
+      "step": 1444
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.957824081037812,
+      "learning_rate": 1.9973549783450563e-06,
+      "loss": 0.0188,
+      "step": 1445
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.134533925904978,
+      "learning_rate": 1.9909603882282156e-06,
+      "loss": 0.023,
+      "step": 1446
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7392234352149661,
+      "learning_rate": 1.984573505139758e-06,
+      "loss": 0.0185,
+      "step": 1447
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8419016776814038,
+      "learning_rate": 1.9781943454383583e-06,
+      "loss": 0.0292,
+      "step": 1448
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5482575841766507,
+      "learning_rate": 1.9718229254629018e-06,
+      "loss": 0.0201,
+      "step": 1449
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7895825808166198,
+      "learning_rate": 1.9654592615324542e-06,
+      "loss": 0.0261,
+      "step": 1450
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5391985624677404,
+      "learning_rate": 1.9591033699462165e-06,
+      "loss": 0.0166,
+      "step": 1451
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.0509339173777723,
+      "learning_rate": 1.9527552669834797e-06,
+      "loss": 0.0317,
+      "step": 1452
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6975292110442908,
+      "learning_rate": 1.9464149689035912e-06,
+      "loss": 0.0259,
+      "step": 1453
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6687890100923896,
+      "learning_rate": 1.940082491945902e-06,
+      "loss": 0.0177,
+      "step": 1454
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.086040887866882,
+      "learning_rate": 1.933757852329734e-06,
+      "loss": 0.0285,
+      "step": 1455
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6219525438540332,
+      "learning_rate": 1.927441066254334e-06,
+      "loss": 0.0237,
+      "step": 1456
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7987299511140524,
+      "learning_rate": 1.921132149898834e-06,
+      "loss": 0.0152,
+      "step": 1457
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8747595635173462,
+      "learning_rate": 1.9148311194222084e-06,
+      "loss": 0.0267,
+      "step": 1458
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.1033645254183688,
+      "learning_rate": 1.908537990963238e-06,
+      "loss": 0.0252,
+      "step": 1459
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.1394276850699414,
+      "learning_rate": 1.9022527806404583e-06,
+      "loss": 0.0399,
+      "step": 1460
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.443076629359654,
+      "learning_rate": 1.8959755045521283e-06,
+      "loss": 0.0169,
+      "step": 1461
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6779496548512952,
+      "learning_rate": 1.8897061787761823e-06,
+      "loss": 0.0272,
+      "step": 1462
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.9373551720667195,
+      "learning_rate": 1.883444819370193e-06,
+      "loss": 0.0384,
+      "step": 1463
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.645241163161573,
+      "learning_rate": 1.8771914423713277e-06,
+      "loss": 0.0186,
+      "step": 1464
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.4385818493699251,
+      "learning_rate": 1.8709460637963123e-06,
+      "loss": 0.0137,
+      "step": 1465
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.0003881903051037,
+      "learning_rate": 1.8647086996413816e-06,
+      "loss": 0.0257,
+      "step": 1466
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5380060560641039,
+      "learning_rate": 1.8584793658822452e-06,
+      "loss": 0.0245,
+      "step": 1467
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5642531276125784,
+      "learning_rate": 1.8522580784740452e-06,
+      "loss": 0.0202,
+      "step": 1468
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5922481529196789,
+      "learning_rate": 1.8460448533513126e-06,
+      "loss": 0.0236,
+      "step": 1469
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3489071504947676,
+      "learning_rate": 1.8398397064279282e-06,
+      "loss": 0.0146,
+      "step": 1470
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.46384690248928895,
+      "learning_rate": 1.833642653597088e-06,
+      "loss": 0.0188,
+      "step": 1471
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.9116631997505819,
+      "learning_rate": 1.8274537107312502e-06,
+      "loss": 0.0182,
+      "step": 1472
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.40347487957056644,
+      "learning_rate": 1.8212728936821034e-06,
+      "loss": 0.0099,
+      "step": 1473
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.9256491401658001,
+      "learning_rate": 1.8151002182805226e-06,
+      "loss": 0.0196,
+      "step": 1474
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.378903999596482,
+      "learning_rate": 1.8089357003365316e-06,
+      "loss": 0.0243,
+      "step": 1475
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.026642187164877,
+      "learning_rate": 1.802779355639257e-06,
+      "loss": 0.0085,
+      "step": 1476
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8339934893740829,
+      "learning_rate": 1.7966311999568974e-06,
+      "loss": 0.0338,
+      "step": 1477
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.1253234493067137,
+      "learning_rate": 1.7904912490366723e-06,
+      "loss": 0.0295,
+      "step": 1478
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.4327414591473575,
+      "learning_rate": 1.7843595186047863e-06,
+      "loss": 0.0271,
+      "step": 1479
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.671546053350665,
+      "learning_rate": 1.7782360243663905e-06,
+      "loss": 0.0206,
+      "step": 1480
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7710182655596616,
+      "learning_rate": 1.7721207820055402e-06,
+      "loss": 0.0232,
+      "step": 1481
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 25.15180406399679,
+      "learning_rate": 1.7660138071851545e-06,
+      "loss": 0.0232,
+      "step": 1482
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8164567044306239,
+      "learning_rate": 1.7599151155469802e-06,
+      "loss": 0.0243,
+      "step": 1483
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6379600019853553,
+      "learning_rate": 1.753824722711544e-06,
+      "loss": 0.02,
+      "step": 1484
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8499432372174188,
+      "learning_rate": 1.7477426442781198e-06,
+      "loss": 0.0261,
+      "step": 1485
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6261917549682076,
+      "learning_rate": 1.7416688958246847e-06,
+      "loss": 0.0193,
+      "step": 1486
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5082059473411364,
+      "learning_rate": 1.7356034929078803e-06,
+      "loss": 0.0111,
+      "step": 1487
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.4353187166263084,
+      "learning_rate": 1.7295464510629718e-06,
+      "loss": 0.0193,
+      "step": 1488
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.5191009047621435,
+      "learning_rate": 1.7234977858038137e-06,
+      "loss": 0.0315,
+      "step": 1489
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6488737261899189,
+      "learning_rate": 1.7174575126228016e-06,
+      "loss": 0.0227,
+      "step": 1490
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9264160824660079,
+      "learning_rate": 1.711425646990838e-06,
+      "loss": 0.0254,
+      "step": 1491
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.4176930071798314,
+      "learning_rate": 1.7054022043572882e-06,
+      "loss": 0.0153,
+      "step": 1492
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3590710184553117,
+      "learning_rate": 1.6993872001499461e-06,
+      "loss": 0.0193,
+      "step": 1493
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.9707860340678295,
+      "learning_rate": 1.6933806497749955e-06,
+      "loss": 0.023,
+      "step": 1494
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.4370541782227388,
+      "learning_rate": 1.6873825686169625e-06,
+      "loss": 0.0171,
+      "step": 1495
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6244044034847467,
+      "learning_rate": 1.681392972038684e-06,
+      "loss": 0.0192,
+      "step": 1496
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8926815180721247,
+      "learning_rate": 1.6754118753812632e-06,
+      "loss": 0.0242,
+      "step": 1497
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7593507350335365,
+      "learning_rate": 1.669439293964033e-06,
+      "loss": 0.0262,
+      "step": 1498
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5578039243484335,
+      "learning_rate": 1.6634752430845196e-06,
+      "loss": 0.0297,
+      "step": 1499
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.584385071528929,
+      "learning_rate": 1.6575197380183965e-06,
+      "loss": 0.0143,
+      "step": 1500
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.56402011021067,
+      "learning_rate": 1.6515727940194493e-06,
+      "loss": 0.0195,
+      "step": 1501
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.44268730746617535,
+      "learning_rate": 1.6456344263195374e-06,
+      "loss": 0.013,
+      "step": 1502
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.4961967681210577,
+      "learning_rate": 1.639704650128552e-06,
+      "loss": 0.0193,
+      "step": 1503
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.970299813916526,
+      "learning_rate": 1.6337834806343783e-06,
+      "loss": 0.019,
+      "step": 1504
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6257331643439866,
+      "learning_rate": 1.6278709330028636e-06,
+      "loss": 0.0206,
+      "step": 1505
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.9686357872123381,
+      "learning_rate": 1.621967022377765e-06,
+      "loss": 0.0284,
+      "step": 1506
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.274058130322033,
+      "learning_rate": 1.6160717638807205e-06,
+      "loss": 0.0226,
+      "step": 1507
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7747602086226845,
+      "learning_rate": 1.6101851726112067e-06,
+      "loss": 0.0116,
+      "step": 1508
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8105945738690109,
+      "learning_rate": 1.6043072636465017e-06,
+      "loss": 0.0217,
+      "step": 1509
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.6793023565390253,
+      "learning_rate": 1.598438052041643e-06,
+      "loss": 0.0285,
+      "step": 1510
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5613810363225675,
+      "learning_rate": 1.5925775528293985e-06,
+      "loss": 0.025,
+      "step": 1511
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.688507515991181,
+      "learning_rate": 1.586725781020214e-06,
+      "loss": 0.0186,
+      "step": 1512
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.3351929076892877,
+      "learning_rate": 1.5808827516021851e-06,
+      "loss": 0.0199,
+      "step": 1513
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.1889487189572696,
+      "learning_rate": 1.5750484795410143e-06,
+      "loss": 0.0307,
+      "step": 1514
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.48865042364173655,
+      "learning_rate": 1.5692229797799752e-06,
+      "loss": 0.0215,
+      "step": 1515
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8347701023901559,
+      "learning_rate": 1.5634062672398697e-06,
+      "loss": 0.0225,
+      "step": 1516
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.800077144695363,
+      "learning_rate": 1.557598356819e-06,
+      "loss": 0.0303,
+      "step": 1517
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3919408000387406,
+      "learning_rate": 1.5517992633931178e-06,
+      "loss": 0.014,
+      "step": 1518
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6177121499869433,
+      "learning_rate": 1.5460090018153923e-06,
+      "loss": 0.0227,
+      "step": 1519
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6349338552380158,
+      "learning_rate": 1.5402275869163736e-06,
+      "loss": 0.0283,
+      "step": 1520
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5126033869999117,
+      "learning_rate": 1.5344550335039521e-06,
+      "loss": 0.0155,
+      "step": 1521
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.856223292939775,
+      "learning_rate": 1.5286913563633194e-06,
+      "loss": 0.017,
+      "step": 1522
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0083250205891625,
+      "learning_rate": 1.522936570256937e-06,
+      "loss": 0.0217,
+      "step": 1523
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.4994108438051356,
+      "learning_rate": 1.517190689924491e-06,
+      "loss": 0.0184,
+      "step": 1524
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.37321497836596046,
+      "learning_rate": 1.5114537300828558e-06,
+      "loss": 0.0173,
+      "step": 1525
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7701752941065383,
+      "learning_rate": 1.505725705426061e-06,
+      "loss": 0.0244,
+      "step": 1526
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1281368029828174,
+      "learning_rate": 1.5000066306252475e-06,
+      "loss": 0.0166,
+      "step": 1527
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7338424153324811,
+      "learning_rate": 1.4942965203286337e-06,
+      "loss": 0.0177,
+      "step": 1528
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.41491260429600013,
+      "learning_rate": 1.4885953891614813e-06,
+      "loss": 0.0153,
+      "step": 1529
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7844248268278589,
+      "learning_rate": 1.482903251726049e-06,
+      "loss": 0.0305,
+      "step": 1530
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8885898820999001,
+      "learning_rate": 1.4772201226015615e-06,
+      "loss": 0.0309,
+      "step": 1531
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.0705376041540378,
+      "learning_rate": 1.4715460163441703e-06,
+      "loss": 0.0492,
+      "step": 1532
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.60630026470556,
+      "learning_rate": 1.4658809474869174e-06,
+      "loss": 0.0163,
+      "step": 1533
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7257005504788903,
+      "learning_rate": 1.4602249305396966e-06,
+      "loss": 0.0146,
+      "step": 1534
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.2429270738071347,
+      "learning_rate": 1.4545779799892179e-06,
+      "loss": 0.0298,
+      "step": 1535
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5803068356342167,
+      "learning_rate": 1.4489401102989698e-06,
+      "loss": 0.0143,
+      "step": 1536
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9334893520475549,
+      "learning_rate": 1.4433113359091805e-06,
+      "loss": 0.029,
+      "step": 1537
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8061341751112961,
+      "learning_rate": 1.4376916712367838e-06,
+      "loss": 0.0186,
+      "step": 1538
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9039562950084497,
+      "learning_rate": 1.4320811306753796e-06,
+      "loss": 0.0212,
+      "step": 1539
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.4714931876115984,
+      "learning_rate": 1.426479728595202e-06,
+      "loss": 0.02,
+      "step": 1540
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7906593826204635,
+      "learning_rate": 1.4208874793430738e-06,
+      "loss": 0.0221,
+      "step": 1541
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7280596395913382,
+      "learning_rate": 1.415304397242378e-06,
+      "loss": 0.0207,
+      "step": 1542
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8117288716060429,
+      "learning_rate": 1.409730496593016e-06,
+      "loss": 0.0312,
+      "step": 1543
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7925630934679432,
+      "learning_rate": 1.4041657916713741e-06,
+      "loss": 0.0181,
+      "step": 1544
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.3841331652094588,
+      "learning_rate": 1.398610296730284e-06,
+      "loss": 0.0262,
+      "step": 1545
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.620345714196815,
+      "learning_rate": 1.3930640259989914e-06,
+      "loss": 0.0215,
+      "step": 1546
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.72655687412436,
+      "learning_rate": 1.3875269936831133e-06,
+      "loss": 0.0246,
+      "step": 1547
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.6574460872584007,
+      "learning_rate": 1.3819992139646037e-06,
+      "loss": 0.0222,
+      "step": 1548
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7069930906531742,
+      "learning_rate": 1.37648070100172e-06,
+      "loss": 0.0266,
+      "step": 1549
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6845802946247612,
+      "learning_rate": 1.3709714689289844e-06,
+      "loss": 0.0196,
+      "step": 1550
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6832692791063089,
+      "learning_rate": 1.3654715318571455e-06,
+      "loss": 0.0161,
+      "step": 1551
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8309536486895222,
+      "learning_rate": 1.3599809038731503e-06,
+      "loss": 0.026,
+      "step": 1552
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.861610800620917,
+      "learning_rate": 1.354499599040097e-06,
+      "loss": 0.0218,
+      "step": 1553
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.793696990792352,
+      "learning_rate": 1.3490276313972073e-06,
+      "loss": 0.0232,
+      "step": 1554
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6407327606080412,
+      "learning_rate": 1.3435650149597862e-06,
+      "loss": 0.0265,
+      "step": 1555
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8759165356024202,
+      "learning_rate": 1.3381117637191887e-06,
+      "loss": 0.0225,
+      "step": 1556
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.0577346876127711,
+      "learning_rate": 1.3326678916427803e-06,
+      "loss": 0.0298,
+      "step": 1557
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.4542805258478709,
+      "learning_rate": 1.3272334126739094e-06,
+      "loss": 0.0174,
+      "step": 1558
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5182354308636529,
+      "learning_rate": 1.3218083407318606e-06,
+      "loss": 0.0179,
+      "step": 1559
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.3954805565498662,
+      "learning_rate": 1.3163926897118252e-06,
+      "loss": 0.0192,
+      "step": 1560
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.811580137556446,
+      "learning_rate": 1.310986473484866e-06,
+      "loss": 0.0182,
+      "step": 1561
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.1921238444026916,
+      "learning_rate": 1.3055897058978806e-06,
+      "loss": 0.0358,
+      "step": 1562
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.0037836419531394,
+      "learning_rate": 1.300202400773563e-06,
+      "loss": 0.0114,
+      "step": 1563
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.9513953738579717,
+      "learning_rate": 1.2948245719103774e-06,
+      "loss": 0.0268,
+      "step": 1564
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.2881756130443442,
+      "learning_rate": 1.2894562330825106e-06,
+      "loss": 0.022,
+      "step": 1565
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.9456261923512207,
+      "learning_rate": 1.2840973980398446e-06,
+      "loss": 0.0251,
+      "step": 1566
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7087238217511729,
+      "learning_rate": 1.27874808050792e-06,
+      "loss": 0.0205,
+      "step": 1567
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8087893817399272,
+      "learning_rate": 1.273408294187899e-06,
+      "loss": 0.0249,
+      "step": 1568
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.2547387247611936,
+      "learning_rate": 1.2680780527565313e-06,
+      "loss": 0.024,
+      "step": 1569
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.0368957302101323,
+      "learning_rate": 1.2627573698661228e-06,
+      "loss": 0.0374,
+      "step": 1570
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.5301292447138974,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.0243,
+      "step": 1571
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.46242128257381343,
+      "learning_rate": 1.2521447341949494e-06,
+      "loss": 0.0174,
+      "step": 1572
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.0290998698478622,
+      "learning_rate": 1.24685280859624e-06,
+      "loss": 0.0329,
+      "step": 1573
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7094916079841542,
+      "learning_rate": 1.2415704959025321e-06,
+      "loss": 0.0197,
+      "step": 1574
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.0464668946048912,
+      "learning_rate": 1.236297809643368e-06,
+      "loss": 0.0259,
+      "step": 1575
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.2412781545389158,
+      "learning_rate": 1.2310347633236402e-06,
+      "loss": 0.0296,
+      "step": 1576
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.3720843586226656,
+      "learning_rate": 1.225781370423541e-06,
+      "loss": 0.0398,
+      "step": 1577
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.011296703595368,
+      "learning_rate": 1.2205376443985444e-06,
+      "loss": 0.0193,
+      "step": 1578
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.5876467993872176,
+      "learning_rate": 1.2153035986793627e-06,
+      "loss": 0.0331,
+      "step": 1579
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5603223195065772,
+      "learning_rate": 1.2100792466719118e-06,
+      "loss": 0.0208,
+      "step": 1580
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6448719403345857,
+      "learning_rate": 1.2048646017572857e-06,
+      "loss": 0.0222,
+      "step": 1581
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.28342050570494437,
+      "learning_rate": 1.1996596772917091e-06,
+      "loss": 0.0092,
+      "step": 1582
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.39879018942234634,
+      "learning_rate": 1.1944644866065125e-06,
+      "loss": 0.0176,
+      "step": 1583
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.42845470718418127,
+      "learning_rate": 1.1892790430080952e-06,
+      "loss": 0.0163,
+      "step": 1584
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.504022504721399,
+      "learning_rate": 1.1841033597778905e-06,
+      "loss": 0.0155,
+      "step": 1585
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8398692539657486,
+      "learning_rate": 1.1789374501723328e-06,
+      "loss": 0.0227,
+      "step": 1586
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0945442042275597,
+      "learning_rate": 1.173781327422826e-06,
+      "loss": 0.0203,
+      "step": 1587
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7836709704029713,
+      "learning_rate": 1.1686350047357036e-06,
+      "loss": 0.0323,
+      "step": 1588
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6375367175543545,
+      "learning_rate": 1.1634984952921996e-06,
+      "loss": 0.0196,
+      "step": 1589
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7287547471933316,
+      "learning_rate": 1.1583718122484128e-06,
+      "loss": 0.0208,
+      "step": 1590
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7912568103894504,
+      "learning_rate": 1.1532549687352752e-06,
+      "loss": 0.021,
+      "step": 1591
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5528882367856399,
+      "learning_rate": 1.1481479778585142e-06,
+      "loss": 0.0193,
+      "step": 1592
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8246650683849523,
+      "learning_rate": 1.1430508526986261e-06,
+      "loss": 0.0216,
+      "step": 1593
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.3077247280046171,
+      "learning_rate": 1.137963606310834e-06,
+      "loss": 0.0232,
+      "step": 1594
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.4102947057515644,
+      "learning_rate": 1.132886251725061e-06,
+      "loss": 0.0227,
+      "step": 1595
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.6409286483577383,
+      "learning_rate": 1.127818801945893e-06,
+      "loss": 0.0209,
+      "step": 1596
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6933263716604913,
+      "learning_rate": 1.1227612699525475e-06,
+      "loss": 0.0255,
+      "step": 1597
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6940363977524191,
+      "learning_rate": 1.1177136686988383e-06,
+      "loss": 0.0205,
+      "step": 1598
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7791018362615008,
+      "learning_rate": 1.1126760111131474e-06,
+      "loss": 0.0186,
+      "step": 1599
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.473217425311028,
+      "learning_rate": 1.1076483100983843e-06,
+      "loss": 0.0193,
+      "step": 1600
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.859716563176724,
+      "learning_rate": 1.1026305785319585e-06,
+      "loss": 0.0276,
+      "step": 1601
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.547744977616838,
+      "learning_rate": 1.0976228292657447e-06,
+      "loss": 0.0231,
+      "step": 1602
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7192227614672417,
+      "learning_rate": 1.0926250751260492e-06,
+      "loss": 0.0217,
+      "step": 1603
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.4437076760782891,
+      "learning_rate": 1.0876373289135778e-06,
+      "loss": 0.0214,
+      "step": 1604
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6198481679854476,
+      "learning_rate": 1.0826596034034066e-06,
+      "loss": 0.0274,
+      "step": 1605
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6905795095781041,
+      "learning_rate": 1.0776919113449407e-06,
+      "loss": 0.0253,
+      "step": 1606
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.9346647468986687,
+      "learning_rate": 1.0727342654618905e-06,
+      "loss": 0.0297,
+      "step": 1607
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6448428595540737,
+      "learning_rate": 1.0677866784522317e-06,
+      "loss": 0.0211,
+      "step": 1608
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7632216839916288,
+      "learning_rate": 1.0628491629881794e-06,
+      "loss": 0.0285,
+      "step": 1609
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.043944368613153,
+      "learning_rate": 1.0579217317161494e-06,
+      "loss": 0.0236,
+      "step": 1610
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6420776815388155,
+      "learning_rate": 1.0530043972567339e-06,
+      "loss": 0.0286,
+      "step": 1611
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.36874244760640196,
+      "learning_rate": 1.0480971722046602e-06,
+      "loss": 0.0119,
+      "step": 1612
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.698553180562803,
+      "learning_rate": 1.0432000691287619e-06,
+      "loss": 0.0232,
+      "step": 1613
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.4035736417059915,
+      "learning_rate": 1.0383131005719505e-06,
+      "loss": 0.0126,
+      "step": 1614
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.516220342713733,
+      "learning_rate": 1.0334362790511776e-06,
+      "loss": 0.0182,
+      "step": 1615
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.39365931838262347,
+      "learning_rate": 1.0285696170574045e-06,
+      "loss": 0.0126,
+      "step": 1616
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5480992689342562,
+      "learning_rate": 1.023713127055575e-06,
+      "loss": 0.0179,
+      "step": 1617
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5196346398987512,
+      "learning_rate": 1.0188668214845765e-06,
+      "loss": 0.0155,
+      "step": 1618
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.4489352579236505,
+      "learning_rate": 1.0140307127572125e-06,
+      "loss": 0.0184,
+      "step": 1619
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.690103014170062,
+      "learning_rate": 1.009204813260164e-06,
+      "loss": 0.0202,
+      "step": 1620
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.3570709210532113,
+      "learning_rate": 1.004389135353972e-06,
+      "loss": 0.0365,
+      "step": 1621
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.0257681719369882,
+      "learning_rate": 9.995836913729918e-07,
+      "loss": 0.0237,
+      "step": 1622
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6168406563119275,
+      "learning_rate": 9.947884936253666e-07,
+      "loss": 0.0196,
+      "step": 1623
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7692543213723014,
+      "learning_rate": 9.90003554392997e-07,
+      "loss": 0.0232,
+      "step": 1624
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.2328193733615247,
+      "learning_rate": 9.852288859315096e-07,
+      "loss": 0.0431,
+      "step": 1625
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.4613027425912049,
+      "learning_rate": 9.804645004702208e-07,
+      "loss": 0.0248,
+      "step": 1626
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7094673929023848,
+      "learning_rate": 9.757104102121152e-07,
+      "loss": 0.0189,
+      "step": 1627
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7652661275955153,
+      "learning_rate": 9.709666273338037e-07,
+      "loss": 0.0227,
+      "step": 1628
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6515555057917235,
+      "learning_rate": 9.662331639854977e-07,
+      "loss": 0.0214,
+      "step": 1629
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.4895831342399959,
+      "learning_rate": 9.615100322909787e-07,
+      "loss": 0.018,
+      "step": 1630
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.609954132677896,
+      "learning_rate": 9.567972443475648e-07,
+      "loss": 0.0203,
+      "step": 1631
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5067204919211488,
+      "learning_rate": 9.520948122260792e-07,
+      "loss": 0.0214,
+      "step": 1632
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.2597206502873675,
+      "learning_rate": 9.474027479708254e-07,
+      "loss": 0.0206,
+      "step": 1633
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3841045305323135,
+      "learning_rate": 9.427210635995482e-07,
+      "loss": 0.0152,
+      "step": 1634
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5821817358914164,
+      "learning_rate": 9.380497711034065e-07,
+      "loss": 0.0203,
+      "step": 1635
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6100463174092755,
+      "learning_rate": 9.33388882446944e-07,
+      "loss": 0.0184,
+      "step": 1636
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9490671043170982,
+      "learning_rate": 9.287384095680558e-07,
+      "loss": 0.0257,
+      "step": 1637
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.41866063189682606,
+      "learning_rate": 9.240983643779589e-07,
+      "loss": 0.0099,
+      "step": 1638
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.916192123980324,
+      "learning_rate": 9.19468758761165e-07,
+      "loss": 0.0196,
+      "step": 1639
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6291804775621946,
+      "learning_rate": 9.148496045754441e-07,
+      "loss": 0.0219,
+      "step": 1640
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.46506938594026603,
+      "learning_rate": 9.10240913651797e-07,
+      "loss": 0.0209,
+      "step": 1641
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3763574887343969,
+      "learning_rate": 9.056426977944272e-07,
+      "loss": 0.0175,
+      "step": 1642
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6263963200007582,
+      "learning_rate": 9.010549687807058e-07,
+      "loss": 0.026,
+      "step": 1643
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8782893052468342,
+      "learning_rate": 8.964777383611445e-07,
+      "loss": 0.0147,
+      "step": 1644
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8499726766140078,
+      "learning_rate": 8.919110182593682e-07,
+      "loss": 0.0307,
+      "step": 1645
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.870144300950776,
+      "learning_rate": 8.873548201720788e-07,
+      "loss": 0.0259,
+      "step": 1646
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7272528604695497,
+      "learning_rate": 8.828091557690288e-07,
+      "loss": 0.0296,
+      "step": 1647
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.676027817512814,
+      "learning_rate": 8.78274036692991e-07,
+      "loss": 0.0284,
+      "step": 1648
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.856038483122214,
+      "learning_rate": 8.73749474559728e-07,
+      "loss": 0.0297,
+      "step": 1649
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.399332352098762,
+      "learning_rate": 8.692354809579634e-07,
+      "loss": 0.0443,
+      "step": 1650
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8513838144339223,
+      "learning_rate": 8.647320674493531e-07,
+      "loss": 0.0233,
+      "step": 1651
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.2022612364336753,
+      "learning_rate": 8.602392455684522e-07,
+      "loss": 0.0288,
+      "step": 1652
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6318734527708079,
+      "learning_rate": 8.557570268226889e-07,
+      "loss": 0.0227,
+      "step": 1653
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.4986624707680987,
+      "learning_rate": 8.512854226923328e-07,
+      "loss": 0.012,
+      "step": 1654
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.9490939870975297,
+      "learning_rate": 8.468244446304664e-07,
+      "loss": 0.0232,
+      "step": 1655
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5687247694401628,
+      "learning_rate": 8.423741040629557e-07,
+      "loss": 0.0171,
+      "step": 1656
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.9342752487631834,
+      "learning_rate": 8.379344123884231e-07,
+      "loss": 0.0256,
+      "step": 1657
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.4259907942214256,
+      "learning_rate": 8.335053809782129e-07,
+      "loss": 0.0208,
+      "step": 1658
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.9361781880015428,
+      "learning_rate": 8.29087021176368e-07,
+      "loss": 0.0143,
+      "step": 1659
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8441919046857431,
+      "learning_rate": 8.246793442995954e-07,
+      "loss": 0.0167,
+      "step": 1660
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6969608754714764,
+      "learning_rate": 8.20282361637243e-07,
+      "loss": 0.0171,
+      "step": 1661
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8202955595934445,
+      "learning_rate": 8.158960844512654e-07,
+      "loss": 0.0195,
+      "step": 1662
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.3779435824059567,
+      "learning_rate": 8.115205239761986e-07,
+      "loss": 0.029,
+      "step": 1663
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.1595061577911956,
+      "learning_rate": 8.071556914191298e-07,
+      "loss": 0.0151,
+      "step": 1664
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3886150974844245,
+      "learning_rate": 8.028015979596681e-07,
+      "loss": 0.0128,
+      "step": 1665
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.715937707945813,
+      "learning_rate": 7.984582547499176e-07,
+      "loss": 0.0219,
+      "step": 1666
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8299951514667157,
+      "learning_rate": 7.941256729144464e-07,
+      "loss": 0.0256,
+      "step": 1667
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.101475048961094,
+      "learning_rate": 7.898038635502631e-07,
+      "loss": 0.0303,
+      "step": 1668
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.6129852664782461,
+      "learning_rate": 7.854928377267812e-07,
+      "loss": 0.023,
+      "step": 1669
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.654917809207751,
+      "learning_rate": 7.81192606485795e-07,
+      "loss": 0.0233,
+      "step": 1670
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7485343464875281,
+      "learning_rate": 7.769031808414523e-07,
+      "loss": 0.0197,
+      "step": 1671
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.4487323641792147,
+      "learning_rate": 7.726245717802233e-07,
+      "loss": 0.0375,
+      "step": 1672
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3217437634939568,
+      "learning_rate": 7.68356790260873e-07,
+      "loss": 0.0122,
+      "step": 1673
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.4868540639820885,
+      "learning_rate": 7.640998472144373e-07,
+      "loss": 0.0196,
+      "step": 1674
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8908143071660273,
+      "learning_rate": 7.59853753544188e-07,
+      "loss": 0.0222,
+      "step": 1675
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9843729139133108,
+      "learning_rate": 7.556185201256105e-07,
+      "loss": 0.0158,
+      "step": 1676
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8812079648965719,
+      "learning_rate": 7.513941578063732e-07,
+      "loss": 0.023,
+      "step": 1677
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.041052471945282,
+      "learning_rate": 7.471806774063e-07,
+      "loss": 0.0214,
+      "step": 1678
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.36991466567386605,
+      "learning_rate": 7.429780897173427e-07,
+      "loss": 0.015,
+      "step": 1679
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3146801134651237,
+      "learning_rate": 7.387864055035571e-07,
+      "loss": 0.013,
+      "step": 1680
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5791780869947971,
+      "learning_rate": 7.346056355010683e-07,
+      "loss": 0.0205,
+      "step": 1681
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.4843910836561715,
+      "learning_rate": 7.304357904180475e-07,
+      "loss": 0.0183,
+      "step": 1682
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9352327802951779,
+      "learning_rate": 7.262768809346849e-07,
+      "loss": 0.0207,
+      "step": 1683
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6355901282179665,
+      "learning_rate": 7.221289177031609e-07,
+      "loss": 0.0215,
+      "step": 1684
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.827734270635811,
+      "learning_rate": 7.179919113476192e-07,
+      "loss": 0.0574,
+      "step": 1685
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.26441919617238413,
+      "learning_rate": 7.138658724641417e-07,
+      "loss": 0.0091,
+      "step": 1686
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.9709198286425481,
+      "learning_rate": 7.097508116207169e-07,
+      "loss": 0.0261,
+      "step": 1687
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3166787822548412,
+      "learning_rate": 7.056467393572158e-07,
+      "loss": 0.0178,
+      "step": 1688
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.721525302234989,
+      "learning_rate": 7.015536661853656e-07,
+      "loss": 0.0241,
+      "step": 1689
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5768351297093026,
+      "learning_rate": 6.974716025887207e-07,
+      "loss": 0.0194,
+      "step": 1690
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3058055162169507,
+      "learning_rate": 6.934005590226361e-07,
+      "loss": 0.0137,
+      "step": 1691
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6765565581741332,
+      "learning_rate": 6.893405459142439e-07,
+      "loss": 0.02,
+      "step": 1692
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.46982186330771497,
+      "learning_rate": 6.85291573662421e-07,
+      "loss": 0.0157,
+      "step": 1693
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8793025805347607,
+      "learning_rate": 6.812536526377673e-07,
+      "loss": 0.0214,
+      "step": 1694
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6662828745104178,
+      "learning_rate": 6.772267931825765e-07,
+      "loss": 0.0226,
+      "step": 1695
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.491202037016575,
+      "learning_rate": 6.732110056108105e-07,
+      "loss": 0.02,
+      "step": 1696
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.676251852333797,
+      "learning_rate": 6.692063002080717e-07,
+      "loss": 0.0303,
+      "step": 1697
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6695814761840161,
+      "learning_rate": 6.652126872315812e-07,
+      "loss": 0.0236,
+      "step": 1698
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.33574986126203843,
+      "learning_rate": 6.612301769101464e-07,
+      "loss": 0.012,
+      "step": 1699
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.4367587595934483,
+      "learning_rate": 6.572587794441381e-07,
+      "loss": 0.0172,
+      "step": 1700
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.62403592781587,
+      "learning_rate": 6.532985050054635e-07,
+      "loss": 0.0208,
+      "step": 1701
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.0886268911836297,
+      "learning_rate": 6.493493637375414e-07,
+      "loss": 0.0266,
+      "step": 1702
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7851741637225625,
+      "learning_rate": 6.45411365755273e-07,
+      "loss": 0.0186,
+      "step": 1703
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.39687201481838075,
+      "learning_rate": 6.414845211450243e-07,
+      "loss": 0.0168,
+      "step": 1704
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.858159783352387,
+      "learning_rate": 6.375688399645863e-07,
+      "loss": 0.0226,
+      "step": 1705
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.569237008902663,
+      "learning_rate": 6.336643322431624e-07,
+      "loss": 0.0224,
+      "step": 1706
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.092683438598888,
+      "learning_rate": 6.297710079813346e-07,
+      "loss": 0.0296,
+      "step": 1707
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7351588601818313,
+      "learning_rate": 6.258888771510441e-07,
+      "loss": 0.0195,
+      "step": 1708
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6404335577191063,
+      "learning_rate": 6.220179496955608e-07,
+      "loss": 0.0295,
+      "step": 1709
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.170165515233352,
+      "learning_rate": 6.181582355294579e-07,
+      "loss": 0.0175,
+      "step": 1710
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.9322068050386567,
+      "learning_rate": 6.143097445385904e-07,
+      "loss": 0.0202,
+      "step": 1711
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.7698760246873417,
+      "learning_rate": 6.104724865800665e-07,
+      "loss": 0.0218,
+      "step": 1712
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7113173705764911,
+      "learning_rate": 6.066464714822224e-07,
+      "loss": 0.017,
+      "step": 1713
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.42364594714097703,
+      "learning_rate": 6.028317090446006e-07,
+      "loss": 0.0146,
+      "step": 1714
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.3595453398378823,
+      "learning_rate": 5.990282090379201e-07,
+      "loss": 0.0182,
+      "step": 1715
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6382144279458921,
+      "learning_rate": 5.952359812040548e-07,
+      "loss": 0.0266,
+      "step": 1716
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.2344671026355398,
+      "learning_rate": 5.91455035256005e-07,
+      "loss": 0.0266,
+      "step": 1717
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.44343690963147425,
+      "learning_rate": 5.876853808778782e-07,
+      "loss": 0.0164,
+      "step": 1718
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.3139772726167909,
+      "learning_rate": 5.839270277248565e-07,
+      "loss": 0.0297,
+      "step": 1719
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5404404381299035,
+      "learning_rate": 5.801799854231826e-07,
+      "loss": 0.0218,
+      "step": 1720
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9120055408801463,
+      "learning_rate": 5.764442635701229e-07,
+      "loss": 0.0263,
+      "step": 1721
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7776723247223076,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.0166,
+      "step": 1722
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.9105654312944207,
+      "learning_rate": 5.690068194539217e-07,
+      "loss": 0.0259,
+      "step": 1723
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5182676413757116,
+      "learning_rate": 5.653051162402445e-07,
+      "loss": 0.021,
+      "step": 1724
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9792005660232671,
+      "learning_rate": 5.616147715740611e-07,
+      "loss": 0.0161,
+      "step": 1725
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.37941653477897735,
+      "learning_rate": 5.579357949074221e-07,
+      "loss": 0.0119,
+      "step": 1726
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.1030201633413343,
+      "learning_rate": 5.542681956632601e-07,
+      "loss": 0.0253,
+      "step": 1727
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5322457268616182,
+      "learning_rate": 5.506119832353662e-07,
+      "loss": 0.0201,
+      "step": 1728
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.787967539652566,
+      "learning_rate": 5.469671669883675e-07,
+      "loss": 0.0146,
+      "step": 1729
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.3041562263159943,
+      "learning_rate": 5.433337562577018e-07,
+      "loss": 0.0244,
+      "step": 1730
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.33548978851264033,
+      "learning_rate": 5.397117603495927e-07,
+      "loss": 0.0163,
+      "step": 1731
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2796190944664051,
+      "learning_rate": 5.361011885410311e-07,
+      "loss": 0.0142,
+      "step": 1732
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.4321550120605657,
+      "learning_rate": 5.325020500797434e-07,
+      "loss": 0.0251,
+      "step": 1733
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.2378411684030717,
+      "learning_rate": 5.289143541841735e-07,
+      "loss": 0.0239,
+      "step": 1734
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6006956440056487,
+      "learning_rate": 5.253381100434574e-07,
+      "loss": 0.0171,
+      "step": 1735
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8446344159688172,
+      "learning_rate": 5.217733268173996e-07,
+      "loss": 0.0167,
+      "step": 1736
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3860698594231629,
+      "learning_rate": 5.182200136364491e-07,
+      "loss": 0.0116,
+      "step": 1737
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6743659201298501,
+      "learning_rate": 5.146781796016798e-07,
+      "loss": 0.0228,
+      "step": 1738
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6578276056168991,
+      "learning_rate": 5.111478337847603e-07,
+      "loss": 0.0145,
+      "step": 1739
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.4444967706503058,
+      "learning_rate": 5.076289852279375e-07,
+      "loss": 0.0199,
+      "step": 1740
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8905759201198322,
+      "learning_rate": 5.041216429440088e-07,
+      "loss": 0.0202,
+      "step": 1741
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8652719179398571,
+      "learning_rate": 5.006258159163007e-07,
+      "loss": 0.026,
+      "step": 1742
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.920438619981302,
+      "learning_rate": 4.971415130986457e-07,
+      "loss": 0.0256,
+      "step": 1743
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.4690467940337577,
+      "learning_rate": 4.936687434153619e-07,
+      "loss": 0.0139,
+      "step": 1744
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.0857169067280574,
+      "learning_rate": 4.902075157612241e-07,
+      "loss": 0.0268,
+      "step": 1745
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.9071954914058828,
+      "learning_rate": 4.867578390014466e-07,
+      "loss": 0.021,
+      "step": 1746
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.0027348201423516,
+      "learning_rate": 4.833197219716595e-07,
+      "loss": 0.0242,
+      "step": 1747
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7208333147985326,
+      "learning_rate": 4.798931734778794e-07,
+      "loss": 0.0225,
+      "step": 1748
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3340296826164514,
+      "learning_rate": 4.764782022965014e-07,
+      "loss": 0.014,
+      "step": 1749
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.41348222258487555,
+      "learning_rate": 4.730748171742611e-07,
+      "loss": 0.0234,
+      "step": 1750
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.783608172436412,
+      "learning_rate": 4.696830268282204e-07,
+      "loss": 0.0207,
+      "step": 1751
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5594126597283506,
+      "learning_rate": 4.6630283994574475e-07,
+      "loss": 0.0188,
+      "step": 1752
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.4130436506101321,
+      "learning_rate": 4.629342651844787e-07,
+      "loss": 0.0149,
+      "step": 1753
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.47117699815033703,
+      "learning_rate": 4.595773111723245e-07,
+      "loss": 0.0157,
+      "step": 1754
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.06468324325383,
+      "learning_rate": 4.562319865074222e-07,
+      "loss": 0.0218,
+      "step": 1755
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.4979626901475594,
+      "learning_rate": 4.528982997581233e-07,
+      "loss": 0.0201,
+      "step": 1756
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.988935035787326,
+      "learning_rate": 4.4957625946297266e-07,
+      "loss": 0.0178,
+      "step": 1757
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.637273291889344,
+      "learning_rate": 4.462658741306847e-07,
+      "loss": 0.016,
+      "step": 1758
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7542039279222552,
+      "learning_rate": 4.4296715224012187e-07,
+      "loss": 0.0169,
+      "step": 1759
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7968353487232488,
+      "learning_rate": 4.3968010224027247e-07,
+      "loss": 0.0147,
+      "step": 1760
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9266624597530224,
+      "learning_rate": 4.364047325502324e-07,
+      "loss": 0.0193,
+      "step": 1761
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.4242221780268279,
+      "learning_rate": 4.331410515591783e-07,
+      "loss": 0.0136,
+      "step": 1762
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.2405700148688243,
+      "learning_rate": 4.298890676263495e-07,
+      "loss": 0.0238,
+      "step": 1763
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.236346074001916,
+      "learning_rate": 4.2664878908102556e-07,
+      "loss": 0.0355,
+      "step": 1764
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8651345063698643,
+      "learning_rate": 4.2342022422250553e-07,
+      "loss": 0.0258,
+      "step": 1765
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3813222697994349,
+      "learning_rate": 4.2020338132008454e-07,
+      "loss": 0.0121,
+      "step": 1766
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7925799529634308,
+      "learning_rate": 4.1699826861303804e-07,
+      "loss": 0.0249,
+      "step": 1767
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9035223172963347,
+      "learning_rate": 4.138048943105938e-07,
+      "loss": 0.0237,
+      "step": 1768
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7287711613306799,
+      "learning_rate": 4.106232665919152e-07,
+      "loss": 0.0109,
+      "step": 1769
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6004627965162954,
+      "learning_rate": 4.0745339360607927e-07,
+      "loss": 0.0136,
+      "step": 1770
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.5499328026562983,
+      "learning_rate": 4.042952834720548e-07,
+      "loss": 0.0203,
+      "step": 1771
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.44565309449455925,
+      "learning_rate": 4.011489442786831e-07,
+      "loss": 0.0147,
+      "step": 1772
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3121101306415952,
+      "learning_rate": 3.9801438408465895e-07,
+      "loss": 0.0136,
+      "step": 1773
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9336561135203006,
+      "learning_rate": 3.9489161091850413e-07,
+      "loss": 0.0228,
+      "step": 1774
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.1439685088962397,
+      "learning_rate": 3.917806327785517e-07,
+      "loss": 0.0194,
+      "step": 1775
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7667463909847458,
+      "learning_rate": 3.886814576329245e-07,
+      "loss": 0.0343,
+      "step": 1776
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.39481524338438234,
+      "learning_rate": 3.855940934195146e-07,
+      "loss": 0.0182,
+      "step": 1777
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.4682956312455625,
+      "learning_rate": 3.825185480459614e-07,
+      "loss": 0.0142,
+      "step": 1778
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.49595668042956637,
+      "learning_rate": 3.794548293896355e-07,
+      "loss": 0.0132,
+      "step": 1779
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.21011637441168096,
+      "learning_rate": 3.7640294529761424e-07,
+      "loss": 0.0081,
+      "step": 1780
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.79580408555057,
+      "learning_rate": 3.7336290358666206e-07,
+      "loss": 0.0291,
+      "step": 1781
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.4241950547346609,
+      "learning_rate": 3.703347120432138e-07,
+      "loss": 0.0164,
+      "step": 1782
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.5231212609907914,
+      "learning_rate": 3.6731837842335085e-07,
+      "loss": 0.0233,
+      "step": 1783
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.9115936016118895,
+      "learning_rate": 3.6431391045278263e-07,
+      "loss": 0.028,
+      "step": 1784
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7705099493152656,
+      "learning_rate": 3.6132131582683086e-07,
+      "loss": 0.0178,
+      "step": 1785
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7642604555944872,
+      "learning_rate": 3.583406022104019e-07,
+      "loss": 0.0203,
+      "step": 1786
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3350098895459638,
+      "learning_rate": 3.5537177723797335e-07,
+      "loss": 0.0106,
+      "step": 1787
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6748864749922056,
+      "learning_rate": 3.52414848513572e-07,
+      "loss": 0.0232,
+      "step": 1788
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7794439927916365,
+      "learning_rate": 3.4946982361075524e-07,
+      "loss": 0.0146,
+      "step": 1789
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.0614980884129646,
+      "learning_rate": 3.4653671007259084e-07,
+      "loss": 0.0259,
+      "step": 1790
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.4421989854447395,
+      "learning_rate": 3.436155154116383e-07,
+      "loss": 0.0122,
+      "step": 1791
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9290405002699073,
+      "learning_rate": 3.407062471099298e-07,
+      "loss": 0.021,
+      "step": 1792
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.1172006958050862,
+      "learning_rate": 3.3780891261895043e-07,
+      "loss": 0.0281,
+      "step": 1793
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.227579498950002,
+      "learning_rate": 3.349235193596184e-07,
+      "loss": 0.0226,
+      "step": 1794
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7512967509133619,
+      "learning_rate": 3.320500747222677e-07,
+      "loss": 0.0172,
+      "step": 1795
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.4290770474862749,
+      "learning_rate": 3.291885860666294e-07,
+      "loss": 0.0173,
+      "step": 1796
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.250875314305571,
+      "learning_rate": 3.263390607218103e-07,
+      "loss": 0.025,
+      "step": 1797
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.944211346524395,
+      "learning_rate": 3.235015059862767e-07,
+      "loss": 0.0241,
+      "step": 1798
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.4028666512919997,
+      "learning_rate": 3.206759291278333e-07,
+      "loss": 0.0133,
+      "step": 1799
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8746495137789382,
+      "learning_rate": 3.178623373836076e-07,
+      "loss": 0.0209,
+      "step": 1800
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.915974115336977,
+      "learning_rate": 3.1506073796002734e-07,
+      "loss": 0.0196,
+      "step": 1801
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.713113489833011,
+      "learning_rate": 3.1227113803280863e-07,
+      "loss": 0.0423,
+      "step": 1802
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.5203902674476263,
+      "learning_rate": 3.0949354474692937e-07,
+      "loss": 0.0155,
+      "step": 1803
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.37043619740618705,
+      "learning_rate": 3.0672796521661663e-07,
+      "loss": 0.0131,
+      "step": 1804
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.600749381569276,
+      "learning_rate": 3.0397440652532585e-07,
+      "loss": 0.0246,
+      "step": 1805
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.5272728545861523,
+      "learning_rate": 3.0123287572572545e-07,
+      "loss": 0.0168,
+      "step": 1806
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.255654485553308,
+      "learning_rate": 2.985033798396736e-07,
+      "loss": 0.0196,
+      "step": 1807
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.9078529994683149,
+      "learning_rate": 2.9578592585820856e-07,
+      "loss": 0.0237,
+      "step": 1808
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.40880978967914217,
+      "learning_rate": 2.930805207415205e-07,
+      "loss": 0.0138,
+      "step": 1809
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8889728807411933,
+      "learning_rate": 2.9038717141894266e-07,
+      "loss": 0.023,
+      "step": 1810
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.887903401671168,
+      "learning_rate": 2.8770588478892805e-07,
+      "loss": 0.0218,
+      "step": 1811
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.5935263702090745,
+      "learning_rate": 2.850366677190336e-07,
+      "loss": 0.0127,
+      "step": 1812
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8674794845738173,
+      "learning_rate": 2.823795270459029e-07,
+      "loss": 0.0148,
+      "step": 1813
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.4074621926720314,
+      "learning_rate": 2.797344695752491e-07,
+      "loss": 0.0364,
+      "step": 1814
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.0369490988278562,
+      "learning_rate": 2.771015020818363e-07,
+      "loss": 0.0206,
+      "step": 1815
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.42829931436334806,
+      "learning_rate": 2.7448063130946224e-07,
+      "loss": 0.0153,
+      "step": 1816
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.23602416053185918,
+      "learning_rate": 2.718718639709411e-07,
+      "loss": 0.0074,
+      "step": 1817
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.629591449198259,
+      "learning_rate": 2.69275206748088e-07,
+      "loss": 0.0207,
+      "step": 1818
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7405988875650557,
+      "learning_rate": 2.666906662916985e-07,
+      "loss": 0.0154,
+      "step": 1819
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.461232966739589,
+      "learning_rate": 2.641182492215361e-07,
+      "loss": 0.0116,
+      "step": 1820
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7075154031682652,
+      "learning_rate": 2.615579621263109e-07,
+      "loss": 0.0211,
+      "step": 1821
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.1371845028463687,
+      "learning_rate": 2.590098115636658e-07,
+      "loss": 0.0316,
+      "step": 1822
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6774639388623163,
+      "learning_rate": 2.5647380406015665e-07,
+      "loss": 0.0187,
+      "step": 1823
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9274279811411574,
+      "learning_rate": 2.5394994611123934e-07,
+      "loss": 0.0179,
+      "step": 1824
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.306866211141696,
+      "learning_rate": 2.514382441812502e-07,
+      "loss": 0.0184,
+      "step": 1825
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.603304233210114,
+      "learning_rate": 2.489387047033909e-07,
+      "loss": 0.015,
+      "step": 1826
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.5328595053040174,
+      "learning_rate": 2.464513340797114e-07,
+      "loss": 0.0226,
+      "step": 1827
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.866341356192597,
+      "learning_rate": 2.439761386810935e-07,
+      "loss": 0.0172,
+      "step": 1828
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.5610663210635402,
+      "learning_rate": 2.4151312484723465e-07,
+      "loss": 0.0133,
+      "step": 1829
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2493186264229105,
+      "learning_rate": 2.39062298886632e-07,
+      "loss": 0.0128,
+      "step": 1830
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6844264402500051,
+      "learning_rate": 2.3662366707656537e-07,
+      "loss": 0.0176,
+      "step": 1831
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.4233840001854265,
+      "learning_rate": 2.341972356630845e-07,
+      "loss": 0.0144,
+      "step": 1832
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7240753632477378,
+      "learning_rate": 2.3178301086098532e-07,
+      "loss": 0.0143,
+      "step": 1833
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.866915589742617,
+      "learning_rate": 2.293809988538037e-07,
+      "loss": 0.0246,
+      "step": 1834
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.1967068684214617,
+      "learning_rate": 2.2699120579379174e-07,
+      "loss": 0.0261,
+      "step": 1835
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.1879307349406338,
+      "learning_rate": 2.246136378019087e-07,
+      "loss": 0.0268,
+      "step": 1836
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.5166219459046876,
+      "learning_rate": 2.2224830096779837e-07,
+      "loss": 0.0195,
+      "step": 1837
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.1229789289094003,
+      "learning_rate": 2.198952013497796e-07,
+      "loss": 0.0213,
+      "step": 1838
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8698757679037399,
+      "learning_rate": 2.175543449748263e-07,
+      "loss": 0.0171,
+      "step": 1839
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.0318428935613762,
+      "learning_rate": 2.1522573783855473e-07,
+      "loss": 0.0094,
+      "step": 1840
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.4064528369110108,
+      "learning_rate": 2.129093859052067e-07,
+      "loss": 0.03,
+      "step": 1841
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.961657681832751,
+      "learning_rate": 2.106052951076365e-07,
+      "loss": 0.0211,
+      "step": 1842
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.9449980636138309,
+      "learning_rate": 2.083134713472923e-07,
+      "loss": 0.0192,
+      "step": 1843
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7800742218320303,
+      "learning_rate": 2.0603392049420357e-07,
+      "loss": 0.0211,
+      "step": 1844
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.0808053808935367,
+      "learning_rate": 2.0376664838696546e-07,
+      "loss": 0.0109,
+      "step": 1845
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6351185948935967,
+      "learning_rate": 2.0151166083272222e-07,
+      "loss": 0.0137,
+      "step": 1846
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.5968180495011286,
+      "learning_rate": 1.9926896360715542e-07,
+      "loss": 0.0161,
+      "step": 1847
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.5208311636465429,
+      "learning_rate": 1.9703856245446795e-07,
+      "loss": 0.0146,
+      "step": 1848
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.5894146185232907,
+      "learning_rate": 1.948204630873668e-07,
+      "loss": 0.0181,
+      "step": 1849
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2825054473544255,
+      "learning_rate": 1.9261467118705246e-07,
+      "loss": 0.0106,
+      "step": 1850
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2878943348884669,
+      "learning_rate": 1.9042119240320067e-07,
+      "loss": 0.0103,
+      "step": 1851
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.5635184171884943,
+      "learning_rate": 1.8824003235395128e-07,
+      "loss": 0.0223,
+      "step": 1852
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.5714917480875857,
+      "learning_rate": 1.8607119662589045e-07,
+      "loss": 0.0214,
+      "step": 1853
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.5819822583931356,
+      "learning_rate": 1.8391469077404134e-07,
+      "loss": 0.0222,
+      "step": 1854
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7490706862618906,
+      "learning_rate": 1.8177052032184285e-07,
+      "loss": 0.014,
+      "step": 1855
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.9170654518824781,
+      "learning_rate": 1.7963869076114193e-07,
+      "loss": 0.018,
+      "step": 1856
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.9597599706808101,
+      "learning_rate": 1.7751920755217532e-07,
+      "loss": 0.0235,
+      "step": 1857
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.8393265995727257,
+      "learning_rate": 1.7541207612355894e-07,
+      "loss": 0.0237,
+      "step": 1858
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.5720599297711036,
+      "learning_rate": 1.7331730187227003e-07,
+      "loss": 0.0159,
+      "step": 1859
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.38535609643181973,
+      "learning_rate": 1.7123489016363793e-07,
+      "loss": 0.0123,
+      "step": 1860
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6143963637658444,
+      "learning_rate": 1.6916484633132558e-07,
+      "loss": 0.0224,
+      "step": 1861
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.5189176558087583,
+      "learning_rate": 1.671071756773196e-07,
+      "loss": 0.0171,
+      "step": 1862
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6668235969578423,
+      "learning_rate": 1.6506188347191477e-07,
+      "loss": 0.0215,
+      "step": 1863
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.1154353763845224,
+      "learning_rate": 1.6302897495370175e-07,
+      "loss": 0.0282,
+      "step": 1864
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6912898158788077,
+      "learning_rate": 1.61008455329551e-07,
+      "loss": 0.0214,
+      "step": 1865
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.755595253610713,
+      "learning_rate": 1.59000329774604e-07,
+      "loss": 0.0243,
+      "step": 1866
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.43189963611385784,
+      "learning_rate": 1.5700460343225644e-07,
+      "loss": 0.0144,
+      "step": 1867
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7980037217430461,
+      "learning_rate": 1.5502128141414496e-07,
+      "loss": 0.0232,
+      "step": 1868
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.42317850471550683,
+      "learning_rate": 1.5305036880013612e-07,
+      "loss": 0.0144,
+      "step": 1869
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.3440586152342501,
+      "learning_rate": 1.5109187063831243e-07,
+      "loss": 0.0151,
+      "step": 1870
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9689038545074031,
+      "learning_rate": 1.4914579194495794e-07,
+      "loss": 0.025,
+      "step": 1871
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.0026706796663856,
+      "learning_rate": 1.4721213770454988e-07,
+      "loss": 0.0193,
+      "step": 1872
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.3776628922159566,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.0323,
+      "step": 1873
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.1160811515237832,
+      "learning_rate": 1.4338212236134518e-07,
+      "loss": 0.0182,
+      "step": 1874
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.27113768389532256,
+      "learning_rate": 1.414857710683365e-07,
+      "loss": 0.0141,
+      "step": 1875
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.4976224971778871,
+      "learning_rate": 1.3960186384782025e-07,
+      "loss": 0.0174,
+      "step": 1876
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.5627076451492086,
+      "learning_rate": 1.3773040552503447e-07,
+      "loss": 0.0139,
+      "step": 1877
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6144894015224014,
+      "learning_rate": 1.3587140089332984e-07,
+      "loss": 0.0235,
+      "step": 1878
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.37920150421252663,
+      "learning_rate": 1.3402485471415871e-07,
+      "loss": 0.0115,
+      "step": 1879
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.3543116999305515,
+      "learning_rate": 1.3219077171706563e-07,
+      "loss": 0.0114,
+      "step": 1880
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6932990480913824,
+      "learning_rate": 1.303691565996712e-07,
+      "loss": 0.0214,
+      "step": 1881
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.48280266582056625,
+      "learning_rate": 1.2856001402766384e-07,
+      "loss": 0.0123,
+      "step": 1882
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.3600213634007312,
+      "learning_rate": 1.267633486347858e-07,
+      "loss": 0.018,
+      "step": 1883
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6233047924767592,
+      "learning_rate": 1.2497916502282104e-07,
+      "loss": 0.0118,
+      "step": 1884
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6557622650641682,
+      "learning_rate": 1.232074677615841e-07,
+      "loss": 0.0144,
+      "step": 1885
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.68680008251895,
+      "learning_rate": 1.2144826138890897e-07,
+      "loss": 0.014,
+      "step": 1886
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.6753520125550132,
+      "learning_rate": 1.1970155041063636e-07,
+      "loss": 0.0185,
+      "step": 1887
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8061089749043806,
+      "learning_rate": 1.1796733930060256e-07,
+      "loss": 0.027,
+      "step": 1888
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.9511108204414935,
+      "learning_rate": 1.1624563250062836e-07,
+      "loss": 0.0249,
+      "step": 1889
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.5887898728183794,
+      "learning_rate": 1.1453643442050744e-07,
+      "loss": 0.0211,
+      "step": 1890
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.4905414507949429,
+      "learning_rate": 1.1283974943799403e-07,
+      "loss": 0.0138,
+      "step": 1891
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.9247168896017007,
+      "learning_rate": 1.1115558189879416e-07,
+      "loss": 0.0122,
+      "step": 1892
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6215487472298206,
+      "learning_rate": 1.0948393611655229e-07,
+      "loss": 0.0163,
+      "step": 1893
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7473599642814582,
+      "learning_rate": 1.0782481637284014e-07,
+      "loss": 0.0157,
+      "step": 1894
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.0402363260233884,
+      "learning_rate": 1.0617822691714796e-07,
+      "loss": 0.0153,
+      "step": 1895
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.5761737409888579,
+      "learning_rate": 1.0454417196687216e-07,
+      "loss": 0.0186,
+      "step": 1896
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.5818796082089626,
+      "learning_rate": 1.0292265570730431e-07,
+      "loss": 0.0173,
+      "step": 1897
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.4937356890575318,
+      "learning_rate": 1.0131368229161997e-07,
+      "loss": 0.0116,
+      "step": 1898
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.595595287827361,
+      "learning_rate": 9.971725584086933e-08,
+      "loss": 0.0182,
+      "step": 1899
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.704730771852314,
+      "learning_rate": 9.813338044396714e-08,
+      "loss": 0.0166,
+      "step": 1900
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9696721850434132,
+      "learning_rate": 9.656206015768e-08,
+      "loss": 0.0162,
+      "step": 1901
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6552313791421823,
+      "learning_rate": 9.500329900661742e-08,
+      "loss": 0.0185,
+      "step": 1902
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6329810385027012,
+      "learning_rate": 9.345710098322247e-08,
+      "loss": 0.0141,
+      "step": 1903
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.47133660028286867,
+      "learning_rate": 9.192347004775781e-08,
+      "loss": 0.0258,
+      "step": 1904
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.37229356852218276,
+      "learning_rate": 9.040241012830131e-08,
+      "loss": 0.0172,
+      "step": 1905
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.55541532428076,
+      "learning_rate": 8.889392512072992e-08,
+      "loss": 0.0225,
+      "step": 1906
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.5276375856301501,
+      "learning_rate": 8.739801888871468e-08,
+      "loss": 0.0211,
+      "step": 1907
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.4956939691451788,
+      "learning_rate": 8.591469526370799e-08,
+      "loss": 0.0156,
+      "step": 1908
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7859967160273817,
+      "learning_rate": 8.444395804493411e-08,
+      "loss": 0.0197,
+      "step": 1909
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.5770278583139444,
+      "learning_rate": 8.298581099937975e-08,
+      "loss": 0.0161,
+      "step": 1910
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1811644683909457,
+      "learning_rate": 8.154025786178577e-08,
+      "loss": 0.0281,
+      "step": 1911
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.48198193891981583,
+      "learning_rate": 8.010730233463493e-08,
+      "loss": 0.0212,
+      "step": 1912
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8232670905597198,
+      "learning_rate": 7.86869480881447e-08,
+      "loss": 0.027,
+      "step": 1913
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.7847367471690805,
+      "learning_rate": 7.727919876025669e-08,
+      "loss": 0.0224,
+      "step": 1914
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.3673488997507827,
+      "learning_rate": 7.588405795662779e-08,
+      "loss": 0.0147,
+      "step": 1915
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.5712286156889304,
+      "learning_rate": 7.450152925062015e-08,
+      "loss": 0.016,
+      "step": 1916
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.584801234518443,
+      "learning_rate": 7.3131616183294e-08,
+      "loss": 0.0221,
+      "step": 1917
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.1547907543640876,
+      "learning_rate": 7.177432226339542e-08,
+      "loss": 0.0326,
+      "step": 1918
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.5996020279426367,
+      "learning_rate": 7.042965096735076e-08,
+      "loss": 0.0173,
+      "step": 1919
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9630804667184417,
+      "learning_rate": 6.909760573925561e-08,
+      "loss": 0.0194,
+      "step": 1920
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7923608884650656,
+      "learning_rate": 6.777818999086582e-08,
+      "loss": 0.0242,
+      "step": 1921
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6136669614491637,
+      "learning_rate": 6.647140710159039e-08,
+      "loss": 0.0231,
+      "step": 1922
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.444098778295829,
+      "learning_rate": 6.51772604184825e-08,
+      "loss": 0.0227,
+      "step": 1923
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7236056147422347,
+      "learning_rate": 6.389575325622787e-08,
+      "loss": 0.0245,
+      "step": 1924
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.5610244487794093,
+      "learning_rate": 6.262688889714152e-08,
+      "loss": 0.0189,
+      "step": 1925
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6533403737022699,
+      "learning_rate": 6.137067059115431e-08,
+      "loss": 0.0223,
+      "step": 1926
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.2250995109120701,
+      "learning_rate": 6.012710155580858e-08,
+      "loss": 0.0266,
+      "step": 1927
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.5322808366260482,
+      "learning_rate": 5.889618497624649e-08,
+      "loss": 0.0142,
+      "step": 1928
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6284323476920674,
+      "learning_rate": 5.767792400520556e-08,
+      "loss": 0.0219,
+      "step": 1929
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.5611338062761357,
+      "learning_rate": 5.647232176300754e-08,
+      "loss": 0.0192,
+      "step": 1930
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.5508596623052354,
+      "learning_rate": 5.5279381337551286e-08,
+      "loss": 0.0186,
+      "step": 1931
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6227099224632395,
+      "learning_rate": 5.409910578430488e-08,
+      "loss": 0.0132,
+      "step": 1932
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.5567745537204065,
+      "learning_rate": 5.2931498126298495e-08,
+      "loss": 0.0097,
+      "step": 1933
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.5576063453236663,
+      "learning_rate": 5.177656135411657e-08,
+      "loss": 0.0161,
+      "step": 1934
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.3370801447420126,
+      "learning_rate": 5.063429842588841e-08,
+      "loss": 0.0265,
+      "step": 1935
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7544241187243449,
+      "learning_rate": 4.950471226728371e-08,
+      "loss": 0.0182,
+      "step": 1936
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.5253352786844406,
+      "learning_rate": 4.838780577150093e-08,
+      "loss": 0.0174,
+      "step": 1937
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7944907580697882,
+      "learning_rate": 4.728358179926451e-08,
+      "loss": 0.0189,
+      "step": 1938
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.9651599809552647,
+      "learning_rate": 4.619204317881376e-08,
+      "loss": 0.0278,
+      "step": 1939
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6170478992323329,
+      "learning_rate": 4.511319270589731e-08,
+      "loss": 0.0187,
+      "step": 1940
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5413410426193749,
+      "learning_rate": 4.404703314376646e-08,
+      "loss": 0.0212,
+      "step": 1941
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.3773729241544756,
+      "learning_rate": 4.299356722316683e-08,
+      "loss": 0.0157,
+      "step": 1942
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5028399673629241,
+      "learning_rate": 4.1952797642331736e-08,
+      "loss": 0.0161,
+      "step": 1943
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7421325043346971,
+      "learning_rate": 4.092472706697603e-08,
+      "loss": 0.02,
+      "step": 1944
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5828564139036714,
+      "learning_rate": 3.99093581302884e-08,
+      "loss": 0.0159,
+      "step": 1945
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.3703638920114571,
+      "learning_rate": 3.890669343292464e-08,
+      "loss": 0.0153,
+      "step": 1946
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7208034989055797,
+      "learning_rate": 3.791673554300157e-08,
+      "loss": 0.0166,
+      "step": 1947
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7710647513091973,
+      "learning_rate": 3.6939486996090953e-08,
+      "loss": 0.0229,
+      "step": 1948
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6997720856631476,
+      "learning_rate": 3.597495029521059e-08,
+      "loss": 0.0223,
+      "step": 1949
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.44352812503571115,
+      "learning_rate": 3.5023127910820966e-08,
+      "loss": 0.0106,
+      "step": 1950
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7756413157400694,
+      "learning_rate": 3.408402228081642e-08,
+      "loss": 0.0233,
+      "step": 1951
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.5743501883431086,
+      "learning_rate": 3.315763581052067e-08,
+      "loss": 0.0136,
+      "step": 1952
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.662378100288348,
+      "learning_rate": 3.224397087267961e-08,
+      "loss": 0.0172,
+      "step": 1953
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.45464868521172774,
+      "learning_rate": 3.1343029807456296e-08,
+      "loss": 0.0162,
+      "step": 1954
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.3151847873751775,
+      "learning_rate": 3.045481492242319e-08,
+      "loss": 0.0088,
+      "step": 1955
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9179713108290374,
+      "learning_rate": 2.9579328492557734e-08,
+      "loss": 0.0188,
+      "step": 1956
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7442297593300125,
+      "learning_rate": 2.8716572760236205e-08,
+      "loss": 0.0176,
+      "step": 1957
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6467132296202227,
+      "learning_rate": 2.7866549935227638e-08,
+      "loss": 0.0262,
+      "step": 1958
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.4271184977588041,
+      "learning_rate": 2.702926219468882e-08,
+      "loss": 0.0181,
+      "step": 1959
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.34858153463525593,
+      "learning_rate": 2.620471168315819e-08,
+      "loss": 0.0114,
+      "step": 1960
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.5191191049249508,
+      "learning_rate": 2.5392900512549168e-08,
+      "loss": 0.0148,
+      "step": 1961
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7904320889688824,
+      "learning_rate": 2.459383076214794e-08,
+      "loss": 0.0227,
+      "step": 1962
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.40664887186710424,
+      "learning_rate": 2.3807504478604583e-08,
+      "loss": 0.014,
+      "step": 1963
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.404072714467341,
+      "learning_rate": 2.303392367593027e-08,
+      "loss": 0.0302,
+      "step": 1964
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8069295138522137,
+      "learning_rate": 2.2273090335491744e-08,
+      "loss": 0.0241,
+      "step": 1965
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.3797343238694821,
+      "learning_rate": 2.152500640600519e-08,
+      "loss": 0.015,
+      "step": 1966
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8463632953185086,
+      "learning_rate": 2.0789673803530696e-08,
+      "loss": 0.0201,
+      "step": 1967
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.0329783325265296,
+      "learning_rate": 2.006709441147059e-08,
+      "loss": 0.0236,
+      "step": 1968
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.4942200195158771,
+      "learning_rate": 1.9357270080561654e-08,
+      "loss": 0.0204,
+      "step": 1969
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.4577934210496395,
+      "learning_rate": 1.866020262887014e-08,
+      "loss": 0.0214,
+      "step": 1970
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9348702790058298,
+      "learning_rate": 1.7975893841790105e-08,
+      "loss": 0.0175,
+      "step": 1971
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.563106283801156,
+      "learning_rate": 1.7304345472035634e-08,
+      "loss": 0.0142,
+      "step": 1972
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8118174545670076,
+      "learning_rate": 1.6645559239638066e-08,
+      "loss": 0.0139,
+      "step": 1973
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7418989057261591,
+      "learning_rate": 1.5999536831941e-08,
+      "loss": 0.0152,
+      "step": 1974
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.0174044425443258,
+      "learning_rate": 1.536627990359585e-08,
+      "loss": 0.0196,
+      "step": 1975
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.5660161998151887,
+      "learning_rate": 1.474579007655963e-08,
+      "loss": 0.0149,
+      "step": 1976
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6964625660138856,
+      "learning_rate": 1.413806894008718e-08,
+      "loss": 0.0278,
+      "step": 1977
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7449047694433539,
+      "learning_rate": 1.3543118050730053e-08,
+      "loss": 0.0095,
+      "step": 1978
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.5775894499740872,
+      "learning_rate": 1.2960938932329858e-08,
+      "loss": 0.0194,
+      "step": 1979
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.5028520606659571,
+      "learning_rate": 1.2391533076018258e-08,
+      "loss": 0.0185,
+      "step": 1980
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8258924922257007,
+      "learning_rate": 1.1834901940209752e-08,
+      "loss": 0.0302,
+      "step": 1981
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.404199557513869,
+      "learning_rate": 1.129104695059835e-08,
+      "loss": 0.0208,
+      "step": 1982
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7722768594707403,
+      "learning_rate": 1.0759969500155897e-08,
+      "loss": 0.018,
+      "step": 1983
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.2411521975184556,
+      "learning_rate": 1.0241670949127091e-08,
+      "loss": 0.0228,
+      "step": 1984
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.40738726676339176,
+      "learning_rate": 9.73615262502503e-09,
+      "loss": 0.0166,
+      "step": 1985
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.5418543637769685,
+      "learning_rate": 9.243415822629553e-09,
+      "loss": 0.0137,
+      "step": 1986
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.5457350500278516,
+      "learning_rate": 8.763461803983907e-09,
+      "loss": 0.0168,
+      "step": 1987
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.046562261481058,
+      "learning_rate": 8.29629179839031e-09,
+      "loss": 0.0236,
+      "step": 1988
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.5594564303475206,
+      "learning_rate": 7.841907002407723e-09,
+      "loss": 0.0188,
+      "step": 1989
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.4520631765372096,
+      "learning_rate": 7.40030857984686e-09,
+      "loss": 0.0175,
+      "step": 1990
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7392882226317737,
+      "learning_rate": 6.971497661771854e-09,
+      "loss": 0.0201,
+      "step": 1991
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7067878390043856,
+      "learning_rate": 6.555475346491369e-09,
+      "loss": 0.019,
+      "step": 1992
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.4829992263795246,
+      "learning_rate": 6.152242699560273e-09,
+      "loss": 0.023,
+      "step": 1993
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.557053916058929,
+      "learning_rate": 5.761800753775193e-09,
+      "loss": 0.0181,
+      "step": 1994
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.4902359995960106,
+      "learning_rate": 5.384150509171737e-09,
+      "loss": 0.0106,
+      "step": 1995
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6514146661773347,
+      "learning_rate": 5.019292933022279e-09,
+      "loss": 0.022,
+      "step": 1996
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6339899864140751,
+      "learning_rate": 4.6672289598337364e-09,
+      "loss": 0.0182,
+      "step": 1997
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.563678714006939,
+      "learning_rate": 4.327959491344791e-09,
+      "loss": 0.019,
+      "step": 1998
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.46035552482503506,
+      "learning_rate": 4.001485396523119e-09,
+      "loss": 0.0158,
+      "step": 1999
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8784396812135417,
+      "learning_rate": 3.6878075115642785e-09,
+      "loss": 0.0144,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2024,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "total_flos": 85478773166080.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}