diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,27027 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.999221991701245,
+  "eval_steps": 500,
+  "global_step": 3855,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007780082987551867,
+      "grad_norm": 5.940507796289728,
+      "learning_rate": 2.590673575129534e-08,
+      "loss": 0.8978,
+      "step": 1
+    },
+    {
+      "epoch": 0.0015560165975103733,
+      "grad_norm": 6.262486435055598,
+      "learning_rate": 5.181347150259068e-08,
+      "loss": 0.9626,
+      "step": 2
+    },
+    {
+      "epoch": 0.00233402489626556,
+      "grad_norm": 6.176830489713383,
+      "learning_rate": 7.772020725388601e-08,
+      "loss": 0.9644,
+      "step": 3
+    },
+    {
+      "epoch": 0.0031120331950207467,
+      "grad_norm": 5.5620229724063535,
+      "learning_rate": 1.0362694300518136e-07,
+      "loss": 0.8468,
+      "step": 4
+    },
+    {
+      "epoch": 0.0038900414937759337,
+      "grad_norm": 6.0902739542946955,
+      "learning_rate": 1.2953367875647668e-07,
+      "loss": 0.9117,
+      "step": 5
+    },
+    {
+      "epoch": 0.00466804979253112,
+      "grad_norm": 5.660737486665828,
+      "learning_rate": 1.5544041450777202e-07,
+      "loss": 0.8557,
+      "step": 6
+    },
+    {
+      "epoch": 0.005446058091286307,
+      "grad_norm": 6.091239936274589,
+      "learning_rate": 1.8134715025906736e-07,
+      "loss": 0.8768,
+      "step": 7
+    },
+    {
+      "epoch": 0.006224066390041493,
+      "grad_norm": 5.83327665062461,
+      "learning_rate": 2.0725388601036273e-07,
+      "loss": 0.8705,
+      "step": 8
+    },
+    {
+      "epoch": 0.007002074688796681,
+      "grad_norm": 5.99654674436611,
+      "learning_rate": 2.3316062176165804e-07,
+      "loss": 0.9124,
+      "step": 9
+    },
+    {
+      "epoch": 0.007780082987551867,
+      "grad_norm": 5.52441765888137,
+      "learning_rate": 2.5906735751295336e-07,
+      "loss": 0.8718,
+      "step": 10
+    },
+    {
+      "epoch": 0.008558091286307055,
+      "grad_norm": 5.523372478178447,
+      "learning_rate": 2.849740932642487e-07,
+      "loss": 0.7916,
+      "step": 11
+    },
+    {
+      "epoch": 0.00933609958506224,
+      "grad_norm": 5.49766667172653,
+      "learning_rate": 3.1088082901554404e-07,
+      "loss": 0.8221,
+      "step": 12
+    },
+    {
+      "epoch": 0.010114107883817428,
+      "grad_norm": 5.899912558980251,
+      "learning_rate": 3.367875647668394e-07,
+      "loss": 0.9144,
+      "step": 13
+    },
+    {
+      "epoch": 0.010892116182572614,
+      "grad_norm": 6.110583397173363,
+      "learning_rate": 3.626943005181347e-07,
+      "loss": 0.931,
+      "step": 14
+    },
+    {
+      "epoch": 0.011670124481327801,
+      "grad_norm": 5.671081408936153,
+      "learning_rate": 3.886010362694301e-07,
+      "loss": 0.9068,
+      "step": 15
+    },
+    {
+      "epoch": 0.012448132780082987,
+      "grad_norm": 5.877368511202829,
+      "learning_rate": 4.1450777202072546e-07,
+      "loss": 0.9291,
+      "step": 16
+    },
+    {
+      "epoch": 0.013226141078838174,
+      "grad_norm": 5.621198482658771,
+      "learning_rate": 4.404145077720207e-07,
+      "loss": 0.8809,
+      "step": 17
+    },
+    {
+      "epoch": 0.014004149377593362,
+      "grad_norm": 5.336370514976981,
+      "learning_rate": 4.663212435233161e-07,
+      "loss": 0.8652,
+      "step": 18
+    },
+    {
+      "epoch": 0.014782157676348547,
+      "grad_norm": 5.172904002119997,
+      "learning_rate": 4.922279792746115e-07,
+      "loss": 0.8476,
+      "step": 19
+    },
+    {
+      "epoch": 0.015560165975103735,
+      "grad_norm": 4.549547487697753,
+      "learning_rate": 5.181347150259067e-07,
+      "loss": 0.8095,
+      "step": 20
+    },
+    {
+      "epoch": 0.016338174273858922,
+      "grad_norm": 4.280791845729526,
+      "learning_rate": 5.440414507772021e-07,
+      "loss": 0.8135,
+      "step": 21
+    },
+    {
+      "epoch": 0.01711618257261411,
+      "grad_norm": 4.913332254979736,
+      "learning_rate": 5.699481865284974e-07,
+      "loss": 0.8695,
+      "step": 22
+    },
+    {
+      "epoch": 0.017894190871369293,
+      "grad_norm": 4.431123895474749,
+      "learning_rate": 5.958549222797927e-07,
+      "loss": 0.8528,
+      "step": 23
+    },
+    {
+      "epoch": 0.01867219917012448,
+      "grad_norm": 4.485598988844872,
+      "learning_rate": 6.217616580310881e-07,
+      "loss": 0.8569,
+      "step": 24
+    },
+    {
+      "epoch": 0.01945020746887967,
+      "grad_norm": 4.163779683328616,
+      "learning_rate": 6.476683937823834e-07,
+      "loss": 0.7902,
+      "step": 25
+    },
+    {
+      "epoch": 0.020228215767634856,
+      "grad_norm": 3.407588931117682,
+      "learning_rate": 6.735751295336788e-07,
+      "loss": 0.7943,
+      "step": 26
+    },
+    {
+      "epoch": 0.021006224066390043,
+      "grad_norm": 2.7740282219214794,
+      "learning_rate": 6.994818652849742e-07,
+      "loss": 0.7481,
+      "step": 27
+    },
+    {
+      "epoch": 0.021784232365145227,
+      "grad_norm": 2.504536887250622,
+      "learning_rate": 7.253886010362694e-07,
+      "loss": 0.716,
+      "step": 28
+    },
+    {
+      "epoch": 0.022562240663900415,
+      "grad_norm": 2.4828533201465404,
+      "learning_rate": 7.512953367875648e-07,
+      "loss": 0.7533,
+      "step": 29
+    },
+    {
+      "epoch": 0.023340248962655602,
+      "grad_norm": 2.5014974259038834,
+      "learning_rate": 7.772020725388602e-07,
+      "loss": 0.8164,
+      "step": 30
+    },
+    {
+      "epoch": 0.02411825726141079,
+      "grad_norm": 2.525507645110024,
+      "learning_rate": 8.031088082901554e-07,
+      "loss": 0.7366,
+      "step": 31
+    },
+    {
+      "epoch": 0.024896265560165973,
+      "grad_norm": 2.4997963072854357,
+      "learning_rate": 8.290155440414509e-07,
+      "loss": 0.7724,
+      "step": 32
+    },
+    {
+      "epoch": 0.02567427385892116,
+      "grad_norm": 2.0875942435744355,
+      "learning_rate": 8.549222797927462e-07,
+      "loss": 0.7477,
+      "step": 33
+    },
+    {
+      "epoch": 0.026452282157676348,
+      "grad_norm": 2.1229190229453394,
+      "learning_rate": 8.808290155440414e-07,
+      "loss": 0.7287,
+      "step": 34
+    },
+    {
+      "epoch": 0.027230290456431536,
+      "grad_norm": 1.7814700005127322,
+      "learning_rate": 9.067357512953369e-07,
+      "loss": 0.7215,
+      "step": 35
+    },
+    {
+      "epoch": 0.028008298755186723,
+      "grad_norm": 1.704658544446142,
+      "learning_rate": 9.326424870466322e-07,
+      "loss": 0.7474,
+      "step": 36
+    },
+    {
+      "epoch": 0.028786307053941907,
+      "grad_norm": 1.8225943922216221,
+      "learning_rate": 9.585492227979275e-07,
+      "loss": 0.805,
+      "step": 37
+    },
+    {
+      "epoch": 0.029564315352697094,
+      "grad_norm": 1.933348194883269,
+      "learning_rate": 9.84455958549223e-07,
+      "loss": 0.7131,
+      "step": 38
+    },
+    {
+      "epoch": 0.030342323651452282,
+      "grad_norm": 1.996163692348563,
+      "learning_rate": 1.0103626943005183e-06,
+      "loss": 0.684,
+      "step": 39
+    },
+    {
+      "epoch": 0.03112033195020747,
+      "grad_norm": 2.0912166530809277,
+      "learning_rate": 1.0362694300518134e-06,
+      "loss": 0.6735,
+      "step": 40
+    },
+    {
+      "epoch": 0.03189834024896265,
+      "grad_norm": 1.9149255066559672,
+      "learning_rate": 1.062176165803109e-06,
+      "loss": 0.6807,
+      "step": 41
+    },
+    {
+      "epoch": 0.032676348547717844,
+      "grad_norm": 1.8248582962820075,
+      "learning_rate": 1.0880829015544042e-06,
+      "loss": 0.7665,
+      "step": 42
+    },
+    {
+      "epoch": 0.03345435684647303,
+      "grad_norm": 1.8008040422420342,
+      "learning_rate": 1.1139896373056995e-06,
+      "loss": 0.6718,
+      "step": 43
+    },
+    {
+      "epoch": 0.03423236514522822,
+      "grad_norm": 1.683648490743137,
+      "learning_rate": 1.139896373056995e-06,
+      "loss": 0.6614,
+      "step": 44
+    },
+    {
+      "epoch": 0.0350103734439834,
+      "grad_norm": 1.580583083695629,
+      "learning_rate": 1.1658031088082903e-06,
+      "loss": 0.6708,
+      "step": 45
+    },
+    {
+      "epoch": 0.03578838174273859,
+      "grad_norm": 1.3328414529638786,
+      "learning_rate": 1.1917098445595854e-06,
+      "loss": 0.6784,
+      "step": 46
+    },
+    {
+      "epoch": 0.03656639004149378,
+      "grad_norm": 1.0697095176711033,
+      "learning_rate": 1.217616580310881e-06,
+      "loss": 0.6299,
+      "step": 47
+    },
+    {
+      "epoch": 0.03734439834024896,
+      "grad_norm": 1.2116062071755522,
+      "learning_rate": 1.2435233160621762e-06,
+      "loss": 0.6967,
+      "step": 48
+    },
+    {
+      "epoch": 0.03812240663900415,
+      "grad_norm": 1.0977364389074036,
+      "learning_rate": 1.2694300518134717e-06,
+      "loss": 0.6851,
+      "step": 49
+    },
+    {
+      "epoch": 0.03890041493775934,
+      "grad_norm": 1.1705131326832117,
+      "learning_rate": 1.2953367875647669e-06,
+      "loss": 0.6897,
+      "step": 50
+    },
+    {
+      "epoch": 0.03967842323651452,
+      "grad_norm": 1.1148917406181216,
+      "learning_rate": 1.3212435233160623e-06,
+      "loss": 0.6392,
+      "step": 51
+    },
+    {
+      "epoch": 0.04045643153526971,
+      "grad_norm": 1.0248113079511978,
+      "learning_rate": 1.3471502590673576e-06,
+      "loss": 0.624,
+      "step": 52
+    },
+    {
+      "epoch": 0.041234439834024895,
+      "grad_norm": 1.1055377598288203,
+      "learning_rate": 1.3730569948186528e-06,
+      "loss": 0.6561,
+      "step": 53
+    },
+    {
+      "epoch": 0.042012448132780086,
+      "grad_norm": 0.9282761384995964,
+      "learning_rate": 1.3989637305699484e-06,
+      "loss": 0.6557,
+      "step": 54
+    },
+    {
+      "epoch": 0.04279045643153527,
+      "grad_norm": 1.0637950550385509,
+      "learning_rate": 1.4248704663212437e-06,
+      "loss": 0.6356,
+      "step": 55
+    },
+    {
+      "epoch": 0.043568464730290454,
+      "grad_norm": 0.9873528990596028,
+      "learning_rate": 1.4507772020725389e-06,
+      "loss": 0.6493,
+      "step": 56
+    },
+    {
+      "epoch": 0.044346473029045645,
+      "grad_norm": 0.8518392174905262,
+      "learning_rate": 1.4766839378238342e-06,
+      "loss": 0.5707,
+      "step": 57
+    },
+    {
+      "epoch": 0.04512448132780083,
+      "grad_norm": 0.9521428142139002,
+      "learning_rate": 1.5025906735751296e-06,
+      "loss": 0.6512,
+      "step": 58
+    },
+    {
+      "epoch": 0.04590248962655602,
+      "grad_norm": 0.8975542715109404,
+      "learning_rate": 1.5284974093264248e-06,
+      "loss": 0.602,
+      "step": 59
+    },
+    {
+      "epoch": 0.046680497925311204,
+      "grad_norm": 0.7794749619927374,
+      "learning_rate": 1.5544041450777204e-06,
+      "loss": 0.6317,
+      "step": 60
+    },
+    {
+      "epoch": 0.04745850622406639,
+      "grad_norm": 0.8809359084924746,
+      "learning_rate": 1.5803108808290157e-06,
+      "loss": 0.6416,
+      "step": 61
+    },
+    {
+      "epoch": 0.04823651452282158,
+      "grad_norm": 0.7763888271191122,
+      "learning_rate": 1.6062176165803109e-06,
+      "loss": 0.6191,
+      "step": 62
+    },
+    {
+      "epoch": 0.04901452282157676,
+      "grad_norm": 0.7577330687240627,
+      "learning_rate": 1.6321243523316062e-06,
+      "loss": 0.5935,
+      "step": 63
+    },
+    {
+      "epoch": 0.04979253112033195,
+      "grad_norm": 0.8994813140579995,
+      "learning_rate": 1.6580310880829018e-06,
+      "loss": 0.635,
+      "step": 64
+    },
+    {
+      "epoch": 0.05057053941908714,
+      "grad_norm": 0.8325095595191202,
+      "learning_rate": 1.683937823834197e-06,
+      "loss": 0.5952,
+      "step": 65
+    },
+    {
+      "epoch": 0.05134854771784232,
+      "grad_norm": 0.8357821506761094,
+      "learning_rate": 1.7098445595854923e-06,
+      "loss": 0.6123,
+      "step": 66
+    },
+    {
+      "epoch": 0.05212655601659751,
+      "grad_norm": 0.8128235049518621,
+      "learning_rate": 1.7357512953367877e-06,
+      "loss": 0.5579,
+      "step": 67
+    },
+    {
+      "epoch": 0.052904564315352696,
+      "grad_norm": 0.7626899914056018,
+      "learning_rate": 1.7616580310880829e-06,
+      "loss": 0.589,
+      "step": 68
+    },
+    {
+      "epoch": 0.05368257261410788,
+      "grad_norm": 0.6876644117267264,
+      "learning_rate": 1.7875647668393784e-06,
+      "loss": 0.5661,
+      "step": 69
+    },
+    {
+      "epoch": 0.05446058091286307,
+      "grad_norm": 0.790368429566323,
+      "learning_rate": 1.8134715025906738e-06,
+      "loss": 0.5964,
+      "step": 70
+    },
+    {
+      "epoch": 0.055238589211618255,
+      "grad_norm": 0.7141589631332569,
+      "learning_rate": 1.839378238341969e-06,
+      "loss": 0.5888,
+      "step": 71
+    },
+    {
+      "epoch": 0.056016597510373446,
+      "grad_norm": 0.6576831541826109,
+      "learning_rate": 1.8652849740932643e-06,
+      "loss": 0.5852,
+      "step": 72
+    },
+    {
+      "epoch": 0.05679460580912863,
+      "grad_norm": 0.7022881633122707,
+      "learning_rate": 1.8911917098445597e-06,
+      "loss": 0.5739,
+      "step": 73
+    },
+    {
+      "epoch": 0.057572614107883814,
+      "grad_norm": 0.6914182857269667,
+      "learning_rate": 1.917098445595855e-06,
+      "loss": 0.5746,
+      "step": 74
+    },
+    {
+      "epoch": 0.058350622406639005,
+      "grad_norm": 0.9014250242689962,
+      "learning_rate": 1.9430051813471504e-06,
+      "loss": 0.6453,
+      "step": 75
+    },
+    {
+      "epoch": 0.05912863070539419,
+      "grad_norm": 0.640065454756644,
+      "learning_rate": 1.968911917098446e-06,
+      "loss": 0.5661,
+      "step": 76
+    },
+    {
+      "epoch": 0.05990663900414938,
+      "grad_norm": 0.7788257739606065,
+      "learning_rate": 1.994818652849741e-06,
+      "loss": 0.6417,
+      "step": 77
+    },
+    {
+      "epoch": 0.060684647302904564,
+      "grad_norm": 0.7399127988280708,
+      "learning_rate": 2.0207253886010365e-06,
+      "loss": 0.5236,
+      "step": 78
+    },
+    {
+      "epoch": 0.06146265560165975,
+      "grad_norm": 0.7440892749897914,
+      "learning_rate": 2.046632124352332e-06,
+      "loss": 0.5879,
+      "step": 79
+    },
+    {
+      "epoch": 0.06224066390041494,
+      "grad_norm": 0.7675724880444978,
+      "learning_rate": 2.072538860103627e-06,
+      "loss": 0.6051,
+      "step": 80
+    },
+    {
+      "epoch": 0.06301867219917012,
+      "grad_norm": 0.8176084012801272,
+      "learning_rate": 2.0984455958549222e-06,
+      "loss": 0.6009,
+      "step": 81
+    },
+    {
+      "epoch": 0.0637966804979253,
+      "grad_norm": 0.6514927010899955,
+      "learning_rate": 2.124352331606218e-06,
+      "loss": 0.5745,
+      "step": 82
+    },
+    {
+      "epoch": 0.0645746887966805,
+      "grad_norm": 0.765377940944985,
+      "learning_rate": 2.150259067357513e-06,
+      "loss": 0.54,
+      "step": 83
+    },
+    {
+      "epoch": 0.06535269709543569,
+      "grad_norm": 0.7672336171514941,
+      "learning_rate": 2.1761658031088083e-06,
+      "loss": 0.5584,
+      "step": 84
+    },
+    {
+      "epoch": 0.06613070539419087,
+      "grad_norm": 0.7269630785394428,
+      "learning_rate": 2.2020725388601037e-06,
+      "loss": 0.5109,
+      "step": 85
+    },
+    {
+      "epoch": 0.06690871369294606,
+      "grad_norm": 0.7253016112269777,
+      "learning_rate": 2.227979274611399e-06,
+      "loss": 0.6229,
+      "step": 86
+    },
+    {
+      "epoch": 0.06768672199170124,
+      "grad_norm": 0.5987505454227107,
+      "learning_rate": 2.2538860103626944e-06,
+      "loss": 0.5535,
+      "step": 87
+    },
+    {
+      "epoch": 0.06846473029045644,
+      "grad_norm": 0.737821084288819,
+      "learning_rate": 2.27979274611399e-06,
+      "loss": 0.5295,
+      "step": 88
+    },
+    {
+      "epoch": 0.06924273858921162,
+      "grad_norm": 0.7310866953811318,
+      "learning_rate": 2.305699481865285e-06,
+      "loss": 0.6169,
+      "step": 89
+    },
+    {
+      "epoch": 0.0700207468879668,
+      "grad_norm": 0.7213761844121452,
+      "learning_rate": 2.3316062176165805e-06,
+      "loss": 0.5741,
+      "step": 90
+    },
+    {
+      "epoch": 0.07079875518672199,
+      "grad_norm": 0.7180607929900488,
+      "learning_rate": 2.357512953367876e-06,
+      "loss": 0.5652,
+      "step": 91
+    },
+    {
+      "epoch": 0.07157676348547717,
+      "grad_norm": 0.7816158818353988,
+      "learning_rate": 2.383419689119171e-06,
+      "loss": 0.5728,
+      "step": 92
+    },
+    {
+      "epoch": 0.07235477178423237,
+      "grad_norm": 0.6420228089642194,
+      "learning_rate": 2.4093264248704666e-06,
+      "loss": 0.5646,
+      "step": 93
+    },
+    {
+      "epoch": 0.07313278008298756,
+      "grad_norm": 0.7725521233918189,
+      "learning_rate": 2.435233160621762e-06,
+      "loss": 0.5934,
+      "step": 94
+    },
+    {
+      "epoch": 0.07391078838174274,
+      "grad_norm": 0.6757366809503692,
+      "learning_rate": 2.461139896373057e-06,
+      "loss": 0.5858,
+      "step": 95
+    },
+    {
+      "epoch": 0.07468879668049792,
+      "grad_norm": 0.7179603560037259,
+      "learning_rate": 2.4870466321243523e-06,
+      "loss": 0.5544,
+      "step": 96
+    },
+    {
+      "epoch": 0.07546680497925311,
+      "grad_norm": 0.7371176059117633,
+      "learning_rate": 2.512953367875648e-06,
+      "loss": 0.5321,
+      "step": 97
+    },
+    {
+      "epoch": 0.0762448132780083,
+      "grad_norm": 0.7179752428280738,
+      "learning_rate": 2.5388601036269435e-06,
+      "loss": 0.5152,
+      "step": 98
+    },
+    {
+      "epoch": 0.07702282157676349,
+      "grad_norm": 0.679712290390269,
+      "learning_rate": 2.564766839378239e-06,
+      "loss": 0.5392,
+      "step": 99
+    },
+    {
+      "epoch": 0.07780082987551867,
+      "grad_norm": 0.5632393506925155,
+      "learning_rate": 2.5906735751295338e-06,
+      "loss": 0.5501,
+      "step": 100
+    },
+    {
+      "epoch": 0.07857883817427386,
+      "grad_norm": 0.6600242332837734,
+      "learning_rate": 2.616580310880829e-06,
+      "loss": 0.5726,
+      "step": 101
+    },
+    {
+      "epoch": 0.07935684647302904,
+      "grad_norm": 0.6464487920663716,
+      "learning_rate": 2.6424870466321245e-06,
+      "loss": 0.561,
+      "step": 102
+    },
+    {
+      "epoch": 0.08013485477178424,
+      "grad_norm": 0.7298755534842148,
+      "learning_rate": 2.66839378238342e-06,
+      "loss": 0.5586,
+      "step": 103
+    },
+    {
+      "epoch": 0.08091286307053942,
+      "grad_norm": 0.6878255864724004,
+      "learning_rate": 2.6943005181347152e-06,
+      "loss": 0.6106,
+      "step": 104
+    },
+    {
+      "epoch": 0.08169087136929461,
+      "grad_norm": 0.6820267434902035,
+      "learning_rate": 2.7202072538860106e-06,
+      "loss": 0.5958,
+      "step": 105
+    },
+    {
+      "epoch": 0.08246887966804979,
+      "grad_norm": 0.7021030436404904,
+      "learning_rate": 2.7461139896373056e-06,
+      "loss": 0.5374,
+      "step": 106
+    },
+    {
+      "epoch": 0.08324688796680497,
+      "grad_norm": 0.770890884966827,
+      "learning_rate": 2.772020725388601e-06,
+      "loss": 0.5839,
+      "step": 107
+    },
+    {
+      "epoch": 0.08402489626556017,
+      "grad_norm": 0.6095671724419239,
+      "learning_rate": 2.7979274611398967e-06,
+      "loss": 0.5197,
+      "step": 108
+    },
+    {
+      "epoch": 0.08480290456431536,
+      "grad_norm": 0.7944548036771061,
+      "learning_rate": 2.823834196891192e-06,
+      "loss": 0.5903,
+      "step": 109
+    },
+    {
+      "epoch": 0.08558091286307054,
+      "grad_norm": 0.6001976550458133,
+      "learning_rate": 2.8497409326424875e-06,
+      "loss": 0.5322,
+      "step": 110
+    },
+    {
+      "epoch": 0.08635892116182572,
+      "grad_norm": 0.6650401501176635,
+      "learning_rate": 2.875647668393783e-06,
+      "loss": 0.5632,
+      "step": 111
+    },
+    {
+      "epoch": 0.08713692946058091,
+      "grad_norm": 0.7076631766589855,
+      "learning_rate": 2.9015544041450778e-06,
+      "loss": 0.5349,
+      "step": 112
+    },
+    {
+      "epoch": 0.0879149377593361,
+      "grad_norm": 0.7109203586044982,
+      "learning_rate": 2.927461139896373e-06,
+      "loss": 0.5901,
+      "step": 113
+    },
+    {
+      "epoch": 0.08869294605809129,
+      "grad_norm": 0.6790587542882515,
+      "learning_rate": 2.9533678756476685e-06,
+      "loss": 0.5369,
+      "step": 114
+    },
+    {
+      "epoch": 0.08947095435684647,
+      "grad_norm": 0.7010723577950845,
+      "learning_rate": 2.979274611398964e-06,
+      "loss": 0.5262,
+      "step": 115
+    },
+    {
+      "epoch": 0.09024896265560166,
+      "grad_norm": 0.6646782507939125,
+      "learning_rate": 3.0051813471502592e-06,
+      "loss": 0.5487,
+      "step": 116
+    },
+    {
+      "epoch": 0.09102697095435684,
+      "grad_norm": 0.7332680418285585,
+      "learning_rate": 3.031088082901555e-06,
+      "loss": 0.5995,
+      "step": 117
+    },
+    {
+      "epoch": 0.09180497925311204,
+      "grad_norm": 0.6604024291231017,
+      "learning_rate": 3.0569948186528495e-06,
+      "loss": 0.5513,
+      "step": 118
+    },
+    {
+      "epoch": 0.09258298755186722,
+      "grad_norm": 0.6235005444696374,
+      "learning_rate": 3.0829015544041453e-06,
+      "loss": 0.5578,
+      "step": 119
+    },
+    {
+      "epoch": 0.09336099585062241,
+      "grad_norm": 0.6695771874348048,
+      "learning_rate": 3.1088082901554407e-06,
+      "loss": 0.5141,
+      "step": 120
+    },
+    {
+      "epoch": 0.09413900414937759,
+      "grad_norm": 0.6174506445364933,
+      "learning_rate": 3.134715025906736e-06,
+      "loss": 0.4948,
+      "step": 121
+    },
+    {
+      "epoch": 0.09491701244813278,
+      "grad_norm": 0.7056115751314509,
+      "learning_rate": 3.1606217616580314e-06,
+      "loss": 0.523,
+      "step": 122
+    },
+    {
+      "epoch": 0.09569502074688796,
+      "grad_norm": 0.6827683777257929,
+      "learning_rate": 3.186528497409327e-06,
+      "loss": 0.5363,
+      "step": 123
+    },
+    {
+      "epoch": 0.09647302904564316,
+      "grad_norm": 0.5881002024151596,
+      "learning_rate": 3.2124352331606218e-06,
+      "loss": 0.5418,
+      "step": 124
+    },
+    {
+      "epoch": 0.09725103734439834,
+      "grad_norm": 0.6266537767952944,
+      "learning_rate": 3.238341968911917e-06,
+      "loss": 0.4766,
+      "step": 125
+    },
+    {
+      "epoch": 0.09802904564315353,
+      "grad_norm": 0.8125020664954572,
+      "learning_rate": 3.2642487046632125e-06,
+      "loss": 0.5571,
+      "step": 126
+    },
+    {
+      "epoch": 0.09880705394190871,
+      "grad_norm": 0.6049890965410883,
+      "learning_rate": 3.2901554404145083e-06,
+      "loss": 0.5666,
+      "step": 127
+    },
+    {
+      "epoch": 0.0995850622406639,
+      "grad_norm": 0.7042566189780375,
+      "learning_rate": 3.3160621761658036e-06,
+      "loss": 0.5741,
+      "step": 128
+    },
+    {
+      "epoch": 0.10036307053941909,
+      "grad_norm": 0.7444030507716393,
+      "learning_rate": 3.341968911917099e-06,
+      "loss": 0.5655,
+      "step": 129
+    },
+    {
+      "epoch": 0.10114107883817428,
+      "grad_norm": 0.6907129550037663,
+      "learning_rate": 3.367875647668394e-06,
+      "loss": 0.5114,
+      "step": 130
+    },
+    {
+      "epoch": 0.10191908713692946,
+      "grad_norm": 0.7391765076508863,
+      "learning_rate": 3.3937823834196893e-06,
+      "loss": 0.5838,
+      "step": 131
+    },
+    {
+      "epoch": 0.10269709543568464,
+      "grad_norm": 0.8098361451910274,
+      "learning_rate": 3.4196891191709847e-06,
+      "loss": 0.5952,
+      "step": 132
+    },
+    {
+      "epoch": 0.10347510373443983,
+      "grad_norm": 0.6979321112725825,
+      "learning_rate": 3.44559585492228e-06,
+      "loss": 0.5506,
+      "step": 133
+    },
+    {
+      "epoch": 0.10425311203319503,
+      "grad_norm": 0.746283163999384,
+      "learning_rate": 3.4715025906735754e-06,
+      "loss": 0.5636,
+      "step": 134
+    },
+    {
+      "epoch": 0.10503112033195021,
+      "grad_norm": 0.709593961177084,
+      "learning_rate": 3.497409326424871e-06,
+      "loss": 0.5383,
+      "step": 135
+    },
+    {
+      "epoch": 0.10580912863070539,
+      "grad_norm": 0.6798991606111219,
+      "learning_rate": 3.5233160621761657e-06,
+      "loss": 0.5473,
+      "step": 136
+    },
+    {
+      "epoch": 0.10658713692946058,
+      "grad_norm": 0.6317225855903414,
+      "learning_rate": 3.549222797927461e-06,
+      "loss": 0.5226,
+      "step": 137
+    },
+    {
+      "epoch": 0.10736514522821576,
+      "grad_norm": 0.710728327147037,
+      "learning_rate": 3.575129533678757e-06,
+      "loss": 0.5451,
+      "step": 138
+    },
+    {
+      "epoch": 0.10814315352697096,
+      "grad_norm": 0.6819998180015413,
+      "learning_rate": 3.6010362694300523e-06,
+      "loss": 0.5197,
+      "step": 139
+    },
+    {
+      "epoch": 0.10892116182572614,
+      "grad_norm": 0.6658648163499558,
+      "learning_rate": 3.6269430051813476e-06,
+      "loss": 0.5238,
+      "step": 140
+    },
+    {
+      "epoch": 0.10969917012448133,
+      "grad_norm": 0.5997333821694552,
+      "learning_rate": 3.652849740932643e-06,
+      "loss": 0.5325,
+      "step": 141
+    },
+    {
+      "epoch": 0.11047717842323651,
+      "grad_norm": 0.6874765942751304,
+      "learning_rate": 3.678756476683938e-06,
+      "loss": 0.5488,
+      "step": 142
+    },
+    {
+      "epoch": 0.1112551867219917,
+      "grad_norm": 0.7017266243860996,
+      "learning_rate": 3.7046632124352333e-06,
+      "loss": 0.5086,
+      "step": 143
+    },
+    {
+      "epoch": 0.11203319502074689,
+      "grad_norm": 0.6679540847774085,
+      "learning_rate": 3.7305699481865287e-06,
+      "loss": 0.4895,
+      "step": 144
+    },
+    {
+      "epoch": 0.11281120331950208,
+      "grad_norm": 0.7119629440706817,
+      "learning_rate": 3.756476683937824e-06,
+      "loss": 0.5336,
+      "step": 145
+    },
+    {
+      "epoch": 0.11358921161825726,
+      "grad_norm": 0.6884733220418674,
+      "learning_rate": 3.7823834196891194e-06,
+      "loss": 0.569,
+      "step": 146
+    },
+    {
+      "epoch": 0.11436721991701244,
+      "grad_norm": 0.659427701761204,
+      "learning_rate": 3.808290155440415e-06,
+      "loss": 0.5235,
+      "step": 147
+    },
+    {
+      "epoch": 0.11514522821576763,
+      "grad_norm": 0.6663455405224277,
+      "learning_rate": 3.83419689119171e-06,
+      "loss": 0.5005,
+      "step": 148
+    },
+    {
+      "epoch": 0.11592323651452283,
+      "grad_norm": 0.666752560069172,
+      "learning_rate": 3.860103626943005e-06,
+      "loss": 0.5496,
+      "step": 149
+    },
+    {
+      "epoch": 0.11670124481327801,
+      "grad_norm": 0.7377086954906804,
+      "learning_rate": 3.886010362694301e-06,
+      "loss": 0.547,
+      "step": 150
+    },
+    {
+      "epoch": 0.1174792531120332,
+      "grad_norm": 0.5715473596148781,
+      "learning_rate": 3.911917098445596e-06,
+      "loss": 0.5041,
+      "step": 151
+    },
+    {
+      "epoch": 0.11825726141078838,
+      "grad_norm": 0.7150439638416728,
+      "learning_rate": 3.937823834196892e-06,
+      "loss": 0.5459,
+      "step": 152
+    },
+    {
+      "epoch": 0.11903526970954356,
+      "grad_norm": 0.6301305099428113,
+      "learning_rate": 3.963730569948187e-06,
+      "loss": 0.5449,
+      "step": 153
+    },
+    {
+      "epoch": 0.11981327800829876,
+      "grad_norm": 0.6543251120735123,
+      "learning_rate": 3.989637305699482e-06,
+      "loss": 0.5333,
+      "step": 154
+    },
+    {
+      "epoch": 0.12059128630705394,
+      "grad_norm": 0.6570622257061327,
+      "learning_rate": 4.015544041450777e-06,
+      "loss": 0.5174,
+      "step": 155
+    },
+    {
+      "epoch": 0.12136929460580913,
+      "grad_norm": 0.7738000665731307,
+      "learning_rate": 4.041450777202073e-06,
+      "loss": 0.5489,
+      "step": 156
+    },
+    {
+      "epoch": 0.12214730290456431,
+      "grad_norm": 0.6956455451596079,
+      "learning_rate": 4.067357512953368e-06,
+      "loss": 0.5514,
+      "step": 157
+    },
+    {
+      "epoch": 0.1229253112033195,
+      "grad_norm": 0.7514141145613829,
+      "learning_rate": 4.093264248704664e-06,
+      "loss": 0.5204,
+      "step": 158
+    },
+    {
+      "epoch": 0.1237033195020747,
+      "grad_norm": 0.7605104347266205,
+      "learning_rate": 4.119170984455959e-06,
+      "loss": 0.5331,
+      "step": 159
+    },
+    {
+      "epoch": 0.12448132780082988,
+      "grad_norm": 0.6954359596323758,
+      "learning_rate": 4.145077720207254e-06,
+      "loss": 0.5108,
+      "step": 160
+    },
+    {
+      "epoch": 0.12525933609958506,
+      "grad_norm": 0.6587932051413148,
+      "learning_rate": 4.1709844559585495e-06,
+      "loss": 0.5043,
+      "step": 161
+    },
+    {
+      "epoch": 0.12603734439834025,
+      "grad_norm": 0.7864570035188176,
+      "learning_rate": 4.1968911917098444e-06,
+      "loss": 0.5123,
+      "step": 162
+    },
+    {
+      "epoch": 0.12681535269709543,
+      "grad_norm": 0.7326232096837338,
+      "learning_rate": 4.22279792746114e-06,
+      "loss": 0.5244,
+      "step": 163
+    },
+    {
+      "epoch": 0.1275933609958506,
+      "grad_norm": 0.6785939435211489,
+      "learning_rate": 4.248704663212436e-06,
+      "loss": 0.5104,
+      "step": 164
+    },
+    {
+      "epoch": 0.1283713692946058,
+      "grad_norm": 0.6661890398583603,
+      "learning_rate": 4.274611398963731e-06,
+      "loss": 0.5137,
+      "step": 165
+    },
+    {
+      "epoch": 0.129149377593361,
+      "grad_norm": 0.727199255368865,
+      "learning_rate": 4.300518134715026e-06,
+      "loss": 0.5307,
+      "step": 166
+    },
+    {
+      "epoch": 0.1299273858921162,
+      "grad_norm": 0.6855816949908273,
+      "learning_rate": 4.326424870466322e-06,
+      "loss": 0.5202,
+      "step": 167
+    },
+    {
+      "epoch": 0.13070539419087138,
+      "grad_norm": 0.7347883304090294,
+      "learning_rate": 4.352331606217617e-06,
+      "loss": 0.5671,
+      "step": 168
+    },
+    {
+      "epoch": 0.13148340248962656,
+      "grad_norm": 0.7035765210591745,
+      "learning_rate": 4.3782383419689124e-06,
+      "loss": 0.559,
+      "step": 169
+    },
+    {
+      "epoch": 0.13226141078838174,
+      "grad_norm": 0.7022287735791105,
+      "learning_rate": 4.404145077720207e-06,
+      "loss": 0.5141,
+      "step": 170
+    },
+    {
+      "epoch": 0.13303941908713693,
+      "grad_norm": 0.6827045744516606,
+      "learning_rate": 4.430051813471503e-06,
+      "loss": 0.534,
+      "step": 171
+    },
+    {
+      "epoch": 0.1338174273858921,
+      "grad_norm": 0.859943481741687,
+      "learning_rate": 4.455958549222798e-06,
+      "loss": 0.5437,
+      "step": 172
+    },
+    {
+      "epoch": 0.1345954356846473,
+      "grad_norm": 0.7083318353436859,
+      "learning_rate": 4.481865284974093e-06,
+      "loss": 0.504,
+      "step": 173
+    },
+    {
+      "epoch": 0.13537344398340248,
+      "grad_norm": 0.6773718862744089,
+      "learning_rate": 4.507772020725389e-06,
+      "loss": 0.5441,
+      "step": 174
+    },
+    {
+      "epoch": 0.13615145228215766,
+      "grad_norm": 0.6898906596109363,
+      "learning_rate": 4.533678756476685e-06,
+      "loss": 0.5503,
+      "step": 175
+    },
+    {
+      "epoch": 0.13692946058091288,
+      "grad_norm": 0.8183657293327374,
+      "learning_rate": 4.55958549222798e-06,
+      "loss": 0.506,
+      "step": 176
+    },
+    {
+      "epoch": 0.13770746887966806,
+      "grad_norm": 0.6649574979945729,
+      "learning_rate": 4.585492227979275e-06,
+      "loss": 0.5336,
+      "step": 177
+    },
+    {
+      "epoch": 0.13848547717842324,
+      "grad_norm": 0.6062588762204482,
+      "learning_rate": 4.61139896373057e-06,
+      "loss": 0.5568,
+      "step": 178
+    },
+    {
+      "epoch": 0.13926348547717843,
+      "grad_norm": 0.6976072584104026,
+      "learning_rate": 4.637305699481865e-06,
+      "loss": 0.5593,
+      "step": 179
+    },
+    {
+      "epoch": 0.1400414937759336,
+      "grad_norm": 0.799970304502699,
+      "learning_rate": 4.663212435233161e-06,
+      "loss": 0.5144,
+      "step": 180
+    },
+    {
+      "epoch": 0.1408195020746888,
+      "grad_norm": 0.747181756263449,
+      "learning_rate": 4.689119170984456e-06,
+      "loss": 0.5074,
+      "step": 181
+    },
+    {
+      "epoch": 0.14159751037344398,
+      "grad_norm": 0.6670330957509855,
+      "learning_rate": 4.715025906735752e-06,
+      "loss": 0.5202,
+      "step": 182
+    },
+    {
+      "epoch": 0.14237551867219916,
+      "grad_norm": 0.8682487440548523,
+      "learning_rate": 4.740932642487048e-06,
+      "loss": 0.517,
+      "step": 183
+    },
+    {
+      "epoch": 0.14315352697095435,
+      "grad_norm": 0.7212300278067795,
+      "learning_rate": 4.766839378238342e-06,
+      "loss": 0.5886,
+      "step": 184
+    },
+    {
+      "epoch": 0.14393153526970953,
+      "grad_norm": 0.6447026565731573,
+      "learning_rate": 4.7927461139896375e-06,
+      "loss": 0.5007,
+      "step": 185
+    },
+    {
+      "epoch": 0.14470954356846474,
+      "grad_norm": 0.6053235510777885,
+      "learning_rate": 4.818652849740933e-06,
+      "loss": 0.501,
+      "step": 186
+    },
+    {
+      "epoch": 0.14548755186721993,
+      "grad_norm": 0.6605850317764244,
+      "learning_rate": 4.844559585492228e-06,
+      "loss": 0.5319,
+      "step": 187
+    },
+    {
+      "epoch": 0.1462655601659751,
+      "grad_norm": 0.6513892977456262,
+      "learning_rate": 4.870466321243524e-06,
+      "loss": 0.5441,
+      "step": 188
+    },
+    {
+      "epoch": 0.1470435684647303,
+      "grad_norm": 0.7140564066086946,
+      "learning_rate": 4.896373056994819e-06,
+      "loss": 0.5194,
+      "step": 189
+    },
+    {
+      "epoch": 0.14782157676348548,
+      "grad_norm": 0.5927699466941649,
+      "learning_rate": 4.922279792746114e-06,
+      "loss": 0.5213,
+      "step": 190
+    },
+    {
+      "epoch": 0.14859958506224066,
+      "grad_norm": 0.7205413896804751,
+      "learning_rate": 4.94818652849741e-06,
+      "loss": 0.5366,
+      "step": 191
+    },
+    {
+      "epoch": 0.14937759336099585,
+      "grad_norm": 0.6054833789411084,
+      "learning_rate": 4.974093264248705e-06,
+      "loss": 0.4887,
+      "step": 192
+    },
+    {
+      "epoch": 0.15015560165975103,
+      "grad_norm": 0.6783527685315859,
+      "learning_rate": 5e-06,
+      "loss": 0.5392,
+      "step": 193
+    },
+    {
+      "epoch": 0.15093360995850622,
+      "grad_norm": 0.6093872881062327,
+      "learning_rate": 5.025906735751296e-06,
+      "loss": 0.4867,
+      "step": 194
+    },
+    {
+      "epoch": 0.1517116182572614,
+      "grad_norm": 0.7294217886497267,
+      "learning_rate": 5.051813471502591e-06,
+      "loss": 0.5544,
+      "step": 195
+    },
+    {
+      "epoch": 0.1524896265560166,
+      "grad_norm": 0.6763215409304602,
+      "learning_rate": 5.077720207253887e-06,
+      "loss": 0.521,
+      "step": 196
+    },
+    {
+      "epoch": 0.1532676348547718,
+      "grad_norm": 0.7589261150615131,
+      "learning_rate": 5.103626943005182e-06,
+      "loss": 0.5432,
+      "step": 197
+    },
+    {
+      "epoch": 0.15404564315352698,
+      "grad_norm": 0.6385484340997387,
+      "learning_rate": 5.129533678756478e-06,
+      "loss": 0.5067,
+      "step": 198
+    },
+    {
+      "epoch": 0.15482365145228216,
+      "grad_norm": 0.6763914569786738,
+      "learning_rate": 5.155440414507773e-06,
+      "loss": 0.5006,
+      "step": 199
+    },
+    {
+      "epoch": 0.15560165975103735,
+      "grad_norm": 0.7117538246155725,
+      "learning_rate": 5.1813471502590676e-06,
+      "loss": 0.5358,
+      "step": 200
+    },
+    {
+      "epoch": 0.15637966804979253,
+      "grad_norm": 0.6616225142818721,
+      "learning_rate": 5.2072538860103625e-06,
+      "loss": 0.5352,
+      "step": 201
+    },
+    {
+      "epoch": 0.15715767634854771,
+      "grad_norm": 0.6427807402055402,
+      "learning_rate": 5.233160621761658e-06,
+      "loss": 0.4901,
+      "step": 202
+    },
+    {
+      "epoch": 0.1579356846473029,
+      "grad_norm": 0.6345895578809747,
+      "learning_rate": 5.259067357512953e-06,
+      "loss": 0.5301,
+      "step": 203
+    },
+    {
+      "epoch": 0.15871369294605808,
+      "grad_norm": 0.6784309124357183,
+      "learning_rate": 5.284974093264249e-06,
+      "loss": 0.4897,
+      "step": 204
+    },
+    {
+      "epoch": 0.15949170124481327,
+      "grad_norm": 0.6566025184684325,
+      "learning_rate": 5.310880829015545e-06,
+      "loss": 0.4856,
+      "step": 205
+    },
+    {
+      "epoch": 0.16026970954356848,
+      "grad_norm": 0.6942676633246562,
+      "learning_rate": 5.33678756476684e-06,
+      "loss": 0.5529,
+      "step": 206
+    },
+    {
+      "epoch": 0.16104771784232366,
+      "grad_norm": 0.7206093233649432,
+      "learning_rate": 5.3626943005181356e-06,
+      "loss": 0.5207,
+      "step": 207
+    },
+    {
+      "epoch": 0.16182572614107885,
+      "grad_norm": 0.703021379565035,
+      "learning_rate": 5.3886010362694305e-06,
+      "loss": 0.5478,
+      "step": 208
+    },
+    {
+      "epoch": 0.16260373443983403,
+      "grad_norm": 0.7018795125173153,
+      "learning_rate": 5.414507772020726e-06,
+      "loss": 0.4937,
+      "step": 209
+    },
+    {
+      "epoch": 0.16338174273858921,
+      "grad_norm": 0.5911728178650582,
+      "learning_rate": 5.440414507772021e-06,
+      "loss": 0.4511,
+      "step": 210
+    },
+    {
+      "epoch": 0.1641597510373444,
+      "grad_norm": 0.6435369959956146,
+      "learning_rate": 5.466321243523317e-06,
+      "loss": 0.4788,
+      "step": 211
+    },
+    {
+      "epoch": 0.16493775933609958,
+      "grad_norm": 0.6475708880490412,
+      "learning_rate": 5.492227979274611e-06,
+      "loss": 0.4925,
+      "step": 212
+    },
+    {
+      "epoch": 0.16571576763485477,
+      "grad_norm": 0.6529736978416963,
+      "learning_rate": 5.518134715025907e-06,
+      "loss": 0.5325,
+      "step": 213
+    },
+    {
+      "epoch": 0.16649377593360995,
+      "grad_norm": 0.6182326914335551,
+      "learning_rate": 5.544041450777202e-06,
+      "loss": 0.4793,
+      "step": 214
+    },
+    {
+      "epoch": 0.16727178423236513,
+      "grad_norm": 0.6482535831269363,
+      "learning_rate": 5.569948186528498e-06,
+      "loss": 0.5215,
+      "step": 215
+    },
+    {
+      "epoch": 0.16804979253112035,
+      "grad_norm": 0.6508933490059878,
+      "learning_rate": 5.5958549222797934e-06,
+      "loss": 0.4948,
+      "step": 216
+    },
+    {
+      "epoch": 0.16882780082987553,
+      "grad_norm": 0.7173284582450842,
+      "learning_rate": 5.621761658031088e-06,
+      "loss": 0.4902,
+      "step": 217
+    },
+    {
+      "epoch": 0.1696058091286307,
+      "grad_norm": 0.6969152852533653,
+      "learning_rate": 5.647668393782384e-06,
+      "loss": 0.5082,
+      "step": 218
+    },
+    {
+      "epoch": 0.1703838174273859,
+      "grad_norm": 0.7707718652821987,
+      "learning_rate": 5.673575129533679e-06,
+      "loss": 0.5139,
+      "step": 219
+    },
+    {
+      "epoch": 0.17116182572614108,
+      "grad_norm": 0.7821698701333843,
+      "learning_rate": 5.699481865284975e-06,
+      "loss": 0.5308,
+      "step": 220
+    },
+    {
+      "epoch": 0.17193983402489627,
+      "grad_norm": 0.6248967957571196,
+      "learning_rate": 5.72538860103627e-06,
+      "loss": 0.5075,
+      "step": 221
+    },
+    {
+      "epoch": 0.17271784232365145,
+      "grad_norm": 0.7205185320492927,
+      "learning_rate": 5.751295336787566e-06,
+      "loss": 0.5117,
+      "step": 222
+    },
+    {
+      "epoch": 0.17349585062240663,
+      "grad_norm": 0.720462140751566,
+      "learning_rate": 5.7772020725388614e-06,
+      "loss": 0.5445,
+      "step": 223
+    },
+    {
+      "epoch": 0.17427385892116182,
+      "grad_norm": 0.6936616237565669,
+      "learning_rate": 5.8031088082901555e-06,
+      "loss": 0.5439,
+      "step": 224
+    },
+    {
+      "epoch": 0.175051867219917,
+      "grad_norm": 0.8240754607416161,
+      "learning_rate": 5.8290155440414505e-06,
+      "loss": 0.5465,
+      "step": 225
+    },
+    {
+      "epoch": 0.1758298755186722,
+      "grad_norm": 0.7031179596715919,
+      "learning_rate": 5.854922279792746e-06,
+      "loss": 0.5232,
+      "step": 226
+    },
+    {
+      "epoch": 0.1766078838174274,
+      "grad_norm": 0.8321019671527347,
+      "learning_rate": 5.880829015544042e-06,
+      "loss": 0.5391,
+      "step": 227
+    },
+    {
+      "epoch": 0.17738589211618258,
+      "grad_norm": 0.6777935280377142,
+      "learning_rate": 5.906735751295337e-06,
+      "loss": 0.4805,
+      "step": 228
+    },
+    {
+      "epoch": 0.17816390041493776,
+      "grad_norm": 0.6205267959225454,
+      "learning_rate": 5.932642487046633e-06,
+      "loss": 0.484,
+      "step": 229
+    },
+    {
+      "epoch": 0.17894190871369295,
+      "grad_norm": 0.6368978888500513,
+      "learning_rate": 5.958549222797928e-06,
+      "loss": 0.4782,
+      "step": 230
+    },
+    {
+      "epoch": 0.17971991701244813,
+      "grad_norm": 0.7096406850327143,
+      "learning_rate": 5.9844559585492235e-06,
+      "loss": 0.5005,
+      "step": 231
+    },
+    {
+      "epoch": 0.18049792531120332,
+      "grad_norm": 0.6951348296169272,
+      "learning_rate": 6.0103626943005185e-06,
+      "loss": 0.5048,
+      "step": 232
+    },
+    {
+      "epoch": 0.1812759336099585,
+      "grad_norm": 0.6417761035180208,
+      "learning_rate": 6.036269430051814e-06,
+      "loss": 0.5432,
+      "step": 233
+    },
+    {
+      "epoch": 0.18205394190871368,
+      "grad_norm": 0.6185580434712037,
+      "learning_rate": 6.06217616580311e-06,
+      "loss": 0.4626,
+      "step": 234
+    },
+    {
+      "epoch": 0.18283195020746887,
+      "grad_norm": 0.7081019384804975,
+      "learning_rate": 6.088082901554405e-06,
+      "loss": 0.4693,
+      "step": 235
+    },
+    {
+      "epoch": 0.18360995850622408,
+      "grad_norm": 0.6508380028543874,
+      "learning_rate": 6.113989637305699e-06,
+      "loss": 0.5166,
+      "step": 236
+    },
+    {
+      "epoch": 0.18438796680497926,
+      "grad_norm": 0.7885579968103317,
+      "learning_rate": 6.139896373056995e-06,
+      "loss": 0.538,
+      "step": 237
+    },
+    {
+      "epoch": 0.18516597510373445,
+      "grad_norm": 0.7138888593917573,
+      "learning_rate": 6.165803108808291e-06,
+      "loss": 0.5041,
+      "step": 238
+    },
+    {
+      "epoch": 0.18594398340248963,
+      "grad_norm": 0.7040121666065993,
+      "learning_rate": 6.191709844559586e-06,
+      "loss": 0.5337,
+      "step": 239
+    },
+    {
+      "epoch": 0.18672199170124482,
+      "grad_norm": 0.6226846647348617,
+      "learning_rate": 6.217616580310881e-06,
+      "loss": 0.504,
+      "step": 240
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 0.7488967072258025,
+      "learning_rate": 6.243523316062176e-06,
+      "loss": 0.558,
+      "step": 241
+    },
+    {
+      "epoch": 0.18827800829875518,
+      "grad_norm": 0.6841569869424846,
+      "learning_rate": 6.269430051813472e-06,
+      "loss": 0.4842,
+      "step": 242
+    },
+    {
+      "epoch": 0.18905601659751037,
+      "grad_norm": 0.7064757777380942,
+      "learning_rate": 6.295336787564768e-06,
+      "loss": 0.4963,
+      "step": 243
+    },
+    {
+      "epoch": 0.18983402489626555,
+      "grad_norm": 0.653955784257911,
+      "learning_rate": 6.321243523316063e-06,
+      "loss": 0.5024,
+      "step": 244
+    },
+    {
+      "epoch": 0.19061203319502074,
+      "grad_norm": 0.7631400410011139,
+      "learning_rate": 6.347150259067359e-06,
+      "loss": 0.5358,
+      "step": 245
+    },
+    {
+      "epoch": 0.19139004149377592,
+      "grad_norm": 0.7967240131444799,
+      "learning_rate": 6.373056994818654e-06,
+      "loss": 0.5304,
+      "step": 246
+    },
+    {
+      "epoch": 0.19216804979253113,
+      "grad_norm": 0.7032352980745357,
+      "learning_rate": 6.398963730569949e-06,
+      "loss": 0.5427,
+      "step": 247
+    },
+    {
+      "epoch": 0.19294605809128632,
+      "grad_norm": 0.6861608267108283,
+      "learning_rate": 6.4248704663212435e-06,
+      "loss": 0.4858,
+      "step": 248
+    },
+    {
+      "epoch": 0.1937240663900415,
+      "grad_norm": 0.7361250824218776,
+      "learning_rate": 6.450777202072539e-06,
+      "loss": 0.5172,
+      "step": 249
+    },
+    {
+      "epoch": 0.19450207468879668,
+      "grad_norm": 0.9094306929430372,
+      "learning_rate": 6.476683937823834e-06,
+      "loss": 0.5224,
+      "step": 250
+    },
+    {
+      "epoch": 0.19528008298755187,
+      "grad_norm": 0.6545818356449009,
+      "learning_rate": 6.50259067357513e-06,
+      "loss": 0.4852,
+      "step": 251
+    },
+    {
+      "epoch": 0.19605809128630705,
+      "grad_norm": 0.7050472920407246,
+      "learning_rate": 6.528497409326425e-06,
+      "loss": 0.5027,
+      "step": 252
+    },
+    {
+      "epoch": 0.19683609958506224,
+      "grad_norm": 0.9133128724096569,
+      "learning_rate": 6.554404145077721e-06,
+      "loss": 0.5116,
+      "step": 253
+    },
+    {
+      "epoch": 0.19761410788381742,
+      "grad_norm": 0.6895247277177783,
+      "learning_rate": 6.5803108808290166e-06,
+      "loss": 0.5031,
+      "step": 254
+    },
+    {
+      "epoch": 0.1983921161825726,
+      "grad_norm": 0.6603599282756638,
+      "learning_rate": 6.6062176165803115e-06,
+      "loss": 0.4908,
+      "step": 255
+    },
+    {
+      "epoch": 0.1991701244813278,
+      "grad_norm": 0.7385641401628485,
+      "learning_rate": 6.632124352331607e-06,
+      "loss": 0.4962,
+      "step": 256
+    },
+    {
+      "epoch": 0.199948132780083,
+      "grad_norm": 0.7819901901268208,
+      "learning_rate": 6.658031088082902e-06,
+      "loss": 0.5263,
+      "step": 257
+    },
+    {
+      "epoch": 0.20072614107883818,
+      "grad_norm": 0.694702916817804,
+      "learning_rate": 6.683937823834198e-06,
+      "loss": 0.5123,
+      "step": 258
+    },
+    {
+      "epoch": 0.20150414937759337,
+      "grad_norm": 0.7621890872810972,
+      "learning_rate": 6.709844559585493e-06,
+      "loss": 0.5217,
+      "step": 259
+    },
+    {
+      "epoch": 0.20228215767634855,
+      "grad_norm": 0.7792956190717889,
+      "learning_rate": 6.735751295336788e-06,
+      "loss": 0.553,
+      "step": 260
+    },
+    {
+      "epoch": 0.20306016597510373,
+      "grad_norm": 0.6913258813284858,
+      "learning_rate": 6.761658031088083e-06,
+      "loss": 0.4993,
+      "step": 261
+    },
+    {
+      "epoch": 0.20383817427385892,
+      "grad_norm": 0.7527093045988372,
+      "learning_rate": 6.787564766839379e-06,
+      "loss": 0.514,
+      "step": 262
+    },
+    {
+      "epoch": 0.2046161825726141,
+      "grad_norm": 0.7191759021316405,
+      "learning_rate": 6.813471502590674e-06,
+      "loss": 0.527,
+      "step": 263
+    },
+    {
+      "epoch": 0.2053941908713693,
+      "grad_norm": 0.6609399239077988,
+      "learning_rate": 6.839378238341969e-06,
+      "loss": 0.5003,
+      "step": 264
+    },
+    {
+      "epoch": 0.20617219917012447,
+      "grad_norm": 0.8686218851022207,
+      "learning_rate": 6.865284974093265e-06,
+      "loss": 0.5539,
+      "step": 265
+    },
+    {
+      "epoch": 0.20695020746887965,
+      "grad_norm": 0.6510989934359599,
+      "learning_rate": 6.89119170984456e-06,
+      "loss": 0.5016,
+      "step": 266
+    },
+    {
+      "epoch": 0.20772821576763487,
+      "grad_norm": 0.8061736886187104,
+      "learning_rate": 6.917098445595856e-06,
+      "loss": 0.5184,
+      "step": 267
+    },
+    {
+      "epoch": 0.20850622406639005,
+      "grad_norm": 0.7020448208732175,
+      "learning_rate": 6.943005181347151e-06,
+      "loss": 0.5411,
+      "step": 268
+    },
+    {
+      "epoch": 0.20928423236514523,
+      "grad_norm": 0.6723724619191563,
+      "learning_rate": 6.968911917098447e-06,
+      "loss": 0.4844,
+      "step": 269
+    },
+    {
+      "epoch": 0.21006224066390042,
+      "grad_norm": 0.7028816660530403,
+      "learning_rate": 6.994818652849742e-06,
+      "loss": 0.5205,
+      "step": 270
+    },
+    {
+      "epoch": 0.2108402489626556,
+      "grad_norm": 0.6865811110766856,
+      "learning_rate": 7.020725388601037e-06,
+      "loss": 0.4869,
+      "step": 271
+    },
+    {
+      "epoch": 0.21161825726141079,
+      "grad_norm": 0.7646319322658366,
+      "learning_rate": 7.0466321243523315e-06,
+      "loss": 0.5269,
+      "step": 272
+    },
+    {
+      "epoch": 0.21239626556016597,
+      "grad_norm": 0.6175637509850409,
+      "learning_rate": 7.072538860103627e-06,
+      "loss": 0.4989,
+      "step": 273
+    },
+    {
+      "epoch": 0.21317427385892115,
+      "grad_norm": 0.6822910024262075,
+      "learning_rate": 7.098445595854922e-06,
+      "loss": 0.5533,
+      "step": 274
+    },
+    {
+      "epoch": 0.21395228215767634,
+      "grad_norm": 0.6225698401209366,
+      "learning_rate": 7.124352331606218e-06,
+      "loss": 0.4861,
+      "step": 275
+    },
+    {
+      "epoch": 0.21473029045643152,
+      "grad_norm": 0.8133289220357584,
+      "learning_rate": 7.150259067357514e-06,
+      "loss": 0.4917,
+      "step": 276
+    },
+    {
+      "epoch": 0.21550829875518673,
+      "grad_norm": 0.7153606376826523,
+      "learning_rate": 7.176165803108809e-06,
+      "loss": 0.4876,
+      "step": 277
+    },
+    {
+      "epoch": 0.21628630705394192,
+      "grad_norm": 0.6792501882846864,
+      "learning_rate": 7.2020725388601045e-06,
+      "loss": 0.4925,
+      "step": 278
+    },
+    {
+      "epoch": 0.2170643153526971,
+      "grad_norm": 0.6690862271598986,
+      "learning_rate": 7.2279792746113995e-06,
+      "loss": 0.4977,
+      "step": 279
+    },
+    {
+      "epoch": 0.21784232365145229,
+      "grad_norm": 0.6714243576824012,
+      "learning_rate": 7.253886010362695e-06,
+      "loss": 0.5118,
+      "step": 280
+    },
+    {
+      "epoch": 0.21862033195020747,
+      "grad_norm": 0.7542255514786808,
+      "learning_rate": 7.27979274611399e-06,
+      "loss": 0.5028,
+      "step": 281
+    },
+    {
+      "epoch": 0.21939834024896265,
+      "grad_norm": 0.6758011331146274,
+      "learning_rate": 7.305699481865286e-06,
+      "loss": 0.5347,
+      "step": 282
+    },
+    {
+      "epoch": 0.22017634854771784,
+      "grad_norm": 0.6991994991040403,
+      "learning_rate": 7.331606217616582e-06,
+      "loss": 0.496,
+      "step": 283
+    },
+    {
+      "epoch": 0.22095435684647302,
+      "grad_norm": 0.7370920448247584,
+      "learning_rate": 7.357512953367876e-06,
+      "loss": 0.483,
+      "step": 284
+    },
+    {
+      "epoch": 0.2217323651452282,
+      "grad_norm": 0.6939777610993035,
+      "learning_rate": 7.383419689119171e-06,
+      "loss": 0.4922,
+      "step": 285
+    },
+    {
+      "epoch": 0.2225103734439834,
+      "grad_norm": 0.6501622233048342,
+      "learning_rate": 7.409326424870467e-06,
+      "loss": 0.5051,
+      "step": 286
+    },
+    {
+      "epoch": 0.2232883817427386,
+      "grad_norm": 0.9074622140562277,
+      "learning_rate": 7.435233160621762e-06,
+      "loss": 0.5487,
+      "step": 287
+    },
+    {
+      "epoch": 0.22406639004149378,
+      "grad_norm": 0.777626395099288,
+      "learning_rate": 7.461139896373057e-06,
+      "loss": 0.4751,
+      "step": 288
+    },
+    {
+      "epoch": 0.22484439834024897,
+      "grad_norm": 0.7377934168160158,
+      "learning_rate": 7.487046632124353e-06,
+      "loss": 0.4806,
+      "step": 289
+    },
+    {
+      "epoch": 0.22562240663900415,
+      "grad_norm": 0.6969781412651022,
+      "learning_rate": 7.512953367875648e-06,
+      "loss": 0.4575,
+      "step": 290
+    },
+    {
+      "epoch": 0.22640041493775934,
+      "grad_norm": 0.8493229555992157,
+      "learning_rate": 7.538860103626944e-06,
+      "loss": 0.5054,
+      "step": 291
+    },
+    {
+      "epoch": 0.22717842323651452,
+      "grad_norm": 0.7712316162006523,
+      "learning_rate": 7.564766839378239e-06,
+      "loss": 0.519,
+      "step": 292
+    },
+    {
+      "epoch": 0.2279564315352697,
+      "grad_norm": 0.6419655455765019,
+      "learning_rate": 7.590673575129535e-06,
+      "loss": 0.4804,
+      "step": 293
+    },
+    {
+      "epoch": 0.2287344398340249,
+      "grad_norm": 0.7928546237512729,
+      "learning_rate": 7.61658031088083e-06,
+      "loss": 0.4809,
+      "step": 294
+    },
+    {
+      "epoch": 0.22951244813278007,
+      "grad_norm": 0.8361464375156841,
+      "learning_rate": 7.642487046632126e-06,
+      "loss": 0.5058,
+      "step": 295
+    },
+    {
+      "epoch": 0.23029045643153526,
+      "grad_norm": 0.692878149227977,
+      "learning_rate": 7.66839378238342e-06,
+      "loss": 0.4541,
+      "step": 296
+    },
+    {
+      "epoch": 0.23106846473029047,
+      "grad_norm": 0.7741285238399388,
+      "learning_rate": 7.694300518134716e-06,
+      "loss": 0.5085,
+      "step": 297
+    },
+    {
+      "epoch": 0.23184647302904565,
+      "grad_norm": 0.7077306857429437,
+      "learning_rate": 7.72020725388601e-06,
+      "loss": 0.5264,
+      "step": 298
+    },
+    {
+      "epoch": 0.23262448132780084,
+      "grad_norm": 0.7506969542550662,
+      "learning_rate": 7.746113989637306e-06,
+      "loss": 0.4724,
+      "step": 299
+    },
+    {
+      "epoch": 0.23340248962655602,
+      "grad_norm": 0.63678357936984,
+      "learning_rate": 7.772020725388602e-06,
+      "loss": 0.4658,
+      "step": 300
+    },
+    {
+      "epoch": 0.2341804979253112,
+      "grad_norm": 0.7165627191542565,
+      "learning_rate": 7.797927461139898e-06,
+      "loss": 0.5147,
+      "step": 301
+    },
+    {
+      "epoch": 0.2349585062240664,
+      "grad_norm": 0.645851322879614,
+      "learning_rate": 7.823834196891192e-06,
+      "loss": 0.4508,
+      "step": 302
+    },
+    {
+      "epoch": 0.23573651452282157,
+      "grad_norm": 0.8362682305342171,
+      "learning_rate": 7.849740932642487e-06,
+      "loss": 0.4576,
+      "step": 303
+    },
+    {
+      "epoch": 0.23651452282157676,
+      "grad_norm": 0.7707812213156408,
+      "learning_rate": 7.875647668393783e-06,
+      "loss": 0.5184,
+      "step": 304
+    },
+    {
+      "epoch": 0.23729253112033194,
+      "grad_norm": 0.6390963624244069,
+      "learning_rate": 7.901554404145079e-06,
+      "loss": 0.4761,
+      "step": 305
+    },
+    {
+      "epoch": 0.23807053941908712,
+      "grad_norm": 0.7657573461580098,
+      "learning_rate": 7.927461139896375e-06,
+      "loss": 0.5647,
+      "step": 306
+    },
+    {
+      "epoch": 0.23884854771784234,
+      "grad_norm": 0.7581538956079235,
+      "learning_rate": 7.953367875647669e-06,
+      "loss": 0.4772,
+      "step": 307
+    },
+    {
+      "epoch": 0.23962655601659752,
+      "grad_norm": 0.7803750342133872,
+      "learning_rate": 7.979274611398965e-06,
+      "loss": 0.5257,
+      "step": 308
+    },
+    {
+      "epoch": 0.2404045643153527,
+      "grad_norm": 0.6861626062044475,
+      "learning_rate": 8.005181347150259e-06,
+      "loss": 0.4873,
+      "step": 309
+    },
+    {
+      "epoch": 0.2411825726141079,
+      "grad_norm": 0.8081957471853651,
+      "learning_rate": 8.031088082901555e-06,
+      "loss": 0.5036,
+      "step": 310
+    },
+    {
+      "epoch": 0.24196058091286307,
+      "grad_norm": 0.6453951466859804,
+      "learning_rate": 8.05699481865285e-06,
+      "loss": 0.4772,
+      "step": 311
+    },
+    {
+      "epoch": 0.24273858921161826,
+      "grad_norm": 0.7494456578945177,
+      "learning_rate": 8.082901554404146e-06,
+      "loss": 0.503,
+      "step": 312
+    },
+    {
+      "epoch": 0.24351659751037344,
+      "grad_norm": 0.6963102182625279,
+      "learning_rate": 8.10880829015544e-06,
+      "loss": 0.5047,
+      "step": 313
+    },
+    {
+      "epoch": 0.24429460580912862,
+      "grad_norm": 0.7945598936304281,
+      "learning_rate": 8.134715025906736e-06,
+      "loss": 0.5529,
+      "step": 314
+    },
+    {
+      "epoch": 0.2450726141078838,
+      "grad_norm": 0.8058410213185938,
+      "learning_rate": 8.160621761658032e-06,
+      "loss": 0.4779,
+      "step": 315
+    },
+    {
+      "epoch": 0.245850622406639,
+      "grad_norm": 0.6628804293970346,
+      "learning_rate": 8.186528497409328e-06,
+      "loss": 0.4615,
+      "step": 316
+    },
+    {
+      "epoch": 0.2466286307053942,
+      "grad_norm": 0.6341983874151081,
+      "learning_rate": 8.212435233160623e-06,
+      "loss": 0.4719,
+      "step": 317
+    },
+    {
+      "epoch": 0.2474066390041494,
+      "grad_norm": 0.7720104691151563,
+      "learning_rate": 8.238341968911918e-06,
+      "loss": 0.4626,
+      "step": 318
+    },
+    {
+      "epoch": 0.24818464730290457,
+      "grad_norm": 0.7198721190202337,
+      "learning_rate": 8.264248704663213e-06,
+      "loss": 0.4897,
+      "step": 319
+    },
+    {
+      "epoch": 0.24896265560165975,
+      "grad_norm": 0.6312974949559387,
+      "learning_rate": 8.290155440414507e-06,
+      "loss": 0.4693,
+      "step": 320
+    },
+    {
+      "epoch": 0.24974066390041494,
+      "grad_norm": 0.726928053328491,
+      "learning_rate": 8.316062176165803e-06,
+      "loss": 0.4722,
+      "step": 321
+    },
+    {
+      "epoch": 0.2505186721991701,
+      "grad_norm": 0.7889739371590359,
+      "learning_rate": 8.341968911917099e-06,
+      "loss": 0.4839,
+      "step": 322
+    },
+    {
+      "epoch": 0.2512966804979253,
+      "grad_norm": 0.6021286083556413,
+      "learning_rate": 8.367875647668395e-06,
+      "loss": 0.4994,
+      "step": 323
+    },
+    {
+      "epoch": 0.2520746887966805,
+      "grad_norm": 0.8627146587078978,
+      "learning_rate": 8.393782383419689e-06,
+      "loss": 0.4908,
+      "step": 324
+    },
+    {
+      "epoch": 0.2528526970954357,
+      "grad_norm": 0.7465727401005439,
+      "learning_rate": 8.419689119170985e-06,
+      "loss": 0.4998,
+      "step": 325
+    },
+    {
+      "epoch": 0.25363070539419086,
+      "grad_norm": 0.7159955936287686,
+      "learning_rate": 8.44559585492228e-06,
+      "loss": 0.4944,
+      "step": 326
+    },
+    {
+      "epoch": 0.25440871369294604,
+      "grad_norm": 0.7817848031937663,
+      "learning_rate": 8.471502590673576e-06,
+      "loss": 0.5182,
+      "step": 327
+    },
+    {
+      "epoch": 0.2551867219917012,
+      "grad_norm": 0.9116899056587449,
+      "learning_rate": 8.497409326424872e-06,
+      "loss": 0.5003,
+      "step": 328
+    },
+    {
+      "epoch": 0.2559647302904564,
+      "grad_norm": 0.7125951683917336,
+      "learning_rate": 8.523316062176166e-06,
+      "loss": 0.495,
+      "step": 329
+    },
+    {
+      "epoch": 0.2567427385892116,
+      "grad_norm": 0.7865158164894225,
+      "learning_rate": 8.549222797927462e-06,
+      "loss": 0.5377,
+      "step": 330
+    },
+    {
+      "epoch": 0.2575207468879668,
+      "grad_norm": 0.7153205949490207,
+      "learning_rate": 8.575129533678758e-06,
+      "loss": 0.4625,
+      "step": 331
+    },
+    {
+      "epoch": 0.258298755186722,
+      "grad_norm": 0.7044789916763718,
+      "learning_rate": 8.601036269430052e-06,
+      "loss": 0.502,
+      "step": 332
+    },
+    {
+      "epoch": 0.2590767634854772,
+      "grad_norm": 0.6528642523307281,
+      "learning_rate": 8.626943005181348e-06,
+      "loss": 0.46,
+      "step": 333
+    },
+    {
+      "epoch": 0.2598547717842324,
+      "grad_norm": 0.6390643435461368,
+      "learning_rate": 8.652849740932643e-06,
+      "loss": 0.5,
+      "step": 334
+    },
+    {
+      "epoch": 0.26063278008298757,
+      "grad_norm": 0.7348798654672425,
+      "learning_rate": 8.678756476683938e-06,
+      "loss": 0.4979,
+      "step": 335
+    },
+    {
+      "epoch": 0.26141078838174275,
+      "grad_norm": 0.6800203302881397,
+      "learning_rate": 8.704663212435233e-06,
+      "loss": 0.4742,
+      "step": 336
+    },
+    {
+      "epoch": 0.26218879668049794,
+      "grad_norm": 0.6919917832892066,
+      "learning_rate": 8.730569948186529e-06,
+      "loss": 0.5157,
+      "step": 337
+    },
+    {
+      "epoch": 0.2629668049792531,
+      "grad_norm": 0.7652512548203049,
+      "learning_rate": 8.756476683937825e-06,
+      "loss": 0.5478,
+      "step": 338
+    },
+    {
+      "epoch": 0.2637448132780083,
+      "grad_norm": 0.7117363010858194,
+      "learning_rate": 8.78238341968912e-06,
+      "loss": 0.4958,
+      "step": 339
+    },
+    {
+      "epoch": 0.2645228215767635,
+      "grad_norm": 0.8430432887750338,
+      "learning_rate": 8.808290155440415e-06,
+      "loss": 0.5014,
+      "step": 340
+    },
+    {
+      "epoch": 0.2653008298755187,
+      "grad_norm": 0.7415290151255904,
+      "learning_rate": 8.83419689119171e-06,
+      "loss": 0.4657,
+      "step": 341
+    },
+    {
+      "epoch": 0.26607883817427386,
+      "grad_norm": 0.781477352059538,
+      "learning_rate": 8.860103626943006e-06,
+      "loss": 0.4989,
+      "step": 342
+    },
+    {
+      "epoch": 0.26685684647302904,
+      "grad_norm": 0.7849821674411382,
+      "learning_rate": 8.886010362694302e-06,
+      "loss": 0.5358,
+      "step": 343
+    },
+    {
+      "epoch": 0.2676348547717842,
+      "grad_norm": 0.6707644115480477,
+      "learning_rate": 8.911917098445596e-06,
+      "loss": 0.4336,
+      "step": 344
+    },
+    {
+      "epoch": 0.2684128630705394,
+      "grad_norm": 0.7865952240646091,
+      "learning_rate": 8.937823834196892e-06,
+      "loss": 0.5575,
+      "step": 345
+    },
+    {
+      "epoch": 0.2691908713692946,
+      "grad_norm": 0.6906622596970327,
+      "learning_rate": 8.963730569948186e-06,
+      "loss": 0.473,
+      "step": 346
+    },
+    {
+      "epoch": 0.2699688796680498,
+      "grad_norm": 0.675878315912764,
+      "learning_rate": 8.989637305699482e-06,
+      "loss": 0.495,
+      "step": 347
+    },
+    {
+      "epoch": 0.27074688796680496,
+      "grad_norm": 0.6745466525824272,
+      "learning_rate": 9.015544041450778e-06,
+      "loss": 0.5035,
+      "step": 348
+    },
+    {
+      "epoch": 0.27152489626556015,
+      "grad_norm": 0.7092771145168886,
+      "learning_rate": 9.041450777202073e-06,
+      "loss": 0.4743,
+      "step": 349
+    },
+    {
+      "epoch": 0.27230290456431533,
+      "grad_norm": 0.6641046910967568,
+      "learning_rate": 9.06735751295337e-06,
+      "loss": 0.4758,
+      "step": 350
+    },
+    {
+      "epoch": 0.2730809128630705,
+      "grad_norm": 0.5897562426173721,
+      "learning_rate": 9.093264248704663e-06,
+      "loss": 0.4825,
+      "step": 351
+    },
+    {
+      "epoch": 0.27385892116182575,
+      "grad_norm": 0.6774127670418412,
+      "learning_rate": 9.11917098445596e-06,
+      "loss": 0.5101,
+      "step": 352
+    },
+    {
+      "epoch": 0.27463692946058094,
+      "grad_norm": 0.7361000542419546,
+      "learning_rate": 9.145077720207255e-06,
+      "loss": 0.5014,
+      "step": 353
+    },
+    {
+      "epoch": 0.2754149377593361,
+      "grad_norm": 0.6180894181067251,
+      "learning_rate": 9.17098445595855e-06,
+      "loss": 0.4561,
+      "step": 354
+    },
+    {
+      "epoch": 0.2761929460580913,
+      "grad_norm": 0.6822213978612627,
+      "learning_rate": 9.196891191709847e-06,
+      "loss": 0.4879,
+      "step": 355
+    },
+    {
+      "epoch": 0.2769709543568465,
+      "grad_norm": 0.7058157604308153,
+      "learning_rate": 9.22279792746114e-06,
+      "loss": 0.5218,
+      "step": 356
+    },
+    {
+      "epoch": 0.27774896265560167,
+      "grad_norm": 0.7620157956178831,
+      "learning_rate": 9.248704663212435e-06,
+      "loss": 0.492,
+      "step": 357
+    },
+    {
+      "epoch": 0.27852697095435686,
+      "grad_norm": 0.661563967335311,
+      "learning_rate": 9.27461139896373e-06,
+      "loss": 0.5134,
+      "step": 358
+    },
+    {
+      "epoch": 0.27930497925311204,
+      "grad_norm": 0.7196004549552166,
+      "learning_rate": 9.300518134715026e-06,
+      "loss": 0.5045,
+      "step": 359
+    },
+    {
+      "epoch": 0.2800829875518672,
+      "grad_norm": 0.6963434670523871,
+      "learning_rate": 9.326424870466322e-06,
+      "loss": 0.4844,
+      "step": 360
+    },
+    {
+      "epoch": 0.2808609958506224,
+      "grad_norm": 0.6352492437919303,
+      "learning_rate": 9.352331606217618e-06,
+      "loss": 0.4578,
+      "step": 361
+    },
+    {
+      "epoch": 0.2816390041493776,
+      "grad_norm": 0.6590954062596068,
+      "learning_rate": 9.378238341968912e-06,
+      "loss": 0.5065,
+      "step": 362
+    },
+    {
+      "epoch": 0.2824170124481328,
+      "grad_norm": 0.7167764176330585,
+      "learning_rate": 9.404145077720208e-06,
+      "loss": 0.474,
+      "step": 363
+    },
+    {
+      "epoch": 0.28319502074688796,
+      "grad_norm": 0.7242653294912278,
+      "learning_rate": 9.430051813471504e-06,
+      "loss": 0.5109,
+      "step": 364
+    },
+    {
+      "epoch": 0.28397302904564314,
+      "grad_norm": 0.6144536519152191,
+      "learning_rate": 9.4559585492228e-06,
+      "loss": 0.4483,
+      "step": 365
+    },
+    {
+      "epoch": 0.28475103734439833,
+      "grad_norm": 0.5676537600994896,
+      "learning_rate": 9.481865284974095e-06,
+      "loss": 0.4605,
+      "step": 366
+    },
+    {
+      "epoch": 0.2855290456431535,
+      "grad_norm": 0.6433629995473887,
+      "learning_rate": 9.50777202072539e-06,
+      "loss": 0.4811,
+      "step": 367
+    },
+    {
+      "epoch": 0.2863070539419087,
+      "grad_norm": 0.7246192070040728,
+      "learning_rate": 9.533678756476683e-06,
+      "loss": 0.5413,
+      "step": 368
+    },
+    {
+      "epoch": 0.2870850622406639,
+      "grad_norm": 0.741046895020594,
+      "learning_rate": 9.559585492227979e-06,
+      "loss": 0.5687,
+      "step": 369
+    },
+    {
+      "epoch": 0.28786307053941906,
+      "grad_norm": 0.6809915572151082,
+      "learning_rate": 9.585492227979275e-06,
+      "loss": 0.5312,
+      "step": 370
+    },
+    {
+      "epoch": 0.28864107883817425,
+      "grad_norm": 0.7605418082980792,
+      "learning_rate": 9.61139896373057e-06,
+      "loss": 0.5436,
+      "step": 371
+    },
+    {
+      "epoch": 0.2894190871369295,
+      "grad_norm": 0.6540849963572976,
+      "learning_rate": 9.637305699481867e-06,
+      "loss": 0.4864,
+      "step": 372
+    },
+    {
+      "epoch": 0.29019709543568467,
+      "grad_norm": 0.6606894395875399,
+      "learning_rate": 9.66321243523316e-06,
+      "loss": 0.4899,
+      "step": 373
+    },
+    {
+      "epoch": 0.29097510373443985,
+      "grad_norm": 0.6882900762353602,
+      "learning_rate": 9.689119170984456e-06,
+      "loss": 0.5123,
+      "step": 374
+    },
+    {
+      "epoch": 0.29175311203319504,
+      "grad_norm": 0.6947256789015608,
+      "learning_rate": 9.715025906735752e-06,
+      "loss": 0.4905,
+      "step": 375
+    },
+    {
+      "epoch": 0.2925311203319502,
+      "grad_norm": 0.7354981240787,
+      "learning_rate": 9.740932642487048e-06,
+      "loss": 0.4801,
+      "step": 376
+    },
+    {
+      "epoch": 0.2933091286307054,
+      "grad_norm": 0.6204479060266902,
+      "learning_rate": 9.766839378238344e-06,
+      "loss": 0.474,
+      "step": 377
+    },
+    {
+      "epoch": 0.2940871369294606,
+      "grad_norm": 0.7037366540868862,
+      "learning_rate": 9.792746113989638e-06,
+      "loss": 0.5007,
+      "step": 378
+    },
+    {
+      "epoch": 0.2948651452282158,
+      "grad_norm": 0.6460040832293884,
+      "learning_rate": 9.818652849740934e-06,
+      "loss": 0.4432,
+      "step": 379
+    },
+    {
+      "epoch": 0.29564315352697096,
+      "grad_norm": 0.6145735522811818,
+      "learning_rate": 9.844559585492228e-06,
+      "loss": 0.4751,
+      "step": 380
+    },
+    {
+      "epoch": 0.29642116182572614,
+      "grad_norm": 0.6005912810893309,
+      "learning_rate": 9.870466321243524e-06,
+      "loss": 0.4853,
+      "step": 381
+    },
+    {
+      "epoch": 0.2971991701244813,
+      "grad_norm": 0.7525349516494815,
+      "learning_rate": 9.89637305699482e-06,
+      "loss": 0.5571,
+      "step": 382
+    },
+    {
+      "epoch": 0.2979771784232365,
+      "grad_norm": 0.7067720305114756,
+      "learning_rate": 9.922279792746115e-06,
+      "loss": 0.485,
+      "step": 383
+    },
+    {
+      "epoch": 0.2987551867219917,
+      "grad_norm": 0.6717944552881441,
+      "learning_rate": 9.94818652849741e-06,
+      "loss": 0.4702,
+      "step": 384
+    },
+    {
+      "epoch": 0.2995331950207469,
+      "grad_norm": 0.728007694313093,
+      "learning_rate": 9.974093264248705e-06,
+      "loss": 0.5008,
+      "step": 385
+    },
+    {
+      "epoch": 0.30031120331950206,
+      "grad_norm": 0.7035447683484868,
+      "learning_rate": 1e-05,
+      "loss": 0.4517,
+      "step": 386
+    },
+    {
+      "epoch": 0.30108921161825725,
+      "grad_norm": 0.7263675922462193,
+      "learning_rate": 9.99999794963526e-06,
+      "loss": 0.4903,
+      "step": 387
+    },
+    {
+      "epoch": 0.30186721991701243,
+      "grad_norm": 0.6218890783199106,
+      "learning_rate": 9.999991798542723e-06,
+      "loss": 0.4756,
+      "step": 388
+    },
+    {
+      "epoch": 0.3026452282157676,
+      "grad_norm": 0.7017988098258009,
+      "learning_rate": 9.999981546727432e-06,
+      "loss": 0.4597,
+      "step": 389
+    },
+    {
+      "epoch": 0.3034232365145228,
+      "grad_norm": 0.6829358335339049,
+      "learning_rate": 9.999967194197793e-06,
+      "loss": 0.482,
+      "step": 390
+    },
+    {
+      "epoch": 0.304201244813278,
+      "grad_norm": 0.6519241463221046,
+      "learning_rate": 9.999948740965581e-06,
+      "loss": 0.489,
+      "step": 391
+    },
+    {
+      "epoch": 0.3049792531120332,
+      "grad_norm": 0.6234598553401526,
+      "learning_rate": 9.99992618704593e-06,
+      "loss": 0.481,
+      "step": 392
+    },
+    {
+      "epoch": 0.3057572614107884,
+      "grad_norm": 0.6842266222992247,
+      "learning_rate": 9.999899532457336e-06,
+      "loss": 0.4705,
+      "step": 393
+    },
+    {
+      "epoch": 0.3065352697095436,
+      "grad_norm": 0.5883256338332109,
+      "learning_rate": 9.99986877722166e-06,
+      "loss": 0.4663,
+      "step": 394
+    },
+    {
+      "epoch": 0.3073132780082988,
+      "grad_norm": 0.6360675009282013,
+      "learning_rate": 9.999833921364126e-06,
+      "loss": 0.4736,
+      "step": 395
+    },
+    {
+      "epoch": 0.30809128630705396,
+      "grad_norm": 0.5810228982040917,
+      "learning_rate": 9.99979496491332e-06,
+      "loss": 0.4383,
+      "step": 396
+    },
+    {
+      "epoch": 0.30886929460580914,
+      "grad_norm": 0.6329066021904639,
+      "learning_rate": 9.999751907901194e-06,
+      "loss": 0.4625,
+      "step": 397
+    },
+    {
+      "epoch": 0.3096473029045643,
+      "grad_norm": 0.5983361600752913,
+      "learning_rate": 9.99970475036306e-06,
+      "loss": 0.5008,
+      "step": 398
+    },
+    {
+      "epoch": 0.3104253112033195,
+      "grad_norm": 0.7359492664756181,
+      "learning_rate": 9.999653492337592e-06,
+      "loss": 0.5257,
+      "step": 399
+    },
+    {
+      "epoch": 0.3112033195020747,
+      "grad_norm": 0.6870280919548796,
+      "learning_rate": 9.999598133866833e-06,
+      "loss": 0.5136,
+      "step": 400
+    },
+    {
+      "epoch": 0.3119813278008299,
+      "grad_norm": 0.764345643521496,
+      "learning_rate": 9.99953867499618e-06,
+      "loss": 0.5153,
+      "step": 401
+    },
+    {
+      "epoch": 0.31275933609958506,
+      "grad_norm": 0.6752257594252361,
+      "learning_rate": 9.999475115774404e-06,
+      "loss": 0.4548,
+      "step": 402
+    },
+    {
+      "epoch": 0.31353734439834025,
+      "grad_norm": 0.6732212611900142,
+      "learning_rate": 9.99940745625363e-06,
+      "loss": 0.4636,
+      "step": 403
+    },
+    {
+      "epoch": 0.31431535269709543,
+      "grad_norm": 0.6895409309370628,
+      "learning_rate": 9.999335696489346e-06,
+      "loss": 0.4975,
+      "step": 404
+    },
+    {
+      "epoch": 0.3150933609958506,
+      "grad_norm": 0.7101101470929858,
+      "learning_rate": 9.999259836540408e-06,
+      "loss": 0.514,
+      "step": 405
+    },
+    {
+      "epoch": 0.3158713692946058,
+      "grad_norm": 0.7264370790667433,
+      "learning_rate": 9.999179876469035e-06,
+      "loss": 0.483,
+      "step": 406
+    },
+    {
+      "epoch": 0.316649377593361,
+      "grad_norm": 0.7001368699445795,
+      "learning_rate": 9.9990958163408e-06,
+      "loss": 0.4927,
+      "step": 407
+    },
+    {
+      "epoch": 0.31742738589211617,
+      "grad_norm": 0.6514825822126968,
+      "learning_rate": 9.999007656224649e-06,
+      "loss": 0.469,
+      "step": 408
+    },
+    {
+      "epoch": 0.31820539419087135,
+      "grad_norm": 0.7128376114444276,
+      "learning_rate": 9.998915396192884e-06,
+      "loss": 0.5055,
+      "step": 409
+    },
+    {
+      "epoch": 0.31898340248962653,
+      "grad_norm": 0.5762991833853229,
+      "learning_rate": 9.998819036321173e-06,
+      "loss": 0.4658,
+      "step": 410
+    },
+    {
+      "epoch": 0.3197614107883817,
+      "grad_norm": 0.683334616476337,
+      "learning_rate": 9.998718576688544e-06,
+      "loss": 0.5019,
+      "step": 411
+    },
+    {
+      "epoch": 0.32053941908713696,
+      "grad_norm": 0.6465214190073462,
+      "learning_rate": 9.99861401737739e-06,
+      "loss": 0.4783,
+      "step": 412
+    },
+    {
+      "epoch": 0.32131742738589214,
+      "grad_norm": 0.6421351076209845,
+      "learning_rate": 9.998505358473465e-06,
+      "loss": 0.5224,
+      "step": 413
+    },
+    {
+      "epoch": 0.3220954356846473,
+      "grad_norm": 0.7184213295170522,
+      "learning_rate": 9.998392600065882e-06,
+      "loss": 0.5174,
+      "step": 414
+    },
+    {
+      "epoch": 0.3228734439834025,
+      "grad_norm": 0.8401251050364933,
+      "learning_rate": 9.998275742247123e-06,
+      "loss": 0.4551,
+      "step": 415
+    },
+    {
+      "epoch": 0.3236514522821577,
+      "grad_norm": 0.7558021263943571,
+      "learning_rate": 9.998154785113027e-06,
+      "loss": 0.4835,
+      "step": 416
+    },
+    {
+      "epoch": 0.3244294605809129,
+      "grad_norm": 0.6638217167521114,
+      "learning_rate": 9.998029728762796e-06,
+      "loss": 0.4703,
+      "step": 417
+    },
+    {
+      "epoch": 0.32520746887966806,
+      "grad_norm": 0.6214882669349648,
+      "learning_rate": 9.997900573298995e-06,
+      "loss": 0.4574,
+      "step": 418
+    },
+    {
+      "epoch": 0.32598547717842324,
+      "grad_norm": 0.6285486027557498,
+      "learning_rate": 9.997767318827549e-06,
+      "loss": 0.4865,
+      "step": 419
+    },
+    {
+      "epoch": 0.32676348547717843,
+      "grad_norm": 0.6926938180181356,
+      "learning_rate": 9.997629965457748e-06,
+      "loss": 0.4873,
+      "step": 420
+    },
+    {
+      "epoch": 0.3275414937759336,
+      "grad_norm": 0.7571079459360025,
+      "learning_rate": 9.997488513302243e-06,
+      "loss": 0.504,
+      "step": 421
+    },
+    {
+      "epoch": 0.3283195020746888,
+      "grad_norm": 0.6669268834677519,
+      "learning_rate": 9.997342962477043e-06,
+      "loss": 0.5246,
+      "step": 422
+    },
+    {
+      "epoch": 0.329097510373444,
+      "grad_norm": 0.665368485106633,
+      "learning_rate": 9.997193313101522e-06,
+      "loss": 0.4772,
+      "step": 423
+    },
+    {
+      "epoch": 0.32987551867219916,
+      "grad_norm": 0.7141030280552456,
+      "learning_rate": 9.997039565298413e-06,
+      "loss": 0.503,
+      "step": 424
+    },
+    {
+      "epoch": 0.33065352697095435,
+      "grad_norm": 0.7140248600343714,
+      "learning_rate": 9.996881719193812e-06,
+      "loss": 0.4925,
+      "step": 425
+    },
+    {
+      "epoch": 0.33143153526970953,
+      "grad_norm": 0.6571249620633155,
+      "learning_rate": 9.996719774917176e-06,
+      "loss": 0.4832,
+      "step": 426
+    },
+    {
+      "epoch": 0.3322095435684647,
+      "grad_norm": 0.7729195626997015,
+      "learning_rate": 9.996553732601324e-06,
+      "loss": 0.4635,
+      "step": 427
+    },
+    {
+      "epoch": 0.3329875518672199,
+      "grad_norm": 0.7198339943109044,
+      "learning_rate": 9.996383592382434e-06,
+      "loss": 0.4878,
+      "step": 428
+    },
+    {
+      "epoch": 0.3337655601659751,
+      "grad_norm": 0.6518771533428243,
+      "learning_rate": 9.996209354400046e-06,
+      "loss": 0.5116,
+      "step": 429
+    },
+    {
+      "epoch": 0.33454356846473027,
+      "grad_norm": 0.7835043550827697,
+      "learning_rate": 9.99603101879706e-06,
+      "loss": 0.4889,
+      "step": 430
+    },
+    {
+      "epoch": 0.33532157676348545,
+      "grad_norm": 0.7058838935988369,
+      "learning_rate": 9.995848585719739e-06,
+      "loss": 0.4924,
+      "step": 431
+    },
+    {
+      "epoch": 0.3360995850622407,
+      "grad_norm": 0.696639149728877,
+      "learning_rate": 9.995662055317703e-06,
+      "loss": 0.5111,
+      "step": 432
+    },
+    {
+      "epoch": 0.3368775933609959,
+      "grad_norm": 0.7397160839743832,
+      "learning_rate": 9.995471427743934e-06,
+      "loss": 0.4745,
+      "step": 433
+    },
+    {
+      "epoch": 0.33765560165975106,
+      "grad_norm": 0.7862789779539546,
+      "learning_rate": 9.995276703154775e-06,
+      "loss": 0.4892,
+      "step": 434
+    },
+    {
+      "epoch": 0.33843360995850624,
+      "grad_norm": 0.7481436005033683,
+      "learning_rate": 9.99507788170993e-06,
+      "loss": 0.5279,
+      "step": 435
+    },
+    {
+      "epoch": 0.3392116182572614,
+      "grad_norm": 0.672168280856431,
+      "learning_rate": 9.994874963572458e-06,
+      "loss": 0.4658,
+      "step": 436
+    },
+    {
+      "epoch": 0.3399896265560166,
+      "grad_norm": 0.7253108842253342,
+      "learning_rate": 9.994667948908785e-06,
+      "loss": 0.4779,
+      "step": 437
+    },
+    {
+      "epoch": 0.3407676348547718,
+      "grad_norm": 0.7622375493762868,
+      "learning_rate": 9.994456837888693e-06,
+      "loss": 0.5072,
+      "step": 438
+    },
+    {
+      "epoch": 0.341545643153527,
+      "grad_norm": 0.6971885592727213,
+      "learning_rate": 9.99424163068532e-06,
+      "loss": 0.4656,
+      "step": 439
+    },
+    {
+      "epoch": 0.34232365145228216,
+      "grad_norm": 0.7168085688770621,
+      "learning_rate": 9.994022327475172e-06,
+      "loss": 0.4725,
+      "step": 440
+    },
+    {
+      "epoch": 0.34310165975103735,
+      "grad_norm": 0.705634830271896,
+      "learning_rate": 9.993798928438108e-06,
+      "loss": 0.4996,
+      "step": 441
+    },
+    {
+      "epoch": 0.34387966804979253,
+      "grad_norm": 0.6565555730131806,
+      "learning_rate": 9.993571433757347e-06,
+      "loss": 0.5067,
+      "step": 442
+    },
+    {
+      "epoch": 0.3446576763485477,
+      "grad_norm": 0.7420805679383415,
+      "learning_rate": 9.993339843619467e-06,
+      "loss": 0.4815,
+      "step": 443
+    },
+    {
+      "epoch": 0.3454356846473029,
+      "grad_norm": 0.6633509939216528,
+      "learning_rate": 9.993104158214408e-06,
+      "loss": 0.4401,
+      "step": 444
+    },
+    {
+      "epoch": 0.3462136929460581,
+      "grad_norm": 0.690468248414658,
+      "learning_rate": 9.992864377735468e-06,
+      "loss": 0.4869,
+      "step": 445
+    },
+    {
+      "epoch": 0.34699170124481327,
+      "grad_norm": 0.7472767432742858,
+      "learning_rate": 9.992620502379296e-06,
+      "loss": 0.4796,
+      "step": 446
+    },
+    {
+      "epoch": 0.34776970954356845,
+      "grad_norm": 0.7402459725749144,
+      "learning_rate": 9.992372532345909e-06,
+      "loss": 0.5036,
+      "step": 447
+    },
+    {
+      "epoch": 0.34854771784232363,
+      "grad_norm": 0.7094599111167943,
+      "learning_rate": 9.992120467838679e-06,
+      "loss": 0.4857,
+      "step": 448
+    },
+    {
+      "epoch": 0.3493257261410788,
+      "grad_norm": 0.639665276872879,
+      "learning_rate": 9.991864309064336e-06,
+      "loss": 0.4684,
+      "step": 449
+    },
+    {
+      "epoch": 0.350103734439834,
+      "grad_norm": 0.6394979818812997,
+      "learning_rate": 9.991604056232966e-06,
+      "loss": 0.4965,
+      "step": 450
+    },
+    {
+      "epoch": 0.3508817427385892,
+      "grad_norm": 0.7512258856605484,
+      "learning_rate": 9.991339709558015e-06,
+      "loss": 0.5132,
+      "step": 451
+    },
+    {
+      "epoch": 0.3516597510373444,
+      "grad_norm": 0.6487035003133411,
+      "learning_rate": 9.991071269256286e-06,
+      "loss": 0.4608,
+      "step": 452
+    },
+    {
+      "epoch": 0.3524377593360996,
+      "grad_norm": 0.7271617105593843,
+      "learning_rate": 9.990798735547937e-06,
+      "loss": 0.5323,
+      "step": 453
+    },
+    {
+      "epoch": 0.3532157676348548,
+      "grad_norm": 0.6725884543169285,
+      "learning_rate": 9.990522108656489e-06,
+      "loss": 0.4854,
+      "step": 454
+    },
+    {
+      "epoch": 0.35399377593361,
+      "grad_norm": 0.6971381716863619,
+      "learning_rate": 9.990241388808814e-06,
+      "loss": 0.5117,
+      "step": 455
+    },
+    {
+      "epoch": 0.35477178423236516,
+      "grad_norm": 0.7038212208400333,
+      "learning_rate": 9.989956576235144e-06,
+      "loss": 0.4801,
+      "step": 456
+    },
+    {
+      "epoch": 0.35554979253112035,
+      "grad_norm": 0.6883350327263039,
+      "learning_rate": 9.989667671169068e-06,
+      "loss": 0.4749,
+      "step": 457
+    },
+    {
+      "epoch": 0.35632780082987553,
+      "grad_norm": 0.6429090794057343,
+      "learning_rate": 9.989374673847529e-06,
+      "loss": 0.4673,
+      "step": 458
+    },
+    {
+      "epoch": 0.3571058091286307,
+      "grad_norm": 0.7354686246741704,
+      "learning_rate": 9.989077584510828e-06,
+      "loss": 0.4722,
+      "step": 459
+    },
+    {
+      "epoch": 0.3578838174273859,
+      "grad_norm": 0.7030846972452037,
+      "learning_rate": 9.98877640340262e-06,
+      "loss": 0.5123,
+      "step": 460
+    },
+    {
+      "epoch": 0.3586618257261411,
+      "grad_norm": 0.613855723570881,
+      "learning_rate": 9.988471130769922e-06,
+      "loss": 0.4614,
+      "step": 461
+    },
+    {
+      "epoch": 0.35943983402489627,
+      "grad_norm": 0.6759788930601867,
+      "learning_rate": 9.988161766863095e-06,
+      "loss": 0.47,
+      "step": 462
+    },
+    {
+      "epoch": 0.36021784232365145,
+      "grad_norm": 0.617668831226455,
+      "learning_rate": 9.987848311935869e-06,
+      "loss": 0.4715,
+      "step": 463
+    },
+    {
+      "epoch": 0.36099585062240663,
+      "grad_norm": 0.6390382225188237,
+      "learning_rate": 9.987530766245317e-06,
+      "loss": 0.4682,
+      "step": 464
+    },
+    {
+      "epoch": 0.3617738589211618,
+      "grad_norm": 0.6324433940698506,
+      "learning_rate": 9.987209130051878e-06,
+      "loss": 0.4949,
+      "step": 465
+    },
+    {
+      "epoch": 0.362551867219917,
+      "grad_norm": 0.6526962740122537,
+      "learning_rate": 9.98688340361934e-06,
+      "loss": 0.5103,
+      "step": 466
+    },
+    {
+      "epoch": 0.3633298755186722,
+      "grad_norm": 0.7992730436640798,
+      "learning_rate": 9.986553587214842e-06,
+      "loss": 0.5592,
+      "step": 467
+    },
+    {
+      "epoch": 0.36410788381742737,
+      "grad_norm": 0.6438493710779923,
+      "learning_rate": 9.986219681108886e-06,
+      "loss": 0.4667,
+      "step": 468
+    },
+    {
+      "epoch": 0.36488589211618255,
+      "grad_norm": 0.6451170079024203,
+      "learning_rate": 9.985881685575323e-06,
+      "loss": 0.49,
+      "step": 469
+    },
+    {
+      "epoch": 0.36566390041493774,
+      "grad_norm": 0.64447547319747,
+      "learning_rate": 9.985539600891356e-06,
+      "loss": 0.4815,
+      "step": 470
+    },
+    {
+      "epoch": 0.3664419087136929,
+      "grad_norm": 0.6510684337883991,
+      "learning_rate": 9.985193427337547e-06,
+      "loss": 0.4874,
+      "step": 471
+    },
+    {
+      "epoch": 0.36721991701244816,
+      "grad_norm": 0.7931717711083817,
+      "learning_rate": 9.984843165197807e-06,
+      "loss": 0.496,
+      "step": 472
+    },
+    {
+      "epoch": 0.36799792531120334,
+      "grad_norm": 0.7007690522292864,
+      "learning_rate": 9.984488814759402e-06,
+      "loss": 0.4782,
+      "step": 473
+    },
+    {
+      "epoch": 0.36877593360995853,
+      "grad_norm": 0.5746719721371671,
+      "learning_rate": 9.984130376312956e-06,
+      "loss": 0.4721,
+      "step": 474
+    },
+    {
+      "epoch": 0.3695539419087137,
+      "grad_norm": 0.6664788554625694,
+      "learning_rate": 9.983767850152433e-06,
+      "loss": 0.4796,
+      "step": 475
+    },
+    {
+      "epoch": 0.3703319502074689,
+      "grad_norm": 0.7481029349264973,
+      "learning_rate": 9.983401236575163e-06,
+      "loss": 0.4601,
+      "step": 476
+    },
+    {
+      "epoch": 0.3711099585062241,
+      "grad_norm": 0.7891366125644006,
+      "learning_rate": 9.983030535881821e-06,
+      "loss": 0.5146,
+      "step": 477
+    },
+    {
+      "epoch": 0.37188796680497926,
+      "grad_norm": 0.5782737994191965,
+      "learning_rate": 9.982655748376433e-06,
+      "loss": 0.4535,
+      "step": 478
+    },
+    {
+      "epoch": 0.37266597510373445,
+      "grad_norm": 0.5837112567607768,
+      "learning_rate": 9.982276874366386e-06,
+      "loss": 0.4906,
+      "step": 479
+    },
+    {
+      "epoch": 0.37344398340248963,
+      "grad_norm": 0.6618981466283724,
+      "learning_rate": 9.981893914162403e-06,
+      "loss": 0.4857,
+      "step": 480
+    },
+    {
+      "epoch": 0.3742219917012448,
+      "grad_norm": 0.6086515967447417,
+      "learning_rate": 9.981506868078576e-06,
+      "loss": 0.4358,
+      "step": 481
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.7171132008557342,
+      "learning_rate": 9.981115736432334e-06,
+      "loss": 0.4721,
+      "step": 482
+    },
+    {
+      "epoch": 0.3757780082987552,
+      "grad_norm": 0.6277078837894168,
+      "learning_rate": 9.980720519544464e-06,
+      "loss": 0.4409,
+      "step": 483
+    },
+    {
+      "epoch": 0.37655601659751037,
+      "grad_norm": 0.6632568911253907,
+      "learning_rate": 9.9803212177391e-06,
+      "loss": 0.4826,
+      "step": 484
+    },
+    {
+      "epoch": 0.37733402489626555,
+      "grad_norm": 0.8914482757407495,
+      "learning_rate": 9.979917831343728e-06,
+      "loss": 0.5563,
+      "step": 485
+    },
+    {
+      "epoch": 0.37811203319502074,
+      "grad_norm": 0.6554799168945573,
+      "learning_rate": 9.979510360689186e-06,
+      "loss": 0.4902,
+      "step": 486
+    },
+    {
+      "epoch": 0.3788900414937759,
+      "grad_norm": 0.6603992662415222,
+      "learning_rate": 9.979098806109657e-06,
+      "loss": 0.493,
+      "step": 487
+    },
+    {
+      "epoch": 0.3796680497925311,
+      "grad_norm": 0.6994986525446696,
+      "learning_rate": 9.978683167942676e-06,
+      "loss": 0.5326,
+      "step": 488
+    },
+    {
+      "epoch": 0.3804460580912863,
+      "grad_norm": 0.6876413850894957,
+      "learning_rate": 9.978263446529127e-06,
+      "loss": 0.4807,
+      "step": 489
+    },
+    {
+      "epoch": 0.38122406639004147,
+      "grad_norm": 0.6063184853454566,
+      "learning_rate": 9.977839642213243e-06,
+      "loss": 0.4409,
+      "step": 490
+    },
+    {
+      "epoch": 0.38200207468879666,
+      "grad_norm": 0.5879424642775147,
+      "learning_rate": 9.977411755342607e-06,
+      "loss": 0.4686,
+      "step": 491
+    },
+    {
+      "epoch": 0.38278008298755184,
+      "grad_norm": 0.5755391022053246,
+      "learning_rate": 9.976979786268146e-06,
+      "loss": 0.4345,
+      "step": 492
+    },
+    {
+      "epoch": 0.3835580912863071,
+      "grad_norm": 0.7017615933470893,
+      "learning_rate": 9.97654373534414e-06,
+      "loss": 0.488,
+      "step": 493
+    },
+    {
+      "epoch": 0.38433609958506226,
+      "grad_norm": 0.6153896074995053,
+      "learning_rate": 9.976103602928214e-06,
+      "loss": 0.4675,
+      "step": 494
+    },
+    {
+      "epoch": 0.38511410788381745,
+      "grad_norm": 0.700172913841235,
+      "learning_rate": 9.975659389381338e-06,
+      "loss": 0.5068,
+      "step": 495
+    },
+    {
+      "epoch": 0.38589211618257263,
+      "grad_norm": 0.6394640447151264,
+      "learning_rate": 9.975211095067835e-06,
+      "loss": 0.4562,
+      "step": 496
+    },
+    {
+      "epoch": 0.3866701244813278,
+      "grad_norm": 0.6507697509464065,
+      "learning_rate": 9.97475872035537e-06,
+      "loss": 0.4339,
+      "step": 497
+    },
+    {
+      "epoch": 0.387448132780083,
+      "grad_norm": 0.6631924769086148,
+      "learning_rate": 9.974302265614959e-06,
+      "loss": 0.5165,
+      "step": 498
+    },
+    {
+      "epoch": 0.3882261410788382,
+      "grad_norm": 0.7001277652412472,
+      "learning_rate": 9.973841731220958e-06,
+      "loss": 0.5096,
+      "step": 499
+    },
+    {
+      "epoch": 0.38900414937759337,
+      "grad_norm": 0.6821441865184117,
+      "learning_rate": 9.973377117551074e-06,
+      "loss": 0.4897,
+      "step": 500
+    },
+    {
+      "epoch": 0.38978215767634855,
+      "grad_norm": 0.664085760379377,
+      "learning_rate": 9.972908424986358e-06,
+      "loss": 0.4895,
+      "step": 501
+    },
+    {
+      "epoch": 0.39056016597510373,
+      "grad_norm": 0.6120585742696022,
+      "learning_rate": 9.972435653911206e-06,
+      "loss": 0.5044,
+      "step": 502
+    },
+    {
+      "epoch": 0.3913381742738589,
+      "grad_norm": 0.6084364603952789,
+      "learning_rate": 9.971958804713359e-06,
+      "loss": 0.4542,
+      "step": 503
+    },
+    {
+      "epoch": 0.3921161825726141,
+      "grad_norm": 0.6518988947492096,
+      "learning_rate": 9.971477877783904e-06,
+      "loss": 0.5209,
+      "step": 504
+    },
+    {
+      "epoch": 0.3928941908713693,
+      "grad_norm": 0.6334222384617313,
+      "learning_rate": 9.970992873517272e-06,
+      "loss": 0.4538,
+      "step": 505
+    },
+    {
+      "epoch": 0.39367219917012447,
+      "grad_norm": 0.6578268595333242,
+      "learning_rate": 9.970503792311233e-06,
+      "loss": 0.4916,
+      "step": 506
+    },
+    {
+      "epoch": 0.39445020746887965,
+      "grad_norm": 0.6106633611643024,
+      "learning_rate": 9.97001063456691e-06,
+      "loss": 0.4738,
+      "step": 507
+    },
+    {
+      "epoch": 0.39522821576763484,
+      "grad_norm": 0.6107955117658455,
+      "learning_rate": 9.96951340068876e-06,
+      "loss": 0.4297,
+      "step": 508
+    },
+    {
+      "epoch": 0.39600622406639,
+      "grad_norm": 0.6744974172828271,
+      "learning_rate": 9.96901209108459e-06,
+      "loss": 0.4877,
+      "step": 509
+    },
+    {
+      "epoch": 0.3967842323651452,
+      "grad_norm": 0.7349615410004952,
+      "learning_rate": 9.968506706165548e-06,
+      "loss": 0.4766,
+      "step": 510
+    },
+    {
+      "epoch": 0.3975622406639004,
+      "grad_norm": 0.6483392857855375,
+      "learning_rate": 9.967997246346121e-06,
+      "loss": 0.4917,
+      "step": 511
+    },
+    {
+      "epoch": 0.3983402489626556,
+      "grad_norm": 0.6933915795224606,
+      "learning_rate": 9.96748371204414e-06,
+      "loss": 0.4717,
+      "step": 512
+    },
+    {
+      "epoch": 0.3991182572614108,
+      "grad_norm": 0.6763693757730737,
+      "learning_rate": 9.96696610368078e-06,
+      "loss": 0.4844,
+      "step": 513
+    },
+    {
+      "epoch": 0.399896265560166,
+      "grad_norm": 0.6474661546522922,
+      "learning_rate": 9.966444421680554e-06,
+      "loss": 0.4573,
+      "step": 514
+    },
+    {
+      "epoch": 0.4006742738589212,
+      "grad_norm": 0.6413011083091857,
+      "learning_rate": 9.96591866647132e-06,
+      "loss": 0.4497,
+      "step": 515
+    },
+    {
+      "epoch": 0.40145228215767637,
+      "grad_norm": 0.6820846157238268,
+      "learning_rate": 9.96538883848427e-06,
+      "loss": 0.513,
+      "step": 516
+    },
+    {
+      "epoch": 0.40223029045643155,
+      "grad_norm": 0.7050338187275862,
+      "learning_rate": 9.964854938153942e-06,
+      "loss": 0.5025,
+      "step": 517
+    },
+    {
+      "epoch": 0.40300829875518673,
+      "grad_norm": 0.6277630080153636,
+      "learning_rate": 9.964316965918215e-06,
+      "loss": 0.5295,
+      "step": 518
+    },
+    {
+      "epoch": 0.4037863070539419,
+      "grad_norm": 0.6221969035650775,
+      "learning_rate": 9.963774922218299e-06,
+      "loss": 0.4771,
+      "step": 519
+    },
+    {
+      "epoch": 0.4045643153526971,
+      "grad_norm": 0.7100540441214629,
+      "learning_rate": 9.963228807498753e-06,
+      "loss": 0.5253,
+      "step": 520
+    },
+    {
+      "epoch": 0.4053423236514523,
+      "grad_norm": 0.625712712308733,
+      "learning_rate": 9.96267862220747e-06,
+      "loss": 0.4615,
+      "step": 521
+    },
+    {
+      "epoch": 0.40612033195020747,
+      "grad_norm": 0.6064818551024935,
+      "learning_rate": 9.962124366795683e-06,
+      "loss": 0.4549,
+      "step": 522
+    },
+    {
+      "epoch": 0.40689834024896265,
+      "grad_norm": 0.6565222369854272,
+      "learning_rate": 9.961566041717958e-06,
+      "loss": 0.5167,
+      "step": 523
+    },
+    {
+      "epoch": 0.40767634854771784,
+      "grad_norm": 0.633948662212453,
+      "learning_rate": 9.96100364743221e-06,
+      "loss": 0.4598,
+      "step": 524
+    },
+    {
+      "epoch": 0.408454356846473,
+      "grad_norm": 0.5677624492517853,
+      "learning_rate": 9.960437184399679e-06,
+      "loss": 0.4631,
+      "step": 525
+    },
+    {
+      "epoch": 0.4092323651452282,
+      "grad_norm": 0.6725795954207179,
+      "learning_rate": 9.959866653084948e-06,
+      "loss": 0.5235,
+      "step": 526
+    },
+    {
+      "epoch": 0.4100103734439834,
+      "grad_norm": 0.6201479993464989,
+      "learning_rate": 9.959292053955937e-06,
+      "loss": 0.4576,
+      "step": 527
+    },
+    {
+      "epoch": 0.4107883817427386,
+      "grad_norm": 0.74490146123115,
+      "learning_rate": 9.958713387483901e-06,
+      "loss": 0.4791,
+      "step": 528
+    },
+    {
+      "epoch": 0.41156639004149376,
+      "grad_norm": 0.6349228074674828,
+      "learning_rate": 9.958130654143432e-06,
+      "loss": 0.4842,
+      "step": 529
+    },
+    {
+      "epoch": 0.41234439834024894,
+      "grad_norm": 0.5308345496999858,
+      "learning_rate": 9.957543854412454e-06,
+      "loss": 0.4589,
+      "step": 530
+    },
+    {
+      "epoch": 0.4131224066390041,
+      "grad_norm": 0.645724230865435,
+      "learning_rate": 9.95695298877223e-06,
+      "loss": 0.469,
+      "step": 531
+    },
+    {
+      "epoch": 0.4139004149377593,
+      "grad_norm": 0.6738906248199441,
+      "learning_rate": 9.956358057707357e-06,
+      "loss": 0.5007,
+      "step": 532
+    },
+    {
+      "epoch": 0.41467842323651455,
+      "grad_norm": 0.6134865551902325,
+      "learning_rate": 9.955759061705762e-06,
+      "loss": 0.4938,
+      "step": 533
+    },
+    {
+      "epoch": 0.41545643153526973,
+      "grad_norm": 0.689443085069017,
+      "learning_rate": 9.955156001258713e-06,
+      "loss": 0.5182,
+      "step": 534
+    },
+    {
+      "epoch": 0.4162344398340249,
+      "grad_norm": 0.6477954758870604,
+      "learning_rate": 9.954548876860805e-06,
+      "loss": 0.4711,
+      "step": 535
+    },
+    {
+      "epoch": 0.4170124481327801,
+      "grad_norm": 0.5643000236127793,
+      "learning_rate": 9.95393768900997e-06,
+      "loss": 0.4443,
+      "step": 536
+    },
+    {
+      "epoch": 0.4177904564315353,
+      "grad_norm": 0.6477540537104645,
+      "learning_rate": 9.95332243820747e-06,
+      "loss": 0.4627,
+      "step": 537
+    },
+    {
+      "epoch": 0.41856846473029047,
+      "grad_norm": 0.6956480238743482,
+      "learning_rate": 9.952703124957899e-06,
+      "loss": 0.5158,
+      "step": 538
+    },
+    {
+      "epoch": 0.41934647302904565,
+      "grad_norm": 0.6235148438330322,
+      "learning_rate": 9.952079749769187e-06,
+      "loss": 0.5204,
+      "step": 539
+    },
+    {
+      "epoch": 0.42012448132780084,
+      "grad_norm": 0.5678844746858847,
+      "learning_rate": 9.951452313152592e-06,
+      "loss": 0.432,
+      "step": 540
+    },
+    {
+      "epoch": 0.420902489626556,
+      "grad_norm": 0.6456077332884878,
+      "learning_rate": 9.950820815622703e-06,
+      "loss": 0.4574,
+      "step": 541
+    },
+    {
+      "epoch": 0.4216804979253112,
+      "grad_norm": 0.7284177173250788,
+      "learning_rate": 9.95018525769744e-06,
+      "loss": 0.5133,
+      "step": 542
+    },
+    {
+      "epoch": 0.4224585062240664,
+      "grad_norm": 0.6045700138062209,
+      "learning_rate": 9.949545639898055e-06,
+      "loss": 0.4965,
+      "step": 543
+    },
+    {
+      "epoch": 0.42323651452282157,
+      "grad_norm": 0.6016232063664123,
+      "learning_rate": 9.948901962749125e-06,
+      "loss": 0.5081,
+      "step": 544
+    },
+    {
+      "epoch": 0.42401452282157676,
+      "grad_norm": 0.7963635405379935,
+      "learning_rate": 9.94825422677856e-06,
+      "loss": 0.4669,
+      "step": 545
+    },
+    {
+      "epoch": 0.42479253112033194,
+      "grad_norm": 0.7143109659259899,
+      "learning_rate": 9.9476024325176e-06,
+      "loss": 0.4464,
+      "step": 546
+    },
+    {
+      "epoch": 0.4255705394190871,
+      "grad_norm": 0.6616281076927624,
+      "learning_rate": 9.94694658050081e-06,
+      "loss": 0.5348,
+      "step": 547
+    },
+    {
+      "epoch": 0.4263485477178423,
+      "grad_norm": 0.6399018171822406,
+      "learning_rate": 9.946286671266086e-06,
+      "loss": 0.4572,
+      "step": 548
+    },
+    {
+      "epoch": 0.4271265560165975,
+      "grad_norm": 0.7510365887184355,
+      "learning_rate": 9.945622705354646e-06,
+      "loss": 0.4941,
+      "step": 549
+    },
+    {
+      "epoch": 0.4279045643153527,
+      "grad_norm": 0.7021145988080252,
+      "learning_rate": 9.944954683311041e-06,
+      "loss": 0.4581,
+      "step": 550
+    },
+    {
+      "epoch": 0.42868257261410786,
+      "grad_norm": 0.6510982279162588,
+      "learning_rate": 9.944282605683147e-06,
+      "loss": 0.4845,
+      "step": 551
+    },
+    {
+      "epoch": 0.42946058091286304,
+      "grad_norm": 0.7233790475353264,
+      "learning_rate": 9.943606473022165e-06,
+      "loss": 0.4942,
+      "step": 552
+    },
+    {
+      "epoch": 0.4302385892116183,
+      "grad_norm": 0.8984675134627667,
+      "learning_rate": 9.942926285882624e-06,
+      "loss": 0.5167,
+      "step": 553
+    },
+    {
+      "epoch": 0.43101659751037347,
+      "grad_norm": 0.6444079170813586,
+      "learning_rate": 9.942242044822374e-06,
+      "loss": 0.4875,
+      "step": 554
+    },
+    {
+      "epoch": 0.43179460580912865,
+      "grad_norm": 0.6392973928731915,
+      "learning_rate": 9.941553750402595e-06,
+      "loss": 0.4659,
+      "step": 555
+    },
+    {
+      "epoch": 0.43257261410788383,
+      "grad_norm": 0.6299901673904328,
+      "learning_rate": 9.940861403187788e-06,
+      "loss": 0.4418,
+      "step": 556
+    },
+    {
+      "epoch": 0.433350622406639,
+      "grad_norm": 0.6471835610011257,
+      "learning_rate": 9.940165003745779e-06,
+      "loss": 0.4606,
+      "step": 557
+    },
+    {
+      "epoch": 0.4341286307053942,
+      "grad_norm": 0.6259788057683332,
+      "learning_rate": 9.939464552647716e-06,
+      "loss": 0.4542,
+      "step": 558
+    },
+    {
+      "epoch": 0.4349066390041494,
+      "grad_norm": 0.5513211330213582,
+      "learning_rate": 9.93876005046807e-06,
+      "loss": 0.4257,
+      "step": 559
+    },
+    {
+      "epoch": 0.43568464730290457,
+      "grad_norm": 0.591739945451658,
+      "learning_rate": 9.938051497784639e-06,
+      "loss": 0.4393,
+      "step": 560
+    },
+    {
+      "epoch": 0.43646265560165975,
+      "grad_norm": 0.7405172088126226,
+      "learning_rate": 9.937338895178538e-06,
+      "loss": 0.4856,
+      "step": 561
+    },
+    {
+      "epoch": 0.43724066390041494,
+      "grad_norm": 0.5909680636691766,
+      "learning_rate": 9.936622243234206e-06,
+      "loss": 0.4596,
+      "step": 562
+    },
+    {
+      "epoch": 0.4380186721991701,
+      "grad_norm": 0.6057856079056377,
+      "learning_rate": 9.9359015425394e-06,
+      "loss": 0.478,
+      "step": 563
+    },
+    {
+      "epoch": 0.4387966804979253,
+      "grad_norm": 0.6025507763268173,
+      "learning_rate": 9.935176793685202e-06,
+      "loss": 0.4828,
+      "step": 564
+    },
+    {
+      "epoch": 0.4395746887966805,
+      "grad_norm": 0.6409600467857902,
+      "learning_rate": 9.93444799726601e-06,
+      "loss": 0.4754,
+      "step": 565
+    },
+    {
+      "epoch": 0.4403526970954357,
+      "grad_norm": 0.6443880245343295,
+      "learning_rate": 9.933715153879544e-06,
+      "loss": 0.4624,
+      "step": 566
+    },
+    {
+      "epoch": 0.44113070539419086,
+      "grad_norm": 0.5470441693944035,
+      "learning_rate": 9.932978264126843e-06,
+      "loss": 0.4248,
+      "step": 567
+    },
+    {
+      "epoch": 0.44190871369294604,
+      "grad_norm": 0.6295026512692594,
+      "learning_rate": 9.932237328612262e-06,
+      "loss": 0.4942,
+      "step": 568
+    },
+    {
+      "epoch": 0.4426867219917012,
+      "grad_norm": 0.622941307109036,
+      "learning_rate": 9.93149234794348e-06,
+      "loss": 0.5183,
+      "step": 569
+    },
+    {
+      "epoch": 0.4434647302904564,
+      "grad_norm": 0.6708474286302462,
+      "learning_rate": 9.930743322731487e-06,
+      "loss": 0.4698,
+      "step": 570
+    },
+    {
+      "epoch": 0.4442427385892116,
+      "grad_norm": 0.5871207611367272,
+      "learning_rate": 9.929990253590593e-06,
+      "loss": 0.4913,
+      "step": 571
+    },
+    {
+      "epoch": 0.4450207468879668,
+      "grad_norm": 0.6703189828392467,
+      "learning_rate": 9.929233141138427e-06,
+      "loss": 0.5206,
+      "step": 572
+    },
+    {
+      "epoch": 0.445798755186722,
+      "grad_norm": 0.6099418805384298,
+      "learning_rate": 9.92847198599593e-06,
+      "loss": 0.4705,
+      "step": 573
+    },
+    {
+      "epoch": 0.4465767634854772,
+      "grad_norm": 0.5454164954580368,
+      "learning_rate": 9.92770678878736e-06,
+      "loss": 0.4811,
+      "step": 574
+    },
+    {
+      "epoch": 0.4473547717842324,
+      "grad_norm": 0.6608150282279543,
+      "learning_rate": 9.926937550140287e-06,
+      "loss": 0.4667,
+      "step": 575
+    },
+    {
+      "epoch": 0.44813278008298757,
+      "grad_norm": 0.6123479211806143,
+      "learning_rate": 9.926164270685607e-06,
+      "loss": 0.4602,
+      "step": 576
+    },
+    {
+      "epoch": 0.44891078838174275,
+      "grad_norm": 0.6838584529925289,
+      "learning_rate": 9.925386951057516e-06,
+      "loss": 0.4867,
+      "step": 577
+    },
+    {
+      "epoch": 0.44968879668049794,
+      "grad_norm": 0.6696803981320506,
+      "learning_rate": 9.92460559189353e-06,
+      "loss": 0.4731,
+      "step": 578
+    },
+    {
+      "epoch": 0.4504668049792531,
+      "grad_norm": 0.6400480053604497,
+      "learning_rate": 9.923820193834479e-06,
+      "loss": 0.4839,
+      "step": 579
+    },
+    {
+      "epoch": 0.4512448132780083,
+      "grad_norm": 0.7326412973245916,
+      "learning_rate": 9.923030757524504e-06,
+      "loss": 0.5151,
+      "step": 580
+    },
+    {
+      "epoch": 0.4520228215767635,
+      "grad_norm": 0.7439762688034677,
+      "learning_rate": 9.922237283611058e-06,
+      "loss": 0.5608,
+      "step": 581
+    },
+    {
+      "epoch": 0.4528008298755187,
+      "grad_norm": 0.6011090708543252,
+      "learning_rate": 9.921439772744903e-06,
+      "loss": 0.4211,
+      "step": 582
+    },
+    {
+      "epoch": 0.45357883817427386,
+      "grad_norm": 0.5936796129314045,
+      "learning_rate": 9.920638225580117e-06,
+      "loss": 0.4568,
+      "step": 583
+    },
+    {
+      "epoch": 0.45435684647302904,
+      "grad_norm": 0.5620259832223025,
+      "learning_rate": 9.919832642774085e-06,
+      "loss": 0.4937,
+      "step": 584
+    },
+    {
+      "epoch": 0.4551348547717842,
+      "grad_norm": 0.7422435642633326,
+      "learning_rate": 9.919023024987501e-06,
+      "loss": 0.5545,
+      "step": 585
+    },
+    {
+      "epoch": 0.4559128630705394,
+      "grad_norm": 0.6758943903591057,
+      "learning_rate": 9.918209372884373e-06,
+      "loss": 0.508,
+      "step": 586
+    },
+    {
+      "epoch": 0.4566908713692946,
+      "grad_norm": 0.628387179600847,
+      "learning_rate": 9.91739168713201e-06,
+      "loss": 0.4788,
+      "step": 587
+    },
+    {
+      "epoch": 0.4574688796680498,
+      "grad_norm": 0.6249838421221252,
+      "learning_rate": 9.916569968401038e-06,
+      "loss": 0.4441,
+      "step": 588
+    },
+    {
+      "epoch": 0.45824688796680496,
+      "grad_norm": 0.5881388508046675,
+      "learning_rate": 9.915744217365382e-06,
+      "loss": 0.486,
+      "step": 589
+    },
+    {
+      "epoch": 0.45902489626556015,
+      "grad_norm": 0.6412346607897325,
+      "learning_rate": 9.914914434702281e-06,
+      "loss": 0.4961,
+      "step": 590
+    },
+    {
+      "epoch": 0.45980290456431533,
+      "grad_norm": 0.6544511031906262,
+      "learning_rate": 9.91408062109228e-06,
+      "loss": 0.4935,
+      "step": 591
+    },
+    {
+      "epoch": 0.4605809128630705,
+      "grad_norm": 0.6909481917140404,
+      "learning_rate": 9.913242777219221e-06,
+      "loss": 0.4607,
+      "step": 592
+    },
+    {
+      "epoch": 0.46135892116182575,
+      "grad_norm": 0.5409649308990738,
+      "learning_rate": 9.912400903770265e-06,
+      "loss": 0.4549,
+      "step": 593
+    },
+    {
+      "epoch": 0.46213692946058094,
+      "grad_norm": 0.6087873732274134,
+      "learning_rate": 9.911555001435867e-06,
+      "loss": 0.4306,
+      "step": 594
+    },
+    {
+      "epoch": 0.4629149377593361,
+      "grad_norm": 0.6181036191644309,
+      "learning_rate": 9.910705070909792e-06,
+      "loss": 0.4756,
+      "step": 595
+    },
+    {
+      "epoch": 0.4636929460580913,
+      "grad_norm": 0.6658902206056407,
+      "learning_rate": 9.909851112889106e-06,
+      "loss": 0.514,
+      "step": 596
+    },
+    {
+      "epoch": 0.4644709543568465,
+      "grad_norm": 0.6986074129631306,
+      "learning_rate": 9.90899312807418e-06,
+      "loss": 0.494,
+      "step": 597
+    },
+    {
+      "epoch": 0.46524896265560167,
+      "grad_norm": 0.5843454057549798,
+      "learning_rate": 9.90813111716869e-06,
+      "loss": 0.4828,
+      "step": 598
+    },
+    {
+      "epoch": 0.46602697095435686,
+      "grad_norm": 0.6934479553425152,
+      "learning_rate": 9.907265080879603e-06,
+      "loss": 0.4987,
+      "step": 599
+    },
+    {
+      "epoch": 0.46680497925311204,
+      "grad_norm": 0.6195973528295227,
+      "learning_rate": 9.9063950199172e-06,
+      "loss": 0.4919,
+      "step": 600
+    },
+    {
+      "epoch": 0.4675829875518672,
+      "grad_norm": 0.6769285464674736,
+      "learning_rate": 9.90552093499506e-06,
+      "loss": 0.5449,
+      "step": 601
+    },
+    {
+      "epoch": 0.4683609958506224,
+      "grad_norm": 0.6712181218839263,
+      "learning_rate": 9.904642826830056e-06,
+      "loss": 0.4763,
+      "step": 602
+    },
+    {
+      "epoch": 0.4691390041493776,
+      "grad_norm": 0.5446064033915384,
+      "learning_rate": 9.903760696142366e-06,
+      "loss": 0.457,
+      "step": 603
+    },
+    {
+      "epoch": 0.4699170124481328,
+      "grad_norm": 0.6303135141372874,
+      "learning_rate": 9.902874543655466e-06,
+      "loss": 0.4739,
+      "step": 604
+    },
+    {
+      "epoch": 0.47069502074688796,
+      "grad_norm": 0.6151993826953289,
+      "learning_rate": 9.901984370096132e-06,
+      "loss": 0.4438,
+      "step": 605
+    },
+    {
+      "epoch": 0.47147302904564314,
+      "grad_norm": 0.5729409656331761,
+      "learning_rate": 9.901090176194432e-06,
+      "loss": 0.4651,
+      "step": 606
+    },
+    {
+      "epoch": 0.47225103734439833,
+      "grad_norm": 0.5648236572830089,
+      "learning_rate": 9.90019196268374e-06,
+      "loss": 0.4301,
+      "step": 607
+    },
+    {
+      "epoch": 0.4730290456431535,
+      "grad_norm": 0.6681610337401934,
+      "learning_rate": 9.899289730300721e-06,
+      "loss": 0.4959,
+      "step": 608
+    },
+    {
+      "epoch": 0.4738070539419087,
+      "grad_norm": 0.6660422320613597,
+      "learning_rate": 9.898383479785335e-06,
+      "loss": 0.479,
+      "step": 609
+    },
+    {
+      "epoch": 0.4745850622406639,
+      "grad_norm": 0.6132162574987826,
+      "learning_rate": 9.89747321188084e-06,
+      "loss": 0.4663,
+      "step": 610
+    },
+    {
+      "epoch": 0.47536307053941906,
+      "grad_norm": 0.6343423996164013,
+      "learning_rate": 9.896558927333793e-06,
+      "loss": 0.4641,
+      "step": 611
+    },
+    {
+      "epoch": 0.47614107883817425,
+      "grad_norm": 0.7178642670099903,
+      "learning_rate": 9.895640626894036e-06,
+      "loss": 0.4851,
+      "step": 612
+    },
+    {
+      "epoch": 0.4769190871369295,
+      "grad_norm": 0.5894518982491137,
+      "learning_rate": 9.894718311314711e-06,
+      "loss": 0.4324,
+      "step": 613
+    },
+    {
+      "epoch": 0.47769709543568467,
+      "grad_norm": 0.641445083594175,
+      "learning_rate": 9.89379198135225e-06,
+      "loss": 0.4813,
+      "step": 614
+    },
+    {
+      "epoch": 0.47847510373443985,
+      "grad_norm": 0.651402031852471,
+      "learning_rate": 9.89286163776638e-06,
+      "loss": 0.5098,
+      "step": 615
+    },
+    {
+      "epoch": 0.47925311203319504,
+      "grad_norm": 0.7317680732695898,
+      "learning_rate": 9.891927281320118e-06,
+      "loss": 0.4465,
+      "step": 616
+    },
+    {
+      "epoch": 0.4800311203319502,
+      "grad_norm": 0.5783451949319504,
+      "learning_rate": 9.890988912779773e-06,
+      "loss": 0.4709,
+      "step": 617
+    },
+    {
+      "epoch": 0.4808091286307054,
+      "grad_norm": 0.5836222745901241,
+      "learning_rate": 9.890046532914944e-06,
+      "loss": 0.4444,
+      "step": 618
+    },
+    {
+      "epoch": 0.4815871369294606,
+      "grad_norm": 0.7090910358847796,
+      "learning_rate": 9.889100142498518e-06,
+      "loss": 0.4537,
+      "step": 619
+    },
+    {
+      "epoch": 0.4823651452282158,
+      "grad_norm": 0.6359173300096261,
+      "learning_rate": 9.888149742306676e-06,
+      "loss": 0.464,
+      "step": 620
+    },
+    {
+      "epoch": 0.48314315352697096,
+      "grad_norm": 0.5935219145938001,
+      "learning_rate": 9.887195333118887e-06,
+      "loss": 0.4567,
+      "step": 621
+    },
+    {
+      "epoch": 0.48392116182572614,
+      "grad_norm": 0.5464520306965782,
+      "learning_rate": 9.886236915717899e-06,
+      "loss": 0.4479,
+      "step": 622
+    },
+    {
+      "epoch": 0.4846991701244813,
+      "grad_norm": 0.6181887605610861,
+      "learning_rate": 9.88527449088976e-06,
+      "loss": 0.465,
+      "step": 623
+    },
+    {
+      "epoch": 0.4854771784232365,
+      "grad_norm": 0.5912742162881465,
+      "learning_rate": 9.884308059423794e-06,
+      "loss": 0.438,
+      "step": 624
+    },
+    {
+      "epoch": 0.4862551867219917,
+      "grad_norm": 0.6030366275299255,
+      "learning_rate": 9.883337622112622e-06,
+      "loss": 0.4588,
+      "step": 625
+    },
+    {
+      "epoch": 0.4870331950207469,
+      "grad_norm": 0.5489933588579986,
+      "learning_rate": 9.882363179752137e-06,
+      "loss": 0.4574,
+      "step": 626
+    },
+    {
+      "epoch": 0.48781120331950206,
+      "grad_norm": 0.578592584584376,
+      "learning_rate": 9.88138473314153e-06,
+      "loss": 0.4477,
+      "step": 627
+    },
+    {
+      "epoch": 0.48858921161825725,
+      "grad_norm": 0.5602000005780063,
+      "learning_rate": 9.880402283083266e-06,
+      "loss": 0.4457,
+      "step": 628
+    },
+    {
+      "epoch": 0.48936721991701243,
+      "grad_norm": 0.6436940875059616,
+      "learning_rate": 9.879415830383098e-06,
+      "loss": 0.486,
+      "step": 629
+    },
+    {
+      "epoch": 0.4901452282157676,
+      "grad_norm": 0.5383128987864056,
+      "learning_rate": 9.878425375850064e-06,
+      "loss": 0.4005,
+      "step": 630
+    },
+    {
+      "epoch": 0.4909232365145228,
+      "grad_norm": 0.5314126396090607,
+      "learning_rate": 9.877430920296478e-06,
+      "loss": 0.4455,
+      "step": 631
+    },
+    {
+      "epoch": 0.491701244813278,
+      "grad_norm": 0.649818211829273,
+      "learning_rate": 9.87643246453794e-06,
+      "loss": 0.481,
+      "step": 632
+    },
+    {
+      "epoch": 0.4924792531120332,
+      "grad_norm": 0.5691725542405641,
+      "learning_rate": 9.87543000939333e-06,
+      "loss": 0.4896,
+      "step": 633
+    },
+    {
+      "epoch": 0.4932572614107884,
+      "grad_norm": 0.653852443188349,
+      "learning_rate": 9.874423555684806e-06,
+      "loss": 0.4906,
+      "step": 634
+    },
+    {
+      "epoch": 0.4940352697095436,
+      "grad_norm": 0.5695063264090495,
+      "learning_rate": 9.873413104237807e-06,
+      "loss": 0.4958,
+      "step": 635
+    },
+    {
+      "epoch": 0.4948132780082988,
+      "grad_norm": 0.5601647521001227,
+      "learning_rate": 9.872398655881052e-06,
+      "loss": 0.4716,
+      "step": 636
+    },
+    {
+      "epoch": 0.49559128630705396,
+      "grad_norm": 0.5679479384840351,
+      "learning_rate": 9.871380211446536e-06,
+      "loss": 0.4471,
+      "step": 637
+    },
+    {
+      "epoch": 0.49636929460580914,
+      "grad_norm": 0.5941694371497589,
+      "learning_rate": 9.87035777176953e-06,
+      "loss": 0.4951,
+      "step": 638
+    },
+    {
+      "epoch": 0.4971473029045643,
+      "grad_norm": 0.5911730959051029,
+      "learning_rate": 9.869331337688587e-06,
+      "loss": 0.4275,
+      "step": 639
+    },
+    {
+      "epoch": 0.4979253112033195,
+      "grad_norm": 0.5921509467471296,
+      "learning_rate": 9.86830091004553e-06,
+      "loss": 0.4556,
+      "step": 640
+    },
+    {
+      "epoch": 0.4987033195020747,
+      "grad_norm": 0.5842598531684317,
+      "learning_rate": 9.867266489685463e-06,
+      "loss": 0.5039,
+      "step": 641
+    },
+    {
+      "epoch": 0.4994813278008299,
+      "grad_norm": 0.6371594405022414,
+      "learning_rate": 9.866228077456759e-06,
+      "loss": 0.4701,
+      "step": 642
+    },
+    {
+      "epoch": 0.5002593360995851,
+      "grad_norm": 0.6131307748833087,
+      "learning_rate": 9.865185674211069e-06,
+      "loss": 0.4788,
+      "step": 643
+    },
+    {
+      "epoch": 0.5010373443983402,
+      "grad_norm": 0.5465475238482574,
+      "learning_rate": 9.864139280803314e-06,
+      "loss": 0.4867,
+      "step": 644
+    },
+    {
+      "epoch": 0.5018153526970954,
+      "grad_norm": 0.619628112169775,
+      "learning_rate": 9.863088898091691e-06,
+      "loss": 0.4877,
+      "step": 645
+    },
+    {
+      "epoch": 0.5025933609958506,
+      "grad_norm": 0.6042944956976263,
+      "learning_rate": 9.862034526937666e-06,
+      "loss": 0.4632,
+      "step": 646
+    },
+    {
+      "epoch": 0.5033713692946058,
+      "grad_norm": 0.6071717875710957,
+      "learning_rate": 9.860976168205978e-06,
+      "loss": 0.4707,
+      "step": 647
+    },
+    {
+      "epoch": 0.504149377593361,
+      "grad_norm": 0.5709767486072447,
+      "learning_rate": 9.859913822764636e-06,
+      "loss": 0.4368,
+      "step": 648
+    },
+    {
+      "epoch": 0.5049273858921162,
+      "grad_norm": 0.6114016881695511,
+      "learning_rate": 9.858847491484914e-06,
+      "loss": 0.4962,
+      "step": 649
+    },
+    {
+      "epoch": 0.5057053941908713,
+      "grad_norm": 0.6678728965144449,
+      "learning_rate": 9.857777175241367e-06,
+      "loss": 0.5053,
+      "step": 650
+    },
+    {
+      "epoch": 0.5064834024896265,
+      "grad_norm": 0.6700830207889674,
+      "learning_rate": 9.856702874911804e-06,
+      "loss": 0.4838,
+      "step": 651
+    },
+    {
+      "epoch": 0.5072614107883817,
+      "grad_norm": 0.6013339788369468,
+      "learning_rate": 9.855624591377311e-06,
+      "loss": 0.4412,
+      "step": 652
+    },
+    {
+      "epoch": 0.5080394190871369,
+      "grad_norm": 0.6741639079605333,
+      "learning_rate": 9.854542325522236e-06,
+      "loss": 0.5056,
+      "step": 653
+    },
+    {
+      "epoch": 0.5088174273858921,
+      "grad_norm": 0.6711517531235603,
+      "learning_rate": 9.853456078234195e-06,
+      "loss": 0.444,
+      "step": 654
+    },
+    {
+      "epoch": 0.5095954356846473,
+      "grad_norm": 0.5530811469773816,
+      "learning_rate": 9.852365850404072e-06,
+      "loss": 0.4379,
+      "step": 655
+    },
+    {
+      "epoch": 0.5103734439834025,
+      "grad_norm": 0.7505263576430584,
+      "learning_rate": 9.85127164292601e-06,
+      "loss": 0.5175,
+      "step": 656
+    },
+    {
+      "epoch": 0.5111514522821576,
+      "grad_norm": 0.5654566629931249,
+      "learning_rate": 9.850173456697421e-06,
+      "loss": 0.4097,
+      "step": 657
+    },
+    {
+      "epoch": 0.5119294605809128,
+      "grad_norm": 0.6040127759322771,
+      "learning_rate": 9.849071292618976e-06,
+      "loss": 0.4526,
+      "step": 658
+    },
+    {
+      "epoch": 0.512707468879668,
+      "grad_norm": 0.6330562614144929,
+      "learning_rate": 9.84796515159461e-06,
+      "loss": 0.4898,
+      "step": 659
+    },
+    {
+      "epoch": 0.5134854771784232,
+      "grad_norm": 0.6604178747418343,
+      "learning_rate": 9.846855034531523e-06,
+      "loss": 0.4961,
+      "step": 660
+    },
+    {
+      "epoch": 0.5142634854771784,
+      "grad_norm": 0.6744125612677283,
+      "learning_rate": 9.84574094234017e-06,
+      "loss": 0.5009,
+      "step": 661
+    },
+    {
+      "epoch": 0.5150414937759336,
+      "grad_norm": 0.6736235107518872,
+      "learning_rate": 9.84462287593427e-06,
+      "loss": 0.4898,
+      "step": 662
+    },
+    {
+      "epoch": 0.5158195020746889,
+      "grad_norm": 0.561387113077577,
+      "learning_rate": 9.8435008362308e-06,
+      "loss": 0.4316,
+      "step": 663
+    },
+    {
+      "epoch": 0.516597510373444,
+      "grad_norm": 0.6876592980265597,
+      "learning_rate": 9.842374824149998e-06,
+      "loss": 0.5047,
+      "step": 664
+    },
+    {
+      "epoch": 0.5173755186721992,
+      "grad_norm": 0.6633177225412092,
+      "learning_rate": 9.841244840615357e-06,
+      "loss": 0.495,
+      "step": 665
+    },
+    {
+      "epoch": 0.5181535269709544,
+      "grad_norm": 0.7580044295676531,
+      "learning_rate": 9.840110886553629e-06,
+      "loss": 0.4798,
+      "step": 666
+    },
+    {
+      "epoch": 0.5189315352697096,
+      "grad_norm": 0.7066929236806533,
+      "learning_rate": 9.83897296289482e-06,
+      "loss": 0.4509,
+      "step": 667
+    },
+    {
+      "epoch": 0.5197095435684648,
+      "grad_norm": 0.6386371905694869,
+      "learning_rate": 9.837831070572197e-06,
+      "loss": 0.48,
+      "step": 668
+    },
+    {
+      "epoch": 0.52048755186722,
+      "grad_norm": 0.6514003163110603,
+      "learning_rate": 9.836685210522274e-06,
+      "loss": 0.4683,
+      "step": 669
+    },
+    {
+      "epoch": 0.5212655601659751,
+      "grad_norm": 0.7068429721888508,
+      "learning_rate": 9.835535383684826e-06,
+      "loss": 0.47,
+      "step": 670
+    },
+    {
+      "epoch": 0.5220435684647303,
+      "grad_norm": 0.7635482711520384,
+      "learning_rate": 9.834381591002877e-06,
+      "loss": 0.4963,
+      "step": 671
+    },
+    {
+      "epoch": 0.5228215767634855,
+      "grad_norm": 0.626358363389767,
+      "learning_rate": 9.833223833422709e-06,
+      "loss": 0.4617,
+      "step": 672
+    },
+    {
+      "epoch": 0.5235995850622407,
+      "grad_norm": 0.5838768132850389,
+      "learning_rate": 9.832062111893848e-06,
+      "loss": 0.4636,
+      "step": 673
+    },
+    {
+      "epoch": 0.5243775933609959,
+      "grad_norm": 0.6520299068197517,
+      "learning_rate": 9.830896427369076e-06,
+      "loss": 0.4659,
+      "step": 674
+    },
+    {
+      "epoch": 0.5251556016597511,
+      "grad_norm": 0.6880160677808856,
+      "learning_rate": 9.829726780804427e-06,
+      "loss": 0.4916,
+      "step": 675
+    },
+    {
+      "epoch": 0.5259336099585062,
+      "grad_norm": 0.6154792289224942,
+      "learning_rate": 9.82855317315918e-06,
+      "loss": 0.4595,
+      "step": 676
+    },
+    {
+      "epoch": 0.5267116182572614,
+      "grad_norm": 0.6273741638406989,
+      "learning_rate": 9.827375605395864e-06,
+      "loss": 0.4467,
+      "step": 677
+    },
+    {
+      "epoch": 0.5274896265560166,
+      "grad_norm": 0.619108215540595,
+      "learning_rate": 9.826194078480257e-06,
+      "loss": 0.4552,
+      "step": 678
+    },
+    {
+      "epoch": 0.5282676348547718,
+      "grad_norm": 0.5585568899692093,
+      "learning_rate": 9.825008593381383e-06,
+      "loss": 0.458,
+      "step": 679
+    },
+    {
+      "epoch": 0.529045643153527,
+      "grad_norm": 0.5553824421207891,
+      "learning_rate": 9.823819151071512e-06,
+      "loss": 0.4644,
+      "step": 680
+    },
+    {
+      "epoch": 0.5298236514522822,
+      "grad_norm": 0.5658018778442839,
+      "learning_rate": 9.822625752526162e-06,
+      "loss": 0.4649,
+      "step": 681
+    },
+    {
+      "epoch": 0.5306016597510373,
+      "grad_norm": 0.5564388778263013,
+      "learning_rate": 9.821428398724096e-06,
+      "loss": 0.4292,
+      "step": 682
+    },
+    {
+      "epoch": 0.5313796680497925,
+      "grad_norm": 0.5321173856785804,
+      "learning_rate": 9.820227090647311e-06,
+      "loss": 0.4246,
+      "step": 683
+    },
+    {
+      "epoch": 0.5321576763485477,
+      "grad_norm": 0.6487565497478155,
+      "learning_rate": 9.819021829281064e-06,
+      "loss": 0.4642,
+      "step": 684
+    },
+    {
+      "epoch": 0.5329356846473029,
+      "grad_norm": 0.6589539221491146,
+      "learning_rate": 9.81781261561384e-06,
+      "loss": 0.4658,
+      "step": 685
+    },
+    {
+      "epoch": 0.5337136929460581,
+      "grad_norm": 0.5730186828775367,
+      "learning_rate": 9.816599450637372e-06,
+      "loss": 0.4997,
+      "step": 686
+    },
+    {
+      "epoch": 0.5344917012448133,
+      "grad_norm": 0.5871535208681111,
+      "learning_rate": 9.81538233534663e-06,
+      "loss": 0.4989,
+      "step": 687
+    },
+    {
+      "epoch": 0.5352697095435685,
+      "grad_norm": 0.6118922042340182,
+      "learning_rate": 9.814161270739831e-06,
+      "loss": 0.5086,
+      "step": 688
+    },
+    {
+      "epoch": 0.5360477178423236,
+      "grad_norm": 0.561631134721975,
+      "learning_rate": 9.81293625781842e-06,
+      "loss": 0.4906,
+      "step": 689
+    },
+    {
+      "epoch": 0.5368257261410788,
+      "grad_norm": 0.5864752672307849,
+      "learning_rate": 9.811707297587091e-06,
+      "loss": 0.544,
+      "step": 690
+    },
+    {
+      "epoch": 0.537603734439834,
+      "grad_norm": 0.5501713897613124,
+      "learning_rate": 9.81047439105377e-06,
+      "loss": 0.4337,
+      "step": 691
+    },
+    {
+      "epoch": 0.5383817427385892,
+      "grad_norm": 0.6281008306927428,
+      "learning_rate": 9.80923753922962e-06,
+      "loss": 0.4567,
+      "step": 692
+    },
+    {
+      "epoch": 0.5391597510373444,
+      "grad_norm": 0.6216913391403109,
+      "learning_rate": 9.807996743129038e-06,
+      "loss": 0.4577,
+      "step": 693
+    },
+    {
+      "epoch": 0.5399377593360996,
+      "grad_norm": 0.560853709888438,
+      "learning_rate": 9.806752003769659e-06,
+      "loss": 0.5206,
+      "step": 694
+    },
+    {
+      "epoch": 0.5407157676348547,
+      "grad_norm": 0.5617482922856936,
+      "learning_rate": 9.805503322172352e-06,
+      "loss": 0.4598,
+      "step": 695
+    },
+    {
+      "epoch": 0.5414937759336099,
+      "grad_norm": 0.5992812257010258,
+      "learning_rate": 9.804250699361215e-06,
+      "loss": 0.4409,
+      "step": 696
+    },
+    {
+      "epoch": 0.5422717842323651,
+      "grad_norm": 0.6285488136504948,
+      "learning_rate": 9.802994136363586e-06,
+      "loss": 0.4796,
+      "step": 697
+    },
+    {
+      "epoch": 0.5430497925311203,
+      "grad_norm": 0.6198887636217946,
+      "learning_rate": 9.801733634210027e-06,
+      "loss": 0.502,
+      "step": 698
+    },
+    {
+      "epoch": 0.5438278008298755,
+      "grad_norm": 0.6811854033202964,
+      "learning_rate": 9.800469193934335e-06,
+      "loss": 0.4689,
+      "step": 699
+    },
+    {
+      "epoch": 0.5446058091286307,
+      "grad_norm": 0.5962332761819187,
+      "learning_rate": 9.799200816573533e-06,
+      "loss": 0.4902,
+      "step": 700
+    },
+    {
+      "epoch": 0.5453838174273858,
+      "grad_norm": 0.6537278025988176,
+      "learning_rate": 9.797928503167879e-06,
+      "loss": 0.5155,
+      "step": 701
+    },
+    {
+      "epoch": 0.546161825726141,
+      "grad_norm": 0.6247438770386486,
+      "learning_rate": 9.796652254760854e-06,
+      "loss": 0.4691,
+      "step": 702
+    },
+    {
+      "epoch": 0.5469398340248963,
+      "grad_norm": 0.6644036739125115,
+      "learning_rate": 9.795372072399168e-06,
+      "loss": 0.4474,
+      "step": 703
+    },
+    {
+      "epoch": 0.5477178423236515,
+      "grad_norm": 0.6078866559927201,
+      "learning_rate": 9.794087957132756e-06,
+      "loss": 0.4406,
+      "step": 704
+    },
+    {
+      "epoch": 0.5484958506224067,
+      "grad_norm": 0.5910481521406526,
+      "learning_rate": 9.792799910014781e-06,
+      "loss": 0.4795,
+      "step": 705
+    },
+    {
+      "epoch": 0.5492738589211619,
+      "grad_norm": 0.5693283868504624,
+      "learning_rate": 9.79150793210163e-06,
+      "loss": 0.4706,
+      "step": 706
+    },
+    {
+      "epoch": 0.5500518672199171,
+      "grad_norm": 0.5516939966491541,
+      "learning_rate": 9.790212024452912e-06,
+      "loss": 0.4623,
+      "step": 707
+    },
+    {
+      "epoch": 0.5508298755186722,
+      "grad_norm": 0.7098072207878008,
+      "learning_rate": 9.788912188131462e-06,
+      "loss": 0.4803,
+      "step": 708
+    },
+    {
+      "epoch": 0.5516078838174274,
+      "grad_norm": 0.6152577625079146,
+      "learning_rate": 9.787608424203336e-06,
+      "loss": 0.4918,
+      "step": 709
+    },
+    {
+      "epoch": 0.5523858921161826,
+      "grad_norm": 0.5619015683129449,
+      "learning_rate": 9.786300733737808e-06,
+      "loss": 0.475,
+      "step": 710
+    },
+    {
+      "epoch": 0.5531639004149378,
+      "grad_norm": 0.5691504912229283,
+      "learning_rate": 9.784989117807375e-06,
+      "loss": 0.4728,
+      "step": 711
+    },
+    {
+      "epoch": 0.553941908713693,
+      "grad_norm": 0.6009919444920512,
+      "learning_rate": 9.783673577487757e-06,
+      "loss": 0.4315,
+      "step": 712
+    },
+    {
+      "epoch": 0.5547199170124482,
+      "grad_norm": 0.6441309563035225,
+      "learning_rate": 9.782354113857886e-06,
+      "loss": 0.504,
+      "step": 713
+    },
+    {
+      "epoch": 0.5554979253112033,
+      "grad_norm": 0.6310063060784749,
+      "learning_rate": 9.781030727999916e-06,
+      "loss": 0.4856,
+      "step": 714
+    },
+    {
+      "epoch": 0.5562759336099585,
+      "grad_norm": 0.5457556071660656,
+      "learning_rate": 9.779703420999214e-06,
+      "loss": 0.4625,
+      "step": 715
+    },
+    {
+      "epoch": 0.5570539419087137,
+      "grad_norm": 0.5488756496782097,
+      "learning_rate": 9.778372193944368e-06,
+      "loss": 0.4583,
+      "step": 716
+    },
+    {
+      "epoch": 0.5578319502074689,
+      "grad_norm": 0.6190872424185517,
+      "learning_rate": 9.777037047927178e-06,
+      "loss": 0.4668,
+      "step": 717
+    },
+    {
+      "epoch": 0.5586099585062241,
+      "grad_norm": 0.5676108559783992,
+      "learning_rate": 9.775697984042658e-06,
+      "loss": 0.4634,
+      "step": 718
+    },
+    {
+      "epoch": 0.5593879668049793,
+      "grad_norm": 0.5540243901332765,
+      "learning_rate": 9.774355003389037e-06,
+      "loss": 0.4739,
+      "step": 719
+    },
+    {
+      "epoch": 0.5601659751037344,
+      "grad_norm": 0.7518402777104115,
+      "learning_rate": 9.773008107067753e-06,
+      "loss": 0.5375,
+      "step": 720
+    },
+    {
+      "epoch": 0.5609439834024896,
+      "grad_norm": 0.6600964517903705,
+      "learning_rate": 9.771657296183458e-06,
+      "loss": 0.477,
+      "step": 721
+    },
+    {
+      "epoch": 0.5617219917012448,
+      "grad_norm": 0.6306088420099737,
+      "learning_rate": 9.770302571844015e-06,
+      "loss": 0.4888,
+      "step": 722
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.5427250598317943,
+      "learning_rate": 9.768943935160495e-06,
+      "loss": 0.4144,
+      "step": 723
+    },
+    {
+      "epoch": 0.5632780082987552,
+      "grad_norm": 0.6738267381870775,
+      "learning_rate": 9.76758138724718e-06,
+      "loss": 0.4706,
+      "step": 724
+    },
+    {
+      "epoch": 0.5640560165975104,
+      "grad_norm": 0.6415517884780721,
+      "learning_rate": 9.766214929221555e-06,
+      "loss": 0.4696,
+      "step": 725
+    },
+    {
+      "epoch": 0.5648340248962656,
+      "grad_norm": 0.5985020972068458,
+      "learning_rate": 9.764844562204316e-06,
+      "loss": 0.4561,
+      "step": 726
+    },
+    {
+      "epoch": 0.5656120331950207,
+      "grad_norm": 0.659244069052128,
+      "learning_rate": 9.763470287319366e-06,
+      "loss": 0.4853,
+      "step": 727
+    },
+    {
+      "epoch": 0.5663900414937759,
+      "grad_norm": 0.6202313434053371,
+      "learning_rate": 9.762092105693809e-06,
+      "loss": 0.4813,
+      "step": 728
+    },
+    {
+      "epoch": 0.5671680497925311,
+      "grad_norm": 0.5631074293241833,
+      "learning_rate": 9.760710018457953e-06,
+      "loss": 0.4634,
+      "step": 729
+    },
+    {
+      "epoch": 0.5679460580912863,
+      "grad_norm": 0.5832036568621398,
+      "learning_rate": 9.759324026745313e-06,
+      "loss": 0.4554,
+      "step": 730
+    },
+    {
+      "epoch": 0.5687240663900415,
+      "grad_norm": 0.6083079101054238,
+      "learning_rate": 9.757934131692608e-06,
+      "loss": 0.4391,
+      "step": 731
+    },
+    {
+      "epoch": 0.5695020746887967,
+      "grad_norm": 0.5788445852396586,
+      "learning_rate": 9.75654033443975e-06,
+      "loss": 0.4735,
+      "step": 732
+    },
+    {
+      "epoch": 0.5702800829875518,
+      "grad_norm": 0.6616658280287437,
+      "learning_rate": 9.755142636129857e-06,
+      "loss": 0.4879,
+      "step": 733
+    },
+    {
+      "epoch": 0.571058091286307,
+      "grad_norm": 0.5828179925719302,
+      "learning_rate": 9.753741037909248e-06,
+      "loss": 0.4125,
+      "step": 734
+    },
+    {
+      "epoch": 0.5718360995850622,
+      "grad_norm": 0.6167617373761645,
+      "learning_rate": 9.752335540927436e-06,
+      "loss": 0.4619,
+      "step": 735
+    },
+    {
+      "epoch": 0.5726141078838174,
+      "grad_norm": 0.6297533010469978,
+      "learning_rate": 9.750926146337132e-06,
+      "loss": 0.4386,
+      "step": 736
+    },
+    {
+      "epoch": 0.5733921161825726,
+      "grad_norm": 0.5523974928333167,
+      "learning_rate": 9.749512855294248e-06,
+      "loss": 0.4604,
+      "step": 737
+    },
+    {
+      "epoch": 0.5741701244813278,
+      "grad_norm": 0.5979997175296589,
+      "learning_rate": 9.748095668957887e-06,
+      "loss": 0.4817,
+      "step": 738
+    },
+    {
+      "epoch": 0.5749481327800829,
+      "grad_norm": 0.6670919835889857,
+      "learning_rate": 9.74667458849035e-06,
+      "loss": 0.4501,
+      "step": 739
+    },
+    {
+      "epoch": 0.5757261410788381,
+      "grad_norm": 0.644102821034043,
+      "learning_rate": 9.74524961505713e-06,
+      "loss": 0.5224,
+      "step": 740
+    },
+    {
+      "epoch": 0.5765041493775933,
+      "grad_norm": 0.6020598572676638,
+      "learning_rate": 9.743820749826914e-06,
+      "loss": 0.4837,
+      "step": 741
+    },
+    {
+      "epoch": 0.5772821576763485,
+      "grad_norm": 0.7627863564982239,
+      "learning_rate": 9.742387993971576e-06,
+      "loss": 0.5386,
+      "step": 742
+    },
+    {
+      "epoch": 0.5780601659751037,
+      "grad_norm": 0.640023135419784,
+      "learning_rate": 9.74095134866619e-06,
+      "loss": 0.5054,
+      "step": 743
+    },
+    {
+      "epoch": 0.578838174273859,
+      "grad_norm": 0.6172928388881246,
+      "learning_rate": 9.73951081508901e-06,
+      "loss": 0.4753,
+      "step": 744
+    },
+    {
+      "epoch": 0.5796161825726142,
+      "grad_norm": 0.5864972650095904,
+      "learning_rate": 9.738066394421488e-06,
+      "loss": 0.5098,
+      "step": 745
+    },
+    {
+      "epoch": 0.5803941908713693,
+      "grad_norm": 0.5454152983915318,
+      "learning_rate": 9.736618087848257e-06,
+      "loss": 0.4854,
+      "step": 746
+    },
+    {
+      "epoch": 0.5811721991701245,
+      "grad_norm": 0.585562201549239,
+      "learning_rate": 9.73516589655714e-06,
+      "loss": 0.4758,
+      "step": 747
+    },
+    {
+      "epoch": 0.5819502074688797,
+      "grad_norm": 0.6107967125692332,
+      "learning_rate": 9.733709821739149e-06,
+      "loss": 0.4827,
+      "step": 748
+    },
+    {
+      "epoch": 0.5827282157676349,
+      "grad_norm": 0.5734055580946169,
+      "learning_rate": 9.732249864588471e-06,
+      "loss": 0.428,
+      "step": 749
+    },
+    {
+      "epoch": 0.5835062240663901,
+      "grad_norm": 0.5300209335045187,
+      "learning_rate": 9.73078602630249e-06,
+      "loss": 0.4459,
+      "step": 750
+    },
+    {
+      "epoch": 0.5842842323651453,
+      "grad_norm": 0.6037430092108411,
+      "learning_rate": 9.729318308081762e-06,
+      "loss": 0.4981,
+      "step": 751
+    },
+    {
+      "epoch": 0.5850622406639004,
+      "grad_norm": 0.6498589483287653,
+      "learning_rate": 9.727846711130037e-06,
+      "loss": 0.4528,
+      "step": 752
+    },
+    {
+      "epoch": 0.5858402489626556,
+      "grad_norm": 0.5796146711990907,
+      "learning_rate": 9.726371236654232e-06,
+      "loss": 0.4569,
+      "step": 753
+    },
+    {
+      "epoch": 0.5866182572614108,
+      "grad_norm": 0.5633169459772102,
+      "learning_rate": 9.724891885864456e-06,
+      "loss": 0.4799,
+      "step": 754
+    },
+    {
+      "epoch": 0.587396265560166,
+      "grad_norm": 0.595664548424352,
+      "learning_rate": 9.723408659973989e-06,
+      "loss": 0.4314,
+      "step": 755
+    },
+    {
+      "epoch": 0.5881742738589212,
+      "grad_norm": 0.6562418766006404,
+      "learning_rate": 9.721921560199297e-06,
+      "loss": 0.4745,
+      "step": 756
+    },
+    {
+      "epoch": 0.5889522821576764,
+      "grad_norm": 0.7033352564128407,
+      "learning_rate": 9.720430587760014e-06,
+      "loss": 0.4765,
+      "step": 757
+    },
+    {
+      "epoch": 0.5897302904564315,
+      "grad_norm": 0.5565094692962064,
+      "learning_rate": 9.71893574387896e-06,
+      "loss": 0.4272,
+      "step": 758
+    },
+    {
+      "epoch": 0.5905082987551867,
+      "grad_norm": 0.6825815330757248,
+      "learning_rate": 9.71743702978212e-06,
+      "loss": 0.466,
+      "step": 759
+    },
+    {
+      "epoch": 0.5912863070539419,
+      "grad_norm": 0.6638384056326533,
+      "learning_rate": 9.71593444669866e-06,
+      "loss": 0.4516,
+      "step": 760
+    },
+    {
+      "epoch": 0.5920643153526971,
+      "grad_norm": 0.595630921177449,
+      "learning_rate": 9.71442799586092e-06,
+      "loss": 0.442,
+      "step": 761
+    },
+    {
+      "epoch": 0.5928423236514523,
+      "grad_norm": 0.5635475094467342,
+      "learning_rate": 9.712917678504405e-06,
+      "loss": 0.4595,
+      "step": 762
+    },
+    {
+      "epoch": 0.5936203319502075,
+      "grad_norm": 0.5777129654543609,
+      "learning_rate": 9.7114034958678e-06,
+      "loss": 0.471,
+      "step": 763
+    },
+    {
+      "epoch": 0.5943983402489627,
+      "grad_norm": 0.6375432349241724,
+      "learning_rate": 9.709885449192952e-06,
+      "loss": 0.4616,
+      "step": 764
+    },
+    {
+      "epoch": 0.5951763485477178,
+      "grad_norm": 0.6584987778164693,
+      "learning_rate": 9.708363539724883e-06,
+      "loss": 0.513,
+      "step": 765
+    },
+    {
+      "epoch": 0.595954356846473,
+      "grad_norm": 0.5676109930146178,
+      "learning_rate": 9.70683776871178e-06,
+      "loss": 0.4468,
+      "step": 766
+    },
+    {
+      "epoch": 0.5967323651452282,
+      "grad_norm": 0.6217992282197027,
+      "learning_rate": 9.705308137404997e-06,
+      "loss": 0.4871,
+      "step": 767
+    },
+    {
+      "epoch": 0.5975103734439834,
+      "grad_norm": 0.6348220952243604,
+      "learning_rate": 9.703774647059057e-06,
+      "loss": 0.4788,
+      "step": 768
+    },
+    {
+      "epoch": 0.5982883817427386,
+      "grad_norm": 0.6623474494330938,
+      "learning_rate": 9.702237298931645e-06,
+      "loss": 0.4909,
+      "step": 769
+    },
+    {
+      "epoch": 0.5990663900414938,
+      "grad_norm": 0.5617920020174203,
+      "learning_rate": 9.700696094283608e-06,
+      "loss": 0.4583,
+      "step": 770
+    },
+    {
+      "epoch": 0.5998443983402489,
+      "grad_norm": 0.6012979629603882,
+      "learning_rate": 9.699151034378965e-06,
+      "loss": 0.4465,
+      "step": 771
+    },
+    {
+      "epoch": 0.6006224066390041,
+      "grad_norm": 0.5747961613319648,
+      "learning_rate": 9.697602120484883e-06,
+      "loss": 0.4847,
+      "step": 772
+    },
+    {
+      "epoch": 0.6014004149377593,
+      "grad_norm": 0.5510081404586327,
+      "learning_rate": 9.696049353871701e-06,
+      "loss": 0.4625,
+      "step": 773
+    },
+    {
+      "epoch": 0.6021784232365145,
+      "grad_norm": 0.5433348504460747,
+      "learning_rate": 9.694492735812917e-06,
+      "loss": 0.4979,
+      "step": 774
+    },
+    {
+      "epoch": 0.6029564315352697,
+      "grad_norm": 0.6634099024093211,
+      "learning_rate": 9.69293226758518e-06,
+      "loss": 0.4889,
+      "step": 775
+    },
+    {
+      "epoch": 0.6037344398340249,
+      "grad_norm": 0.5678552008776818,
+      "learning_rate": 9.691367950468303e-06,
+      "loss": 0.4653,
+      "step": 776
+    },
+    {
+      "epoch": 0.60451244813278,
+      "grad_norm": 0.6453396020813985,
+      "learning_rate": 9.689799785745256e-06,
+      "loss": 0.5215,
+      "step": 777
+    },
+    {
+      "epoch": 0.6052904564315352,
+      "grad_norm": 0.5698574505309918,
+      "learning_rate": 9.688227774702162e-06,
+      "loss": 0.4449,
+      "step": 778
+    },
+    {
+      "epoch": 0.6060684647302904,
+      "grad_norm": 0.6185073931050156,
+      "learning_rate": 9.686651918628298e-06,
+      "loss": 0.4837,
+      "step": 779
+    },
+    {
+      "epoch": 0.6068464730290456,
+      "grad_norm": 0.5391506445997648,
+      "learning_rate": 9.685072218816098e-06,
+      "loss": 0.4577,
+      "step": 780
+    },
+    {
+      "epoch": 0.6076244813278008,
+      "grad_norm": 0.5792597283715439,
+      "learning_rate": 9.683488676561143e-06,
+      "loss": 0.4776,
+      "step": 781
+    },
+    {
+      "epoch": 0.608402489626556,
+      "grad_norm": 0.49389094768277225,
+      "learning_rate": 9.681901293162175e-06,
+      "loss": 0.4539,
+      "step": 782
+    },
+    {
+      "epoch": 0.6091804979253111,
+      "grad_norm": 0.5752331462776296,
+      "learning_rate": 9.680310069921072e-06,
+      "loss": 0.5012,
+      "step": 783
+    },
+    {
+      "epoch": 0.6099585062240664,
+      "grad_norm": 0.5890520777513659,
+      "learning_rate": 9.678715008142875e-06,
+      "loss": 0.4752,
+      "step": 784
+    },
+    {
+      "epoch": 0.6107365145228216,
+      "grad_norm": 0.5581468997868002,
+      "learning_rate": 9.677116109135767e-06,
+      "loss": 0.472,
+      "step": 785
+    },
+    {
+      "epoch": 0.6115145228215768,
+      "grad_norm": 0.6768407095358085,
+      "learning_rate": 9.675513374211077e-06,
+      "loss": 0.5051,
+      "step": 786
+    },
+    {
+      "epoch": 0.612292531120332,
+      "grad_norm": 0.5601930217619304,
+      "learning_rate": 9.673906804683279e-06,
+      "loss": 0.4572,
+      "step": 787
+    },
+    {
+      "epoch": 0.6130705394190872,
+      "grad_norm": 0.575741653112418,
+      "learning_rate": 9.672296401869999e-06,
+      "loss": 0.4756,
+      "step": 788
+    },
+    {
+      "epoch": 0.6138485477178424,
+      "grad_norm": 0.6381038346742645,
+      "learning_rate": 9.670682167092e-06,
+      "loss": 0.4542,
+      "step": 789
+    },
+    {
+      "epoch": 0.6146265560165975,
+      "grad_norm": 0.5303999644298575,
+      "learning_rate": 9.66906410167319e-06,
+      "loss": 0.4049,
+      "step": 790
+    },
+    {
+      "epoch": 0.6154045643153527,
+      "grad_norm": 0.5802991195090851,
+      "learning_rate": 9.667442206940618e-06,
+      "loss": 0.4562,
+      "step": 791
+    },
+    {
+      "epoch": 0.6161825726141079,
+      "grad_norm": 0.5278688785312275,
+      "learning_rate": 9.665816484224477e-06,
+      "loss": 0.4545,
+      "step": 792
+    },
+    {
+      "epoch": 0.6169605809128631,
+      "grad_norm": 0.6996245704183648,
+      "learning_rate": 9.664186934858093e-06,
+      "loss": 0.4857,
+      "step": 793
+    },
+    {
+      "epoch": 0.6177385892116183,
+      "grad_norm": 0.6378772423002602,
+      "learning_rate": 9.66255356017794e-06,
+      "loss": 0.4863,
+      "step": 794
+    },
+    {
+      "epoch": 0.6185165975103735,
+      "grad_norm": 0.6165109730125118,
+      "learning_rate": 9.660916361523616e-06,
+      "loss": 0.4566,
+      "step": 795
+    },
+    {
+      "epoch": 0.6192946058091287,
+      "grad_norm": 0.5270705513134124,
+      "learning_rate": 9.659275340237869e-06,
+      "loss": 0.434,
+      "step": 796
+    },
+    {
+      "epoch": 0.6200726141078838,
+      "grad_norm": 0.5619472206674515,
+      "learning_rate": 9.65763049766657e-06,
+      "loss": 0.4607,
+      "step": 797
+    },
+    {
+      "epoch": 0.620850622406639,
+      "grad_norm": 0.6144720495931283,
+      "learning_rate": 9.655981835158736e-06,
+      "loss": 0.4357,
+      "step": 798
+    },
+    {
+      "epoch": 0.6216286307053942,
+      "grad_norm": 0.6201498032536459,
+      "learning_rate": 9.654329354066507e-06,
+      "loss": 0.4803,
+      "step": 799
+    },
+    {
+      "epoch": 0.6224066390041494,
+      "grad_norm": 0.6410618692792788,
+      "learning_rate": 9.65267305574516e-06,
+      "loss": 0.4694,
+      "step": 800
+    },
+    {
+      "epoch": 0.6231846473029046,
+      "grad_norm": 0.7122900847483166,
+      "learning_rate": 9.651012941553102e-06,
+      "loss": 0.4996,
+      "step": 801
+    },
+    {
+      "epoch": 0.6239626556016598,
+      "grad_norm": 0.579820359095041,
+      "learning_rate": 9.649349012851868e-06,
+      "loss": 0.4907,
+      "step": 802
+    },
+    {
+      "epoch": 0.6247406639004149,
+      "grad_norm": 0.588847016473329,
+      "learning_rate": 9.647681271006122e-06,
+      "loss": 0.453,
+      "step": 803
+    },
+    {
+      "epoch": 0.6255186721991701,
+      "grad_norm": 0.5630980187325512,
+      "learning_rate": 9.646009717383655e-06,
+      "loss": 0.4661,
+      "step": 804
+    },
+    {
+      "epoch": 0.6262966804979253,
+      "grad_norm": 0.5809259318187365,
+      "learning_rate": 9.644334353355386e-06,
+      "loss": 0.4668,
+      "step": 805
+    },
+    {
+      "epoch": 0.6270746887966805,
+      "grad_norm": 0.61530580256758,
+      "learning_rate": 9.642655180295357e-06,
+      "loss": 0.4402,
+      "step": 806
+    },
+    {
+      "epoch": 0.6278526970954357,
+      "grad_norm": 0.6705809828860775,
+      "learning_rate": 9.640972199580735e-06,
+      "loss": 0.4817,
+      "step": 807
+    },
+    {
+      "epoch": 0.6286307053941909,
+      "grad_norm": 0.6325036540547536,
+      "learning_rate": 9.639285412591812e-06,
+      "loss": 0.4448,
+      "step": 808
+    },
+    {
+      "epoch": 0.629408713692946,
+      "grad_norm": 0.5449547674864282,
+      "learning_rate": 9.637594820711996e-06,
+      "loss": 0.4557,
+      "step": 809
+    },
+    {
+      "epoch": 0.6301867219917012,
+      "grad_norm": 0.6057670293186024,
+      "learning_rate": 9.635900425327822e-06,
+      "loss": 0.4718,
+      "step": 810
+    },
+    {
+      "epoch": 0.6309647302904564,
+      "grad_norm": 0.6203522269836137,
+      "learning_rate": 9.634202227828938e-06,
+      "loss": 0.4632,
+      "step": 811
+    },
+    {
+      "epoch": 0.6317427385892116,
+      "grad_norm": 0.6897385444868245,
+      "learning_rate": 9.632500229608118e-06,
+      "loss": 0.5385,
+      "step": 812
+    },
+    {
+      "epoch": 0.6325207468879668,
+      "grad_norm": 0.6272344763984108,
+      "learning_rate": 9.630794432061243e-06,
+      "loss": 0.4383,
+      "step": 813
+    },
+    {
+      "epoch": 0.633298755186722,
+      "grad_norm": 0.6898853608882402,
+      "learning_rate": 9.629084836587324e-06,
+      "loss": 0.4603,
+      "step": 814
+    },
+    {
+      "epoch": 0.6340767634854771,
+      "grad_norm": 0.6658695415087533,
+      "learning_rate": 9.62737144458847e-06,
+      "loss": 0.477,
+      "step": 815
+    },
+    {
+      "epoch": 0.6348547717842323,
+      "grad_norm": 0.7169618367294606,
+      "learning_rate": 9.625654257469919e-06,
+      "loss": 0.452,
+      "step": 816
+    },
+    {
+      "epoch": 0.6356327800829875,
+      "grad_norm": 0.6397281084344514,
+      "learning_rate": 9.623933276640012e-06,
+      "loss": 0.4658,
+      "step": 817
+    },
+    {
+      "epoch": 0.6364107883817427,
+      "grad_norm": 0.6185751117960256,
+      "learning_rate": 9.622208503510204e-06,
+      "loss": 0.4805,
+      "step": 818
+    },
+    {
+      "epoch": 0.6371887966804979,
+      "grad_norm": 0.7622166723300784,
+      "learning_rate": 9.620479939495062e-06,
+      "loss": 0.4586,
+      "step": 819
+    },
+    {
+      "epoch": 0.6379668049792531,
+      "grad_norm": 0.6807976855395602,
+      "learning_rate": 9.618747586012262e-06,
+      "loss": 0.4485,
+      "step": 820
+    },
+    {
+      "epoch": 0.6387448132780082,
+      "grad_norm": 0.5950234332707601,
+      "learning_rate": 9.617011444482582e-06,
+      "loss": 0.4555,
+      "step": 821
+    },
+    {
+      "epoch": 0.6395228215767634,
+      "grad_norm": 0.547184750526114,
+      "learning_rate": 9.615271516329916e-06,
+      "loss": 0.4714,
+      "step": 822
+    },
+    {
+      "epoch": 0.6403008298755186,
+      "grad_norm": 0.5989033451796816,
+      "learning_rate": 9.613527802981255e-06,
+      "loss": 0.4518,
+      "step": 823
+    },
+    {
+      "epoch": 0.6410788381742739,
+      "grad_norm": 0.6126959503885495,
+      "learning_rate": 9.611780305866703e-06,
+      "loss": 0.4758,
+      "step": 824
+    },
+    {
+      "epoch": 0.6418568464730291,
+      "grad_norm": 0.6108699622262126,
+      "learning_rate": 9.610029026419458e-06,
+      "loss": 0.5086,
+      "step": 825
+    },
+    {
+      "epoch": 0.6426348547717843,
+      "grad_norm": 0.657967054606472,
+      "learning_rate": 9.608273966075827e-06,
+      "loss": 0.4739,
+      "step": 826
+    },
+    {
+      "epoch": 0.6434128630705395,
+      "grad_norm": 0.6204756067050131,
+      "learning_rate": 9.606515126275216e-06,
+      "loss": 0.4238,
+      "step": 827
+    },
+    {
+      "epoch": 0.6441908713692946,
+      "grad_norm": 0.55874484124918,
+      "learning_rate": 9.60475250846013e-06,
+      "loss": 0.5012,
+      "step": 828
+    },
+    {
+      "epoch": 0.6449688796680498,
+      "grad_norm": 0.5742393116114516,
+      "learning_rate": 9.602986114076169e-06,
+      "loss": 0.4664,
+      "step": 829
+    },
+    {
+      "epoch": 0.645746887966805,
+      "grad_norm": 0.6614309707715372,
+      "learning_rate": 9.60121594457204e-06,
+      "loss": 0.4708,
+      "step": 830
+    },
+    {
+      "epoch": 0.6465248962655602,
+      "grad_norm": 0.5879756612576721,
+      "learning_rate": 9.599442001399537e-06,
+      "loss": 0.4532,
+      "step": 831
+    },
+    {
+      "epoch": 0.6473029045643154,
+      "grad_norm": 0.6078404830051827,
+      "learning_rate": 9.597664286013552e-06,
+      "loss": 0.4767,
+      "step": 832
+    },
+    {
+      "epoch": 0.6480809128630706,
+      "grad_norm": 0.5783987366241182,
+      "learning_rate": 9.595882799872072e-06,
+      "loss": 0.4749,
+      "step": 833
+    },
+    {
+      "epoch": 0.6488589211618258,
+      "grad_norm": 0.5826648288214097,
+      "learning_rate": 9.594097544436177e-06,
+      "loss": 0.4862,
+      "step": 834
+    },
+    {
+      "epoch": 0.6496369294605809,
+      "grad_norm": 0.5831102665115827,
+      "learning_rate": 9.592308521170033e-06,
+      "loss": 0.4837,
+      "step": 835
+    },
+    {
+      "epoch": 0.6504149377593361,
+      "grad_norm": 0.48969089422146966,
+      "learning_rate": 9.590515731540903e-06,
+      "loss": 0.4345,
+      "step": 836
+    },
+    {
+      "epoch": 0.6511929460580913,
+      "grad_norm": 0.5710144251785657,
+      "learning_rate": 9.588719177019138e-06,
+      "loss": 0.4513,
+      "step": 837
+    },
+    {
+      "epoch": 0.6519709543568465,
+      "grad_norm": 0.5794740608070689,
+      "learning_rate": 9.586918859078166e-06,
+      "loss": 0.4167,
+      "step": 838
+    },
+    {
+      "epoch": 0.6527489626556017,
+      "grad_norm": 0.6075973535456305,
+      "learning_rate": 9.585114779194521e-06,
+      "loss": 0.5207,
+      "step": 839
+    },
+    {
+      "epoch": 0.6535269709543569,
+      "grad_norm": 0.6129190479610004,
+      "learning_rate": 9.583306938847807e-06,
+      "loss": 0.4727,
+      "step": 840
+    },
+    {
+      "epoch": 0.654304979253112,
+      "grad_norm": 0.6345691104036318,
+      "learning_rate": 9.581495339520717e-06,
+      "loss": 0.4446,
+      "step": 841
+    },
+    {
+      "epoch": 0.6550829875518672,
+      "grad_norm": 0.6569872139264525,
+      "learning_rate": 9.579679982699026e-06,
+      "loss": 0.4908,
+      "step": 842
+    },
+    {
+      "epoch": 0.6558609958506224,
+      "grad_norm": 0.6964550598517922,
+      "learning_rate": 9.577860869871592e-06,
+      "loss": 0.5231,
+      "step": 843
+    },
+    {
+      "epoch": 0.6566390041493776,
+      "grad_norm": 0.7449171336621091,
+      "learning_rate": 9.576038002530355e-06,
+      "loss": 0.5117,
+      "step": 844
+    },
+    {
+      "epoch": 0.6574170124481328,
+      "grad_norm": 0.5714850833460681,
+      "learning_rate": 9.574211382170328e-06,
+      "loss": 0.4319,
+      "step": 845
+    },
+    {
+      "epoch": 0.658195020746888,
+      "grad_norm": 0.5389025913152753,
+      "learning_rate": 9.57238101028961e-06,
+      "loss": 0.4585,
+      "step": 846
+    },
+    {
+      "epoch": 0.6589730290456431,
+      "grad_norm": 0.5898593540212483,
+      "learning_rate": 9.570546888389372e-06,
+      "loss": 0.4689,
+      "step": 847
+    },
+    {
+      "epoch": 0.6597510373443983,
+      "grad_norm": 0.6820455559247703,
+      "learning_rate": 9.56870901797386e-06,
+      "loss": 0.4898,
+      "step": 848
+    },
+    {
+      "epoch": 0.6605290456431535,
+      "grad_norm": 0.6779155586076264,
+      "learning_rate": 9.566867400550396e-06,
+      "loss": 0.4351,
+      "step": 849
+    },
+    {
+      "epoch": 0.6613070539419087,
+      "grad_norm": 0.6863892224715249,
+      "learning_rate": 9.565022037629379e-06,
+      "loss": 0.4875,
+      "step": 850
+    },
+    {
+      "epoch": 0.6620850622406639,
+      "grad_norm": 0.5681432471066956,
+      "learning_rate": 9.56317293072427e-06,
+      "loss": 0.4489,
+      "step": 851
+    },
+    {
+      "epoch": 0.6628630705394191,
+      "grad_norm": 0.6331011047907313,
+      "learning_rate": 9.56132008135161e-06,
+      "loss": 0.4807,
+      "step": 852
+    },
+    {
+      "epoch": 0.6636410788381742,
+      "grad_norm": 0.5648308193616103,
+      "learning_rate": 9.559463491031006e-06,
+      "loss": 0.3958,
+      "step": 853
+    },
+    {
+      "epoch": 0.6644190871369294,
+      "grad_norm": 0.6008664307698174,
+      "learning_rate": 9.557603161285132e-06,
+      "loss": 0.4638,
+      "step": 854
+    },
+    {
+      "epoch": 0.6651970954356846,
+      "grad_norm": 0.6160144881973248,
+      "learning_rate": 9.555739093639728e-06,
+      "loss": 0.4627,
+      "step": 855
+    },
+    {
+      "epoch": 0.6659751037344398,
+      "grad_norm": 0.5659005276159228,
+      "learning_rate": 9.553871289623605e-06,
+      "loss": 0.4378,
+      "step": 856
+    },
+    {
+      "epoch": 0.666753112033195,
+      "grad_norm": 0.5955498580226579,
+      "learning_rate": 9.551999750768632e-06,
+      "loss": 0.5078,
+      "step": 857
+    },
+    {
+      "epoch": 0.6675311203319502,
+      "grad_norm": 0.54075576471204,
+      "learning_rate": 9.550124478609744e-06,
+      "loss": 0.4281,
+      "step": 858
+    },
+    {
+      "epoch": 0.6683091286307054,
+      "grad_norm": 0.627825076606112,
+      "learning_rate": 9.54824547468494e-06,
+      "loss": 0.4686,
+      "step": 859
+    },
+    {
+      "epoch": 0.6690871369294605,
+      "grad_norm": 0.5447970861254994,
+      "learning_rate": 9.546362740535276e-06,
+      "loss": 0.4568,
+      "step": 860
+    },
+    {
+      "epoch": 0.6698651452282157,
+      "grad_norm": 0.7141173558124233,
+      "learning_rate": 9.544476277704869e-06,
+      "loss": 0.4748,
+      "step": 861
+    },
+    {
+      "epoch": 0.6706431535269709,
+      "grad_norm": 0.719870845291295,
+      "learning_rate": 9.542586087740893e-06,
+      "loss": 0.5006,
+      "step": 862
+    },
+    {
+      "epoch": 0.6714211618257261,
+      "grad_norm": 0.6094355534228667,
+      "learning_rate": 9.54069217219358e-06,
+      "loss": 0.4957,
+      "step": 863
+    },
+    {
+      "epoch": 0.6721991701244814,
+      "grad_norm": 0.5927091817777506,
+      "learning_rate": 9.538794532616216e-06,
+      "loss": 0.4241,
+      "step": 864
+    },
+    {
+      "epoch": 0.6729771784232366,
+      "grad_norm": 0.5433179048104608,
+      "learning_rate": 9.536893170565145e-06,
+      "loss": 0.4042,
+      "step": 865
+    },
+    {
+      "epoch": 0.6737551867219918,
+      "grad_norm": 0.6229505412901453,
+      "learning_rate": 9.53498808759976e-06,
+      "loss": 0.4574,
+      "step": 866
+    },
+    {
+      "epoch": 0.6745331950207469,
+      "grad_norm": 0.617830949640727,
+      "learning_rate": 9.533079285282505e-06,
+      "loss": 0.4675,
+      "step": 867
+    },
+    {
+      "epoch": 0.6753112033195021,
+      "grad_norm": 0.5880387544589057,
+      "learning_rate": 9.531166765178878e-06,
+      "loss": 0.47,
+      "step": 868
+    },
+    {
+      "epoch": 0.6760892116182573,
+      "grad_norm": 0.6055521078691127,
+      "learning_rate": 9.529250528857425e-06,
+      "loss": 0.4503,
+      "step": 869
+    },
+    {
+      "epoch": 0.6768672199170125,
+      "grad_norm": 0.5480595126262697,
+      "learning_rate": 9.52733057788974e-06,
+      "loss": 0.4511,
+      "step": 870
+    },
+    {
+      "epoch": 0.6776452282157677,
+      "grad_norm": 0.6787929635960235,
+      "learning_rate": 9.52540691385046e-06,
+      "loss": 0.4511,
+      "step": 871
+    },
+    {
+      "epoch": 0.6784232365145229,
+      "grad_norm": 0.638717237141898,
+      "learning_rate": 9.523479538317274e-06,
+      "loss": 0.4822,
+      "step": 872
+    },
+    {
+      "epoch": 0.679201244813278,
+      "grad_norm": 0.6092211435464282,
+      "learning_rate": 9.521548452870908e-06,
+      "loss": 0.4897,
+      "step": 873
+    },
+    {
+      "epoch": 0.6799792531120332,
+      "grad_norm": 0.6177679284317467,
+      "learning_rate": 9.519613659095134e-06,
+      "loss": 0.4619,
+      "step": 874
+    },
+    {
+      "epoch": 0.6807572614107884,
+      "grad_norm": 0.5977880166825478,
+      "learning_rate": 9.517675158576767e-06,
+      "loss": 0.465,
+      "step": 875
+    },
+    {
+      "epoch": 0.6815352697095436,
+      "grad_norm": 0.6254477021763162,
+      "learning_rate": 9.515732952905659e-06,
+      "loss": 0.4552,
+      "step": 876
+    },
+    {
+      "epoch": 0.6823132780082988,
+      "grad_norm": 0.633251328226882,
+      "learning_rate": 9.513787043674702e-06,
+      "loss": 0.4937,
+      "step": 877
+    },
+    {
+      "epoch": 0.683091286307054,
+      "grad_norm": 0.7102110419748336,
+      "learning_rate": 9.511837432479828e-06,
+      "loss": 0.5006,
+      "step": 878
+    },
+    {
+      "epoch": 0.6838692946058091,
+      "grad_norm": 0.561376019412783,
+      "learning_rate": 9.509884120919997e-06,
+      "loss": 0.4709,
+      "step": 879
+    },
+    {
+      "epoch": 0.6846473029045643,
+      "grad_norm": 0.565536817130948,
+      "learning_rate": 9.507927110597215e-06,
+      "loss": 0.4491,
+      "step": 880
+    },
+    {
+      "epoch": 0.6854253112033195,
+      "grad_norm": 0.6004910412516672,
+      "learning_rate": 9.505966403116514e-06,
+      "loss": 0.4346,
+      "step": 881
+    },
+    {
+      "epoch": 0.6862033195020747,
+      "grad_norm": 0.5838999320620706,
+      "learning_rate": 9.504002000085958e-06,
+      "loss": 0.4508,
+      "step": 882
+    },
+    {
+      "epoch": 0.6869813278008299,
+      "grad_norm": 0.6223564422059608,
+      "learning_rate": 9.50203390311665e-06,
+      "loss": 0.515,
+      "step": 883
+    },
+    {
+      "epoch": 0.6877593360995851,
+      "grad_norm": 0.5939234709444422,
+      "learning_rate": 9.50006211382271e-06,
+      "loss": 0.4154,
+      "step": 884
+    },
+    {
+      "epoch": 0.6885373443983402,
+      "grad_norm": 0.6710670589548863,
+      "learning_rate": 9.498086633821295e-06,
+      "loss": 0.4589,
+      "step": 885
+    },
+    {
+      "epoch": 0.6893153526970954,
+      "grad_norm": 0.6161701580025164,
+      "learning_rate": 9.496107464732588e-06,
+      "loss": 0.4755,
+      "step": 886
+    },
+    {
+      "epoch": 0.6900933609958506,
+      "grad_norm": 0.5762652652960636,
+      "learning_rate": 9.494124608179797e-06,
+      "loss": 0.4611,
+      "step": 887
+    },
+    {
+      "epoch": 0.6908713692946058,
+      "grad_norm": 0.6045799150208313,
+      "learning_rate": 9.492138065789153e-06,
+      "loss": 0.421,
+      "step": 888
+    },
+    {
+      "epoch": 0.691649377593361,
+      "grad_norm": 0.6816106131176702,
+      "learning_rate": 9.49014783918991e-06,
+      "loss": 0.4752,
+      "step": 889
+    },
+    {
+      "epoch": 0.6924273858921162,
+      "grad_norm": 0.7067525030485823,
+      "learning_rate": 9.488153930014344e-06,
+      "loss": 0.4487,
+      "step": 890
+    },
+    {
+      "epoch": 0.6932053941908713,
+      "grad_norm": 0.6735442055631788,
+      "learning_rate": 9.48615633989775e-06,
+      "loss": 0.4776,
+      "step": 891
+    },
+    {
+      "epoch": 0.6939834024896265,
+      "grad_norm": 0.6528878912839085,
+      "learning_rate": 9.484155070478447e-06,
+      "loss": 0.5098,
+      "step": 892
+    },
+    {
+      "epoch": 0.6947614107883817,
+      "grad_norm": 0.7178012175880448,
+      "learning_rate": 9.482150123397766e-06,
+      "loss": 0.461,
+      "step": 893
+    },
+    {
+      "epoch": 0.6955394190871369,
+      "grad_norm": 0.6629316575488973,
+      "learning_rate": 9.480141500300056e-06,
+      "loss": 0.5082,
+      "step": 894
+    },
+    {
+      "epoch": 0.6963174273858921,
+      "grad_norm": 0.520819337818098,
+      "learning_rate": 9.47812920283268e-06,
+      "loss": 0.4164,
+      "step": 895
+    },
+    {
+      "epoch": 0.6970954356846473,
+      "grad_norm": 0.6746097021837089,
+      "learning_rate": 9.476113232646018e-06,
+      "loss": 0.4695,
+      "step": 896
+    },
+    {
+      "epoch": 0.6978734439834025,
+      "grad_norm": 0.6795057480533362,
+      "learning_rate": 9.474093591393458e-06,
+      "loss": 0.4784,
+      "step": 897
+    },
+    {
+      "epoch": 0.6986514522821576,
+      "grad_norm": 0.6406476429868452,
+      "learning_rate": 9.4720702807314e-06,
+      "loss": 0.447,
+      "step": 898
+    },
+    {
+      "epoch": 0.6994294605809128,
+      "grad_norm": 0.5552348981024319,
+      "learning_rate": 9.470043302319256e-06,
+      "loss": 0.469,
+      "step": 899
+    },
+    {
+      "epoch": 0.700207468879668,
+      "grad_norm": 0.7446516304557894,
+      "learning_rate": 9.46801265781944e-06,
+      "loss": 0.4836,
+      "step": 900
+    },
+    {
+      "epoch": 0.7009854771784232,
+      "grad_norm": 0.5676593609592274,
+      "learning_rate": 9.465978348897383e-06,
+      "loss": 0.4314,
+      "step": 901
+    },
+    {
+      "epoch": 0.7017634854771784,
+      "grad_norm": 0.5865067280206147,
+      "learning_rate": 9.46394037722151e-06,
+      "loss": 0.4883,
+      "step": 902
+    },
+    {
+      "epoch": 0.7025414937759336,
+      "grad_norm": 0.7048010642728125,
+      "learning_rate": 9.461898744463256e-06,
+      "loss": 0.471,
+      "step": 903
+    },
+    {
+      "epoch": 0.7033195020746889,
+      "grad_norm": 0.5319467284300705,
+      "learning_rate": 9.45985345229706e-06,
+      "loss": 0.479,
+      "step": 904
+    },
+    {
+      "epoch": 0.704097510373444,
+      "grad_norm": 0.6192970153802269,
+      "learning_rate": 9.457804502400358e-06,
+      "loss": 0.4493,
+      "step": 905
+    },
+    {
+      "epoch": 0.7048755186721992,
+      "grad_norm": 0.579647929889083,
+      "learning_rate": 9.455751896453587e-06,
+      "loss": 0.4811,
+      "step": 906
+    },
+    {
+      "epoch": 0.7056535269709544,
+      "grad_norm": 0.5739821861596629,
+      "learning_rate": 9.453695636140185e-06,
+      "loss": 0.4671,
+      "step": 907
+    },
+    {
+      "epoch": 0.7064315352697096,
+      "grad_norm": 0.5716113430542653,
+      "learning_rate": 9.451635723146586e-06,
+      "loss": 0.443,
+      "step": 908
+    },
+    {
+      "epoch": 0.7072095435684648,
+      "grad_norm": 0.49312327935085026,
+      "learning_rate": 9.449572159162219e-06,
+      "loss": 0.404,
+      "step": 909
+    },
+    {
+      "epoch": 0.70798755186722,
+      "grad_norm": 0.5658247468562768,
+      "learning_rate": 9.447504945879505e-06,
+      "loss": 0.4535,
+      "step": 910
+    },
+    {
+      "epoch": 0.7087655601659751,
+      "grad_norm": 0.7453390027277186,
+      "learning_rate": 9.445434084993863e-06,
+      "loss": 0.5409,
+      "step": 911
+    },
+    {
+      "epoch": 0.7095435684647303,
+      "grad_norm": 0.6131635118908613,
+      "learning_rate": 9.4433595782037e-06,
+      "loss": 0.4431,
+      "step": 912
+    },
+    {
+      "epoch": 0.7103215767634855,
+      "grad_norm": 0.535939715922106,
+      "learning_rate": 9.441281427210415e-06,
+      "loss": 0.4472,
+      "step": 913
+    },
+    {
+      "epoch": 0.7110995850622407,
+      "grad_norm": 0.6113866328722033,
+      "learning_rate": 9.439199633718394e-06,
+      "loss": 0.4707,
+      "step": 914
+    },
+    {
+      "epoch": 0.7118775933609959,
+      "grad_norm": 0.5882939344803524,
+      "learning_rate": 9.437114199435013e-06,
+      "loss": 0.4746,
+      "step": 915
+    },
+    {
+      "epoch": 0.7126556016597511,
+      "grad_norm": 0.6666652131110754,
+      "learning_rate": 9.435025126070631e-06,
+      "loss": 0.497,
+      "step": 916
+    },
+    {
+      "epoch": 0.7134336099585062,
+      "grad_norm": 0.6179212159128344,
+      "learning_rate": 9.432932415338592e-06,
+      "loss": 0.4638,
+      "step": 917
+    },
+    {
+      "epoch": 0.7142116182572614,
+      "grad_norm": 0.5947989320371424,
+      "learning_rate": 9.430836068955227e-06,
+      "loss": 0.4719,
+      "step": 918
+    },
+    {
+      "epoch": 0.7149896265560166,
+      "grad_norm": 0.6690964779262584,
+      "learning_rate": 9.428736088639843e-06,
+      "loss": 0.5258,
+      "step": 919
+    },
+    {
+      "epoch": 0.7157676348547718,
+      "grad_norm": 0.611588046554929,
+      "learning_rate": 9.426632476114732e-06,
+      "loss": 0.4199,
+      "step": 920
+    },
+    {
+      "epoch": 0.716545643153527,
+      "grad_norm": 0.5875531905522247,
+      "learning_rate": 9.424525233105164e-06,
+      "loss": 0.4668,
+      "step": 921
+    },
+    {
+      "epoch": 0.7173236514522822,
+      "grad_norm": 0.6090933058701035,
+      "learning_rate": 9.422414361339383e-06,
+      "loss": 0.4319,
+      "step": 922
+    },
+    {
+      "epoch": 0.7181016597510373,
+      "grad_norm": 0.594248661920813,
+      "learning_rate": 9.420299862548615e-06,
+      "loss": 0.4421,
+      "step": 923
+    },
+    {
+      "epoch": 0.7188796680497925,
+      "grad_norm": 0.6206787441452869,
+      "learning_rate": 9.418181738467054e-06,
+      "loss": 0.4537,
+      "step": 924
+    },
+    {
+      "epoch": 0.7196576763485477,
+      "grad_norm": 0.5860510862249836,
+      "learning_rate": 9.416059990831871e-06,
+      "loss": 0.4327,
+      "step": 925
+    },
+    {
+      "epoch": 0.7204356846473029,
+      "grad_norm": 0.5492780328156327,
+      "learning_rate": 9.413934621383212e-06,
+      "loss": 0.4216,
+      "step": 926
+    },
+    {
+      "epoch": 0.7212136929460581,
+      "grad_norm": 0.5969418616275369,
+      "learning_rate": 9.411805631864189e-06,
+      "loss": 0.4448,
+      "step": 927
+    },
+    {
+      "epoch": 0.7219917012448133,
+      "grad_norm": 0.5728528175513504,
+      "learning_rate": 9.409673024020882e-06,
+      "loss": 0.4901,
+      "step": 928
+    },
+    {
+      "epoch": 0.7227697095435685,
+      "grad_norm": 0.6070394941942485,
+      "learning_rate": 9.40753679960234e-06,
+      "loss": 0.441,
+      "step": 929
+    },
+    {
+      "epoch": 0.7235477178423236,
+      "grad_norm": 0.6401206257098296,
+      "learning_rate": 9.405396960360584e-06,
+      "loss": 0.5181,
+      "step": 930
+    },
+    {
+      "epoch": 0.7243257261410788,
+      "grad_norm": 0.5324358587713556,
+      "learning_rate": 9.403253508050589e-06,
+      "loss": 0.4659,
+      "step": 931
+    },
+    {
+      "epoch": 0.725103734439834,
+      "grad_norm": 0.5853095423064312,
+      "learning_rate": 9.401106444430302e-06,
+      "loss": 0.477,
+      "step": 932
+    },
+    {
+      "epoch": 0.7258817427385892,
+      "grad_norm": 0.6007845954080344,
+      "learning_rate": 9.398955771260623e-06,
+      "loss": 0.4645,
+      "step": 933
+    },
+    {
+      "epoch": 0.7266597510373444,
+      "grad_norm": 0.6122723226092205,
+      "learning_rate": 9.396801490305425e-06,
+      "loss": 0.4714,
+      "step": 934
+    },
+    {
+      "epoch": 0.7274377593360996,
+      "grad_norm": 0.6210362449514901,
+      "learning_rate": 9.394643603331528e-06,
+      "loss": 0.5004,
+      "step": 935
+    },
+    {
+      "epoch": 0.7282157676348547,
+      "grad_norm": 0.6300478630965026,
+      "learning_rate": 9.392482112108716e-06,
+      "loss": 0.4854,
+      "step": 936
+    },
+    {
+      "epoch": 0.7289937759336099,
+      "grad_norm": 0.6472471023219617,
+      "learning_rate": 9.390317018409726e-06,
+      "loss": 0.4577,
+      "step": 937
+    },
+    {
+      "epoch": 0.7297717842323651,
+      "grad_norm": 0.6072988578983813,
+      "learning_rate": 9.38814832401025e-06,
+      "loss": 0.4915,
+      "step": 938
+    },
+    {
+      "epoch": 0.7305497925311203,
+      "grad_norm": 0.6282772649395366,
+      "learning_rate": 9.385976030688938e-06,
+      "loss": 0.4403,
+      "step": 939
+    },
+    {
+      "epoch": 0.7313278008298755,
+      "grad_norm": 0.6946243011909601,
+      "learning_rate": 9.383800140227383e-06,
+      "loss": 0.5069,
+      "step": 940
+    },
+    {
+      "epoch": 0.7321058091286307,
+      "grad_norm": 0.6261051561320691,
+      "learning_rate": 9.381620654410135e-06,
+      "loss": 0.4525,
+      "step": 941
+    },
+    {
+      "epoch": 0.7328838174273858,
+      "grad_norm": 0.5842014811490717,
+      "learning_rate": 9.379437575024688e-06,
+      "loss": 0.4577,
+      "step": 942
+    },
+    {
+      "epoch": 0.733661825726141,
+      "grad_norm": 0.5913115086007845,
+      "learning_rate": 9.37725090386149e-06,
+      "loss": 0.4676,
+      "step": 943
+    },
+    {
+      "epoch": 0.7344398340248963,
+      "grad_norm": 0.6360983588395872,
+      "learning_rate": 9.375060642713925e-06,
+      "loss": 0.4655,
+      "step": 944
+    },
+    {
+      "epoch": 0.7352178423236515,
+      "grad_norm": 0.5596942012987094,
+      "learning_rate": 9.37286679337833e-06,
+      "loss": 0.4834,
+      "step": 945
+    },
+    {
+      "epoch": 0.7359958506224067,
+      "grad_norm": 0.6008377584636402,
+      "learning_rate": 9.37066935765398e-06,
+      "loss": 0.4551,
+      "step": 946
+    },
+    {
+      "epoch": 0.7367738589211619,
+      "grad_norm": 0.5921800067425539,
+      "learning_rate": 9.368468337343093e-06,
+      "loss": 0.4612,
+      "step": 947
+    },
+    {
+      "epoch": 0.7375518672199171,
+      "grad_norm": 0.5663317345901085,
+      "learning_rate": 9.36626373425083e-06,
+      "loss": 0.4098,
+      "step": 948
+    },
+    {
+      "epoch": 0.7383298755186722,
+      "grad_norm": 0.5566562548091446,
+      "learning_rate": 9.364055550185283e-06,
+      "loss": 0.4645,
+      "step": 949
+    },
+    {
+      "epoch": 0.7391078838174274,
+      "grad_norm": 0.66750577988827,
+      "learning_rate": 9.361843786957486e-06,
+      "loss": 0.4857,
+      "step": 950
+    },
+    {
+      "epoch": 0.7398858921161826,
+      "grad_norm": 0.6690924194330922,
+      "learning_rate": 9.35962844638141e-06,
+      "loss": 0.4699,
+      "step": 951
+    },
+    {
+      "epoch": 0.7406639004149378,
+      "grad_norm": 0.5934468346539318,
+      "learning_rate": 9.357409530273954e-06,
+      "loss": 0.4345,
+      "step": 952
+    },
+    {
+      "epoch": 0.741441908713693,
+      "grad_norm": 0.629607438692411,
+      "learning_rate": 9.355187040454957e-06,
+      "loss": 0.451,
+      "step": 953
+    },
+    {
+      "epoch": 0.7422199170124482,
+      "grad_norm": 0.6009604552932272,
+      "learning_rate": 9.35296097874718e-06,
+      "loss": 0.4548,
+      "step": 954
+    },
+    {
+      "epoch": 0.7429979253112033,
+      "grad_norm": 0.6075804939136963,
+      "learning_rate": 9.35073134697632e-06,
+      "loss": 0.4736,
+      "step": 955
+    },
+    {
+      "epoch": 0.7437759336099585,
+      "grad_norm": 0.5922375325073594,
+      "learning_rate": 9.348498146971004e-06,
+      "loss": 0.4809,
+      "step": 956
+    },
+    {
+      "epoch": 0.7445539419087137,
+      "grad_norm": 0.5843679108545788,
+      "learning_rate": 9.346261380562778e-06,
+      "loss": 0.4845,
+      "step": 957
+    },
+    {
+      "epoch": 0.7453319502074689,
+      "grad_norm": 0.5717404045383747,
+      "learning_rate": 9.344021049586118e-06,
+      "loss": 0.4468,
+      "step": 958
+    },
+    {
+      "epoch": 0.7461099585062241,
+      "grad_norm": 0.5317009805744096,
+      "learning_rate": 9.341777155878423e-06,
+      "loss": 0.4359,
+      "step": 959
+    },
+    {
+      "epoch": 0.7468879668049793,
+      "grad_norm": 0.5230044234234553,
+      "learning_rate": 9.33952970128001e-06,
+      "loss": 0.4353,
+      "step": 960
+    },
+    {
+      "epoch": 0.7476659751037344,
+      "grad_norm": 0.5756570304706152,
+      "learning_rate": 9.337278687634124e-06,
+      "loss": 0.4616,
+      "step": 961
+    },
+    {
+      "epoch": 0.7484439834024896,
+      "grad_norm": 0.604688750216549,
+      "learning_rate": 9.335024116786924e-06,
+      "loss": 0.4398,
+      "step": 962
+    },
+    {
+      "epoch": 0.7492219917012448,
+      "grad_norm": 0.6051181102271095,
+      "learning_rate": 9.332765990587482e-06,
+      "loss": 0.4825,
+      "step": 963
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5576765477130864,
+      "learning_rate": 9.330504310887798e-06,
+      "loss": 0.47,
+      "step": 964
+    },
+    {
+      "epoch": 0.7507780082987552,
+      "grad_norm": 0.6114847494119985,
+      "learning_rate": 9.328239079542774e-06,
+      "loss": 0.4645,
+      "step": 965
+    },
+    {
+      "epoch": 0.7515560165975104,
+      "grad_norm": 0.6083805721848677,
+      "learning_rate": 9.325970298410232e-06,
+      "loss": 0.4382,
+      "step": 966
+    },
+    {
+      "epoch": 0.7523340248962656,
+      "grad_norm": 0.6215014212342788,
+      "learning_rate": 9.323697969350904e-06,
+      "loss": 0.4771,
+      "step": 967
+    },
+    {
+      "epoch": 0.7531120331950207,
+      "grad_norm": 0.6240704041287471,
+      "learning_rate": 9.321422094228431e-06,
+      "loss": 0.4762,
+      "step": 968
+    },
+    {
+      "epoch": 0.7538900414937759,
+      "grad_norm": 0.5943795358623963,
+      "learning_rate": 9.319142674909363e-06,
+      "loss": 0.4406,
+      "step": 969
+    },
+    {
+      "epoch": 0.7546680497925311,
+      "grad_norm": 0.5593356799794046,
+      "learning_rate": 9.316859713263155e-06,
+      "loss": 0.4448,
+      "step": 970
+    },
+    {
+      "epoch": 0.7554460580912863,
+      "grad_norm": 0.5697377296288101,
+      "learning_rate": 9.31457321116217e-06,
+      "loss": 0.4858,
+      "step": 971
+    },
+    {
+      "epoch": 0.7562240663900415,
+      "grad_norm": 0.5343167836172215,
+      "learning_rate": 9.312283170481673e-06,
+      "loss": 0.4525,
+      "step": 972
+    },
+    {
+      "epoch": 0.7570020746887967,
+      "grad_norm": 0.5603041794503046,
+      "learning_rate": 9.309989593099833e-06,
+      "loss": 0.4308,
+      "step": 973
+    },
+    {
+      "epoch": 0.7577800829875518,
+      "grad_norm": 0.6886874679849514,
+      "learning_rate": 9.307692480897715e-06,
+      "loss": 0.4887,
+      "step": 974
+    },
+    {
+      "epoch": 0.758558091286307,
+      "grad_norm": 0.621337453896326,
+      "learning_rate": 9.305391835759287e-06,
+      "loss": 0.4602,
+      "step": 975
+    },
+    {
+      "epoch": 0.7593360995850622,
+      "grad_norm": 0.632497522299684,
+      "learning_rate": 9.303087659571415e-06,
+      "loss": 0.4578,
+      "step": 976
+    },
+    {
+      "epoch": 0.7601141078838174,
+      "grad_norm": 0.6133759357410938,
+      "learning_rate": 9.300779954223859e-06,
+      "loss": 0.4485,
+      "step": 977
+    },
+    {
+      "epoch": 0.7608921161825726,
+      "grad_norm": 0.6212191977645002,
+      "learning_rate": 9.298468721609272e-06,
+      "loss": 0.4711,
+      "step": 978
+    },
+    {
+      "epoch": 0.7616701244813278,
+      "grad_norm": 0.5570192330768642,
+      "learning_rate": 9.296153963623206e-06,
+      "loss": 0.4571,
+      "step": 979
+    },
+    {
+      "epoch": 0.7624481327800829,
+      "grad_norm": 0.6235356562285591,
+      "learning_rate": 9.2938356821641e-06,
+      "loss": 0.4392,
+      "step": 980
+    },
+    {
+      "epoch": 0.7632261410788381,
+      "grad_norm": 0.614080268404056,
+      "learning_rate": 9.29151387913328e-06,
+      "loss": 0.448,
+      "step": 981
+    },
+    {
+      "epoch": 0.7640041493775933,
+      "grad_norm": 0.6048147948823029,
+      "learning_rate": 9.289188556434962e-06,
+      "loss": 0.4623,
+      "step": 982
+    },
+    {
+      "epoch": 0.7647821576763485,
+      "grad_norm": 0.5681283094319545,
+      "learning_rate": 9.286859715976256e-06,
+      "loss": 0.4749,
+      "step": 983
+    },
+    {
+      "epoch": 0.7655601659751037,
+      "grad_norm": 0.6128379145719766,
+      "learning_rate": 9.284527359667146e-06,
+      "loss": 0.4879,
+      "step": 984
+    },
+    {
+      "epoch": 0.766338174273859,
+      "grad_norm": 0.6200298152282174,
+      "learning_rate": 9.282191489420505e-06,
+      "loss": 0.4866,
+      "step": 985
+    },
+    {
+      "epoch": 0.7671161825726142,
+      "grad_norm": 0.6846593043376137,
+      "learning_rate": 9.279852107152088e-06,
+      "loss": 0.4926,
+      "step": 986
+    },
+    {
+      "epoch": 0.7678941908713693,
+      "grad_norm": 0.6248588850507624,
+      "learning_rate": 9.27750921478053e-06,
+      "loss": 0.4439,
+      "step": 987
+    },
+    {
+      "epoch": 0.7686721991701245,
+      "grad_norm": 0.5988708460539353,
+      "learning_rate": 9.275162814227346e-06,
+      "loss": 0.4631,
+      "step": 988
+    },
+    {
+      "epoch": 0.7694502074688797,
+      "grad_norm": 0.6106899326454651,
+      "learning_rate": 9.272812907416923e-06,
+      "loss": 0.4692,
+      "step": 989
+    },
+    {
+      "epoch": 0.7702282157676349,
+      "grad_norm": 0.6160507257618659,
+      "learning_rate": 9.270459496276531e-06,
+      "loss": 0.4712,
+      "step": 990
+    },
+    {
+      "epoch": 0.7710062240663901,
+      "grad_norm": 0.5839718595707181,
+      "learning_rate": 9.268102582736308e-06,
+      "loss": 0.4963,
+      "step": 991
+    },
+    {
+      "epoch": 0.7717842323651453,
+      "grad_norm": 0.5321471781879094,
+      "learning_rate": 9.26574216872927e-06,
+      "loss": 0.4777,
+      "step": 992
+    },
+    {
+      "epoch": 0.7725622406639004,
+      "grad_norm": 0.5541463682968509,
+      "learning_rate": 9.263378256191298e-06,
+      "loss": 0.4576,
+      "step": 993
+    },
+    {
+      "epoch": 0.7733402489626556,
+      "grad_norm": 0.6553777949176746,
+      "learning_rate": 9.261010847061147e-06,
+      "loss": 0.4671,
+      "step": 994
+    },
+    {
+      "epoch": 0.7741182572614108,
+      "grad_norm": 0.6373116287814153,
+      "learning_rate": 9.258639943280437e-06,
+      "loss": 0.4676,
+      "step": 995
+    },
+    {
+      "epoch": 0.774896265560166,
+      "grad_norm": 0.6106063556424047,
+      "learning_rate": 9.256265546793653e-06,
+      "loss": 0.5068,
+      "step": 996
+    },
+    {
+      "epoch": 0.7756742738589212,
+      "grad_norm": 0.6052143878172509,
+      "learning_rate": 9.253887659548151e-06,
+      "loss": 0.4774,
+      "step": 997
+    },
+    {
+      "epoch": 0.7764522821576764,
+      "grad_norm": 0.6118629233019967,
+      "learning_rate": 9.251506283494142e-06,
+      "loss": 0.4792,
+      "step": 998
+    },
+    {
+      "epoch": 0.7772302904564315,
+      "grad_norm": 0.6585187588231466,
+      "learning_rate": 9.249121420584703e-06,
+      "loss": 0.5141,
+      "step": 999
+    },
+    {
+      "epoch": 0.7780082987551867,
+      "grad_norm": 0.5948997577882585,
+      "learning_rate": 9.246733072775769e-06,
+      "loss": 0.4209,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7787863070539419,
+      "grad_norm": 0.5639740736596206,
+      "learning_rate": 9.244341242026135e-06,
+      "loss": 0.4479,
+      "step": 1001
+    },
+    {
+      "epoch": 0.7795643153526971,
+      "grad_norm": 0.6391546913873629,
+      "learning_rate": 9.241945930297448e-06,
+      "loss": 0.4416,
+      "step": 1002
+    },
+    {
+      "epoch": 0.7803423236514523,
+      "grad_norm": 0.6056095370457599,
+      "learning_rate": 9.239547139554218e-06,
+      "loss": 0.4557,
+      "step": 1003
+    },
+    {
+      "epoch": 0.7811203319502075,
+      "grad_norm": 0.5697417813161267,
+      "learning_rate": 9.237144871763799e-06,
+      "loss": 0.4496,
+      "step": 1004
+    },
+    {
+      "epoch": 0.7818983402489627,
+      "grad_norm": 0.5437527784310121,
+      "learning_rate": 9.234739128896403e-06,
+      "loss": 0.4393,
+      "step": 1005
+    },
+    {
+      "epoch": 0.7826763485477178,
+      "grad_norm": 0.5845878810788505,
+      "learning_rate": 9.232329912925091e-06,
+      "loss": 0.4496,
+      "step": 1006
+    },
+    {
+      "epoch": 0.783454356846473,
+      "grad_norm": 0.6526399819193138,
+      "learning_rate": 9.22991722582577e-06,
+      "loss": 0.4723,
+      "step": 1007
+    },
+    {
+      "epoch": 0.7842323651452282,
+      "grad_norm": 0.5865224711435765,
+      "learning_rate": 9.227501069577194e-06,
+      "loss": 0.4794,
+      "step": 1008
+    },
+    {
+      "epoch": 0.7850103734439834,
+      "grad_norm": 0.5686451322741624,
+      "learning_rate": 9.225081446160968e-06,
+      "loss": 0.451,
+      "step": 1009
+    },
+    {
+      "epoch": 0.7857883817427386,
+      "grad_norm": 0.6051569034576163,
+      "learning_rate": 9.222658357561534e-06,
+      "loss": 0.4673,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7865663900414938,
+      "grad_norm": 0.6053860946367862,
+      "learning_rate": 9.220231805766175e-06,
+      "loss": 0.4701,
+      "step": 1011
+    },
+    {
+      "epoch": 0.7873443983402489,
+      "grad_norm": 0.6499237106827441,
+      "learning_rate": 9.217801792765023e-06,
+      "loss": 0.4802,
+      "step": 1012
+    },
+    {
+      "epoch": 0.7881224066390041,
+      "grad_norm": 0.6399885030288344,
+      "learning_rate": 9.215368320551039e-06,
+      "loss": 0.5091,
+      "step": 1013
+    },
+    {
+      "epoch": 0.7889004149377593,
+      "grad_norm": 0.6311620737743737,
+      "learning_rate": 9.212931391120028e-06,
+      "loss": 0.4555,
+      "step": 1014
+    },
+    {
+      "epoch": 0.7896784232365145,
+      "grad_norm": 0.6168056106949081,
+      "learning_rate": 9.210491006470627e-06,
+      "loss": 0.4892,
+      "step": 1015
+    },
+    {
+      "epoch": 0.7904564315352697,
+      "grad_norm": 0.6323798416924965,
+      "learning_rate": 9.208047168604306e-06,
+      "loss": 0.478,
+      "step": 1016
+    },
+    {
+      "epoch": 0.7912344398340249,
+      "grad_norm": 0.5415413534817347,
+      "learning_rate": 9.20559987952537e-06,
+      "loss": 0.4484,
+      "step": 1017
+    },
+    {
+      "epoch": 0.79201244813278,
+      "grad_norm": 0.5764970969048904,
+      "learning_rate": 9.203149141240952e-06,
+      "loss": 0.4791,
+      "step": 1018
+    },
+    {
+      "epoch": 0.7927904564315352,
+      "grad_norm": 0.6251666841352921,
+      "learning_rate": 9.200694955761015e-06,
+      "loss": 0.5087,
+      "step": 1019
+    },
+    {
+      "epoch": 0.7935684647302904,
+      "grad_norm": 0.5585113675915873,
+      "learning_rate": 9.19823732509835e-06,
+      "loss": 0.4714,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7943464730290456,
+      "grad_norm": 0.5799467047393477,
+      "learning_rate": 9.195776251268575e-06,
+      "loss": 0.4707,
+      "step": 1021
+    },
+    {
+      "epoch": 0.7951244813278008,
+      "grad_norm": 0.5801203599084489,
+      "learning_rate": 9.193311736290124e-06,
+      "loss": 0.4519,
+      "step": 1022
+    },
+    {
+      "epoch": 0.795902489626556,
+      "grad_norm": 0.5861261479248425,
+      "learning_rate": 9.190843782184265e-06,
+      "loss": 0.4493,
+      "step": 1023
+    },
+    {
+      "epoch": 0.7966804979253111,
+      "grad_norm": 0.6394523568839049,
+      "learning_rate": 9.188372390975074e-06,
+      "loss": 0.4518,
+      "step": 1024
+    },
+    {
+      "epoch": 0.7974585062240664,
+      "grad_norm": 0.5629086466058565,
+      "learning_rate": 9.185897564689457e-06,
+      "loss": 0.4565,
+      "step": 1025
+    },
+    {
+      "epoch": 0.7982365145228216,
+      "grad_norm": 0.5346457718613361,
+      "learning_rate": 9.183419305357132e-06,
+      "loss": 0.4779,
+      "step": 1026
+    },
+    {
+      "epoch": 0.7990145228215768,
+      "grad_norm": 0.5804996061343236,
+      "learning_rate": 9.180937615010631e-06,
+      "loss": 0.4551,
+      "step": 1027
+    },
+    {
+      "epoch": 0.799792531120332,
+      "grad_norm": 0.5762701216146465,
+      "learning_rate": 9.178452495685306e-06,
+      "loss": 0.4735,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8005705394190872,
+      "grad_norm": 0.5529350371775633,
+      "learning_rate": 9.175963949419314e-06,
+      "loss": 0.4631,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8013485477178424,
+      "grad_norm": 0.6216291092939621,
+      "learning_rate": 9.173471978253628e-06,
+      "loss": 0.4897,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8021265560165975,
+      "grad_norm": 0.5595810414433808,
+      "learning_rate": 9.170976584232026e-06,
+      "loss": 0.4472,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8029045643153527,
+      "grad_norm": 0.5975968467374871,
+      "learning_rate": 9.168477769401094e-06,
+      "loss": 0.4566,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8036825726141079,
+      "grad_norm": 0.5211601342325312,
+      "learning_rate": 9.16597553581023e-06,
+      "loss": 0.4445,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8044605809128631,
+      "grad_norm": 0.5433143711243938,
+      "learning_rate": 9.163469885511626e-06,
+      "loss": 0.4547,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8052385892116183,
+      "grad_norm": 0.5666729982767013,
+      "learning_rate": 9.160960820560286e-06,
+      "loss": 0.4398,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8060165975103735,
+      "grad_norm": 0.5827348052286895,
+      "learning_rate": 9.158448343014001e-06,
+      "loss": 0.4547,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8067946058091287,
+      "grad_norm": 0.5781436741198746,
+      "learning_rate": 9.155932454933375e-06,
+      "loss": 0.4603,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8075726141078838,
+      "grad_norm": 0.5414655976565699,
+      "learning_rate": 9.153413158381803e-06,
+      "loss": 0.4455,
+      "step": 1038
+    },
+    {
+      "epoch": 0.808350622406639,
+      "grad_norm": 0.5158075106940401,
+      "learning_rate": 9.150890455425474e-06,
+      "loss": 0.4228,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8091286307053942,
+      "grad_norm": 0.5880792122706849,
+      "learning_rate": 9.148364348133376e-06,
+      "loss": 0.4786,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8099066390041494,
+      "grad_norm": 0.524956967589777,
+      "learning_rate": 9.14583483857728e-06,
+      "loss": 0.4543,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8106846473029046,
+      "grad_norm": 0.5924070868198928,
+      "learning_rate": 9.143301928831756e-06,
+      "loss": 0.4454,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8114626556016598,
+      "grad_norm": 0.605449379186435,
+      "learning_rate": 9.14076562097416e-06,
+      "loss": 0.4766,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8122406639004149,
+      "grad_norm": 0.542690756432596,
+      "learning_rate": 9.138225917084634e-06,
+      "loss": 0.4387,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8130186721991701,
+      "grad_norm": 0.5844163039133797,
+      "learning_rate": 9.135682819246103e-06,
+      "loss": 0.4325,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8137966804979253,
+      "grad_norm": 0.6170793158785922,
+      "learning_rate": 9.133136329544283e-06,
+      "loss": 0.4473,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8145746887966805,
+      "grad_norm": 0.5461191030257552,
+      "learning_rate": 9.130586450067664e-06,
+      "loss": 0.4143,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8153526970954357,
+      "grad_norm": 0.6082016883032599,
+      "learning_rate": 9.12803318290752e-06,
+      "loss": 0.47,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8161307053941909,
+      "grad_norm": 0.5260848639106689,
+      "learning_rate": 9.125476530157903e-06,
+      "loss": 0.4316,
+      "step": 1049
+    },
+    {
+      "epoch": 0.816908713692946,
+      "grad_norm": 0.6687779581424161,
+      "learning_rate": 9.122916493915643e-06,
+      "loss": 0.4647,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8176867219917012,
+      "grad_norm": 0.6374380940043036,
+      "learning_rate": 9.120353076280338e-06,
+      "loss": 0.4449,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8184647302904564,
+      "grad_norm": 0.7318166174031484,
+      "learning_rate": 9.117786279354367e-06,
+      "loss": 0.4536,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8192427385892116,
+      "grad_norm": 0.6305482509266764,
+      "learning_rate": 9.11521610524288e-06,
+      "loss": 0.4815,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8200207468879668,
+      "grad_norm": 0.6222012109911158,
+      "learning_rate": 9.112642556053794e-06,
+      "loss": 0.4684,
+      "step": 1054
+    },
+    {
+      "epoch": 0.820798755186722,
+      "grad_norm": 0.6242984148008559,
+      "learning_rate": 9.110065633897793e-06,
+      "loss": 0.4899,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8215767634854771,
+      "grad_norm": 0.6784114824813955,
+      "learning_rate": 9.10748534088833e-06,
+      "loss": 0.4638,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8223547717842323,
+      "grad_norm": 0.5780372430602528,
+      "learning_rate": 9.104901679141623e-06,
+      "loss": 0.4732,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8231327800829875,
+      "grad_norm": 0.5825383151270402,
+      "learning_rate": 9.10231465077665e-06,
+      "loss": 0.4379,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8239107883817427,
+      "grad_norm": 0.6531404158498082,
+      "learning_rate": 9.09972425791515e-06,
+      "loss": 0.4584,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8246887966804979,
+      "grad_norm": 0.5747827310334281,
+      "learning_rate": 9.097130502681626e-06,
+      "loss": 0.4566,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8254668049792531,
+      "grad_norm": 0.5273643728756384,
+      "learning_rate": 9.094533387203337e-06,
+      "loss": 0.4567,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8262448132780082,
+      "grad_norm": 0.600226829077028,
+      "learning_rate": 9.091932913610294e-06,
+      "loss": 0.4475,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8270228215767634,
+      "grad_norm": 0.5878471841610988,
+      "learning_rate": 9.089329084035264e-06,
+      "loss": 0.4303,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8278008298755186,
+      "grad_norm": 0.5882036815951942,
+      "learning_rate": 9.08672190061377e-06,
+      "loss": 0.4464,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8285788381742739,
+      "grad_norm": 0.5437174359538157,
+      "learning_rate": 9.084111365484078e-06,
+      "loss": 0.4666,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8293568464730291,
+      "grad_norm": 0.6384131103276415,
+      "learning_rate": 9.081497480787213e-06,
+      "loss": 0.4924,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8301348547717843,
+      "grad_norm": 0.6122314632672708,
+      "learning_rate": 9.078880248666941e-06,
+      "loss": 0.4347,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8309128630705395,
+      "grad_norm": 0.6039263526937011,
+      "learning_rate": 9.076259671269771e-06,
+      "loss": 0.4972,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8316908713692946,
+      "grad_norm": 0.5890236320520699,
+      "learning_rate": 9.07363575074496e-06,
+      "loss": 0.4573,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8324688796680498,
+      "grad_norm": 0.591237345983886,
+      "learning_rate": 9.071008489244508e-06,
+      "loss": 0.4711,
+      "step": 1070
+    },
+    {
+      "epoch": 0.833246887966805,
+      "grad_norm": 0.6207853312311854,
+      "learning_rate": 9.068377888923153e-06,
+      "loss": 0.4655,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8340248962655602,
+      "grad_norm": 0.5653357727994602,
+      "learning_rate": 9.065743951938365e-06,
+      "loss": 0.458,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8348029045643154,
+      "grad_norm": 0.5796492058483245,
+      "learning_rate": 9.063106680450362e-06,
+      "loss": 0.4319,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8355809128630706,
+      "grad_norm": 0.6315746921480735,
+      "learning_rate": 9.060466076622092e-06,
+      "loss": 0.5093,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8363589211618258,
+      "grad_norm": 0.6622189751053903,
+      "learning_rate": 9.05782214261923e-06,
+      "loss": 0.4855,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8371369294605809,
+      "grad_norm": 0.5147682499271341,
+      "learning_rate": 9.055174880610194e-06,
+      "loss": 0.4563,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8379149377593361,
+      "grad_norm": 0.6415920429563793,
+      "learning_rate": 9.05252429276612e-06,
+      "loss": 0.5036,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8386929460580913,
+      "grad_norm": 0.5381156830170406,
+      "learning_rate": 9.04987038126088e-06,
+      "loss": 0.4441,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8394709543568465,
+      "grad_norm": 0.4976800634348349,
+      "learning_rate": 9.047213148271069e-06,
+      "loss": 0.4391,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8402489626556017,
+      "grad_norm": 0.5876495407257842,
+      "learning_rate": 9.044552595976004e-06,
+      "loss": 0.469,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8410269709543569,
+      "grad_norm": 0.5467437424784772,
+      "learning_rate": 9.041888726557724e-06,
+      "loss": 0.4311,
+      "step": 1081
+    },
+    {
+      "epoch": 0.841804979253112,
+      "grad_norm": 0.5254078427528449,
+      "learning_rate": 9.039221542200996e-06,
+      "loss": 0.4586,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8425829875518672,
+      "grad_norm": 0.544304623244372,
+      "learning_rate": 9.036551045093293e-06,
+      "loss": 0.4691,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8433609958506224,
+      "grad_norm": 0.6652930828993748,
+      "learning_rate": 9.03387723742482e-06,
+      "loss": 0.483,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8441390041493776,
+      "grad_norm": 0.5991029585292743,
+      "learning_rate": 9.031200121388486e-06,
+      "loss": 0.5123,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8449170124481328,
+      "grad_norm": 0.567205175114081,
+      "learning_rate": 9.028519699179914e-06,
+      "loss": 0.4421,
+      "step": 1086
+    },
+    {
+      "epoch": 0.845695020746888,
+      "grad_norm": 0.5481983106558831,
+      "learning_rate": 9.025835972997446e-06,
+      "loss": 0.4208,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8464730290456431,
+      "grad_norm": 0.5449522051701893,
+      "learning_rate": 9.023148945042126e-06,
+      "loss": 0.4432,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8472510373443983,
+      "grad_norm": 0.5735339617537626,
+      "learning_rate": 9.02045861751771e-06,
+      "loss": 0.4957,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8480290456431535,
+      "grad_norm": 0.5100676943837129,
+      "learning_rate": 9.017764992630657e-06,
+      "loss": 0.4688,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8488070539419087,
+      "grad_norm": 0.5475101303282949,
+      "learning_rate": 9.015068072590136e-06,
+      "loss": 0.4398,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8495850622406639,
+      "grad_norm": 0.5469156377339702,
+      "learning_rate": 9.012367859608014e-06,
+      "loss": 0.4239,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8503630705394191,
+      "grad_norm": 0.5932690232558361,
+      "learning_rate": 9.009664355898858e-06,
+      "loss": 0.452,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8511410788381742,
+      "grad_norm": 0.6560566668592601,
+      "learning_rate": 9.006957563679936e-06,
+      "loss": 0.4981,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8519190871369294,
+      "grad_norm": 0.5935153973103316,
+      "learning_rate": 9.004247485171214e-06,
+      "loss": 0.4755,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8526970954356846,
+      "grad_norm": 0.7162736653277993,
+      "learning_rate": 9.001534122595347e-06,
+      "loss": 0.4513,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8534751037344398,
+      "grad_norm": 0.5670024698318993,
+      "learning_rate": 8.998817478177694e-06,
+      "loss": 0.4484,
+      "step": 1097
+    },
+    {
+      "epoch": 0.854253112033195,
+      "grad_norm": 0.5426924838581024,
+      "learning_rate": 8.9960975541463e-06,
+      "loss": 0.444,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8550311203319502,
+      "grad_norm": 0.5471150983182225,
+      "learning_rate": 8.993374352731894e-06,
+      "loss": 0.4191,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8558091286307054,
+      "grad_norm": 0.6095500308219233,
+      "learning_rate": 8.990647876167901e-06,
+      "loss": 0.443,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8565871369294605,
+      "grad_norm": 0.5563817922195291,
+      "learning_rate": 8.987918126690433e-06,
+      "loss": 0.465,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8573651452282157,
+      "grad_norm": 0.544769667688187,
+      "learning_rate": 8.985185106538277e-06,
+      "loss": 0.4652,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8581431535269709,
+      "grad_norm": 0.6174251360439055,
+      "learning_rate": 8.982448817952914e-06,
+      "loss": 0.4537,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8589211618257261,
+      "grad_norm": 0.6755110270410458,
+      "learning_rate": 8.979709263178493e-06,
+      "loss": 0.4439,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8596991701244814,
+      "grad_norm": 0.5766277504014591,
+      "learning_rate": 8.976966444461856e-06,
+      "loss": 0.444,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8604771784232366,
+      "grad_norm": 0.5806110102547257,
+      "learning_rate": 8.97422036405251e-06,
+      "loss": 0.4507,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8612551867219918,
+      "grad_norm": 0.5095831763876618,
+      "learning_rate": 8.971471024202643e-06,
+      "loss": 0.4271,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8620331950207469,
+      "grad_norm": 0.6064829040429137,
+      "learning_rate": 8.968718427167113e-06,
+      "loss": 0.4431,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8628112033195021,
+      "grad_norm": 0.63265152308709,
+      "learning_rate": 8.965962575203455e-06,
+      "loss": 0.477,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8635892116182573,
+      "grad_norm": 0.6064365915412281,
+      "learning_rate": 8.963203470571867e-06,
+      "loss": 0.4921,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8643672199170125,
+      "grad_norm": 0.602765269997304,
+      "learning_rate": 8.960441115535216e-06,
+      "loss": 0.4824,
+      "step": 1111
+    },
+    {
+      "epoch": 0.8651452282157677,
+      "grad_norm": 0.6085985352447052,
+      "learning_rate": 8.95767551235904e-06,
+      "loss": 0.4513,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8659232365145229,
+      "grad_norm": 0.671186742521655,
+      "learning_rate": 8.954906663311533e-06,
+      "loss": 0.4834,
+      "step": 1113
+    },
+    {
+      "epoch": 0.866701244813278,
+      "grad_norm": 0.6217742574010019,
+      "learning_rate": 8.952134570663557e-06,
+      "loss": 0.4703,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8674792531120332,
+      "grad_norm": 0.5903565879452011,
+      "learning_rate": 8.949359236688633e-06,
+      "loss": 0.4377,
+      "step": 1115
+    },
+    {
+      "epoch": 0.8682572614107884,
+      "grad_norm": 0.5819716947148461,
+      "learning_rate": 8.94658066366294e-06,
+      "loss": 0.4159,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8690352697095436,
+      "grad_norm": 0.6139967548402401,
+      "learning_rate": 8.943798853865312e-06,
+      "loss": 0.4901,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8698132780082988,
+      "grad_norm": 0.6281065092691911,
+      "learning_rate": 8.941013809577239e-06,
+      "loss": 0.446,
+      "step": 1118
+    },
+    {
+      "epoch": 0.870591286307054,
+      "grad_norm": 0.6174521627191852,
+      "learning_rate": 8.938225533082866e-06,
+      "loss": 0.4578,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8713692946058091,
+      "grad_norm": 0.5783073007700142,
+      "learning_rate": 8.935434026668982e-06,
+      "loss": 0.475,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8721473029045643,
+      "grad_norm": 0.594419964796681,
+      "learning_rate": 8.932639292625035e-06,
+      "loss": 0.4439,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8729253112033195,
+      "grad_norm": 0.6215188049288222,
+      "learning_rate": 8.929841333243108e-06,
+      "loss": 0.5142,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8737033195020747,
+      "grad_norm": 0.5485570861756526,
+      "learning_rate": 8.927040150817942e-06,
+      "loss": 0.4631,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8744813278008299,
+      "grad_norm": 0.6227832441311874,
+      "learning_rate": 8.924235747646911e-06,
+      "loss": 0.4919,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8752593360995851,
+      "grad_norm": 0.52591793828532,
+      "learning_rate": 8.921428126030041e-06,
+      "loss": 0.3921,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8760373443983402,
+      "grad_norm": 0.5910864240480492,
+      "learning_rate": 8.918617288269984e-06,
+      "loss": 0.4621,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8768153526970954,
+      "grad_norm": 0.5619378890015939,
+      "learning_rate": 8.91580323667204e-06,
+      "loss": 0.471,
+      "step": 1127
+    },
+    {
+      "epoch": 0.8775933609958506,
+      "grad_norm": 0.5998572034846,
+      "learning_rate": 8.912985973544143e-06,
+      "loss": 0.5264,
+      "step": 1128
+    },
+    {
+      "epoch": 0.8783713692946058,
+      "grad_norm": 0.5171725626141854,
+      "learning_rate": 8.910165501196856e-06,
+      "loss": 0.4834,
+      "step": 1129
+    },
+    {
+      "epoch": 0.879149377593361,
+      "grad_norm": 0.5953670431814426,
+      "learning_rate": 8.907341821943385e-06,
+      "loss": 0.5118,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8799273858921162,
+      "grad_norm": 0.5690776795015047,
+      "learning_rate": 8.90451493809955e-06,
+      "loss": 0.4526,
+      "step": 1131
+    },
+    {
+      "epoch": 0.8807053941908713,
+      "grad_norm": 0.5734125168158668,
+      "learning_rate": 8.901684851983817e-06,
+      "loss": 0.4389,
+      "step": 1132
+    },
+    {
+      "epoch": 0.8814834024896265,
+      "grad_norm": 0.5857321082849918,
+      "learning_rate": 8.898851565917264e-06,
+      "loss": 0.4225,
+      "step": 1133
+    },
+    {
+      "epoch": 0.8822614107883817,
+      "grad_norm": 0.5650936171986685,
+      "learning_rate": 8.8960150822236e-06,
+      "loss": 0.4374,
+      "step": 1134
+    },
+    {
+      "epoch": 0.8830394190871369,
+      "grad_norm": 0.5722660086082307,
+      "learning_rate": 8.893175403229153e-06,
+      "loss": 0.4877,
+      "step": 1135
+    },
+    {
+      "epoch": 0.8838174273858921,
+      "grad_norm": 0.7043797383159467,
+      "learning_rate": 8.890332531262879e-06,
+      "loss": 0.453,
+      "step": 1136
+    },
+    {
+      "epoch": 0.8845954356846473,
+      "grad_norm": 0.5270797036991951,
+      "learning_rate": 8.887486468656346e-06,
+      "loss": 0.4726,
+      "step": 1137
+    },
+    {
+      "epoch": 0.8853734439834025,
+      "grad_norm": 0.5486309484998437,
+      "learning_rate": 8.884637217743738e-06,
+      "loss": 0.4573,
+      "step": 1138
+    },
+    {
+      "epoch": 0.8861514522821576,
+      "grad_norm": 0.4970889932063969,
+      "learning_rate": 8.881784780861862e-06,
+      "loss": 0.4253,
+      "step": 1139
+    },
+    {
+      "epoch": 0.8869294605809128,
+      "grad_norm": 0.4961616187300532,
+      "learning_rate": 8.878929160350127e-06,
+      "loss": 0.4525,
+      "step": 1140
+    },
+    {
+      "epoch": 0.887707468879668,
+      "grad_norm": 0.5611904448251314,
+      "learning_rate": 8.876070358550559e-06,
+      "loss": 0.4407,
+      "step": 1141
+    },
+    {
+      "epoch": 0.8884854771784232,
+      "grad_norm": 0.5606593882076895,
+      "learning_rate": 8.873208377807796e-06,
+      "loss": 0.4623,
+      "step": 1142
+    },
+    {
+      "epoch": 0.8892634854771784,
+      "grad_norm": 0.5073411815594733,
+      "learning_rate": 8.870343220469076e-06,
+      "loss": 0.4196,
+      "step": 1143
+    },
+    {
+      "epoch": 0.8900414937759336,
+      "grad_norm": 0.6730653861114814,
+      "learning_rate": 8.867474888884249e-06,
+      "loss": 0.4603,
+      "step": 1144
+    },
+    {
+      "epoch": 0.8908195020746889,
+      "grad_norm": 0.6833454359436391,
+      "learning_rate": 8.864603385405764e-06,
+      "loss": 0.4319,
+      "step": 1145
+    },
+    {
+      "epoch": 0.891597510373444,
+      "grad_norm": 0.5491066542441609,
+      "learning_rate": 8.861728712388671e-06,
+      "loss": 0.4353,
+      "step": 1146
+    },
+    {
+      "epoch": 0.8923755186721992,
+      "grad_norm": 0.5612718765838423,
+      "learning_rate": 8.858850872190624e-06,
+      "loss": 0.4545,
+      "step": 1147
+    },
+    {
+      "epoch": 0.8931535269709544,
+      "grad_norm": 0.5576430039990395,
+      "learning_rate": 8.855969867171872e-06,
+      "loss": 0.4341,
+      "step": 1148
+    },
+    {
+      "epoch": 0.8939315352697096,
+      "grad_norm": 0.581967241883703,
+      "learning_rate": 8.853085699695258e-06,
+      "loss": 0.4727,
+      "step": 1149
+    },
+    {
+      "epoch": 0.8947095435684648,
+      "grad_norm": 0.5658052436543853,
+      "learning_rate": 8.850198372126219e-06,
+      "loss": 0.4675,
+      "step": 1150
+    },
+    {
+      "epoch": 0.89548755186722,
+      "grad_norm": 0.5214580135956548,
+      "learning_rate": 8.84730788683279e-06,
+      "loss": 0.4404,
+      "step": 1151
+    },
+    {
+      "epoch": 0.8962655601659751,
+      "grad_norm": 0.5228959363483624,
+      "learning_rate": 8.844414246185584e-06,
+      "loss": 0.448,
+      "step": 1152
+    },
+    {
+      "epoch": 0.8970435684647303,
+      "grad_norm": 0.5180355368500008,
+      "learning_rate": 8.841517452557813e-06,
+      "loss": 0.4441,
+      "step": 1153
+    },
+    {
+      "epoch": 0.8978215767634855,
+      "grad_norm": 0.564589075000934,
+      "learning_rate": 8.838617508325267e-06,
+      "loss": 0.4374,
+      "step": 1154
+    },
+    {
+      "epoch": 0.8985995850622407,
+      "grad_norm": 0.5823534923187391,
+      "learning_rate": 8.835714415866327e-06,
+      "loss": 0.4959,
+      "step": 1155
+    },
+    {
+      "epoch": 0.8993775933609959,
+      "grad_norm": 0.5129042304937075,
+      "learning_rate": 8.83280817756195e-06,
+      "loss": 0.4166,
+      "step": 1156
+    },
+    {
+      "epoch": 0.9001556016597511,
+      "grad_norm": 0.5587628773617983,
+      "learning_rate": 8.829898795795676e-06,
+      "loss": 0.4602,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9009336099585062,
+      "grad_norm": 0.6239816121833376,
+      "learning_rate": 8.826986272953623e-06,
+      "loss": 0.457,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9017116182572614,
+      "grad_norm": 0.572535739856388,
+      "learning_rate": 8.824070611424484e-06,
+      "loss": 0.4343,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9024896265560166,
+      "grad_norm": 0.5892357597953787,
+      "learning_rate": 8.821151813599527e-06,
+      "loss": 0.4645,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9032676348547718,
+      "grad_norm": 0.5327630015424487,
+      "learning_rate": 8.818229881872594e-06,
+      "loss": 0.4243,
+      "step": 1161
+    },
+    {
+      "epoch": 0.904045643153527,
+      "grad_norm": 0.654770699425073,
+      "learning_rate": 8.81530481864009e-06,
+      "loss": 0.5025,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9048236514522822,
+      "grad_norm": 0.5121395576023797,
+      "learning_rate": 8.812376626301001e-06,
+      "loss": 0.4463,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9056016597510373,
+      "grad_norm": 0.5713019335275539,
+      "learning_rate": 8.809445307256865e-06,
+      "loss": 0.4659,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9063796680497925,
+      "grad_norm": 0.5590567385872878,
+      "learning_rate": 8.806510863911795e-06,
+      "loss": 0.447,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9071576763485477,
+      "grad_norm": 0.6092228425013944,
+      "learning_rate": 8.803573298672463e-06,
+      "loss": 0.4458,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9079356846473029,
+      "grad_norm": 0.6145612562654129,
+      "learning_rate": 8.800632613948099e-06,
+      "loss": 0.4328,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9087136929460581,
+      "grad_norm": 0.5807379533194865,
+      "learning_rate": 8.797688812150495e-06,
+      "loss": 0.4429,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9094917012448133,
+      "grad_norm": 0.6180009912754423,
+      "learning_rate": 8.794741895693997e-06,
+      "loss": 0.4683,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9102697095435685,
+      "grad_norm": 0.5855453471252907,
+      "learning_rate": 8.791791866995508e-06,
+      "loss": 0.4385,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9110477178423236,
+      "grad_norm": 0.6105611382521265,
+      "learning_rate": 8.788838728474481e-06,
+      "loss": 0.4413,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9118257261410788,
+      "grad_norm": 0.5633864614132523,
+      "learning_rate": 8.78588248255292e-06,
+      "loss": 0.4414,
+      "step": 1172
+    },
+    {
+      "epoch": 0.912603734439834,
+      "grad_norm": 0.5847278200696867,
+      "learning_rate": 8.782923131655376e-06,
+      "loss": 0.4421,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9133817427385892,
+      "grad_norm": 0.5885337298286076,
+      "learning_rate": 8.779960678208952e-06,
+      "loss": 0.468,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9141597510373444,
+      "grad_norm": 0.5330163380430607,
+      "learning_rate": 8.776995124643292e-06,
+      "loss": 0.4826,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9149377593360996,
+      "grad_norm": 0.5529001669611443,
+      "learning_rate": 8.77402647339058e-06,
+      "loss": 0.5214,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9157157676348547,
+      "grad_norm": 0.6097634251819343,
+      "learning_rate": 8.771054726885546e-06,
+      "loss": 0.4569,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9164937759336099,
+      "grad_norm": 0.5901094787008891,
+      "learning_rate": 8.768079887565453e-06,
+      "loss": 0.4295,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9172717842323651,
+      "grad_norm": 0.5779069217074718,
+      "learning_rate": 8.765101957870105e-06,
+      "loss": 0.4542,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9180497925311203,
+      "grad_norm": 0.6026723268144607,
+      "learning_rate": 8.762120940241839e-06,
+      "loss": 0.4379,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9188278008298755,
+      "grad_norm": 0.6035412628386222,
+      "learning_rate": 8.759136837125522e-06,
+      "loss": 0.466,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9196058091286307,
+      "grad_norm": 0.5999476459796275,
+      "learning_rate": 8.756149650968557e-06,
+      "loss": 0.4659,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9203838174273858,
+      "grad_norm": 0.5649725104651698,
+      "learning_rate": 8.753159384220871e-06,
+      "loss": 0.4627,
+      "step": 1183
+    },
+    {
+      "epoch": 0.921161825726141,
+      "grad_norm": 0.6267055024815074,
+      "learning_rate": 8.750166039334918e-06,
+      "loss": 0.5044,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9219398340248963,
+      "grad_norm": 0.5366404544126924,
+      "learning_rate": 8.74716961876568e-06,
+      "loss": 0.4378,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9227178423236515,
+      "grad_norm": 0.5428310568279734,
+      "learning_rate": 8.744170124970655e-06,
+      "loss": 0.4625,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9234958506224067,
+      "grad_norm": 0.5362783574119684,
+      "learning_rate": 8.741167560409871e-06,
+      "loss": 0.4337,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9242738589211619,
+      "grad_norm": 0.5821977101779944,
+      "learning_rate": 8.738161927545865e-06,
+      "loss": 0.4618,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9250518672199171,
+      "grad_norm": 0.6100424807159776,
+      "learning_rate": 8.735153228843696e-06,
+      "loss": 0.4621,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9258298755186722,
+      "grad_norm": 0.5240308842386767,
+      "learning_rate": 8.732141466770934e-06,
+      "loss": 0.4232,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9266078838174274,
+      "grad_norm": 0.5129125803235188,
+      "learning_rate": 8.729126643797668e-06,
+      "loss": 0.43,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9273858921161826,
+      "grad_norm": 0.49378709359287903,
+      "learning_rate": 8.726108762396486e-06,
+      "loss": 0.4379,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9281639004149378,
+      "grad_norm": 0.5537624239208685,
+      "learning_rate": 8.723087825042495e-06,
+      "loss": 0.4366,
+      "step": 1193
+    },
+    {
+      "epoch": 0.928941908713693,
+      "grad_norm": 0.6654648466187115,
+      "learning_rate": 8.720063834213305e-06,
+      "loss": 0.4567,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9297199170124482,
+      "grad_norm": 0.5189703288230892,
+      "learning_rate": 8.71703679238903e-06,
+      "loss": 0.4134,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9304979253112033,
+      "grad_norm": 0.5587564171619025,
+      "learning_rate": 8.714006702052282e-06,
+      "loss": 0.4753,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9312759336099585,
+      "grad_norm": 0.5574704399086938,
+      "learning_rate": 8.710973565688182e-06,
+      "loss": 0.4006,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9320539419087137,
+      "grad_norm": 0.6502092555686538,
+      "learning_rate": 8.70793738578434e-06,
+      "loss": 0.4921,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9328319502074689,
+      "grad_norm": 0.6562548448159771,
+      "learning_rate": 8.70489816483087e-06,
+      "loss": 0.4548,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9336099585062241,
+      "grad_norm": 0.5350903781268164,
+      "learning_rate": 8.701855905320373e-06,
+      "loss": 0.4714,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9343879668049793,
+      "grad_norm": 0.6400243325411877,
+      "learning_rate": 8.698810609747949e-06,
+      "loss": 0.4558,
+      "step": 1201
+    },
+    {
+      "epoch": 0.9351659751037344,
+      "grad_norm": 0.6281902085117935,
+      "learning_rate": 8.695762280611184e-06,
+      "loss": 0.4281,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9359439834024896,
+      "grad_norm": 0.6002973539861263,
+      "learning_rate": 8.692710920410152e-06,
+      "loss": 0.4269,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9367219917012448,
+      "grad_norm": 0.5535009996878942,
+      "learning_rate": 8.689656531647413e-06,
+      "loss": 0.4438,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.5682025939479307,
+      "learning_rate": 8.686599116828012e-06,
+      "loss": 0.4622,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9382780082987552,
+      "grad_norm": 0.586981565702648,
+      "learning_rate": 8.683538678459475e-06,
+      "loss": 0.4712,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9390560165975104,
+      "grad_norm": 0.623267805255394,
+      "learning_rate": 8.68047521905181e-06,
+      "loss": 0.4177,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9398340248962656,
+      "grad_norm": 0.6042668314519692,
+      "learning_rate": 8.677408741117497e-06,
+      "loss": 0.4303,
+      "step": 1208
+    },
+    {
+      "epoch": 0.9406120331950207,
+      "grad_norm": 0.5983528465176061,
+      "learning_rate": 8.674339247171498e-06,
+      "loss": 0.4796,
+      "step": 1209
+    },
+    {
+      "epoch": 0.9413900414937759,
+      "grad_norm": 0.6987570835110576,
+      "learning_rate": 8.671266739731244e-06,
+      "loss": 0.4906,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9421680497925311,
+      "grad_norm": 0.5715212987543843,
+      "learning_rate": 8.66819122131664e-06,
+      "loss": 0.4632,
+      "step": 1211
+    },
+    {
+      "epoch": 0.9429460580912863,
+      "grad_norm": 0.7121487727486373,
+      "learning_rate": 8.66511269445006e-06,
+      "loss": 0.4877,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9437240663900415,
+      "grad_norm": 0.7306559644861813,
+      "learning_rate": 8.662031161656347e-06,
+      "loss": 0.4503,
+      "step": 1213
+    },
+    {
+      "epoch": 0.9445020746887967,
+      "grad_norm": 0.6143283134877987,
+      "learning_rate": 8.658946625462805e-06,
+      "loss": 0.4444,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9452800829875518,
+      "grad_norm": 0.6691288772098091,
+      "learning_rate": 8.655859088399205e-06,
+      "loss": 0.4575,
+      "step": 1215
+    },
+    {
+      "epoch": 0.946058091286307,
+      "grad_norm": 0.6539303401762822,
+      "learning_rate": 8.652768552997778e-06,
+      "loss": 0.4412,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9468360995850622,
+      "grad_norm": 0.574250119854465,
+      "learning_rate": 8.649675021793212e-06,
+      "loss": 0.4342,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9476141078838174,
+      "grad_norm": 0.5598677149347583,
+      "learning_rate": 8.646578497322655e-06,
+      "loss": 0.4615,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9483921161825726,
+      "grad_norm": 0.6488325167465456,
+      "learning_rate": 8.64347898212571e-06,
+      "loss": 0.4684,
+      "step": 1219
+    },
+    {
+      "epoch": 0.9491701244813278,
+      "grad_norm": 0.5998483817348131,
+      "learning_rate": 8.640376478744428e-06,
+      "loss": 0.4539,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9499481327800829,
+      "grad_norm": 0.6351557530508337,
+      "learning_rate": 8.63727098972332e-06,
+      "loss": 0.5072,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9507261410788381,
+      "grad_norm": 0.5411875238006773,
+      "learning_rate": 8.634162517609337e-06,
+      "loss": 0.4831,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9515041493775933,
+      "grad_norm": 0.49795443442734255,
+      "learning_rate": 8.63105106495188e-06,
+      "loss": 0.4161,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9522821576763485,
+      "grad_norm": 0.6359730804819647,
+      "learning_rate": 8.627936634302794e-06,
+      "loss": 0.4495,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9530601659751037,
+      "grad_norm": 0.5878205488629656,
+      "learning_rate": 8.624819228216364e-06,
+      "loss": 0.4278,
+      "step": 1225
+    },
+    {
+      "epoch": 0.953838174273859,
+      "grad_norm": 0.6077832912854163,
+      "learning_rate": 8.621698849249323e-06,
+      "loss": 0.441,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9546161825726142,
+      "grad_norm": 0.5785016862490937,
+      "learning_rate": 8.618575499960834e-06,
+      "loss": 0.4662,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9553941908713693,
+      "grad_norm": 0.5612119636867512,
+      "learning_rate": 8.615449182912499e-06,
+      "loss": 0.4669,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9561721991701245,
+      "grad_norm": 0.5349062656117417,
+      "learning_rate": 8.612319900668353e-06,
+      "loss": 0.4693,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9569502074688797,
+      "grad_norm": 0.5727734804223958,
+      "learning_rate": 8.609187655794868e-06,
+      "loss": 0.4052,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9577282157676349,
+      "grad_norm": 0.5113605572260594,
+      "learning_rate": 8.606052450860938e-06,
+      "loss": 0.4598,
+      "step": 1231
+    },
+    {
+      "epoch": 0.9585062240663901,
+      "grad_norm": 0.6016243711113223,
+      "learning_rate": 8.60291428843789e-06,
+      "loss": 0.4726,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9592842323651453,
+      "grad_norm": 0.5329511129287781,
+      "learning_rate": 8.599773171099474e-06,
+      "loss": 0.4148,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9600622406639004,
+      "grad_norm": 0.65480613525165,
+      "learning_rate": 8.596629101421868e-06,
+      "loss": 0.4491,
+      "step": 1234
+    },
+    {
+      "epoch": 0.9608402489626556,
+      "grad_norm": 0.6537060215995049,
+      "learning_rate": 8.593482081983662e-06,
+      "loss": 0.4982,
+      "step": 1235
+    },
+    {
+      "epoch": 0.9616182572614108,
+      "grad_norm": 0.6419419125623482,
+      "learning_rate": 8.590332115365874e-06,
+      "loss": 0.4934,
+      "step": 1236
+    },
+    {
+      "epoch": 0.962396265560166,
+      "grad_norm": 0.5844730005644424,
+      "learning_rate": 8.58717920415194e-06,
+      "loss": 0.4352,
+      "step": 1237
+    },
+    {
+      "epoch": 0.9631742738589212,
+      "grad_norm": 0.629237085944135,
+      "learning_rate": 8.584023350927699e-06,
+      "loss": 0.4662,
+      "step": 1238
+    },
+    {
+      "epoch": 0.9639522821576764,
+      "grad_norm": 0.6132394168579092,
+      "learning_rate": 8.580864558281419e-06,
+      "loss": 0.4804,
+      "step": 1239
+    },
+    {
+      "epoch": 0.9647302904564315,
+      "grad_norm": 0.5783137781301194,
+      "learning_rate": 8.577702828803766e-06,
+      "loss": 0.4338,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9655082987551867,
+      "grad_norm": 0.62579135808541,
+      "learning_rate": 8.57453816508782e-06,
+      "loss": 0.4365,
+      "step": 1241
+    },
+    {
+      "epoch": 0.9662863070539419,
+      "grad_norm": 0.5953740256251876,
+      "learning_rate": 8.57137056972907e-06,
+      "loss": 0.4578,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9670643153526971,
+      "grad_norm": 0.5850904341671314,
+      "learning_rate": 8.568200045325403e-06,
+      "loss": 0.4502,
+      "step": 1243
+    },
+    {
+      "epoch": 0.9678423236514523,
+      "grad_norm": 0.5705932688293063,
+      "learning_rate": 8.565026594477112e-06,
+      "loss": 0.465,
+      "step": 1244
+    },
+    {
+      "epoch": 0.9686203319502075,
+      "grad_norm": 0.5807909562462058,
+      "learning_rate": 8.561850219786893e-06,
+      "loss": 0.4632,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9693983402489627,
+      "grad_norm": 0.5382309302909498,
+      "learning_rate": 8.558670923859833e-06,
+      "loss": 0.4085,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9701763485477178,
+      "grad_norm": 0.6096178303993491,
+      "learning_rate": 8.555488709303418e-06,
+      "loss": 0.4775,
+      "step": 1247
+    },
+    {
+      "epoch": 0.970954356846473,
+      "grad_norm": 0.6542706130645093,
+      "learning_rate": 8.552303578727532e-06,
+      "loss": 0.5361,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9717323651452282,
+      "grad_norm": 0.5911844707792128,
+      "learning_rate": 8.549115534744443e-06,
+      "loss": 0.4442,
+      "step": 1249
+    },
+    {
+      "epoch": 0.9725103734439834,
+      "grad_norm": 0.6304960621533435,
+      "learning_rate": 8.545924579968817e-06,
+      "loss": 0.4869,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9732883817427386,
+      "grad_norm": 0.5766675873012993,
+      "learning_rate": 8.542730717017695e-06,
+      "loss": 0.4216,
+      "step": 1251
+    },
+    {
+      "epoch": 0.9740663900414938,
+      "grad_norm": 0.5957932414004596,
+      "learning_rate": 8.539533948510517e-06,
+      "loss": 0.4852,
+      "step": 1252
+    },
+    {
+      "epoch": 0.9748443983402489,
+      "grad_norm": 0.567776836811789,
+      "learning_rate": 8.536334277069098e-06,
+      "loss": 0.4874,
+      "step": 1253
+    },
+    {
+      "epoch": 0.9756224066390041,
+      "grad_norm": 0.6814654387196332,
+      "learning_rate": 8.533131705317635e-06,
+      "loss": 0.514,
+      "step": 1254
+    },
+    {
+      "epoch": 0.9764004149377593,
+      "grad_norm": 0.5109632813588798,
+      "learning_rate": 8.529926235882701e-06,
+      "loss": 0.4357,
+      "step": 1255
+    },
+    {
+      "epoch": 0.9771784232365145,
+      "grad_norm": 0.5540936177710905,
+      "learning_rate": 8.526717871393255e-06,
+      "loss": 0.4563,
+      "step": 1256
+    },
+    {
+      "epoch": 0.9779564315352697,
+      "grad_norm": 0.5133112670203739,
+      "learning_rate": 8.52350661448062e-06,
+      "loss": 0.4597,
+      "step": 1257
+    },
+    {
+      "epoch": 0.9787344398340249,
+      "grad_norm": 0.6013658317835902,
+      "learning_rate": 8.520292467778494e-06,
+      "loss": 0.4373,
+      "step": 1258
+    },
+    {
+      "epoch": 0.97951244813278,
+      "grad_norm": 0.689930329349292,
+      "learning_rate": 8.517075433922948e-06,
+      "loss": 0.5396,
+      "step": 1259
+    },
+    {
+      "epoch": 0.9802904564315352,
+      "grad_norm": 0.6110604409213041,
+      "learning_rate": 8.51385551555242e-06,
+      "loss": 0.4724,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9810684647302904,
+      "grad_norm": 0.5820176448057848,
+      "learning_rate": 8.51063271530771e-06,
+      "loss": 0.4903,
+      "step": 1261
+    },
+    {
+      "epoch": 0.9818464730290456,
+      "grad_norm": 0.6224823479884644,
+      "learning_rate": 8.507407035831989e-06,
+      "loss": 0.4638,
+      "step": 1262
+    },
+    {
+      "epoch": 0.9826244813278008,
+      "grad_norm": 0.5485811691916527,
+      "learning_rate": 8.50417847977078e-06,
+      "loss": 0.4457,
+      "step": 1263
+    },
+    {
+      "epoch": 0.983402489626556,
+      "grad_norm": 0.5618948894165271,
+      "learning_rate": 8.500947049771973e-06,
+      "loss": 0.4256,
+      "step": 1264
+    },
+    {
+      "epoch": 0.9841804979253111,
+      "grad_norm": 0.4998887484845048,
+      "learning_rate": 8.497712748485812e-06,
+      "loss": 0.4476,
+      "step": 1265
+    },
+    {
+      "epoch": 0.9849585062240664,
+      "grad_norm": 0.5241837237377934,
+      "learning_rate": 8.494475578564892e-06,
+      "loss": 0.4317,
+      "step": 1266
+    },
+    {
+      "epoch": 0.9857365145228216,
+      "grad_norm": 0.5885538551696331,
+      "learning_rate": 8.49123554266417e-06,
+      "loss": 0.4703,
+      "step": 1267
+    },
+    {
+      "epoch": 0.9865145228215768,
+      "grad_norm": 0.5826202302092038,
+      "learning_rate": 8.487992643440945e-06,
+      "loss": 0.4652,
+      "step": 1268
+    },
+    {
+      "epoch": 0.987292531120332,
+      "grad_norm": 0.6100846381775894,
+      "learning_rate": 8.484746883554868e-06,
+      "loss": 0.4357,
+      "step": 1269
+    },
+    {
+      "epoch": 0.9880705394190872,
+      "grad_norm": 0.5223728623518132,
+      "learning_rate": 8.481498265667938e-06,
+      "loss": 0.4476,
+      "step": 1270
+    },
+    {
+      "epoch": 0.9888485477178424,
+      "grad_norm": 0.5589492242734003,
+      "learning_rate": 8.478246792444491e-06,
+      "loss": 0.4622,
+      "step": 1271
+    },
+    {
+      "epoch": 0.9896265560165975,
+      "grad_norm": 0.5296121392538022,
+      "learning_rate": 8.474992466551213e-06,
+      "loss": 0.4373,
+      "step": 1272
+    },
+    {
+      "epoch": 0.9904045643153527,
+      "grad_norm": 0.5243700194471825,
+      "learning_rate": 8.471735290657126e-06,
+      "loss": 0.4628,
+      "step": 1273
+    },
+    {
+      "epoch": 0.9911825726141079,
+      "grad_norm": 0.6518285643448622,
+      "learning_rate": 8.468475267433587e-06,
+      "loss": 0.5147,
+      "step": 1274
+    },
+    {
+      "epoch": 0.9919605809128631,
+      "grad_norm": 0.6285997065314599,
+      "learning_rate": 8.465212399554292e-06,
+      "loss": 0.4883,
+      "step": 1275
+    },
+    {
+      "epoch": 0.9927385892116183,
+      "grad_norm": 0.5556960709842628,
+      "learning_rate": 8.46194668969527e-06,
+      "loss": 0.446,
+      "step": 1276
+    },
+    {
+      "epoch": 0.9935165975103735,
+      "grad_norm": 0.6497027291454285,
+      "learning_rate": 8.458678140534877e-06,
+      "loss": 0.5065,
+      "step": 1277
+    },
+    {
+      "epoch": 0.9942946058091287,
+      "grad_norm": 0.5483571541423276,
+      "learning_rate": 8.455406754753804e-06,
+      "loss": 0.427,
+      "step": 1278
+    },
+    {
+      "epoch": 0.9950726141078838,
+      "grad_norm": 0.5747985822385383,
+      "learning_rate": 8.452132535035059e-06,
+      "loss": 0.4589,
+      "step": 1279
+    },
+    {
+      "epoch": 0.995850622406639,
+      "grad_norm": 0.5426103372120343,
+      "learning_rate": 8.448855484063983e-06,
+      "loss": 0.4291,
+      "step": 1280
+    },
+    {
+      "epoch": 0.9966286307053942,
+      "grad_norm": 0.5467210391551138,
+      "learning_rate": 8.445575604528237e-06,
+      "loss": 0.4509,
+      "step": 1281
+    },
+    {
+      "epoch": 0.9974066390041494,
+      "grad_norm": 0.5986824392676836,
+      "learning_rate": 8.442292899117801e-06,
+      "loss": 0.4392,
+      "step": 1282
+    },
+    {
+      "epoch": 0.9981846473029046,
+      "grad_norm": 0.5556624556408858,
+      "learning_rate": 8.43900737052497e-06,
+      "loss": 0.4841,
+      "step": 1283
+    },
+    {
+      "epoch": 0.9989626556016598,
+      "grad_norm": 0.6075900962677953,
+      "learning_rate": 8.435719021444357e-06,
+      "loss": 0.4594,
+      "step": 1284
+    },
+    {
+      "epoch": 0.9997406639004149,
+      "grad_norm": 0.5988032262859158,
+      "learning_rate": 8.43242785457289e-06,
+      "loss": 0.4681,
+      "step": 1285
+    },
+    {
+      "epoch": 1.0005186721991701,
+      "grad_norm": 1.458656265687863,
+      "learning_rate": 8.429133872609806e-06,
+      "loss": 0.813,
+      "step": 1286
+    },
+    {
+      "epoch": 1.0012966804979253,
+      "grad_norm": 0.6240704887901364,
+      "learning_rate": 8.425837078256647e-06,
+      "loss": 0.4318,
+      "step": 1287
+    },
+    {
+      "epoch": 1.0020746887966805,
+      "grad_norm": 0.528305910823561,
+      "learning_rate": 8.422537474217268e-06,
+      "loss": 0.4003,
+      "step": 1288
+    },
+    {
+      "epoch": 1.0028526970954357,
+      "grad_norm": 0.48551169536860367,
+      "learning_rate": 8.419235063197828e-06,
+      "loss": 0.3502,
+      "step": 1289
+    },
+    {
+      "epoch": 1.0036307053941909,
+      "grad_norm": 0.5592835274457795,
+      "learning_rate": 8.415929847906782e-06,
+      "loss": 0.472,
+      "step": 1290
+    },
+    {
+      "epoch": 1.004408713692946,
+      "grad_norm": 0.5924230614868322,
+      "learning_rate": 8.412621831054891e-06,
+      "loss": 0.4413,
+      "step": 1291
+    },
+    {
+      "epoch": 1.0051867219917012,
+      "grad_norm": 0.6304842307011802,
+      "learning_rate": 8.409311015355207e-06,
+      "loss": 0.4271,
+      "step": 1292
+    },
+    {
+      "epoch": 1.0059647302904564,
+      "grad_norm": 0.5629296772700644,
+      "learning_rate": 8.40599740352309e-06,
+      "loss": 0.3873,
+      "step": 1293
+    },
+    {
+      "epoch": 1.0067427385892116,
+      "grad_norm": 0.5496516473999633,
+      "learning_rate": 8.402680998276181e-06,
+      "loss": 0.329,
+      "step": 1294
+    },
+    {
+      "epoch": 1.0075207468879668,
+      "grad_norm": 0.5683875356208078,
+      "learning_rate": 8.399361802334413e-06,
+      "loss": 0.3971,
+      "step": 1295
+    },
+    {
+      "epoch": 1.008298755186722,
+      "grad_norm": 0.5505465139643404,
+      "learning_rate": 8.396039818420016e-06,
+      "loss": 0.3866,
+      "step": 1296
+    },
+    {
+      "epoch": 1.0090767634854771,
+      "grad_norm": 0.6135080411151075,
+      "learning_rate": 8.392715049257498e-06,
+      "loss": 0.4006,
+      "step": 1297
+    },
+    {
+      "epoch": 1.0098547717842323,
+      "grad_norm": 0.615277967084249,
+      "learning_rate": 8.389387497573658e-06,
+      "loss": 0.452,
+      "step": 1298
+    },
+    {
+      "epoch": 1.0106327800829875,
+      "grad_norm": 0.5795392553808958,
+      "learning_rate": 8.38605716609757e-06,
+      "loss": 0.4268,
+      "step": 1299
+    },
+    {
+      "epoch": 1.0114107883817427,
+      "grad_norm": 0.5614167466487714,
+      "learning_rate": 8.382724057560594e-06,
+      "loss": 0.4503,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0121887966804979,
+      "grad_norm": 0.6291911154524816,
+      "learning_rate": 8.379388174696366e-06,
+      "loss": 0.409,
+      "step": 1301
+    },
+    {
+      "epoch": 1.012966804979253,
+      "grad_norm": 0.6865845054713223,
+      "learning_rate": 8.376049520240794e-06,
+      "loss": 0.3684,
+      "step": 1302
+    },
+    {
+      "epoch": 1.0137448132780082,
+      "grad_norm": 0.6583692348957879,
+      "learning_rate": 8.372708096932066e-06,
+      "loss": 0.4153,
+      "step": 1303
+    },
+    {
+      "epoch": 1.0145228215767634,
+      "grad_norm": 0.5452882605815004,
+      "learning_rate": 8.369363907510632e-06,
+      "loss": 0.3605,
+      "step": 1304
+    },
+    {
+      "epoch": 1.0153008298755186,
+      "grad_norm": 0.6682475554417185,
+      "learning_rate": 8.366016954719217e-06,
+      "loss": 0.4509,
+      "step": 1305
+    },
+    {
+      "epoch": 1.0160788381742738,
+      "grad_norm": 0.5600687417818717,
+      "learning_rate": 8.362667241302812e-06,
+      "loss": 0.3733,
+      "step": 1306
+    },
+    {
+      "epoch": 1.016856846473029,
+      "grad_norm": 0.652183046193633,
+      "learning_rate": 8.359314770008668e-06,
+      "loss": 0.4805,
+      "step": 1307
+    },
+    {
+      "epoch": 1.0176348547717842,
+      "grad_norm": 0.5701393772390344,
+      "learning_rate": 8.355959543586303e-06,
+      "loss": 0.3713,
+      "step": 1308
+    },
+    {
+      "epoch": 1.0184128630705394,
+      "grad_norm": 0.5619736698010922,
+      "learning_rate": 8.35260156478749e-06,
+      "loss": 0.4,
+      "step": 1309
+    },
+    {
+      "epoch": 1.0191908713692945,
+      "grad_norm": 0.6112901961721919,
+      "learning_rate": 8.349240836366265e-06,
+      "loss": 0.418,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0199688796680497,
+      "grad_norm": 0.5613290132845083,
+      "learning_rate": 8.345877361078912e-06,
+      "loss": 0.3906,
+      "step": 1311
+    },
+    {
+      "epoch": 1.020746887966805,
+      "grad_norm": 0.5475896749988102,
+      "learning_rate": 8.342511141683974e-06,
+      "loss": 0.3945,
+      "step": 1312
+    },
+    {
+      "epoch": 1.02152489626556,
+      "grad_norm": 0.5237513482701716,
+      "learning_rate": 8.339142180942239e-06,
+      "loss": 0.4195,
+      "step": 1313
+    },
+    {
+      "epoch": 1.0223029045643153,
+      "grad_norm": 0.5552339413944289,
+      "learning_rate": 8.335770481616749e-06,
+      "loss": 0.416,
+      "step": 1314
+    },
+    {
+      "epoch": 1.0230809128630705,
+      "grad_norm": 0.6350616837090877,
+      "learning_rate": 8.332396046472789e-06,
+      "loss": 0.4028,
+      "step": 1315
+    },
+    {
+      "epoch": 1.0238589211618256,
+      "grad_norm": 0.6054053113425739,
+      "learning_rate": 8.329018878277887e-06,
+      "loss": 0.4054,
+      "step": 1316
+    },
+    {
+      "epoch": 1.0246369294605808,
+      "grad_norm": 0.5120792361205516,
+      "learning_rate": 8.325638979801815e-06,
+      "loss": 0.3983,
+      "step": 1317
+    },
+    {
+      "epoch": 1.025414937759336,
+      "grad_norm": 0.5614377816882914,
+      "learning_rate": 8.322256353816582e-06,
+      "loss": 0.3953,
+      "step": 1318
+    },
+    {
+      "epoch": 1.0261929460580912,
+      "grad_norm": 0.6538418864733029,
+      "learning_rate": 8.318871003096436e-06,
+      "loss": 0.429,
+      "step": 1319
+    },
+    {
+      "epoch": 1.0269709543568464,
+      "grad_norm": 0.6415737191815186,
+      "learning_rate": 8.315482930417855e-06,
+      "loss": 0.463,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0277489626556016,
+      "grad_norm": 0.5222246746173295,
+      "learning_rate": 8.312092138559558e-06,
+      "loss": 0.4017,
+      "step": 1321
+    },
+    {
+      "epoch": 1.0285269709543567,
+      "grad_norm": 0.5771916071433528,
+      "learning_rate": 8.308698630302485e-06,
+      "loss": 0.4455,
+      "step": 1322
+    },
+    {
+      "epoch": 1.029304979253112,
+      "grad_norm": 0.6354903883914541,
+      "learning_rate": 8.30530240842981e-06,
+      "loss": 0.3647,
+      "step": 1323
+    },
+    {
+      "epoch": 1.0300829875518671,
+      "grad_norm": 0.686466428181603,
+      "learning_rate": 8.30190347572693e-06,
+      "loss": 0.4391,
+      "step": 1324
+    },
+    {
+      "epoch": 1.0308609958506225,
+      "grad_norm": 0.5278534859421102,
+      "learning_rate": 8.298501834981465e-06,
+      "loss": 0.3836,
+      "step": 1325
+    },
+    {
+      "epoch": 1.0316390041493775,
+      "grad_norm": 0.6496490003361728,
+      "learning_rate": 8.295097488983257e-06,
+      "loss": 0.415,
+      "step": 1326
+    },
+    {
+      "epoch": 1.0324170124481329,
+      "grad_norm": 0.5049353602575389,
+      "learning_rate": 8.291690440524367e-06,
+      "loss": 0.3712,
+      "step": 1327
+    },
+    {
+      "epoch": 1.033195020746888,
+      "grad_norm": 0.5530966409270982,
+      "learning_rate": 8.28828069239907e-06,
+      "loss": 0.4154,
+      "step": 1328
+    },
+    {
+      "epoch": 1.0339730290456433,
+      "grad_norm": 0.5133102854813661,
+      "learning_rate": 8.28486824740386e-06,
+      "loss": 0.3979,
+      "step": 1329
+    },
+    {
+      "epoch": 1.0347510373443984,
+      "grad_norm": 0.4840449064795839,
+      "learning_rate": 8.281453108337437e-06,
+      "loss": 0.3043,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0355290456431536,
+      "grad_norm": 0.6057723010502458,
+      "learning_rate": 8.278035278000715e-06,
+      "loss": 0.4334,
+      "step": 1331
+    },
+    {
+      "epoch": 1.0363070539419088,
+      "grad_norm": 0.5256582180538489,
+      "learning_rate": 8.274614759196813e-06,
+      "loss": 0.3785,
+      "step": 1332
+    },
+    {
+      "epoch": 1.037085062240664,
+      "grad_norm": 0.5768788068965635,
+      "learning_rate": 8.271191554731056e-06,
+      "loss": 0.4251,
+      "step": 1333
+    },
+    {
+      "epoch": 1.0378630705394192,
+      "grad_norm": 0.5476227604793121,
+      "learning_rate": 8.267765667410968e-06,
+      "loss": 0.3797,
+      "step": 1334
+    },
+    {
+      "epoch": 1.0386410788381744,
+      "grad_norm": 0.5886196383364174,
+      "learning_rate": 8.26433710004628e-06,
+      "loss": 0.3934,
+      "step": 1335
+    },
+    {
+      "epoch": 1.0394190871369295,
+      "grad_norm": 0.6480007387260713,
+      "learning_rate": 8.260905855448917e-06,
+      "loss": 0.4095,
+      "step": 1336
+    },
+    {
+      "epoch": 1.0401970954356847,
+      "grad_norm": 0.6336180600979558,
+      "learning_rate": 8.257471936432998e-06,
+      "loss": 0.3947,
+      "step": 1337
+    },
+    {
+      "epoch": 1.04097510373444,
+      "grad_norm": 0.6368063669567644,
+      "learning_rate": 8.25403534581484e-06,
+      "loss": 0.4234,
+      "step": 1338
+    },
+    {
+      "epoch": 1.041753112033195,
+      "grad_norm": 0.5499306749733116,
+      "learning_rate": 8.250596086412947e-06,
+      "loss": 0.3878,
+      "step": 1339
+    },
+    {
+      "epoch": 1.0425311203319503,
+      "grad_norm": 0.5708899977435344,
+      "learning_rate": 8.247154161048016e-06,
+      "loss": 0.3951,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0433091286307055,
+      "grad_norm": 0.5748820954165432,
+      "learning_rate": 8.243709572542923e-06,
+      "loss": 0.4646,
+      "step": 1341
+    },
+    {
+      "epoch": 1.0440871369294606,
+      "grad_norm": 0.6252957537692352,
+      "learning_rate": 8.24026232372274e-06,
+      "loss": 0.4107,
+      "step": 1342
+    },
+    {
+      "epoch": 1.0448651452282158,
+      "grad_norm": 0.5951110860606619,
+      "learning_rate": 8.236812417414709e-06,
+      "loss": 0.4622,
+      "step": 1343
+    },
+    {
+      "epoch": 1.045643153526971,
+      "grad_norm": 0.5033097845201155,
+      "learning_rate": 8.233359856448258e-06,
+      "loss": 0.3779,
+      "step": 1344
+    },
+    {
+      "epoch": 1.0464211618257262,
+      "grad_norm": 0.5456044681552226,
+      "learning_rate": 8.22990464365499e-06,
+      "loss": 0.435,
+      "step": 1345
+    },
+    {
+      "epoch": 1.0471991701244814,
+      "grad_norm": 0.5502362374902866,
+      "learning_rate": 8.226446781868684e-06,
+      "loss": 0.3968,
+      "step": 1346
+    },
+    {
+      "epoch": 1.0479771784232366,
+      "grad_norm": 0.5119037471417479,
+      "learning_rate": 8.222986273925292e-06,
+      "loss": 0.3769,
+      "step": 1347
+    },
+    {
+      "epoch": 1.0487551867219918,
+      "grad_norm": 0.5283081049324071,
+      "learning_rate": 8.219523122662933e-06,
+      "loss": 0.3384,
+      "step": 1348
+    },
+    {
+      "epoch": 1.049533195020747,
+      "grad_norm": 0.6202669560638924,
+      "learning_rate": 8.216057330921901e-06,
+      "loss": 0.451,
+      "step": 1349
+    },
+    {
+      "epoch": 1.0503112033195021,
+      "grad_norm": 0.5671247194576388,
+      "learning_rate": 8.212588901544644e-06,
+      "loss": 0.4221,
+      "step": 1350
+    },
+    {
+      "epoch": 1.0510892116182573,
+      "grad_norm": 0.5840637310161602,
+      "learning_rate": 8.209117837375786e-06,
+      "loss": 0.4189,
+      "step": 1351
+    },
+    {
+      "epoch": 1.0518672199170125,
+      "grad_norm": 0.5615055718313914,
+      "learning_rate": 8.205644141262104e-06,
+      "loss": 0.3534,
+      "step": 1352
+    },
+    {
+      "epoch": 1.0526452282157677,
+      "grad_norm": 0.5032457439741671,
+      "learning_rate": 8.202167816052535e-06,
+      "loss": 0.4051,
+      "step": 1353
+    },
+    {
+      "epoch": 1.0534232365145229,
+      "grad_norm": 0.651198587491215,
+      "learning_rate": 8.198688864598175e-06,
+      "loss": 0.4758,
+      "step": 1354
+    },
+    {
+      "epoch": 1.054201244813278,
+      "grad_norm": 0.5584028264016039,
+      "learning_rate": 8.195207289752266e-06,
+      "loss": 0.3885,
+      "step": 1355
+    },
+    {
+      "epoch": 1.0549792531120332,
+      "grad_norm": 0.603754670403591,
+      "learning_rate": 8.191723094370217e-06,
+      "loss": 0.3996,
+      "step": 1356
+    },
+    {
+      "epoch": 1.0557572614107884,
+      "grad_norm": 0.5137350073758362,
+      "learning_rate": 8.188236281309568e-06,
+      "loss": 0.3572,
+      "step": 1357
+    },
+    {
+      "epoch": 1.0565352697095436,
+      "grad_norm": 0.6048946071788108,
+      "learning_rate": 8.184746853430018e-06,
+      "loss": 0.4307,
+      "step": 1358
+    },
+    {
+      "epoch": 1.0573132780082988,
+      "grad_norm": 0.5669804938596207,
+      "learning_rate": 8.181254813593406e-06,
+      "loss": 0.3967,
+      "step": 1359
+    },
+    {
+      "epoch": 1.058091286307054,
+      "grad_norm": 0.5333322597117555,
+      "learning_rate": 8.177760164663715e-06,
+      "loss": 0.4015,
+      "step": 1360
+    },
+    {
+      "epoch": 1.0588692946058091,
+      "grad_norm": 0.5661451246270239,
+      "learning_rate": 8.174262909507067e-06,
+      "loss": 0.4353,
+      "step": 1361
+    },
+    {
+      "epoch": 1.0596473029045643,
+      "grad_norm": 0.5691854467393215,
+      "learning_rate": 8.170763050991719e-06,
+      "loss": 0.3754,
+      "step": 1362
+    },
+    {
+      "epoch": 1.0604253112033195,
+      "grad_norm": 0.6295628917964932,
+      "learning_rate": 8.167260591988069e-06,
+      "loss": 0.4078,
+      "step": 1363
+    },
+    {
+      "epoch": 1.0612033195020747,
+      "grad_norm": 0.5454232311255419,
+      "learning_rate": 8.163755535368643e-06,
+      "loss": 0.4075,
+      "step": 1364
+    },
+    {
+      "epoch": 1.0619813278008299,
+      "grad_norm": 0.6257227275396263,
+      "learning_rate": 8.1602478840081e-06,
+      "loss": 0.3732,
+      "step": 1365
+    },
+    {
+      "epoch": 1.062759336099585,
+      "grad_norm": 0.6100577346283473,
+      "learning_rate": 8.156737640783224e-06,
+      "loss": 0.4014,
+      "step": 1366
+    },
+    {
+      "epoch": 1.0635373443983402,
+      "grad_norm": 0.5693103498723098,
+      "learning_rate": 8.153224808572926e-06,
+      "loss": 0.3737,
+      "step": 1367
+    },
+    {
+      "epoch": 1.0643153526970954,
+      "grad_norm": 0.5714640967243104,
+      "learning_rate": 8.149709390258242e-06,
+      "loss": 0.4046,
+      "step": 1368
+    },
+    {
+      "epoch": 1.0650933609958506,
+      "grad_norm": 0.5251962237135068,
+      "learning_rate": 8.146191388722329e-06,
+      "loss": 0.3557,
+      "step": 1369
+    },
+    {
+      "epoch": 1.0658713692946058,
+      "grad_norm": 0.585610980907101,
+      "learning_rate": 8.142670806850462e-06,
+      "loss": 0.437,
+      "step": 1370
+    },
+    {
+      "epoch": 1.066649377593361,
+      "grad_norm": 0.5481603371135878,
+      "learning_rate": 8.139147647530029e-06,
+      "loss": 0.4393,
+      "step": 1371
+    },
+    {
+      "epoch": 1.0674273858921162,
+      "grad_norm": 0.5553260803597959,
+      "learning_rate": 8.135621913650533e-06,
+      "loss": 0.3803,
+      "step": 1372
+    },
+    {
+      "epoch": 1.0682053941908713,
+      "grad_norm": 0.5493888222350103,
+      "learning_rate": 8.132093608103597e-06,
+      "loss": 0.4283,
+      "step": 1373
+    },
+    {
+      "epoch": 1.0689834024896265,
+      "grad_norm": 0.5287526778227691,
+      "learning_rate": 8.128562733782941e-06,
+      "loss": 0.4104,
+      "step": 1374
+    },
+    {
+      "epoch": 1.0697614107883817,
+      "grad_norm": 0.5619646089323247,
+      "learning_rate": 8.125029293584398e-06,
+      "loss": 0.4159,
+      "step": 1375
+    },
+    {
+      "epoch": 1.070539419087137,
+      "grad_norm": 0.5886044929918628,
+      "learning_rate": 8.121493290405906e-06,
+      "loss": 0.4327,
+      "step": 1376
+    },
+    {
+      "epoch": 1.071317427385892,
+      "grad_norm": 0.539368390511097,
+      "learning_rate": 8.117954727147504e-06,
+      "loss": 0.4164,
+      "step": 1377
+    },
+    {
+      "epoch": 1.0720954356846473,
+      "grad_norm": 0.5423226361999143,
+      "learning_rate": 8.114413606711327e-06,
+      "loss": 0.3441,
+      "step": 1378
+    },
+    {
+      "epoch": 1.0728734439834025,
+      "grad_norm": 0.6224247145522701,
+      "learning_rate": 8.110869932001613e-06,
+      "loss": 0.4617,
+      "step": 1379
+    },
+    {
+      "epoch": 1.0736514522821576,
+      "grad_norm": 0.5581368965145439,
+      "learning_rate": 8.10732370592469e-06,
+      "loss": 0.4177,
+      "step": 1380
+    },
+    {
+      "epoch": 1.0744294605809128,
+      "grad_norm": 0.5607983531161876,
+      "learning_rate": 8.103774931388983e-06,
+      "loss": 0.3943,
+      "step": 1381
+    },
+    {
+      "epoch": 1.075207468879668,
+      "grad_norm": 0.640142137155787,
+      "learning_rate": 8.100223611305003e-06,
+      "loss": 0.4549,
+      "step": 1382
+    },
+    {
+      "epoch": 1.0759854771784232,
+      "grad_norm": 0.5367508994196175,
+      "learning_rate": 8.096669748585352e-06,
+      "loss": 0.3699,
+      "step": 1383
+    },
+    {
+      "epoch": 1.0767634854771784,
+      "grad_norm": 0.5862247715778217,
+      "learning_rate": 8.093113346144716e-06,
+      "loss": 0.3761,
+      "step": 1384
+    },
+    {
+      "epoch": 1.0775414937759336,
+      "grad_norm": 0.5600692685242706,
+      "learning_rate": 8.089554406899863e-06,
+      "loss": 0.4221,
+      "step": 1385
+    },
+    {
+      "epoch": 1.0783195020746887,
+      "grad_norm": 0.5305242304112296,
+      "learning_rate": 8.085992933769644e-06,
+      "loss": 0.3469,
+      "step": 1386
+    },
+    {
+      "epoch": 1.079097510373444,
+      "grad_norm": 0.6185923453111936,
+      "learning_rate": 8.082428929674983e-06,
+      "loss": 0.4357,
+      "step": 1387
+    },
+    {
+      "epoch": 1.079875518672199,
+      "grad_norm": 0.6156333034671444,
+      "learning_rate": 8.078862397538889e-06,
+      "loss": 0.4297,
+      "step": 1388
+    },
+    {
+      "epoch": 1.0806535269709543,
+      "grad_norm": 0.5063964675157749,
+      "learning_rate": 8.075293340286431e-06,
+      "loss": 0.3371,
+      "step": 1389
+    },
+    {
+      "epoch": 1.0814315352697095,
+      "grad_norm": 0.46860598389125563,
+      "learning_rate": 8.071721760844763e-06,
+      "loss": 0.3784,
+      "step": 1390
+    },
+    {
+      "epoch": 1.0822095435684647,
+      "grad_norm": 0.5737836058343903,
+      "learning_rate": 8.0681476621431e-06,
+      "loss": 0.3851,
+      "step": 1391
+    },
+    {
+      "epoch": 1.0829875518672198,
+      "grad_norm": 0.6459359461198065,
+      "learning_rate": 8.064571047112724e-06,
+      "loss": 0.3863,
+      "step": 1392
+    },
+    {
+      "epoch": 1.083765560165975,
+      "grad_norm": 0.5573998996482928,
+      "learning_rate": 8.06099191868698e-06,
+      "loss": 0.4041,
+      "step": 1393
+    },
+    {
+      "epoch": 1.0845435684647302,
+      "grad_norm": 0.5969322231046646,
+      "learning_rate": 8.057410279801277e-06,
+      "loss": 0.4657,
+      "step": 1394
+    },
+    {
+      "epoch": 1.0853215767634854,
+      "grad_norm": 0.5562912642186019,
+      "learning_rate": 8.05382613339308e-06,
+      "loss": 0.4248,
+      "step": 1395
+    },
+    {
+      "epoch": 1.0860995850622406,
+      "grad_norm": 0.6625760570571028,
+      "learning_rate": 8.05023948240191e-06,
+      "loss": 0.3616,
+      "step": 1396
+    },
+    {
+      "epoch": 1.0868775933609958,
+      "grad_norm": 0.5630947407247578,
+      "learning_rate": 8.046650329769349e-06,
+      "loss": 0.4528,
+      "step": 1397
+    },
+    {
+      "epoch": 1.087655601659751,
+      "grad_norm": 0.5364169969918922,
+      "learning_rate": 8.043058678439026e-06,
+      "loss": 0.3672,
+      "step": 1398
+    },
+    {
+      "epoch": 1.0884336099585061,
+      "grad_norm": 0.600851294910548,
+      "learning_rate": 8.039464531356614e-06,
+      "loss": 0.3816,
+      "step": 1399
+    },
+    {
+      "epoch": 1.0892116182572613,
+      "grad_norm": 0.7174268533985041,
+      "learning_rate": 8.035867891469843e-06,
+      "loss": 0.4403,
+      "step": 1400
+    },
+    {
+      "epoch": 1.0899896265560165,
+      "grad_norm": 0.5592657837011809,
+      "learning_rate": 8.032268761728476e-06,
+      "loss": 0.3945,
+      "step": 1401
+    },
+    {
+      "epoch": 1.0907676348547717,
+      "grad_norm": 0.5679806238218321,
+      "learning_rate": 8.028667145084331e-06,
+      "loss": 0.3852,
+      "step": 1402
+    },
+    {
+      "epoch": 1.0915456431535269,
+      "grad_norm": 0.5464241759004352,
+      "learning_rate": 8.025063044491257e-06,
+      "loss": 0.3796,
+      "step": 1403
+    },
+    {
+      "epoch": 1.0923236514522823,
+      "grad_norm": 0.6503734851773109,
+      "learning_rate": 8.021456462905141e-06,
+      "loss": 0.4093,
+      "step": 1404
+    },
+    {
+      "epoch": 1.0931016597510372,
+      "grad_norm": 0.5819604530030845,
+      "learning_rate": 8.017847403283907e-06,
+      "loss": 0.4674,
+      "step": 1405
+    },
+    {
+      "epoch": 1.0938796680497926,
+      "grad_norm": 0.5676491006982237,
+      "learning_rate": 8.014235868587511e-06,
+      "loss": 0.4484,
+      "step": 1406
+    },
+    {
+      "epoch": 1.0946576763485476,
+      "grad_norm": 0.5564835071103597,
+      "learning_rate": 8.01062186177794e-06,
+      "loss": 0.3769,
+      "step": 1407
+    },
+    {
+      "epoch": 1.095435684647303,
+      "grad_norm": 0.5307200221717499,
+      "learning_rate": 8.007005385819201e-06,
+      "loss": 0.3707,
+      "step": 1408
+    },
+    {
+      "epoch": 1.0962136929460582,
+      "grad_norm": 0.5587289075445787,
+      "learning_rate": 8.003386443677337e-06,
+      "loss": 0.3901,
+      "step": 1409
+    },
+    {
+      "epoch": 1.0969917012448134,
+      "grad_norm": 0.5440984559144222,
+      "learning_rate": 7.999765038320406e-06,
+      "loss": 0.4277,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0977697095435686,
+      "grad_norm": 0.45815581053656024,
+      "learning_rate": 7.996141172718492e-06,
+      "loss": 0.3392,
+      "step": 1411
+    },
+    {
+      "epoch": 1.0985477178423237,
+      "grad_norm": 0.5689723879218066,
+      "learning_rate": 7.99251484984369e-06,
+      "loss": 0.3852,
+      "step": 1412
+    },
+    {
+      "epoch": 1.099325726141079,
+      "grad_norm": 0.6058551185357035,
+      "learning_rate": 7.988886072670117e-06,
+      "loss": 0.4148,
+      "step": 1413
+    },
+    {
+      "epoch": 1.1001037344398341,
+      "grad_norm": 0.5271400658893012,
+      "learning_rate": 7.985254844173898e-06,
+      "loss": 0.4011,
+      "step": 1414
+    },
+    {
+      "epoch": 1.1008817427385893,
+      "grad_norm": 0.5476523857384659,
+      "learning_rate": 7.981621167333167e-06,
+      "loss": 0.4181,
+      "step": 1415
+    },
+    {
+      "epoch": 1.1016597510373445,
+      "grad_norm": 0.5982453018292986,
+      "learning_rate": 7.977985045128073e-06,
+      "loss": 0.4617,
+      "step": 1416
+    },
+    {
+      "epoch": 1.1024377593360997,
+      "grad_norm": 0.596902365023028,
+      "learning_rate": 7.97434648054077e-06,
+      "loss": 0.3929,
+      "step": 1417
+    },
+    {
+      "epoch": 1.1032157676348548,
+      "grad_norm": 0.6095209446977282,
+      "learning_rate": 7.970705476555403e-06,
+      "loss": 0.4342,
+      "step": 1418
+    },
+    {
+      "epoch": 1.10399377593361,
+      "grad_norm": 0.6146719336970172,
+      "learning_rate": 7.967062036158134e-06,
+      "loss": 0.446,
+      "step": 1419
+    },
+    {
+      "epoch": 1.1047717842323652,
+      "grad_norm": 0.5531044454718986,
+      "learning_rate": 7.963416162337111e-06,
+      "loss": 0.3548,
+      "step": 1420
+    },
+    {
+      "epoch": 1.1055497925311204,
+      "grad_norm": 0.5860369029104273,
+      "learning_rate": 7.959767858082486e-06,
+      "loss": 0.4243,
+      "step": 1421
+    },
+    {
+      "epoch": 1.1063278008298756,
+      "grad_norm": 0.615855777828817,
+      "learning_rate": 7.956117126386399e-06,
+      "loss": 0.4536,
+      "step": 1422
+    },
+    {
+      "epoch": 1.1071058091286308,
+      "grad_norm": 0.5939993682477374,
+      "learning_rate": 7.952463970242982e-06,
+      "loss": 0.4293,
+      "step": 1423
+    },
+    {
+      "epoch": 1.107883817427386,
+      "grad_norm": 0.5689507331611757,
+      "learning_rate": 7.948808392648358e-06,
+      "loss": 0.357,
+      "step": 1424
+    },
+    {
+      "epoch": 1.1086618257261411,
+      "grad_norm": 0.6118425431965728,
+      "learning_rate": 7.945150396600631e-06,
+      "loss": 0.3963,
+      "step": 1425
+    },
+    {
+      "epoch": 1.1094398340248963,
+      "grad_norm": 0.526180275966419,
+      "learning_rate": 7.941489985099894e-06,
+      "loss": 0.3846,
+      "step": 1426
+    },
+    {
+      "epoch": 1.1102178423236515,
+      "grad_norm": 0.505863656544409,
+      "learning_rate": 7.937827161148217e-06,
+      "loss": 0.3963,
+      "step": 1427
+    },
+    {
+      "epoch": 1.1109958506224067,
+      "grad_norm": 0.5904632880156279,
+      "learning_rate": 7.934161927749652e-06,
+      "loss": 0.4259,
+      "step": 1428
+    },
+    {
+      "epoch": 1.1117738589211619,
+      "grad_norm": 0.47621121662534044,
+      "learning_rate": 7.930494287910224e-06,
+      "loss": 0.338,
+      "step": 1429
+    },
+    {
+      "epoch": 1.112551867219917,
+      "grad_norm": 0.5630148381893848,
+      "learning_rate": 7.92682424463793e-06,
+      "loss": 0.4468,
+      "step": 1430
+    },
+    {
+      "epoch": 1.1133298755186722,
+      "grad_norm": 0.5945547237085325,
+      "learning_rate": 7.923151800942745e-06,
+      "loss": 0.3786,
+      "step": 1431
+    },
+    {
+      "epoch": 1.1141078838174274,
+      "grad_norm": 0.560377513693759,
+      "learning_rate": 7.919476959836608e-06,
+      "loss": 0.3998,
+      "step": 1432
+    },
+    {
+      "epoch": 1.1148858921161826,
+      "grad_norm": 0.5769639945753612,
+      "learning_rate": 7.915799724333423e-06,
+      "loss": 0.4184,
+      "step": 1433
+    },
+    {
+      "epoch": 1.1156639004149378,
+      "grad_norm": 0.49529606117293523,
+      "learning_rate": 7.912120097449062e-06,
+      "loss": 0.3974,
+      "step": 1434
+    },
+    {
+      "epoch": 1.116441908713693,
+      "grad_norm": 0.4998758048588482,
+      "learning_rate": 7.908438082201354e-06,
+      "loss": 0.3717,
+      "step": 1435
+    },
+    {
+      "epoch": 1.1172199170124482,
+      "grad_norm": 0.6329564723237525,
+      "learning_rate": 7.904753681610088e-06,
+      "loss": 0.4024,
+      "step": 1436
+    },
+    {
+      "epoch": 1.1179979253112033,
+      "grad_norm": 0.5849516924514625,
+      "learning_rate": 7.901066898697012e-06,
+      "loss": 0.4206,
+      "step": 1437
+    },
+    {
+      "epoch": 1.1187759336099585,
+      "grad_norm": 0.5584433564151128,
+      "learning_rate": 7.897377736485824e-06,
+      "loss": 0.4002,
+      "step": 1438
+    },
+    {
+      "epoch": 1.1195539419087137,
+      "grad_norm": 0.542587823136344,
+      "learning_rate": 7.893686198002176e-06,
+      "loss": 0.4738,
+      "step": 1439
+    },
+    {
+      "epoch": 1.120331950207469,
+      "grad_norm": 0.5031418747702614,
+      "learning_rate": 7.88999228627367e-06,
+      "loss": 0.3847,
+      "step": 1440
+    },
+    {
+      "epoch": 1.121109958506224,
+      "grad_norm": 0.5044440619477898,
+      "learning_rate": 7.886296004329851e-06,
+      "loss": 0.3732,
+      "step": 1441
+    },
+    {
+      "epoch": 1.1218879668049793,
+      "grad_norm": 0.5662501671844404,
+      "learning_rate": 7.88259735520221e-06,
+      "loss": 0.4227,
+      "step": 1442
+    },
+    {
+      "epoch": 1.1226659751037344,
+      "grad_norm": 0.499229786393981,
+      "learning_rate": 7.878896341924178e-06,
+      "loss": 0.4311,
+      "step": 1443
+    },
+    {
+      "epoch": 1.1234439834024896,
+      "grad_norm": 0.5351163010623946,
+      "learning_rate": 7.875192967531124e-06,
+      "loss": 0.3809,
+      "step": 1444
+    },
+    {
+      "epoch": 1.1242219917012448,
+      "grad_norm": 0.5653799556032734,
+      "learning_rate": 7.871487235060359e-06,
+      "loss": 0.4151,
+      "step": 1445
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.5611894808342737,
+      "learning_rate": 7.867779147551124e-06,
+      "loss": 0.4134,
+      "step": 1446
+    },
+    {
+      "epoch": 1.1257780082987552,
+      "grad_norm": 0.5408281290761537,
+      "learning_rate": 7.864068708044589e-06,
+      "loss": 0.3784,
+      "step": 1447
+    },
+    {
+      "epoch": 1.1265560165975104,
+      "grad_norm": 0.5860399504810941,
+      "learning_rate": 7.860355919583857e-06,
+      "loss": 0.4028,
+      "step": 1448
+    },
+    {
+      "epoch": 1.1273340248962656,
+      "grad_norm": 0.5157554285086084,
+      "learning_rate": 7.856640785213958e-06,
+      "loss": 0.4042,
+      "step": 1449
+    },
+    {
+      "epoch": 1.1281120331950207,
+      "grad_norm": 0.5336850735037264,
+      "learning_rate": 7.85292330798184e-06,
+      "loss": 0.3751,
+      "step": 1450
+    },
+    {
+      "epoch": 1.128890041493776,
+      "grad_norm": 0.48148655559554987,
+      "learning_rate": 7.849203490936384e-06,
+      "loss": 0.3627,
+      "step": 1451
+    },
+    {
+      "epoch": 1.129668049792531,
+      "grad_norm": 0.5173233329111275,
+      "learning_rate": 7.845481337128376e-06,
+      "loss": 0.3776,
+      "step": 1452
+    },
+    {
+      "epoch": 1.1304460580912863,
+      "grad_norm": 0.6251630953369415,
+      "learning_rate": 7.841756849610526e-06,
+      "loss": 0.4426,
+      "step": 1453
+    },
+    {
+      "epoch": 1.1312240663900415,
+      "grad_norm": 0.5768531744288714,
+      "learning_rate": 7.838030031437457e-06,
+      "loss": 0.3952,
+      "step": 1454
+    },
+    {
+      "epoch": 1.1320020746887967,
+      "grad_norm": 0.5335386853387257,
+      "learning_rate": 7.834300885665709e-06,
+      "loss": 0.3424,
+      "step": 1455
+    },
+    {
+      "epoch": 1.1327800829875518,
+      "grad_norm": 0.6675565444807611,
+      "learning_rate": 7.83056941535372e-06,
+      "loss": 0.4826,
+      "step": 1456
+    },
+    {
+      "epoch": 1.133558091286307,
+      "grad_norm": 0.563110369599117,
+      "learning_rate": 7.82683562356184e-06,
+      "loss": 0.415,
+      "step": 1457
+    },
+    {
+      "epoch": 1.1343360995850622,
+      "grad_norm": 0.5112292241098756,
+      "learning_rate": 7.823099513352328e-06,
+      "loss": 0.3459,
+      "step": 1458
+    },
+    {
+      "epoch": 1.1351141078838174,
+      "grad_norm": 0.5381061652698177,
+      "learning_rate": 7.819361087789334e-06,
+      "loss": 0.4223,
+      "step": 1459
+    },
+    {
+      "epoch": 1.1358921161825726,
+      "grad_norm": 0.5478535681308061,
+      "learning_rate": 7.815620349938914e-06,
+      "loss": 0.3937,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1366701244813278,
+      "grad_norm": 0.54326234260466,
+      "learning_rate": 7.811877302869022e-06,
+      "loss": 0.3731,
+      "step": 1461
+    },
+    {
+      "epoch": 1.137448132780083,
+      "grad_norm": 0.5528828396512969,
+      "learning_rate": 7.808131949649499e-06,
+      "loss": 0.3568,
+      "step": 1462
+    },
+    {
+      "epoch": 1.1382261410788381,
+      "grad_norm": 0.5911270261737542,
+      "learning_rate": 7.804384293352081e-06,
+      "loss": 0.4699,
+      "step": 1463
+    },
+    {
+      "epoch": 1.1390041493775933,
+      "grad_norm": 0.49579448777496116,
+      "learning_rate": 7.800634337050394e-06,
+      "loss": 0.3703,
+      "step": 1464
+    },
+    {
+      "epoch": 1.1397821576763485,
+      "grad_norm": 0.5570525567477936,
+      "learning_rate": 7.79688208381995e-06,
+      "loss": 0.4078,
+      "step": 1465
+    },
+    {
+      "epoch": 1.1405601659751037,
+      "grad_norm": 0.5270560207296844,
+      "learning_rate": 7.793127536738143e-06,
+      "loss": 0.3824,
+      "step": 1466
+    },
+    {
+      "epoch": 1.1413381742738589,
+      "grad_norm": 0.5132496868896803,
+      "learning_rate": 7.78937069888425e-06,
+      "loss": 0.3992,
+      "step": 1467
+    },
+    {
+      "epoch": 1.142116182572614,
+      "grad_norm": 0.524156625663267,
+      "learning_rate": 7.785611573339426e-06,
+      "loss": 0.3995,
+      "step": 1468
+    },
+    {
+      "epoch": 1.1428941908713692,
+      "grad_norm": 0.5384709543172357,
+      "learning_rate": 7.781850163186704e-06,
+      "loss": 0.3912,
+      "step": 1469
+    },
+    {
+      "epoch": 1.1436721991701244,
+      "grad_norm": 0.5260962897188535,
+      "learning_rate": 7.778086471510986e-06,
+      "loss": 0.4169,
+      "step": 1470
+    },
+    {
+      "epoch": 1.1444502074688796,
+      "grad_norm": 0.6020023833525956,
+      "learning_rate": 7.774320501399048e-06,
+      "loss": 0.4097,
+      "step": 1471
+    },
+    {
+      "epoch": 1.1452282157676348,
+      "grad_norm": 0.5280420210888803,
+      "learning_rate": 7.770552255939538e-06,
+      "loss": 0.367,
+      "step": 1472
+    },
+    {
+      "epoch": 1.14600622406639,
+      "grad_norm": 0.6156994156600215,
+      "learning_rate": 7.766781738222966e-06,
+      "loss": 0.4709,
+      "step": 1473
+    },
+    {
+      "epoch": 1.1467842323651452,
+      "grad_norm": 0.5178459376983555,
+      "learning_rate": 7.763008951341707e-06,
+      "loss": 0.3609,
+      "step": 1474
+    },
+    {
+      "epoch": 1.1475622406639003,
+      "grad_norm": 0.6403843304078559,
+      "learning_rate": 7.759233898389995e-06,
+      "loss": 0.4284,
+      "step": 1475
+    },
+    {
+      "epoch": 1.1483402489626555,
+      "grad_norm": 0.5935598714701663,
+      "learning_rate": 7.755456582463924e-06,
+      "loss": 0.3273,
+      "step": 1476
+    },
+    {
+      "epoch": 1.1491182572614107,
+      "grad_norm": 0.5797199330301683,
+      "learning_rate": 7.75167700666145e-06,
+      "loss": 0.3918,
+      "step": 1477
+    },
+    {
+      "epoch": 1.1498962655601659,
+      "grad_norm": 0.5902825616263621,
+      "learning_rate": 7.747895174082368e-06,
+      "loss": 0.3713,
+      "step": 1478
+    },
+    {
+      "epoch": 1.150674273858921,
+      "grad_norm": 0.6457853109040355,
+      "learning_rate": 7.744111087828336e-06,
+      "loss": 0.4068,
+      "step": 1479
+    },
+    {
+      "epoch": 1.1514522821576763,
+      "grad_norm": 0.5992143624031135,
+      "learning_rate": 7.740324751002859e-06,
+      "loss": 0.4199,
+      "step": 1480
+    },
+    {
+      "epoch": 1.1522302904564314,
+      "grad_norm": 0.5709469540556855,
+      "learning_rate": 7.736536166711283e-06,
+      "loss": 0.4531,
+      "step": 1481
+    },
+    {
+      "epoch": 1.1530082987551866,
+      "grad_norm": 0.49051530701129054,
+      "learning_rate": 7.732745338060801e-06,
+      "loss": 0.3731,
+      "step": 1482
+    },
+    {
+      "epoch": 1.153786307053942,
+      "grad_norm": 0.5920646238728784,
+      "learning_rate": 7.728952268160445e-06,
+      "loss": 0.4059,
+      "step": 1483
+    },
+    {
+      "epoch": 1.154564315352697,
+      "grad_norm": 0.617866409200694,
+      "learning_rate": 7.725156960121085e-06,
+      "loss": 0.4043,
+      "step": 1484
+    },
+    {
+      "epoch": 1.1553423236514524,
+      "grad_norm": 0.6380121922999537,
+      "learning_rate": 7.72135941705543e-06,
+      "loss": 0.3972,
+      "step": 1485
+    },
+    {
+      "epoch": 1.1561203319502074,
+      "grad_norm": 0.6488232218310072,
+      "learning_rate": 7.717559642078016e-06,
+      "loss": 0.4806,
+      "step": 1486
+    },
+    {
+      "epoch": 1.1568983402489628,
+      "grad_norm": 0.5340719048749198,
+      "learning_rate": 7.713757638305214e-06,
+      "loss": 0.4092,
+      "step": 1487
+    },
+    {
+      "epoch": 1.1576763485477177,
+      "grad_norm": 0.5136721425143437,
+      "learning_rate": 7.709953408855225e-06,
+      "loss": 0.3886,
+      "step": 1488
+    },
+    {
+      "epoch": 1.1584543568464731,
+      "grad_norm": 0.539123278674372,
+      "learning_rate": 7.706146956848068e-06,
+      "loss": 0.3987,
+      "step": 1489
+    },
+    {
+      "epoch": 1.159232365145228,
+      "grad_norm": 0.54294130202285,
+      "learning_rate": 7.702338285405591e-06,
+      "loss": 0.3972,
+      "step": 1490
+    },
+    {
+      "epoch": 1.1600103734439835,
+      "grad_norm": 0.5727244328777308,
+      "learning_rate": 7.698527397651458e-06,
+      "loss": 0.3898,
+      "step": 1491
+    },
+    {
+      "epoch": 1.1607883817427387,
+      "grad_norm": 0.515221203910289,
+      "learning_rate": 7.694714296711155e-06,
+      "loss": 0.4187,
+      "step": 1492
+    },
+    {
+      "epoch": 1.1615663900414939,
+      "grad_norm": 0.511724410011109,
+      "learning_rate": 7.69089898571198e-06,
+      "loss": 0.3641,
+      "step": 1493
+    },
+    {
+      "epoch": 1.162344398340249,
+      "grad_norm": 0.5857335630355712,
+      "learning_rate": 7.687081467783048e-06,
+      "loss": 0.47,
+      "step": 1494
+    },
+    {
+      "epoch": 1.1631224066390042,
+      "grad_norm": 0.5397309867815036,
+      "learning_rate": 7.683261746055278e-06,
+      "loss": 0.3653,
+      "step": 1495
+    },
+    {
+      "epoch": 1.1639004149377594,
+      "grad_norm": 0.491242869395371,
+      "learning_rate": 7.679439823661397e-06,
+      "loss": 0.347,
+      "step": 1496
+    },
+    {
+      "epoch": 1.1646784232365146,
+      "grad_norm": 0.5868532861666693,
+      "learning_rate": 7.675615703735943e-06,
+      "loss": 0.4028,
+      "step": 1497
+    },
+    {
+      "epoch": 1.1654564315352698,
+      "grad_norm": 0.5014726999617953,
+      "learning_rate": 7.671789389415249e-06,
+      "loss": 0.341,
+      "step": 1498
+    },
+    {
+      "epoch": 1.166234439834025,
+      "grad_norm": 0.5676967628093902,
+      "learning_rate": 7.667960883837453e-06,
+      "loss": 0.3654,
+      "step": 1499
+    },
+    {
+      "epoch": 1.1670124481327802,
+      "grad_norm": 0.5588250277653084,
+      "learning_rate": 7.664130190142487e-06,
+      "loss": 0.4514,
+      "step": 1500
+    },
+    {
+      "epoch": 1.1677904564315353,
+      "grad_norm": 0.5138784200355874,
+      "learning_rate": 7.660297311472079e-06,
+      "loss": 0.3422,
+      "step": 1501
+    },
+    {
+      "epoch": 1.1685684647302905,
+      "grad_norm": 0.6218964837473028,
+      "learning_rate": 7.656462250969748e-06,
+      "loss": 0.4103,
+      "step": 1502
+    },
+    {
+      "epoch": 1.1693464730290457,
+      "grad_norm": 0.5558933916358123,
+      "learning_rate": 7.652625011780807e-06,
+      "loss": 0.3718,
+      "step": 1503
+    },
+    {
+      "epoch": 1.170124481327801,
+      "grad_norm": 0.575755937901326,
+      "learning_rate": 7.648785597052346e-06,
+      "loss": 0.3774,
+      "step": 1504
+    },
+    {
+      "epoch": 1.170902489626556,
+      "grad_norm": 0.5412433680999448,
+      "learning_rate": 7.64494400993325e-06,
+      "loss": 0.4142,
+      "step": 1505
+    },
+    {
+      "epoch": 1.1716804979253113,
+      "grad_norm": 0.5730732928099385,
+      "learning_rate": 7.641100253574179e-06,
+      "loss": 0.3886,
+      "step": 1506
+    },
+    {
+      "epoch": 1.1724585062240664,
+      "grad_norm": 0.6463257613953296,
+      "learning_rate": 7.637254331127574e-06,
+      "loss": 0.4464,
+      "step": 1507
+    },
+    {
+      "epoch": 1.1732365145228216,
+      "grad_norm": 0.5172119754236452,
+      "learning_rate": 7.633406245747652e-06,
+      "loss": 0.3613,
+      "step": 1508
+    },
+    {
+      "epoch": 1.1740145228215768,
+      "grad_norm": 0.5905054166587222,
+      "learning_rate": 7.629556000590406e-06,
+      "loss": 0.3891,
+      "step": 1509
+    },
+    {
+      "epoch": 1.174792531120332,
+      "grad_norm": 0.5873872981303458,
+      "learning_rate": 7.6257035988135966e-06,
+      "loss": 0.3908,
+      "step": 1510
+    },
+    {
+      "epoch": 1.1755705394190872,
+      "grad_norm": 0.5070218328759967,
+      "learning_rate": 7.621849043576758e-06,
+      "loss": 0.3495,
+      "step": 1511
+    },
+    {
+      "epoch": 1.1763485477178424,
+      "grad_norm": 0.5726275783513178,
+      "learning_rate": 7.617992338041186e-06,
+      "loss": 0.4276,
+      "step": 1512
+    },
+    {
+      "epoch": 1.1771265560165975,
+      "grad_norm": 0.5550726730014821,
+      "learning_rate": 7.6141334853699415e-06,
+      "loss": 0.3501,
+      "step": 1513
+    },
+    {
+      "epoch": 1.1779045643153527,
+      "grad_norm": 0.5950275316007586,
+      "learning_rate": 7.610272488727848e-06,
+      "loss": 0.4573,
+      "step": 1514
+    },
+    {
+      "epoch": 1.178682572614108,
+      "grad_norm": 0.5801250533108883,
+      "learning_rate": 7.6064093512814854e-06,
+      "loss": 0.388,
+      "step": 1515
+    },
+    {
+      "epoch": 1.179460580912863,
+      "grad_norm": 0.5536482523433435,
+      "learning_rate": 7.602544076199192e-06,
+      "loss": 0.4026,
+      "step": 1516
+    },
+    {
+      "epoch": 1.1802385892116183,
+      "grad_norm": 0.558222849999672,
+      "learning_rate": 7.598676666651053e-06,
+      "loss": 0.3848,
+      "step": 1517
+    },
+    {
+      "epoch": 1.1810165975103735,
+      "grad_norm": 0.6462780276415394,
+      "learning_rate": 7.594807125808911e-06,
+      "loss": 0.4547,
+      "step": 1518
+    },
+    {
+      "epoch": 1.1817946058091287,
+      "grad_norm": 0.6043488262427703,
+      "learning_rate": 7.590935456846353e-06,
+      "loss": 0.4241,
+      "step": 1519
+    },
+    {
+      "epoch": 1.1825726141078838,
+      "grad_norm": 0.5485979177403408,
+      "learning_rate": 7.5870616629387164e-06,
+      "loss": 0.3489,
+      "step": 1520
+    },
+    {
+      "epoch": 1.183350622406639,
+      "grad_norm": 0.6094267276981263,
+      "learning_rate": 7.583185747263072e-06,
+      "loss": 0.3664,
+      "step": 1521
+    },
+    {
+      "epoch": 1.1841286307053942,
+      "grad_norm": 0.7098297874276976,
+      "learning_rate": 7.57930771299824e-06,
+      "loss": 0.4128,
+      "step": 1522
+    },
+    {
+      "epoch": 1.1849066390041494,
+      "grad_norm": 0.498856180869917,
+      "learning_rate": 7.575427563324771e-06,
+      "loss": 0.3143,
+      "step": 1523
+    },
+    {
+      "epoch": 1.1856846473029046,
+      "grad_norm": 0.5687575079276747,
+      "learning_rate": 7.571545301424957e-06,
+      "loss": 0.3739,
+      "step": 1524
+    },
+    {
+      "epoch": 1.1864626556016598,
+      "grad_norm": 0.5983214423315681,
+      "learning_rate": 7.567660930482816e-06,
+      "loss": 0.3726,
+      "step": 1525
+    },
+    {
+      "epoch": 1.187240663900415,
+      "grad_norm": 0.630553504042556,
+      "learning_rate": 7.563774453684102e-06,
+      "loss": 0.3843,
+      "step": 1526
+    },
+    {
+      "epoch": 1.1880186721991701,
+      "grad_norm": 0.579823184423382,
+      "learning_rate": 7.55988587421629e-06,
+      "loss": 0.3747,
+      "step": 1527
+    },
+    {
+      "epoch": 1.1887966804979253,
+      "grad_norm": 0.6585063861771742,
+      "learning_rate": 7.555995195268586e-06,
+      "loss": 0.438,
+      "step": 1528
+    },
+    {
+      "epoch": 1.1895746887966805,
+      "grad_norm": 0.6079330179018797,
+      "learning_rate": 7.552102420031913e-06,
+      "loss": 0.4045,
+      "step": 1529
+    },
+    {
+      "epoch": 1.1903526970954357,
+      "grad_norm": 0.5576360407504127,
+      "learning_rate": 7.548207551698911e-06,
+      "loss": 0.3899,
+      "step": 1530
+    },
+    {
+      "epoch": 1.1911307053941909,
+      "grad_norm": 0.6006625185622774,
+      "learning_rate": 7.544310593463946e-06,
+      "loss": 0.4023,
+      "step": 1531
+    },
+    {
+      "epoch": 1.191908713692946,
+      "grad_norm": 0.6101113801558775,
+      "learning_rate": 7.540411548523089e-06,
+      "loss": 0.4099,
+      "step": 1532
+    },
+    {
+      "epoch": 1.1926867219917012,
+      "grad_norm": 0.6432772709978408,
+      "learning_rate": 7.536510420074125e-06,
+      "loss": 0.3686,
+      "step": 1533
+    },
+    {
+      "epoch": 1.1934647302904564,
+      "grad_norm": 0.5926651477009937,
+      "learning_rate": 7.532607211316552e-06,
+      "loss": 0.4118,
+      "step": 1534
+    },
+    {
+      "epoch": 1.1942427385892116,
+      "grad_norm": 0.5908162889246386,
+      "learning_rate": 7.528701925451567e-06,
+      "loss": 0.3645,
+      "step": 1535
+    },
+    {
+      "epoch": 1.1950207468879668,
+      "grad_norm": 0.5617610820730953,
+      "learning_rate": 7.524794565682077e-06,
+      "loss": 0.4043,
+      "step": 1536
+    },
+    {
+      "epoch": 1.195798755186722,
+      "grad_norm": 0.544071312407292,
+      "learning_rate": 7.520885135212684e-06,
+      "loss": 0.3665,
+      "step": 1537
+    },
+    {
+      "epoch": 1.1965767634854771,
+      "grad_norm": 0.6689236462393179,
+      "learning_rate": 7.516973637249696e-06,
+      "loss": 0.4229,
+      "step": 1538
+    },
+    {
+      "epoch": 1.1973547717842323,
+      "grad_norm": 0.6249261899669233,
+      "learning_rate": 7.5130600750011065e-06,
+      "loss": 0.4734,
+      "step": 1539
+    },
+    {
+      "epoch": 1.1981327800829875,
+      "grad_norm": 0.621978683727849,
+      "learning_rate": 7.50914445167661e-06,
+      "loss": 0.3953,
+      "step": 1540
+    },
+    {
+      "epoch": 1.1989107883817427,
+      "grad_norm": 0.5887987888684486,
+      "learning_rate": 7.50522677048759e-06,
+      "loss": 0.4458,
+      "step": 1541
+    },
+    {
+      "epoch": 1.1996887966804979,
+      "grad_norm": 0.63882630052998,
+      "learning_rate": 7.501307034647116e-06,
+      "loss": 0.3977,
+      "step": 1542
+    },
+    {
+      "epoch": 1.200466804979253,
+      "grad_norm": 0.6171795204612863,
+      "learning_rate": 7.4973852473699424e-06,
+      "loss": 0.4214,
+      "step": 1543
+    },
+    {
+      "epoch": 1.2012448132780082,
+      "grad_norm": 0.4929733475114353,
+      "learning_rate": 7.493461411872508e-06,
+      "loss": 0.3691,
+      "step": 1544
+    },
+    {
+      "epoch": 1.2020228215767634,
+      "grad_norm": 0.5513406736020697,
+      "learning_rate": 7.489535531372929e-06,
+      "loss": 0.4211,
+      "step": 1545
+    },
+    {
+      "epoch": 1.2028008298755186,
+      "grad_norm": 0.528650305984375,
+      "learning_rate": 7.485607609091001e-06,
+      "loss": 0.368,
+      "step": 1546
+    },
+    {
+      "epoch": 1.2035788381742738,
+      "grad_norm": 0.5480543524775933,
+      "learning_rate": 7.481677648248194e-06,
+      "loss": 0.429,
+      "step": 1547
+    },
+    {
+      "epoch": 1.204356846473029,
+      "grad_norm": 0.6575566473158704,
+      "learning_rate": 7.4777456520676496e-06,
+      "loss": 0.45,
+      "step": 1548
+    },
+    {
+      "epoch": 1.2051348547717842,
+      "grad_norm": 0.620300110757438,
+      "learning_rate": 7.473811623774176e-06,
+      "loss": 0.3683,
+      "step": 1549
+    },
+    {
+      "epoch": 1.2059128630705394,
+      "grad_norm": 0.5967536007559099,
+      "learning_rate": 7.469875566594251e-06,
+      "loss": 0.4072,
+      "step": 1550
+    },
+    {
+      "epoch": 1.2066908713692945,
+      "grad_norm": 0.5540643899021762,
+      "learning_rate": 7.46593748375602e-06,
+      "loss": 0.4024,
+      "step": 1551
+    },
+    {
+      "epoch": 1.2074688796680497,
+      "grad_norm": 0.6112346675650694,
+      "learning_rate": 7.461997378489279e-06,
+      "loss": 0.3658,
+      "step": 1552
+    },
+    {
+      "epoch": 1.208246887966805,
+      "grad_norm": 0.5720908714188554,
+      "learning_rate": 7.458055254025491e-06,
+      "loss": 0.4341,
+      "step": 1553
+    },
+    {
+      "epoch": 1.20902489626556,
+      "grad_norm": 0.5479099144476386,
+      "learning_rate": 7.454111113597776e-06,
+      "loss": 0.3614,
+      "step": 1554
+    },
+    {
+      "epoch": 1.2098029045643153,
+      "grad_norm": 0.6052539549856062,
+      "learning_rate": 7.4501649604409046e-06,
+      "loss": 0.4181,
+      "step": 1555
+    },
+    {
+      "epoch": 1.2105809128630705,
+      "grad_norm": 0.5322256946274699,
+      "learning_rate": 7.446216797791294e-06,
+      "loss": 0.3996,
+      "step": 1556
+    },
+    {
+      "epoch": 1.2113589211618256,
+      "grad_norm": 0.6088595874036016,
+      "learning_rate": 7.442266628887016e-06,
+      "loss": 0.4232,
+      "step": 1557
+    },
+    {
+      "epoch": 1.2121369294605808,
+      "grad_norm": 0.5376272400706118,
+      "learning_rate": 7.438314456967787e-06,
+      "loss": 0.3877,
+      "step": 1558
+    },
+    {
+      "epoch": 1.212914937759336,
+      "grad_norm": 0.6651793770058789,
+      "learning_rate": 7.434360285274962e-06,
+      "loss": 0.4365,
+      "step": 1559
+    },
+    {
+      "epoch": 1.2136929460580912,
+      "grad_norm": 0.5621826012152716,
+      "learning_rate": 7.43040411705154e-06,
+      "loss": 0.4616,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2144709543568464,
+      "grad_norm": 0.5560248155796111,
+      "learning_rate": 7.426445955542154e-06,
+      "loss": 0.4011,
+      "step": 1561
+    },
+    {
+      "epoch": 1.2152489626556018,
+      "grad_norm": 0.709002658771397,
+      "learning_rate": 7.422485803993078e-06,
+      "loss": 0.4424,
+      "step": 1562
+    },
+    {
+      "epoch": 1.2160269709543567,
+      "grad_norm": 0.6211031082155304,
+      "learning_rate": 7.418523665652211e-06,
+      "loss": 0.4132,
+      "step": 1563
+    },
+    {
+      "epoch": 1.2168049792531122,
+      "grad_norm": 0.5598581506674439,
+      "learning_rate": 7.414559543769087e-06,
+      "loss": 0.4014,
+      "step": 1564
+    },
+    {
+      "epoch": 1.2175829875518671,
+      "grad_norm": 0.5685835395099509,
+      "learning_rate": 7.41059344159486e-06,
+      "loss": 0.438,
+      "step": 1565
+    },
+    {
+      "epoch": 1.2183609958506225,
+      "grad_norm": 0.5427837685235937,
+      "learning_rate": 7.406625362382316e-06,
+      "loss": 0.4035,
+      "step": 1566
+    },
+    {
+      "epoch": 1.2191390041493775,
+      "grad_norm": 0.6220554860321991,
+      "learning_rate": 7.402655309385856e-06,
+      "loss": 0.422,
+      "step": 1567
+    },
+    {
+      "epoch": 1.2199170124481329,
+      "grad_norm": 0.5761318017873159,
+      "learning_rate": 7.398683285861507e-06,
+      "loss": 0.4189,
+      "step": 1568
+    },
+    {
+      "epoch": 1.2206950207468878,
+      "grad_norm": 0.6131353820439592,
+      "learning_rate": 7.394709295066906e-06,
+      "loss": 0.3951,
+      "step": 1569
+    },
+    {
+      "epoch": 1.2214730290456433,
+      "grad_norm": 0.6025467488252516,
+      "learning_rate": 7.390733340261303e-06,
+      "loss": 0.3714,
+      "step": 1570
+    },
+    {
+      "epoch": 1.2222510373443982,
+      "grad_norm": 0.5221190296023915,
+      "learning_rate": 7.386755424705563e-06,
+      "loss": 0.4285,
+      "step": 1571
+    },
+    {
+      "epoch": 1.2230290456431536,
+      "grad_norm": 0.5318752430849215,
+      "learning_rate": 7.382775551662157e-06,
+      "loss": 0.4053,
+      "step": 1572
+    },
+    {
+      "epoch": 1.2238070539419088,
+      "grad_norm": 0.5308246489367512,
+      "learning_rate": 7.378793724395161e-06,
+      "loss": 0.374,
+      "step": 1573
+    },
+    {
+      "epoch": 1.224585062240664,
+      "grad_norm": 0.4981623799679136,
+      "learning_rate": 7.374809946170254e-06,
+      "loss": 0.3459,
+      "step": 1574
+    },
+    {
+      "epoch": 1.2253630705394192,
+      "grad_norm": 0.5481679222117711,
+      "learning_rate": 7.370824220254716e-06,
+      "loss": 0.3969,
+      "step": 1575
+    },
+    {
+      "epoch": 1.2261410788381744,
+      "grad_norm": 0.5999917637249105,
+      "learning_rate": 7.366836549917425e-06,
+      "loss": 0.442,
+      "step": 1576
+    },
+    {
+      "epoch": 1.2269190871369295,
+      "grad_norm": 0.5188747537680172,
+      "learning_rate": 7.36284693842885e-06,
+      "loss": 0.3529,
+      "step": 1577
+    },
+    {
+      "epoch": 1.2276970954356847,
+      "grad_norm": 0.5826208955285568,
+      "learning_rate": 7.358855389061056e-06,
+      "loss": 0.4119,
+      "step": 1578
+    },
+    {
+      "epoch": 1.22847510373444,
+      "grad_norm": 0.5199911785328973,
+      "learning_rate": 7.354861905087695e-06,
+      "loss": 0.4118,
+      "step": 1579
+    },
+    {
+      "epoch": 1.229253112033195,
+      "grad_norm": 0.5775136614073505,
+      "learning_rate": 7.350866489784006e-06,
+      "loss": 0.413,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2300311203319503,
+      "grad_norm": 0.5694865481207685,
+      "learning_rate": 7.346869146426816e-06,
+      "loss": 0.3715,
+      "step": 1581
+    },
+    {
+      "epoch": 1.2308091286307055,
+      "grad_norm": 0.5218175489880589,
+      "learning_rate": 7.342869878294525e-06,
+      "loss": 0.3923,
+      "step": 1582
+    },
+    {
+      "epoch": 1.2315871369294606,
+      "grad_norm": 0.558567823698048,
+      "learning_rate": 7.33886868866712e-06,
+      "loss": 0.4081,
+      "step": 1583
+    },
+    {
+      "epoch": 1.2323651452282158,
+      "grad_norm": 0.5815994491042404,
+      "learning_rate": 7.334865580826157e-06,
+      "loss": 0.3851,
+      "step": 1584
+    },
+    {
+      "epoch": 1.233143153526971,
+      "grad_norm": 0.591731452981316,
+      "learning_rate": 7.330860558054771e-06,
+      "loss": 0.4274,
+      "step": 1585
+    },
+    {
+      "epoch": 1.2339211618257262,
+      "grad_norm": 0.477888722062954,
+      "learning_rate": 7.326853623637666e-06,
+      "loss": 0.3662,
+      "step": 1586
+    },
+    {
+      "epoch": 1.2346991701244814,
+      "grad_norm": 0.4942110888448413,
+      "learning_rate": 7.322844780861108e-06,
+      "loss": 0.3886,
+      "step": 1587
+    },
+    {
+      "epoch": 1.2354771784232366,
+      "grad_norm": 0.516610757940013,
+      "learning_rate": 7.318834033012938e-06,
+      "loss": 0.4367,
+      "step": 1588
+    },
+    {
+      "epoch": 1.2362551867219918,
+      "grad_norm": 0.5049624137263926,
+      "learning_rate": 7.314821383382552e-06,
+      "loss": 0.3661,
+      "step": 1589
+    },
+    {
+      "epoch": 1.237033195020747,
+      "grad_norm": 0.5078831493955072,
+      "learning_rate": 7.310806835260909e-06,
+      "loss": 0.3873,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2378112033195021,
+      "grad_norm": 0.48002588091590864,
+      "learning_rate": 7.306790391940522e-06,
+      "loss": 0.3715,
+      "step": 1591
+    },
+    {
+      "epoch": 1.2385892116182573,
+      "grad_norm": 0.5949370699311587,
+      "learning_rate": 7.302772056715463e-06,
+      "loss": 0.3969,
+      "step": 1592
+    },
+    {
+      "epoch": 1.2393672199170125,
+      "grad_norm": 0.499214236072551,
+      "learning_rate": 7.298751832881352e-06,
+      "loss": 0.3476,
+      "step": 1593
+    },
+    {
+      "epoch": 1.2401452282157677,
+      "grad_norm": 0.5564798851848611,
+      "learning_rate": 7.2947297237353594e-06,
+      "loss": 0.4093,
+      "step": 1594
+    },
+    {
+      "epoch": 1.2409232365145229,
+      "grad_norm": 0.6065951070929204,
+      "learning_rate": 7.290705732576202e-06,
+      "loss": 0.43,
+      "step": 1595
+    },
+    {
+      "epoch": 1.241701244813278,
+      "grad_norm": 0.6061151925176615,
+      "learning_rate": 7.286679862704139e-06,
+      "loss": 0.4035,
+      "step": 1596
+    },
+    {
+      "epoch": 1.2424792531120332,
+      "grad_norm": 0.5851113679590698,
+      "learning_rate": 7.2826521174209716e-06,
+      "loss": 0.3746,
+      "step": 1597
+    },
+    {
+      "epoch": 1.2432572614107884,
+      "grad_norm": 0.6139572588121467,
+      "learning_rate": 7.278622500030039e-06,
+      "loss": 0.4023,
+      "step": 1598
+    },
+    {
+      "epoch": 1.2440352697095436,
+      "grad_norm": 0.5829405679643073,
+      "learning_rate": 7.274591013836213e-06,
+      "loss": 0.412,
+      "step": 1599
+    },
+    {
+      "epoch": 1.2448132780082988,
+      "grad_norm": 0.5807880160521984,
+      "learning_rate": 7.270557662145902e-06,
+      "loss": 0.4149,
+      "step": 1600
+    },
+    {
+      "epoch": 1.245591286307054,
+      "grad_norm": 0.6448668251403272,
+      "learning_rate": 7.2665224482670415e-06,
+      "loss": 0.4178,
+      "step": 1601
+    },
+    {
+      "epoch": 1.2463692946058091,
+      "grad_norm": 0.540732924907333,
+      "learning_rate": 7.262485375509099e-06,
+      "loss": 0.4008,
+      "step": 1602
+    },
+    {
+      "epoch": 1.2471473029045643,
+      "grad_norm": 0.5916127011155556,
+      "learning_rate": 7.258446447183061e-06,
+      "loss": 0.3956,
+      "step": 1603
+    },
+    {
+      "epoch": 1.2479253112033195,
+      "grad_norm": 0.5960060490897193,
+      "learning_rate": 7.254405666601437e-06,
+      "loss": 0.3891,
+      "step": 1604
+    },
+    {
+      "epoch": 1.2487033195020747,
+      "grad_norm": 0.5425777498450528,
+      "learning_rate": 7.250363037078259e-06,
+      "loss": 0.361,
+      "step": 1605
+    },
+    {
+      "epoch": 1.2494813278008299,
+      "grad_norm": 0.6358166724341331,
+      "learning_rate": 7.24631856192907e-06,
+      "loss": 0.4104,
+      "step": 1606
+    },
+    {
+      "epoch": 1.250259336099585,
+      "grad_norm": 0.5912865710323821,
+      "learning_rate": 7.242272244470934e-06,
+      "loss": 0.4581,
+      "step": 1607
+    },
+    {
+      "epoch": 1.2510373443983402,
+      "grad_norm": 0.5411427861414384,
+      "learning_rate": 7.238224088022416e-06,
+      "loss": 0.3843,
+      "step": 1608
+    },
+    {
+      "epoch": 1.2518153526970954,
+      "grad_norm": 0.6144908309800248,
+      "learning_rate": 7.2341740959035985e-06,
+      "loss": 0.4399,
+      "step": 1609
+    },
+    {
+      "epoch": 1.2525933609958506,
+      "grad_norm": 0.6157951234868635,
+      "learning_rate": 7.2301222714360666e-06,
+      "loss": 0.452,
+      "step": 1610
+    },
+    {
+      "epoch": 1.2533713692946058,
+      "grad_norm": 0.5315761918396622,
+      "learning_rate": 7.226068617942906e-06,
+      "loss": 0.3854,
+      "step": 1611
+    },
+    {
+      "epoch": 1.254149377593361,
+      "grad_norm": 0.6000012075863647,
+      "learning_rate": 7.222013138748704e-06,
+      "loss": 0.4118,
+      "step": 1612
+    },
+    {
+      "epoch": 1.2549273858921162,
+      "grad_norm": 0.6244862972319819,
+      "learning_rate": 7.217955837179545e-06,
+      "loss": 0.4152,
+      "step": 1613
+    },
+    {
+      "epoch": 1.2557053941908713,
+      "grad_norm": 0.5265810663786469,
+      "learning_rate": 7.213896716563009e-06,
+      "loss": 0.3892,
+      "step": 1614
+    },
+    {
+      "epoch": 1.2564834024896265,
+      "grad_norm": 0.5303020797211274,
+      "learning_rate": 7.209835780228167e-06,
+      "loss": 0.3745,
+      "step": 1615
+    },
+    {
+      "epoch": 1.2572614107883817,
+      "grad_norm": 0.5372941968349159,
+      "learning_rate": 7.2057730315055785e-06,
+      "loss": 0.4113,
+      "step": 1616
+    },
+    {
+      "epoch": 1.258039419087137,
+      "grad_norm": 0.6020474325737551,
+      "learning_rate": 7.201708473727291e-06,
+      "loss": 0.3968,
+      "step": 1617
+    },
+    {
+      "epoch": 1.258817427385892,
+      "grad_norm": 0.5975086192148445,
+      "learning_rate": 7.1976421102268346e-06,
+      "loss": 0.3483,
+      "step": 1618
+    },
+    {
+      "epoch": 1.2595954356846473,
+      "grad_norm": 0.5517477205809665,
+      "learning_rate": 7.19357394433922e-06,
+      "loss": 0.3693,
+      "step": 1619
+    },
+    {
+      "epoch": 1.2603734439834025,
+      "grad_norm": 0.6514310170472642,
+      "learning_rate": 7.189503979400939e-06,
+      "loss": 0.4849,
+      "step": 1620
+    },
+    {
+      "epoch": 1.2611514522821576,
+      "grad_norm": 0.5498535088469549,
+      "learning_rate": 7.185432218749955e-06,
+      "loss": 0.3551,
+      "step": 1621
+    },
+    {
+      "epoch": 1.2619294605809128,
+      "grad_norm": 0.6491569716217163,
+      "learning_rate": 7.181358665725704e-06,
+      "loss": 0.4683,
+      "step": 1622
+    },
+    {
+      "epoch": 1.262707468879668,
+      "grad_norm": 0.5449794775608832,
+      "learning_rate": 7.177283323669096e-06,
+      "loss": 0.3901,
+      "step": 1623
+    },
+    {
+      "epoch": 1.2634854771784232,
+      "grad_norm": 0.5395530842305475,
+      "learning_rate": 7.173206195922508e-06,
+      "loss": 0.4029,
+      "step": 1624
+    },
+    {
+      "epoch": 1.2642634854771784,
+      "grad_norm": 0.5628087287160611,
+      "learning_rate": 7.169127285829776e-06,
+      "loss": 0.4311,
+      "step": 1625
+    },
+    {
+      "epoch": 1.2650414937759336,
+      "grad_norm": 0.5858105808893589,
+      "learning_rate": 7.165046596736201e-06,
+      "loss": 0.382,
+      "step": 1626
+    },
+    {
+      "epoch": 1.2658195020746887,
+      "grad_norm": 0.649723213589266,
+      "learning_rate": 7.160964131988547e-06,
+      "loss": 0.4409,
+      "step": 1627
+    },
+    {
+      "epoch": 1.266597510373444,
+      "grad_norm": 0.5234528947090321,
+      "learning_rate": 7.156879894935028e-06,
+      "loss": 0.3435,
+      "step": 1628
+    },
+    {
+      "epoch": 1.267375518672199,
+      "grad_norm": 0.597861512173525,
+      "learning_rate": 7.152793888925316e-06,
+      "loss": 0.4676,
+      "step": 1629
+    },
+    {
+      "epoch": 1.2681535269709543,
+      "grad_norm": 0.5389246546241906,
+      "learning_rate": 7.14870611731053e-06,
+      "loss": 0.3708,
+      "step": 1630
+    },
+    {
+      "epoch": 1.2689315352697095,
+      "grad_norm": 0.5815350763226333,
+      "learning_rate": 7.14461658344324e-06,
+      "loss": 0.401,
+      "step": 1631
+    },
+    {
+      "epoch": 1.2697095435684647,
+      "grad_norm": 0.5708866067002049,
+      "learning_rate": 7.140525290677462e-06,
+      "loss": 0.3841,
+      "step": 1632
+    },
+    {
+      "epoch": 1.2704875518672198,
+      "grad_norm": 0.4994154906061019,
+      "learning_rate": 7.1364322423686514e-06,
+      "loss": 0.3827,
+      "step": 1633
+    },
+    {
+      "epoch": 1.271265560165975,
+      "grad_norm": 0.7405658990396081,
+      "learning_rate": 7.132337441873704e-06,
+      "loss": 0.4556,
+      "step": 1634
+    },
+    {
+      "epoch": 1.2720435684647302,
+      "grad_norm": 0.6300891255695585,
+      "learning_rate": 7.128240892550955e-06,
+      "loss": 0.4112,
+      "step": 1635
+    },
+    {
+      "epoch": 1.2728215767634854,
+      "grad_norm": 0.5659053735808333,
+      "learning_rate": 7.124142597760172e-06,
+      "loss": 0.45,
+      "step": 1636
+    },
+    {
+      "epoch": 1.2735995850622408,
+      "grad_norm": 0.601347728187396,
+      "learning_rate": 7.120042560862556e-06,
+      "loss": 0.4196,
+      "step": 1637
+    },
+    {
+      "epoch": 1.2743775933609958,
+      "grad_norm": 0.5575781367544485,
+      "learning_rate": 7.115940785220735e-06,
+      "loss": 0.4017,
+      "step": 1638
+    },
+    {
+      "epoch": 1.2751556016597512,
+      "grad_norm": 0.4941688534362159,
+      "learning_rate": 7.111837274198762e-06,
+      "loss": 0.3812,
+      "step": 1639
+    },
+    {
+      "epoch": 1.2759336099585061,
+      "grad_norm": 0.5007748554033356,
+      "learning_rate": 7.107732031162116e-06,
+      "loss": 0.3419,
+      "step": 1640
+    },
+    {
+      "epoch": 1.2767116182572615,
+      "grad_norm": 0.6227132268978763,
+      "learning_rate": 7.103625059477693e-06,
+      "loss": 0.4544,
+      "step": 1641
+    },
+    {
+      "epoch": 1.2774896265560165,
+      "grad_norm": 0.5014943428033439,
+      "learning_rate": 7.099516362513812e-06,
+      "loss": 0.3619,
+      "step": 1642
+    },
+    {
+      "epoch": 1.278267634854772,
+      "grad_norm": 0.5554494176248954,
+      "learning_rate": 7.0954059436402035e-06,
+      "loss": 0.3955,
+      "step": 1643
+    },
+    {
+      "epoch": 1.2790456431535269,
+      "grad_norm": 0.5730787410234404,
+      "learning_rate": 7.091293806228009e-06,
+      "loss": 0.4275,
+      "step": 1644
+    },
+    {
+      "epoch": 1.2798236514522823,
+      "grad_norm": 0.5403056136279832,
+      "learning_rate": 7.087179953649782e-06,
+      "loss": 0.3961,
+      "step": 1645
+    },
+    {
+      "epoch": 1.2806016597510372,
+      "grad_norm": 0.5632097092465174,
+      "learning_rate": 7.083064389279484e-06,
+      "loss": 0.4147,
+      "step": 1646
+    },
+    {
+      "epoch": 1.2813796680497926,
+      "grad_norm": 0.6241596949529887,
+      "learning_rate": 7.078947116492474e-06,
+      "loss": 0.3973,
+      "step": 1647
+    },
+    {
+      "epoch": 1.2821576763485476,
+      "grad_norm": 0.6786075442812364,
+      "learning_rate": 7.074828138665518e-06,
+      "loss": 0.4091,
+      "step": 1648
+    },
+    {
+      "epoch": 1.282935684647303,
+      "grad_norm": 0.6003955613452574,
+      "learning_rate": 7.070707459176779e-06,
+      "loss": 0.4017,
+      "step": 1649
+    },
+    {
+      "epoch": 1.283713692946058,
+      "grad_norm": 0.5618788969675224,
+      "learning_rate": 7.0665850814058175e-06,
+      "loss": 0.396,
+      "step": 1650
+    },
+    {
+      "epoch": 1.2844917012448134,
+      "grad_norm": 0.6140919461242613,
+      "learning_rate": 7.062461008733582e-06,
+      "loss": 0.4625,
+      "step": 1651
+    },
+    {
+      "epoch": 1.2852697095435683,
+      "grad_norm": 0.5280490372978015,
+      "learning_rate": 7.058335244542414e-06,
+      "loss": 0.3784,
+      "step": 1652
+    },
+    {
+      "epoch": 1.2860477178423237,
+      "grad_norm": 0.5306166145577029,
+      "learning_rate": 7.054207792216044e-06,
+      "loss": 0.3589,
+      "step": 1653
+    },
+    {
+      "epoch": 1.2868257261410787,
+      "grad_norm": 0.5287219552096547,
+      "learning_rate": 7.0500786551395815e-06,
+      "loss": 0.3824,
+      "step": 1654
+    },
+    {
+      "epoch": 1.2876037344398341,
+      "grad_norm": 0.5123317832132583,
+      "learning_rate": 7.045947836699526e-06,
+      "loss": 0.3954,
+      "step": 1655
+    },
+    {
+      "epoch": 1.288381742738589,
+      "grad_norm": 0.5099284764112317,
+      "learning_rate": 7.041815340283746e-06,
+      "loss": 0.3616,
+      "step": 1656
+    },
+    {
+      "epoch": 1.2891597510373445,
+      "grad_norm": 0.6282920978710018,
+      "learning_rate": 7.037681169281496e-06,
+      "loss": 0.4465,
+      "step": 1657
+    },
+    {
+      "epoch": 1.2899377593360997,
+      "grad_norm": 0.5984466575265501,
+      "learning_rate": 7.0335453270833975e-06,
+      "loss": 0.4299,
+      "step": 1658
+    },
+    {
+      "epoch": 1.2907157676348548,
+      "grad_norm": 0.587822758396702,
+      "learning_rate": 7.029407817081446e-06,
+      "loss": 0.3641,
+      "step": 1659
+    },
+    {
+      "epoch": 1.29149377593361,
+      "grad_norm": 0.6137577423848142,
+      "learning_rate": 7.025268642669001e-06,
+      "loss": 0.4434,
+      "step": 1660
+    },
+    {
+      "epoch": 1.2922717842323652,
+      "grad_norm": 0.5047273193375017,
+      "learning_rate": 7.02112780724079e-06,
+      "loss": 0.3767,
+      "step": 1661
+    },
+    {
+      "epoch": 1.2930497925311204,
+      "grad_norm": 0.514721479117624,
+      "learning_rate": 7.016985314192902e-06,
+      "loss": 0.3652,
+      "step": 1662
+    },
+    {
+      "epoch": 1.2938278008298756,
+      "grad_norm": 0.6132423077514747,
+      "learning_rate": 7.012841166922788e-06,
+      "loss": 0.4149,
+      "step": 1663
+    },
+    {
+      "epoch": 1.2946058091286308,
+      "grad_norm": 0.5195518008522797,
+      "learning_rate": 7.00869536882925e-06,
+      "loss": 0.3746,
+      "step": 1664
+    },
+    {
+      "epoch": 1.295383817427386,
+      "grad_norm": 0.5179506521472323,
+      "learning_rate": 7.00454792331245e-06,
+      "loss": 0.3584,
+      "step": 1665
+    },
+    {
+      "epoch": 1.2961618257261411,
+      "grad_norm": 0.6141244873903942,
+      "learning_rate": 7.000398833773898e-06,
+      "loss": 0.4985,
+      "step": 1666
+    },
+    {
+      "epoch": 1.2969398340248963,
+      "grad_norm": 0.5290408940215875,
+      "learning_rate": 6.996248103616451e-06,
+      "loss": 0.3485,
+      "step": 1667
+    },
+    {
+      "epoch": 1.2977178423236515,
+      "grad_norm": 0.5526450919937296,
+      "learning_rate": 6.992095736244316e-06,
+      "loss": 0.4055,
+      "step": 1668
+    },
+    {
+      "epoch": 1.2984958506224067,
+      "grad_norm": 0.6038775413636117,
+      "learning_rate": 6.987941735063036e-06,
+      "loss": 0.4647,
+      "step": 1669
+    },
+    {
+      "epoch": 1.2992738589211619,
+      "grad_norm": 0.5282123092235617,
+      "learning_rate": 6.983786103479502e-06,
+      "loss": 0.3831,
+      "step": 1670
+    },
+    {
+      "epoch": 1.300051867219917,
+      "grad_norm": 0.5280797909746603,
+      "learning_rate": 6.979628844901937e-06,
+      "loss": 0.4376,
+      "step": 1671
+    },
+    {
+      "epoch": 1.3008298755186722,
+      "grad_norm": 0.5282812069341611,
+      "learning_rate": 6.975469962739901e-06,
+      "loss": 0.4127,
+      "step": 1672
+    },
+    {
+      "epoch": 1.3016078838174274,
+      "grad_norm": 0.5543076074488127,
+      "learning_rate": 6.971309460404278e-06,
+      "loss": 0.3611,
+      "step": 1673
+    },
+    {
+      "epoch": 1.3023858921161826,
+      "grad_norm": 0.5722143604460135,
+      "learning_rate": 6.9671473413072945e-06,
+      "loss": 0.4152,
+      "step": 1674
+    },
+    {
+      "epoch": 1.3031639004149378,
+      "grad_norm": 0.5183991117045863,
+      "learning_rate": 6.962983608862492e-06,
+      "loss": 0.3847,
+      "step": 1675
+    },
+    {
+      "epoch": 1.303941908713693,
+      "grad_norm": 0.5489213152569359,
+      "learning_rate": 6.9588182664847385e-06,
+      "loss": 0.3809,
+      "step": 1676
+    },
+    {
+      "epoch": 1.3047199170124482,
+      "grad_norm": 0.5281228111116878,
+      "learning_rate": 6.954651317590222e-06,
+      "loss": 0.3585,
+      "step": 1677
+    },
+    {
+      "epoch": 1.3054979253112033,
+      "grad_norm": 0.5518097201643423,
+      "learning_rate": 6.95048276559645e-06,
+      "loss": 0.3624,
+      "step": 1678
+    },
+    {
+      "epoch": 1.3062759336099585,
+      "grad_norm": 0.6175895599940456,
+      "learning_rate": 6.9463126139222425e-06,
+      "loss": 0.4336,
+      "step": 1679
+    },
+    {
+      "epoch": 1.3070539419087137,
+      "grad_norm": 0.56043351032771,
+      "learning_rate": 6.942140865987733e-06,
+      "loss": 0.4312,
+      "step": 1680
+    },
+    {
+      "epoch": 1.307831950207469,
+      "grad_norm": 0.5221893745983957,
+      "learning_rate": 6.937967525214362e-06,
+      "loss": 0.3671,
+      "step": 1681
+    },
+    {
+      "epoch": 1.308609958506224,
+      "grad_norm": 0.6153938410337886,
+      "learning_rate": 6.93379259502488e-06,
+      "loss": 0.4186,
+      "step": 1682
+    },
+    {
+      "epoch": 1.3093879668049793,
+      "grad_norm": 0.5773300491139118,
+      "learning_rate": 6.929616078843335e-06,
+      "loss": 0.4058,
+      "step": 1683
+    },
+    {
+      "epoch": 1.3101659751037344,
+      "grad_norm": 0.5296223113557944,
+      "learning_rate": 6.9254379800950845e-06,
+      "loss": 0.4103,
+      "step": 1684
+    },
+    {
+      "epoch": 1.3109439834024896,
+      "grad_norm": 0.5578047416165028,
+      "learning_rate": 6.921258302206775e-06,
+      "loss": 0.3917,
+      "step": 1685
+    },
+    {
+      "epoch": 1.3117219917012448,
+      "grad_norm": 0.5606524455354327,
+      "learning_rate": 6.9170770486063544e-06,
+      "loss": 0.4451,
+      "step": 1686
+    },
+    {
+      "epoch": 1.3125,
+      "grad_norm": 0.5475710305565681,
+      "learning_rate": 6.91289422272306e-06,
+      "loss": 0.3558,
+      "step": 1687
+    },
+    {
+      "epoch": 1.3132780082987552,
+      "grad_norm": 0.5528605456282057,
+      "learning_rate": 6.90870982798742e-06,
+      "loss": 0.4311,
+      "step": 1688
+    },
+    {
+      "epoch": 1.3140560165975104,
+      "grad_norm": 0.4907607747283202,
+      "learning_rate": 6.904523867831248e-06,
+      "loss": 0.3755,
+      "step": 1689
+    },
+    {
+      "epoch": 1.3148340248962656,
+      "grad_norm": 0.5213828213188413,
+      "learning_rate": 6.900336345687641e-06,
+      "loss": 0.3168,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3156120331950207,
+      "grad_norm": 0.6485530936319517,
+      "learning_rate": 6.896147264990979e-06,
+      "loss": 0.4638,
+      "step": 1691
+    },
+    {
+      "epoch": 1.316390041493776,
+      "grad_norm": 0.5456892075717727,
+      "learning_rate": 6.89195662917692e-06,
+      "loss": 0.4025,
+      "step": 1692
+    },
+    {
+      "epoch": 1.317168049792531,
+      "grad_norm": 0.5084710586117943,
+      "learning_rate": 6.887764441682397e-06,
+      "loss": 0.3903,
+      "step": 1693
+    },
+    {
+      "epoch": 1.3179460580912863,
+      "grad_norm": 0.6374501890665484,
+      "learning_rate": 6.8835707059456144e-06,
+      "loss": 0.4214,
+      "step": 1694
+    },
+    {
+      "epoch": 1.3187240663900415,
+      "grad_norm": 0.6179233270237683,
+      "learning_rate": 6.879375425406047e-06,
+      "loss": 0.4128,
+      "step": 1695
+    },
+    {
+      "epoch": 1.3195020746887967,
+      "grad_norm": 0.6233368799560599,
+      "learning_rate": 6.875178603504438e-06,
+      "loss": 0.3901,
+      "step": 1696
+    },
+    {
+      "epoch": 1.3202800829875518,
+      "grad_norm": 0.6131452837742047,
+      "learning_rate": 6.870980243682792e-06,
+      "loss": 0.4439,
+      "step": 1697
+    },
+    {
+      "epoch": 1.321058091286307,
+      "grad_norm": 0.5783244767232769,
+      "learning_rate": 6.8667803493843774e-06,
+      "loss": 0.3734,
+      "step": 1698
+    },
+    {
+      "epoch": 1.3218360995850622,
+      "grad_norm": 0.5416294720472409,
+      "learning_rate": 6.862578924053723e-06,
+      "loss": 0.4008,
+      "step": 1699
+    },
+    {
+      "epoch": 1.3226141078838174,
+      "grad_norm": 0.5513522623278179,
+      "learning_rate": 6.858375971136605e-06,
+      "loss": 0.3867,
+      "step": 1700
+    },
+    {
+      "epoch": 1.3233921161825726,
+      "grad_norm": 0.550562796475914,
+      "learning_rate": 6.854171494080063e-06,
+      "loss": 0.3955,
+      "step": 1701
+    },
+    {
+      "epoch": 1.3241701244813278,
+      "grad_norm": 0.5506597353057833,
+      "learning_rate": 6.849965496332381e-06,
+      "loss": 0.441,
+      "step": 1702
+    },
+    {
+      "epoch": 1.324948132780083,
+      "grad_norm": 0.5788924028964789,
+      "learning_rate": 6.845757981343086e-06,
+      "loss": 0.4355,
+      "step": 1703
+    },
+    {
+      "epoch": 1.3257261410788381,
+      "grad_norm": 0.5629212011556988,
+      "learning_rate": 6.841548952562959e-06,
+      "loss": 0.4256,
+      "step": 1704
+    },
+    {
+      "epoch": 1.3265041493775933,
+      "grad_norm": 0.5357112529848885,
+      "learning_rate": 6.8373384134440164e-06,
+      "loss": 0.3889,
+      "step": 1705
+    },
+    {
+      "epoch": 1.3272821576763485,
+      "grad_norm": 0.6560847219475834,
+      "learning_rate": 6.8331263674395155e-06,
+      "loss": 0.4437,
+      "step": 1706
+    },
+    {
+      "epoch": 1.3280601659751037,
+      "grad_norm": 0.5311958480288121,
+      "learning_rate": 6.828912818003947e-06,
+      "loss": 0.3246,
+      "step": 1707
+    },
+    {
+      "epoch": 1.3288381742738589,
+      "grad_norm": 0.6251990037117102,
+      "learning_rate": 6.824697768593036e-06,
+      "loss": 0.4515,
+      "step": 1708
+    },
+    {
+      "epoch": 1.329616182572614,
+      "grad_norm": 0.5872846132367648,
+      "learning_rate": 6.820481222663739e-06,
+      "loss": 0.4091,
+      "step": 1709
+    },
+    {
+      "epoch": 1.3303941908713692,
+      "grad_norm": 0.5600764348712332,
+      "learning_rate": 6.8162631836742385e-06,
+      "loss": 0.4023,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3311721991701244,
+      "grad_norm": 0.5906304886021444,
+      "learning_rate": 6.812043655083942e-06,
+      "loss": 0.405,
+      "step": 1711
+    },
+    {
+      "epoch": 1.3319502074688796,
+      "grad_norm": 0.7063359668548291,
+      "learning_rate": 6.807822640353479e-06,
+      "loss": 0.4318,
+      "step": 1712
+    },
+    {
+      "epoch": 1.3327282157676348,
+      "grad_norm": 0.5331154249441544,
+      "learning_rate": 6.803600142944696e-06,
+      "loss": 0.3804,
+      "step": 1713
+    },
+    {
+      "epoch": 1.33350622406639,
+      "grad_norm": 0.5764201652162635,
+      "learning_rate": 6.7993761663206585e-06,
+      "loss": 0.4371,
+      "step": 1714
+    },
+    {
+      "epoch": 1.3342842323651452,
+      "grad_norm": 0.6239691665945333,
+      "learning_rate": 6.795150713945644e-06,
+      "loss": 0.3693,
+      "step": 1715
+    },
+    {
+      "epoch": 1.3350622406639003,
+      "grad_norm": 0.594734466145116,
+      "learning_rate": 6.790923789285136e-06,
+      "loss": 0.4038,
+      "step": 1716
+    },
+    {
+      "epoch": 1.3358402489626555,
+      "grad_norm": 0.5697263926998187,
+      "learning_rate": 6.786695395805833e-06,
+      "loss": 0.4269,
+      "step": 1717
+    },
+    {
+      "epoch": 1.336618257261411,
+      "grad_norm": 0.6218981901033386,
+      "learning_rate": 6.782465536975634e-06,
+      "loss": 0.4253,
+      "step": 1718
+    },
+    {
+      "epoch": 1.3373962655601659,
+      "grad_norm": 0.5172525901115875,
+      "learning_rate": 6.778234216263642e-06,
+      "loss": 0.3476,
+      "step": 1719
+    },
+    {
+      "epoch": 1.3381742738589213,
+      "grad_norm": 0.5622644387736258,
+      "learning_rate": 6.774001437140153e-06,
+      "loss": 0.3936,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3389522821576763,
+      "grad_norm": 0.6132526725445708,
+      "learning_rate": 6.769767203076667e-06,
+      "loss": 0.4438,
+      "step": 1721
+    },
+    {
+      "epoch": 1.3397302904564317,
+      "grad_norm": 0.5461791581680095,
+      "learning_rate": 6.765531517545873e-06,
+      "loss": 0.3637,
+      "step": 1722
+    },
+    {
+      "epoch": 1.3405082987551866,
+      "grad_norm": 0.5911748119025947,
+      "learning_rate": 6.761294384021648e-06,
+      "loss": 0.425,
+      "step": 1723
+    },
+    {
+      "epoch": 1.341286307053942,
+      "grad_norm": 0.5179377903968306,
+      "learning_rate": 6.757055805979065e-06,
+      "loss": 0.4175,
+      "step": 1724
+    },
+    {
+      "epoch": 1.342064315352697,
+      "grad_norm": 0.5422473797373768,
+      "learning_rate": 6.752815786894373e-06,
+      "loss": 0.3498,
+      "step": 1725
+    },
+    {
+      "epoch": 1.3428423236514524,
+      "grad_norm": 0.5857542414665832,
+      "learning_rate": 6.748574330245007e-06,
+      "loss": 0.4503,
+      "step": 1726
+    },
+    {
+      "epoch": 1.3436203319502074,
+      "grad_norm": 0.5589937740896187,
+      "learning_rate": 6.7443314395095794e-06,
+      "loss": 0.3738,
+      "step": 1727
+    },
+    {
+      "epoch": 1.3443983402489628,
+      "grad_norm": 0.5251783003317574,
+      "learning_rate": 6.740087118167883e-06,
+      "loss": 0.4106,
+      "step": 1728
+    },
+    {
+      "epoch": 1.3451763485477177,
+      "grad_norm": 0.5479404415550049,
+      "learning_rate": 6.735841369700876e-06,
+      "loss": 0.3779,
+      "step": 1729
+    },
+    {
+      "epoch": 1.3459543568464731,
+      "grad_norm": 0.5576432864594838,
+      "learning_rate": 6.731594197590694e-06,
+      "loss": 0.4174,
+      "step": 1730
+    },
+    {
+      "epoch": 1.346732365145228,
+      "grad_norm": 0.5781265090575366,
+      "learning_rate": 6.727345605320635e-06,
+      "loss": 0.4276,
+      "step": 1731
+    },
+    {
+      "epoch": 1.3475103734439835,
+      "grad_norm": 0.6170486558802624,
+      "learning_rate": 6.723095596375171e-06,
+      "loss": 0.4005,
+      "step": 1732
+    },
+    {
+      "epoch": 1.3482883817427385,
+      "grad_norm": 0.6015823591761444,
+      "learning_rate": 6.718844174239922e-06,
+      "loss": 0.4122,
+      "step": 1733
+    },
+    {
+      "epoch": 1.3490663900414939,
+      "grad_norm": 0.494735248602647,
+      "learning_rate": 6.714591342401679e-06,
+      "loss": 0.4317,
+      "step": 1734
+    },
+    {
+      "epoch": 1.3498443983402488,
+      "grad_norm": 0.6369322498487617,
+      "learning_rate": 6.710337104348384e-06,
+      "loss": 0.411,
+      "step": 1735
+    },
+    {
+      "epoch": 1.3506224066390042,
+      "grad_norm": 0.515715965471999,
+      "learning_rate": 6.706081463569131e-06,
+      "loss": 0.3878,
+      "step": 1736
+    },
+    {
+      "epoch": 1.3514004149377592,
+      "grad_norm": 0.5318943215807422,
+      "learning_rate": 6.701824423554169e-06,
+      "loss": 0.4161,
+      "step": 1737
+    },
+    {
+      "epoch": 1.3521784232365146,
+      "grad_norm": 0.5285019149141635,
+      "learning_rate": 6.6975659877948875e-06,
+      "loss": 0.3977,
+      "step": 1738
+    },
+    {
+      "epoch": 1.3529564315352698,
+      "grad_norm": 0.6043964921715564,
+      "learning_rate": 6.693306159783829e-06,
+      "loss": 0.4016,
+      "step": 1739
+    },
+    {
+      "epoch": 1.353734439834025,
+      "grad_norm": 0.5837204544243133,
+      "learning_rate": 6.689044943014673e-06,
+      "loss": 0.4233,
+      "step": 1740
+    },
+    {
+      "epoch": 1.3545124481327802,
+      "grad_norm": 0.47275334188730284,
+      "learning_rate": 6.684782340982241e-06,
+      "loss": 0.3798,
+      "step": 1741
+    },
+    {
+      "epoch": 1.3552904564315353,
+      "grad_norm": 0.5751845556357185,
+      "learning_rate": 6.680518357182484e-06,
+      "loss": 0.452,
+      "step": 1742
+    },
+    {
+      "epoch": 1.3560684647302905,
+      "grad_norm": 0.5804155878606007,
+      "learning_rate": 6.676252995112494e-06,
+      "loss": 0.3924,
+      "step": 1743
+    },
+    {
+      "epoch": 1.3568464730290457,
+      "grad_norm": 0.6334414911904052,
+      "learning_rate": 6.6719862582704895e-06,
+      "loss": 0.4224,
+      "step": 1744
+    },
+    {
+      "epoch": 1.357624481327801,
+      "grad_norm": 0.5221382833261897,
+      "learning_rate": 6.6677181501558165e-06,
+      "loss": 0.3816,
+      "step": 1745
+    },
+    {
+      "epoch": 1.358402489626556,
+      "grad_norm": 0.5455260207188976,
+      "learning_rate": 6.663448674268947e-06,
+      "loss": 0.4298,
+      "step": 1746
+    },
+    {
+      "epoch": 1.3591804979253113,
+      "grad_norm": 0.5954086724656051,
+      "learning_rate": 6.659177834111476e-06,
+      "loss": 0.4642,
+      "step": 1747
+    },
+    {
+      "epoch": 1.3599585062240664,
+      "grad_norm": 0.49007292361009513,
+      "learning_rate": 6.654905633186112e-06,
+      "loss": 0.3611,
+      "step": 1748
+    },
+    {
+      "epoch": 1.3607365145228216,
+      "grad_norm": 0.4810771991125642,
+      "learning_rate": 6.650632074996685e-06,
+      "loss": 0.325,
+      "step": 1749
+    },
+    {
+      "epoch": 1.3615145228215768,
+      "grad_norm": 0.5961429129693377,
+      "learning_rate": 6.646357163048136e-06,
+      "loss": 0.4087,
+      "step": 1750
+    },
+    {
+      "epoch": 1.362292531120332,
+      "grad_norm": 0.5775982469667138,
+      "learning_rate": 6.642080900846517e-06,
+      "loss": 0.3911,
+      "step": 1751
+    },
+    {
+      "epoch": 1.3630705394190872,
+      "grad_norm": 0.580813940315118,
+      "learning_rate": 6.6378032918989836e-06,
+      "loss": 0.4031,
+      "step": 1752
+    },
+    {
+      "epoch": 1.3638485477178424,
+      "grad_norm": 0.699083866208835,
+      "learning_rate": 6.6335243397138035e-06,
+      "loss": 0.3778,
+      "step": 1753
+    },
+    {
+      "epoch": 1.3646265560165975,
+      "grad_norm": 0.5648914965322726,
+      "learning_rate": 6.629244047800343e-06,
+      "loss": 0.4043,
+      "step": 1754
+    },
+    {
+      "epoch": 1.3654045643153527,
+      "grad_norm": 0.5763177251496391,
+      "learning_rate": 6.62496241966906e-06,
+      "loss": 0.412,
+      "step": 1755
+    },
+    {
+      "epoch": 1.366182572614108,
+      "grad_norm": 0.541987262402631,
+      "learning_rate": 6.6206794588315185e-06,
+      "loss": 0.399,
+      "step": 1756
+    },
+    {
+      "epoch": 1.366960580912863,
+      "grad_norm": 0.623275626272059,
+      "learning_rate": 6.616395168800369e-06,
+      "loss": 0.4118,
+      "step": 1757
+    },
+    {
+      "epoch": 1.3677385892116183,
+      "grad_norm": 0.5943220608862692,
+      "learning_rate": 6.612109553089356e-06,
+      "loss": 0.3532,
+      "step": 1758
+    },
+    {
+      "epoch": 1.3685165975103735,
+      "grad_norm": 0.6101229850286001,
+      "learning_rate": 6.607822615213309e-06,
+      "loss": 0.4613,
+      "step": 1759
+    },
+    {
+      "epoch": 1.3692946058091287,
+      "grad_norm": 0.5119523536009045,
+      "learning_rate": 6.603534358688144e-06,
+      "loss": 0.3889,
+      "step": 1760
+    },
+    {
+      "epoch": 1.3700726141078838,
+      "grad_norm": 0.5780835181717563,
+      "learning_rate": 6.599244787030854e-06,
+      "loss": 0.3657,
+      "step": 1761
+    },
+    {
+      "epoch": 1.370850622406639,
+      "grad_norm": 0.5046277218705573,
+      "learning_rate": 6.594953903759516e-06,
+      "loss": 0.3499,
+      "step": 1762
+    },
+    {
+      "epoch": 1.3716286307053942,
+      "grad_norm": 0.5617743380806962,
+      "learning_rate": 6.590661712393281e-06,
+      "loss": 0.3791,
+      "step": 1763
+    },
+    {
+      "epoch": 1.3724066390041494,
+      "grad_norm": 0.6148249818336358,
+      "learning_rate": 6.586368216452369e-06,
+      "loss": 0.4542,
+      "step": 1764
+    },
+    {
+      "epoch": 1.3731846473029046,
+      "grad_norm": 0.5311303903119958,
+      "learning_rate": 6.582073419458074e-06,
+      "loss": 0.3927,
+      "step": 1765
+    },
+    {
+      "epoch": 1.3739626556016598,
+      "grad_norm": 0.5547447272528735,
+      "learning_rate": 6.577777324932759e-06,
+      "loss": 0.4161,
+      "step": 1766
+    },
+    {
+      "epoch": 1.374740663900415,
+      "grad_norm": 0.5945511000620665,
+      "learning_rate": 6.573479936399844e-06,
+      "loss": 0.3971,
+      "step": 1767
+    },
+    {
+      "epoch": 1.3755186721991701,
+      "grad_norm": 0.5814892824926656,
+      "learning_rate": 6.569181257383818e-06,
+      "loss": 0.3604,
+      "step": 1768
+    },
+    {
+      "epoch": 1.3762966804979253,
+      "grad_norm": 0.5549616638102715,
+      "learning_rate": 6.5648812914102244e-06,
+      "loss": 0.4135,
+      "step": 1769
+    },
+    {
+      "epoch": 1.3770746887966805,
+      "grad_norm": 0.6445528007062602,
+      "learning_rate": 6.560580042005661e-06,
+      "loss": 0.4783,
+      "step": 1770
+    },
+    {
+      "epoch": 1.3778526970954357,
+      "grad_norm": 0.6425021223716602,
+      "learning_rate": 6.5562775126977815e-06,
+      "loss": 0.4345,
+      "step": 1771
+    },
+    {
+      "epoch": 1.3786307053941909,
+      "grad_norm": 0.5331960463818388,
+      "learning_rate": 6.551973707015285e-06,
+      "loss": 0.36,
+      "step": 1772
+    },
+    {
+      "epoch": 1.379408713692946,
+      "grad_norm": 0.6543214194320436,
+      "learning_rate": 6.547668628487924e-06,
+      "loss": 0.4337,
+      "step": 1773
+    },
+    {
+      "epoch": 1.3801867219917012,
+      "grad_norm": 0.6241527318844776,
+      "learning_rate": 6.543362280646488e-06,
+      "loss": 0.4285,
+      "step": 1774
+    },
+    {
+      "epoch": 1.3809647302904564,
+      "grad_norm": 0.6028979719260263,
+      "learning_rate": 6.539054667022813e-06,
+      "loss": 0.4392,
+      "step": 1775
+    },
+    {
+      "epoch": 1.3817427385892116,
+      "grad_norm": 0.5281248486669883,
+      "learning_rate": 6.534745791149768e-06,
+      "loss": 0.399,
+      "step": 1776
+    },
+    {
+      "epoch": 1.3825207468879668,
+      "grad_norm": 0.55997051783221,
+      "learning_rate": 6.530435656561262e-06,
+      "loss": 0.4056,
+      "step": 1777
+    },
+    {
+      "epoch": 1.383298755186722,
+      "grad_norm": 0.5828907184315852,
+      "learning_rate": 6.5261242667922306e-06,
+      "loss": 0.4351,
+      "step": 1778
+    },
+    {
+      "epoch": 1.3840767634854771,
+      "grad_norm": 0.48594708540027676,
+      "learning_rate": 6.5218116253786475e-06,
+      "loss": 0.3682,
+      "step": 1779
+    },
+    {
+      "epoch": 1.3848547717842323,
+      "grad_norm": 0.5026804806371116,
+      "learning_rate": 6.517497735857505e-06,
+      "loss": 0.3562,
+      "step": 1780
+    },
+    {
+      "epoch": 1.3856327800829875,
+      "grad_norm": 0.49756328453809534,
+      "learning_rate": 6.513182601766821e-06,
+      "loss": 0.4168,
+      "step": 1781
+    },
+    {
+      "epoch": 1.3864107883817427,
+      "grad_norm": 0.4941562493681498,
+      "learning_rate": 6.508866226645638e-06,
+      "loss": 0.3951,
+      "step": 1782
+    },
+    {
+      "epoch": 1.3871887966804979,
+      "grad_norm": 0.5620551570745533,
+      "learning_rate": 6.50454861403401e-06,
+      "loss": 0.3849,
+      "step": 1783
+    },
+    {
+      "epoch": 1.387966804979253,
+      "grad_norm": 0.5374618388083449,
+      "learning_rate": 6.500229767473013e-06,
+      "loss": 0.3979,
+      "step": 1784
+    },
+    {
+      "epoch": 1.3887448132780082,
+      "grad_norm": 0.5929453679175324,
+      "learning_rate": 6.4959096905047284e-06,
+      "loss": 0.4291,
+      "step": 1785
+    },
+    {
+      "epoch": 1.3895228215767634,
+      "grad_norm": 0.5462785576777387,
+      "learning_rate": 6.491588386672248e-06,
+      "loss": 0.4169,
+      "step": 1786
+    },
+    {
+      "epoch": 1.3903008298755186,
+      "grad_norm": 0.5804928897830113,
+      "learning_rate": 6.4872658595196755e-06,
+      "loss": 0.4127,
+      "step": 1787
+    },
+    {
+      "epoch": 1.3910788381742738,
+      "grad_norm": 0.6134377099706915,
+      "learning_rate": 6.482942112592113e-06,
+      "loss": 0.4266,
+      "step": 1788
+    },
+    {
+      "epoch": 1.391856846473029,
+      "grad_norm": 0.5460775970890747,
+      "learning_rate": 6.478617149435662e-06,
+      "loss": 0.3623,
+      "step": 1789
+    },
+    {
+      "epoch": 1.3926348547717842,
+      "grad_norm": 0.5817085406128563,
+      "learning_rate": 6.474290973597424e-06,
+      "loss": 0.4498,
+      "step": 1790
+    },
+    {
+      "epoch": 1.3934128630705394,
+      "grad_norm": 0.4732615287627943,
+      "learning_rate": 6.469963588625495e-06,
+      "loss": 0.3577,
+      "step": 1791
+    },
+    {
+      "epoch": 1.3941908713692945,
+      "grad_norm": 0.5042366087327689,
+      "learning_rate": 6.465634998068962e-06,
+      "loss": 0.3937,
+      "step": 1792
+    },
+    {
+      "epoch": 1.3949688796680497,
+      "grad_norm": 0.47440491417372693,
+      "learning_rate": 6.461305205477899e-06,
+      "loss": 0.3529,
+      "step": 1793
+    },
+    {
+      "epoch": 1.395746887966805,
+      "grad_norm": 0.4885184755460271,
+      "learning_rate": 6.45697421440337e-06,
+      "loss": 0.4056,
+      "step": 1794
+    },
+    {
+      "epoch": 1.39652489626556,
+      "grad_norm": 0.5978359067410601,
+      "learning_rate": 6.452642028397418e-06,
+      "loss": 0.413,
+      "step": 1795
+    },
+    {
+      "epoch": 1.3973029045643153,
+      "grad_norm": 0.6224438731967217,
+      "learning_rate": 6.448308651013068e-06,
+      "loss": 0.4408,
+      "step": 1796
+    },
+    {
+      "epoch": 1.3980809128630705,
+      "grad_norm": 0.5216643586839355,
+      "learning_rate": 6.443974085804323e-06,
+      "loss": 0.3622,
+      "step": 1797
+    },
+    {
+      "epoch": 1.3988589211618256,
+      "grad_norm": 0.5344777747439211,
+      "learning_rate": 6.439638336326156e-06,
+      "loss": 0.4185,
+      "step": 1798
+    },
+    {
+      "epoch": 1.399636929460581,
+      "grad_norm": 0.4923750010797027,
+      "learning_rate": 6.435301406134515e-06,
+      "loss": 0.3537,
+      "step": 1799
+    },
+    {
+      "epoch": 1.400414937759336,
+      "grad_norm": 0.5207464500631588,
+      "learning_rate": 6.430963298786318e-06,
+      "loss": 0.3758,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4011929460580914,
+      "grad_norm": 0.5748729531280841,
+      "learning_rate": 6.426624017839444e-06,
+      "loss": 0.3874,
+      "step": 1801
+    },
+    {
+      "epoch": 1.4019709543568464,
+      "grad_norm": 0.5547806650093206,
+      "learning_rate": 6.422283566852737e-06,
+      "loss": 0.4159,
+      "step": 1802
+    },
+    {
+      "epoch": 1.4027489626556018,
+      "grad_norm": 0.5209344548592254,
+      "learning_rate": 6.417941949385999e-06,
+      "loss": 0.3895,
+      "step": 1803
+    },
+    {
+      "epoch": 1.4035269709543567,
+      "grad_norm": 0.5496067634597235,
+      "learning_rate": 6.413599168999991e-06,
+      "loss": 0.4288,
+      "step": 1804
+    },
+    {
+      "epoch": 1.4043049792531122,
+      "grad_norm": 0.614408963439971,
+      "learning_rate": 6.4092552292564256e-06,
+      "loss": 0.4092,
+      "step": 1805
+    },
+    {
+      "epoch": 1.4050829875518671,
+      "grad_norm": 0.673851940120555,
+      "learning_rate": 6.404910133717968e-06,
+      "loss": 0.4196,
+      "step": 1806
+    },
+    {
+      "epoch": 1.4058609958506225,
+      "grad_norm": 0.6189990804693439,
+      "learning_rate": 6.400563885948229e-06,
+      "loss": 0.4208,
+      "step": 1807
+    },
+    {
+      "epoch": 1.4066390041493775,
+      "grad_norm": 0.5464849655887589,
+      "learning_rate": 6.396216489511767e-06,
+      "loss": 0.3425,
+      "step": 1808
+    },
+    {
+      "epoch": 1.4074170124481329,
+      "grad_norm": 0.6321429205427284,
+      "learning_rate": 6.3918679479740816e-06,
+      "loss": 0.4799,
+      "step": 1809
+    },
+    {
+      "epoch": 1.4081950207468878,
+      "grad_norm": 0.5373611207627417,
+      "learning_rate": 6.387518264901613e-06,
+      "loss": 0.4277,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4089730290456433,
+      "grad_norm": 0.5424063422698552,
+      "learning_rate": 6.3831674438617305e-06,
+      "loss": 0.3805,
+      "step": 1811
+    },
+    {
+      "epoch": 1.4097510373443982,
+      "grad_norm": 0.616035266226144,
+      "learning_rate": 6.3788154884227475e-06,
+      "loss": 0.4272,
+      "step": 1812
+    },
+    {
+      "epoch": 1.4105290456431536,
+      "grad_norm": 0.5704555155780068,
+      "learning_rate": 6.374462402153898e-06,
+      "loss": 0.3919,
+      "step": 1813
+    },
+    {
+      "epoch": 1.4113070539419086,
+      "grad_norm": 0.5286436794959385,
+      "learning_rate": 6.370108188625353e-06,
+      "loss": 0.3935,
+      "step": 1814
+    },
+    {
+      "epoch": 1.412085062240664,
+      "grad_norm": 0.5779369210739931,
+      "learning_rate": 6.3657528514082e-06,
+      "loss": 0.4446,
+      "step": 1815
+    },
+    {
+      "epoch": 1.412863070539419,
+      "grad_norm": 0.53617539809806,
+      "learning_rate": 6.361396394074449e-06,
+      "loss": 0.3956,
+      "step": 1816
+    },
+    {
+      "epoch": 1.4136410788381744,
+      "grad_norm": 0.5190862735816862,
+      "learning_rate": 6.3570388201970335e-06,
+      "loss": 0.3714,
+      "step": 1817
+    },
+    {
+      "epoch": 1.4144190871369295,
+      "grad_norm": 0.6551767865066369,
+      "learning_rate": 6.352680133349799e-06,
+      "loss": 0.4355,
+      "step": 1818
+    },
+    {
+      "epoch": 1.4151970954356847,
+      "grad_norm": 0.5493546416390112,
+      "learning_rate": 6.348320337107506e-06,
+      "loss": 0.3579,
+      "step": 1819
+    },
+    {
+      "epoch": 1.41597510373444,
+      "grad_norm": 0.5095709748230268,
+      "learning_rate": 6.343959435045819e-06,
+      "loss": 0.3775,
+      "step": 1820
+    },
+    {
+      "epoch": 1.416753112033195,
+      "grad_norm": 0.5383193671745815,
+      "learning_rate": 6.339597430741319e-06,
+      "loss": 0.3853,
+      "step": 1821
+    },
+    {
+      "epoch": 1.4175311203319503,
+      "grad_norm": 0.5408866643895894,
+      "learning_rate": 6.335234327771482e-06,
+      "loss": 0.3642,
+      "step": 1822
+    },
+    {
+      "epoch": 1.4183091286307055,
+      "grad_norm": 0.5790004823005052,
+      "learning_rate": 6.330870129714694e-06,
+      "loss": 0.4291,
+      "step": 1823
+    },
+    {
+      "epoch": 1.4190871369294606,
+      "grad_norm": 0.5361718947609089,
+      "learning_rate": 6.326504840150229e-06,
+      "loss": 0.4306,
+      "step": 1824
+    },
+    {
+      "epoch": 1.4198651452282158,
+      "grad_norm": 0.6594407571093052,
+      "learning_rate": 6.322138462658264e-06,
+      "loss": 0.4013,
+      "step": 1825
+    },
+    {
+      "epoch": 1.420643153526971,
+      "grad_norm": 0.6050189554727828,
+      "learning_rate": 6.317771000819864e-06,
+      "loss": 0.4086,
+      "step": 1826
+    },
+    {
+      "epoch": 1.4214211618257262,
+      "grad_norm": 0.4959385354164978,
+      "learning_rate": 6.313402458216988e-06,
+      "loss": 0.3463,
+      "step": 1827
+    },
+    {
+      "epoch": 1.4221991701244814,
+      "grad_norm": 0.5703317982372617,
+      "learning_rate": 6.309032838432474e-06,
+      "loss": 0.4198,
+      "step": 1828
+    },
+    {
+      "epoch": 1.4229771784232366,
+      "grad_norm": 0.5645722422018045,
+      "learning_rate": 6.30466214505005e-06,
+      "loss": 0.4237,
+      "step": 1829
+    },
+    {
+      "epoch": 1.4237551867219918,
+      "grad_norm": 0.5478749109218541,
+      "learning_rate": 6.300290381654323e-06,
+      "loss": 0.4193,
+      "step": 1830
+    },
+    {
+      "epoch": 1.424533195020747,
+      "grad_norm": 0.4814912494675584,
+      "learning_rate": 6.295917551830775e-06,
+      "loss": 0.3605,
+      "step": 1831
+    },
+    {
+      "epoch": 1.4253112033195021,
+      "grad_norm": 0.5445529959829698,
+      "learning_rate": 6.291543659165767e-06,
+      "loss": 0.3838,
+      "step": 1832
+    },
+    {
+      "epoch": 1.4260892116182573,
+      "grad_norm": 0.6124478499479935,
+      "learning_rate": 6.287168707246526e-06,
+      "loss": 0.3352,
+      "step": 1833
+    },
+    {
+      "epoch": 1.4268672199170125,
+      "grad_norm": 0.689431091605669,
+      "learning_rate": 6.282792699661151e-06,
+      "loss": 0.4914,
+      "step": 1834
+    },
+    {
+      "epoch": 1.4276452282157677,
+      "grad_norm": 0.5198364733023733,
+      "learning_rate": 6.27841563999861e-06,
+      "loss": 0.3638,
+      "step": 1835
+    },
+    {
+      "epoch": 1.4284232365145229,
+      "grad_norm": 0.4979733892741099,
+      "learning_rate": 6.274037531848729e-06,
+      "loss": 0.3545,
+      "step": 1836
+    },
+    {
+      "epoch": 1.429201244813278,
+      "grad_norm": 0.5854367196134146,
+      "learning_rate": 6.2696583788021934e-06,
+      "loss": 0.391,
+      "step": 1837
+    },
+    {
+      "epoch": 1.4299792531120332,
+      "grad_norm": 0.6419180492957453,
+      "learning_rate": 6.265278184450551e-06,
+      "loss": 0.4536,
+      "step": 1838
+    },
+    {
+      "epoch": 1.4307572614107884,
+      "grad_norm": 0.541438990672762,
+      "learning_rate": 6.260896952386196e-06,
+      "loss": 0.34,
+      "step": 1839
+    },
+    {
+      "epoch": 1.4315352697095436,
+      "grad_norm": 0.6042289891686399,
+      "learning_rate": 6.256514686202382e-06,
+      "loss": 0.4543,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4323132780082988,
+      "grad_norm": 0.5286705824783464,
+      "learning_rate": 6.252131389493204e-06,
+      "loss": 0.3967,
+      "step": 1841
+    },
+    {
+      "epoch": 1.433091286307054,
+      "grad_norm": 0.509105719521019,
+      "learning_rate": 6.247747065853605e-06,
+      "loss": 0.3369,
+      "step": 1842
+    },
+    {
+      "epoch": 1.4338692946058091,
+      "grad_norm": 0.5323942379972877,
+      "learning_rate": 6.2433617188793715e-06,
+      "loss": 0.4104,
+      "step": 1843
+    },
+    {
+      "epoch": 1.4346473029045643,
+      "grad_norm": 0.5393017718525399,
+      "learning_rate": 6.238975352167127e-06,
+      "loss": 0.4298,
+      "step": 1844
+    },
+    {
+      "epoch": 1.4354253112033195,
+      "grad_norm": 0.5327736318982298,
+      "learning_rate": 6.234587969314334e-06,
+      "loss": 0.3787,
+      "step": 1845
+    },
+    {
+      "epoch": 1.4362033195020747,
+      "grad_norm": 0.5408033296477444,
+      "learning_rate": 6.2301995739192824e-06,
+      "loss": 0.3464,
+      "step": 1846
+    },
+    {
+      "epoch": 1.4369813278008299,
+      "grad_norm": 0.5751669197827151,
+      "learning_rate": 6.225810169581099e-06,
+      "loss": 0.4142,
+      "step": 1847
+    },
+    {
+      "epoch": 1.437759336099585,
+      "grad_norm": 0.5952229932149881,
+      "learning_rate": 6.221419759899736e-06,
+      "loss": 0.4276,
+      "step": 1848
+    },
+    {
+      "epoch": 1.4385373443983402,
+      "grad_norm": 0.514374836234326,
+      "learning_rate": 6.2170283484759705e-06,
+      "loss": 0.3678,
+      "step": 1849
+    },
+    {
+      "epoch": 1.4393153526970954,
+      "grad_norm": 0.481273395918956,
+      "learning_rate": 6.2126359389114e-06,
+      "loss": 0.369,
+      "step": 1850
+    },
+    {
+      "epoch": 1.4400933609958506,
+      "grad_norm": 0.5136215406988995,
+      "learning_rate": 6.20824253480844e-06,
+      "loss": 0.4549,
+      "step": 1851
+    },
+    {
+      "epoch": 1.4408713692946058,
+      "grad_norm": 0.5022751161784524,
+      "learning_rate": 6.203848139770324e-06,
+      "loss": 0.3898,
+      "step": 1852
+    },
+    {
+      "epoch": 1.441649377593361,
+      "grad_norm": 0.6068680455774318,
+      "learning_rate": 6.1994527574010965e-06,
+      "loss": 0.4316,
+      "step": 1853
+    },
+    {
+      "epoch": 1.4424273858921162,
+      "grad_norm": 0.5157813261434567,
+      "learning_rate": 6.195056391305613e-06,
+      "loss": 0.3202,
+      "step": 1854
+    },
+    {
+      "epoch": 1.4432053941908713,
+      "grad_norm": 0.7267017826362311,
+      "learning_rate": 6.190659045089534e-06,
+      "loss": 0.4763,
+      "step": 1855
+    },
+    {
+      "epoch": 1.4439834024896265,
+      "grad_norm": 0.6536355684044604,
+      "learning_rate": 6.186260722359327e-06,
+      "loss": 0.4186,
+      "step": 1856
+    },
+    {
+      "epoch": 1.4447614107883817,
+      "grad_norm": 0.6041894744620306,
+      "learning_rate": 6.181861426722256e-06,
+      "loss": 0.4728,
+      "step": 1857
+    },
+    {
+      "epoch": 1.445539419087137,
+      "grad_norm": 0.5322409261440894,
+      "learning_rate": 6.177461161786387e-06,
+      "loss": 0.3767,
+      "step": 1858
+    },
+    {
+      "epoch": 1.446317427385892,
+      "grad_norm": 0.5845043289472616,
+      "learning_rate": 6.173059931160578e-06,
+      "loss": 0.3847,
+      "step": 1859
+    },
+    {
+      "epoch": 1.4470954356846473,
+      "grad_norm": 0.5396741991879973,
+      "learning_rate": 6.16865773845448e-06,
+      "loss": 0.4377,
+      "step": 1860
+    },
+    {
+      "epoch": 1.4478734439834025,
+      "grad_norm": 0.563578462776393,
+      "learning_rate": 6.164254587278534e-06,
+      "loss": 0.3851,
+      "step": 1861
+    },
+    {
+      "epoch": 1.4486514522821576,
+      "grad_norm": 0.5083009654061971,
+      "learning_rate": 6.159850481243968e-06,
+      "loss": 0.3856,
+      "step": 1862
+    },
+    {
+      "epoch": 1.4494294605809128,
+      "grad_norm": 0.5475797889990748,
+      "learning_rate": 6.155445423962788e-06,
+      "loss": 0.3898,
+      "step": 1863
+    },
+    {
+      "epoch": 1.450207468879668,
+      "grad_norm": 0.5194422789664345,
+      "learning_rate": 6.151039419047786e-06,
+      "loss": 0.4212,
+      "step": 1864
+    },
+    {
+      "epoch": 1.4509854771784232,
+      "grad_norm": 0.6162144543729037,
+      "learning_rate": 6.146632470112529e-06,
+      "loss": 0.4396,
+      "step": 1865
+    },
+    {
+      "epoch": 1.4517634854771784,
+      "grad_norm": 0.5460266690280053,
+      "learning_rate": 6.142224580771357e-06,
+      "loss": 0.3663,
+      "step": 1866
+    },
+    {
+      "epoch": 1.4525414937759336,
+      "grad_norm": 0.544666334442976,
+      "learning_rate": 6.137815754639384e-06,
+      "loss": 0.3675,
+      "step": 1867
+    },
+    {
+      "epoch": 1.4533195020746887,
+      "grad_norm": 0.5311770492789724,
+      "learning_rate": 6.133405995332486e-06,
+      "loss": 0.4059,
+      "step": 1868
+    },
+    {
+      "epoch": 1.454097510373444,
+      "grad_norm": 0.5096260462499042,
+      "learning_rate": 6.1289953064673155e-06,
+      "loss": 0.403,
+      "step": 1869
+    },
+    {
+      "epoch": 1.454875518672199,
+      "grad_norm": 0.560081478501465,
+      "learning_rate": 6.124583691661279e-06,
+      "loss": 0.4176,
+      "step": 1870
+    },
+    {
+      "epoch": 1.4556535269709543,
+      "grad_norm": 0.5504193899078467,
+      "learning_rate": 6.12017115453254e-06,
+      "loss": 0.4094,
+      "step": 1871
+    },
+    {
+      "epoch": 1.4564315352697095,
+      "grad_norm": 0.5554094182498059,
+      "learning_rate": 6.1157576987000274e-06,
+      "loss": 0.4075,
+      "step": 1872
+    },
+    {
+      "epoch": 1.4572095435684647,
+      "grad_norm": 0.5384584287474515,
+      "learning_rate": 6.111343327783417e-06,
+      "loss": 0.3949,
+      "step": 1873
+    },
+    {
+      "epoch": 1.4579875518672198,
+      "grad_norm": 0.5320762621327707,
+      "learning_rate": 6.106928045403138e-06,
+      "loss": 0.3768,
+      "step": 1874
+    },
+    {
+      "epoch": 1.458765560165975,
+      "grad_norm": 0.4814529143022474,
+      "learning_rate": 6.102511855180364e-06,
+      "loss": 0.382,
+      "step": 1875
+    },
+    {
+      "epoch": 1.4595435684647302,
+      "grad_norm": 0.5216760729273155,
+      "learning_rate": 6.098094760737017e-06,
+      "loss": 0.4143,
+      "step": 1876
+    },
+    {
+      "epoch": 1.4603215767634854,
+      "grad_norm": 0.46049167747107295,
+      "learning_rate": 6.093676765695759e-06,
+      "loss": 0.3686,
+      "step": 1877
+    },
+    {
+      "epoch": 1.4610995850622408,
+      "grad_norm": 0.585991567899267,
+      "learning_rate": 6.08925787367999e-06,
+      "loss": 0.388,
+      "step": 1878
+    },
+    {
+      "epoch": 1.4618775933609958,
+      "grad_norm": 0.5311248148576917,
+      "learning_rate": 6.084838088313847e-06,
+      "loss": 0.3946,
+      "step": 1879
+    },
+    {
+      "epoch": 1.4626556016597512,
+      "grad_norm": 0.5930242263653599,
+      "learning_rate": 6.080417413222197e-06,
+      "loss": 0.4255,
+      "step": 1880
+    },
+    {
+      "epoch": 1.4634336099585061,
+      "grad_norm": 0.4762842128449552,
+      "learning_rate": 6.07599585203064e-06,
+      "loss": 0.3486,
+      "step": 1881
+    },
+    {
+      "epoch": 1.4642116182572615,
+      "grad_norm": 0.5335029501546144,
+      "learning_rate": 6.0715734083654995e-06,
+      "loss": 0.3978,
+      "step": 1882
+    },
+    {
+      "epoch": 1.4649896265560165,
+      "grad_norm": 0.5768394259165284,
+      "learning_rate": 6.067150085853828e-06,
+      "loss": 0.3972,
+      "step": 1883
+    },
+    {
+      "epoch": 1.465767634854772,
+      "grad_norm": 0.5767831971578576,
+      "learning_rate": 6.062725888123393e-06,
+      "loss": 0.4387,
+      "step": 1884
+    },
+    {
+      "epoch": 1.4665456431535269,
+      "grad_norm": 0.5024508820429034,
+      "learning_rate": 6.058300818802682e-06,
+      "loss": 0.3826,
+      "step": 1885
+    },
+    {
+      "epoch": 1.4673236514522823,
+      "grad_norm": 0.5003198697840447,
+      "learning_rate": 6.053874881520897e-06,
+      "loss": 0.3885,
+      "step": 1886
+    },
+    {
+      "epoch": 1.4681016597510372,
+      "grad_norm": 0.6220977321124136,
+      "learning_rate": 6.049448079907952e-06,
+      "loss": 0.4142,
+      "step": 1887
+    },
+    {
+      "epoch": 1.4688796680497926,
+      "grad_norm": 0.5837231040848196,
+      "learning_rate": 6.0450204175944735e-06,
+      "loss": 0.4528,
+      "step": 1888
+    },
+    {
+      "epoch": 1.4696576763485476,
+      "grad_norm": 0.48260998370114927,
+      "learning_rate": 6.040591898211787e-06,
+      "loss": 0.3624,
+      "step": 1889
+    },
+    {
+      "epoch": 1.470435684647303,
+      "grad_norm": 0.5754787065153876,
+      "learning_rate": 6.036162525391927e-06,
+      "loss": 0.3948,
+      "step": 1890
+    },
+    {
+      "epoch": 1.471213692946058,
+      "grad_norm": 0.5387213653084538,
+      "learning_rate": 6.0317323027676235e-06,
+      "loss": 0.405,
+      "step": 1891
+    },
+    {
+      "epoch": 1.4719917012448134,
+      "grad_norm": 0.5110209450152622,
+      "learning_rate": 6.027301233972307e-06,
+      "loss": 0.3447,
+      "step": 1892
+    },
+    {
+      "epoch": 1.4727697095435683,
+      "grad_norm": 0.5867613793572509,
+      "learning_rate": 6.022869322640098e-06,
+      "loss": 0.4307,
+      "step": 1893
+    },
+    {
+      "epoch": 1.4735477178423237,
+      "grad_norm": 0.5525109085847445,
+      "learning_rate": 6.018436572405813e-06,
+      "loss": 0.325,
+      "step": 1894
+    },
+    {
+      "epoch": 1.4743257261410787,
+      "grad_norm": 0.6343133177012369,
+      "learning_rate": 6.014002986904951e-06,
+      "loss": 0.3789,
+      "step": 1895
+    },
+    {
+      "epoch": 1.4751037344398341,
+      "grad_norm": 0.6705769667782592,
+      "learning_rate": 6.009568569773703e-06,
+      "loss": 0.4892,
+      "step": 1896
+    },
+    {
+      "epoch": 1.475881742738589,
+      "grad_norm": 0.5233978029499535,
+      "learning_rate": 6.005133324648935e-06,
+      "loss": 0.4114,
+      "step": 1897
+    },
+    {
+      "epoch": 1.4766597510373445,
+      "grad_norm": 0.5786321863128031,
+      "learning_rate": 6.000697255168195e-06,
+      "loss": 0.4226,
+      "step": 1898
+    },
+    {
+      "epoch": 1.4774377593360997,
+      "grad_norm": 0.5472847316882402,
+      "learning_rate": 5.996260364969709e-06,
+      "loss": 0.4001,
+      "step": 1899
+    },
+    {
+      "epoch": 1.4782157676348548,
+      "grad_norm": 0.5614212417354801,
+      "learning_rate": 5.9918226576923735e-06,
+      "loss": 0.4231,
+      "step": 1900
+    },
+    {
+      "epoch": 1.47899377593361,
+      "grad_norm": 0.5489100372660577,
+      "learning_rate": 5.987384136975757e-06,
+      "loss": 0.3725,
+      "step": 1901
+    },
+    {
+      "epoch": 1.4797717842323652,
+      "grad_norm": 0.5316483757726137,
+      "learning_rate": 5.982944806460089e-06,
+      "loss": 0.4084,
+      "step": 1902
+    },
+    {
+      "epoch": 1.4805497925311204,
+      "grad_norm": 0.5554385125288969,
+      "learning_rate": 5.978504669786275e-06,
+      "loss": 0.3865,
+      "step": 1903
+    },
+    {
+      "epoch": 1.4813278008298756,
+      "grad_norm": 0.5236410116863706,
+      "learning_rate": 5.97406373059587e-06,
+      "loss": 0.4,
+      "step": 1904
+    },
+    {
+      "epoch": 1.4821058091286308,
+      "grad_norm": 0.5953986831366602,
+      "learning_rate": 5.969621992531097e-06,
+      "loss": 0.4073,
+      "step": 1905
+    },
+    {
+      "epoch": 1.482883817427386,
+      "grad_norm": 0.46922790477957077,
+      "learning_rate": 5.965179459234825e-06,
+      "loss": 0.3147,
+      "step": 1906
+    },
+    {
+      "epoch": 1.4836618257261411,
+      "grad_norm": 0.5490062744913063,
+      "learning_rate": 5.960736134350579e-06,
+      "loss": 0.3865,
+      "step": 1907
+    },
+    {
+      "epoch": 1.4844398340248963,
+      "grad_norm": 0.6219893734838755,
+      "learning_rate": 5.956292021522536e-06,
+      "loss": 0.4524,
+      "step": 1908
+    },
+    {
+      "epoch": 1.4852178423236515,
+      "grad_norm": 0.5482469420225672,
+      "learning_rate": 5.951847124395517e-06,
+      "loss": 0.4076,
+      "step": 1909
+    },
+    {
+      "epoch": 1.4859958506224067,
+      "grad_norm": 0.6867830384158,
+      "learning_rate": 5.947401446614984e-06,
+      "loss": 0.4007,
+      "step": 1910
+    },
+    {
+      "epoch": 1.4867738589211619,
+      "grad_norm": 0.5423480043519175,
+      "learning_rate": 5.942954991827043e-06,
+      "loss": 0.4205,
+      "step": 1911
+    },
+    {
+      "epoch": 1.487551867219917,
+      "grad_norm": 0.5020573921545881,
+      "learning_rate": 5.938507763678436e-06,
+      "loss": 0.3819,
+      "step": 1912
+    },
+    {
+      "epoch": 1.4883298755186722,
+      "grad_norm": 0.6190494004286328,
+      "learning_rate": 5.934059765816537e-06,
+      "loss": 0.4485,
+      "step": 1913
+    },
+    {
+      "epoch": 1.4891078838174274,
+      "grad_norm": 0.5467116273980007,
+      "learning_rate": 5.929611001889357e-06,
+      "loss": 0.4117,
+      "step": 1914
+    },
+    {
+      "epoch": 1.4898858921161826,
+      "grad_norm": 0.5045784871244817,
+      "learning_rate": 5.925161475545528e-06,
+      "loss": 0.3419,
+      "step": 1915
+    },
+    {
+      "epoch": 1.4906639004149378,
+      "grad_norm": 0.5351345059467595,
+      "learning_rate": 5.9207111904343094e-06,
+      "loss": 0.4753,
+      "step": 1916
+    },
+    {
+      "epoch": 1.491441908713693,
+      "grad_norm": 0.5329618729842015,
+      "learning_rate": 5.9162601502055885e-06,
+      "loss": 0.4058,
+      "step": 1917
+    },
+    {
+      "epoch": 1.4922199170124482,
+      "grad_norm": 0.5166349066260123,
+      "learning_rate": 5.911808358509866e-06,
+      "loss": 0.4209,
+      "step": 1918
+    },
+    {
+      "epoch": 1.4929979253112033,
+      "grad_norm": 0.5032761941333065,
+      "learning_rate": 5.9073558189982605e-06,
+      "loss": 0.3941,
+      "step": 1919
+    },
+    {
+      "epoch": 1.4937759336099585,
+      "grad_norm": 0.5392806531390584,
+      "learning_rate": 5.902902535322503e-06,
+      "loss": 0.3456,
+      "step": 1920
+    },
+    {
+      "epoch": 1.4945539419087137,
+      "grad_norm": 0.5626939575430695,
+      "learning_rate": 5.898448511134936e-06,
+      "loss": 0.3602,
+      "step": 1921
+    },
+    {
+      "epoch": 1.495331950207469,
+      "grad_norm": 0.5902801777982899,
+      "learning_rate": 5.893993750088512e-06,
+      "loss": 0.4135,
+      "step": 1922
+    },
+    {
+      "epoch": 1.496109958506224,
+      "grad_norm": 0.6308642222898304,
+      "learning_rate": 5.889538255836781e-06,
+      "loss": 0.4327,
+      "step": 1923
+    },
+    {
+      "epoch": 1.4968879668049793,
+      "grad_norm": 0.6201665413640429,
+      "learning_rate": 5.885082032033901e-06,
+      "loss": 0.3843,
+      "step": 1924
+    },
+    {
+      "epoch": 1.4976659751037344,
+      "grad_norm": 0.639152959219269,
+      "learning_rate": 5.880625082334625e-06,
+      "loss": 0.3779,
+      "step": 1925
+    },
+    {
+      "epoch": 1.4984439834024896,
+      "grad_norm": 0.5915296465952844,
+      "learning_rate": 5.876167410394301e-06,
+      "loss": 0.4499,
+      "step": 1926
+    },
+    {
+      "epoch": 1.4992219917012448,
+      "grad_norm": 0.5012915855925397,
+      "learning_rate": 5.871709019868872e-06,
+      "loss": 0.3484,
+      "step": 1927
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5485830593969527,
+      "learning_rate": 5.867249914414867e-06,
+      "loss": 0.3811,
+      "step": 1928
+    },
+    {
+      "epoch": 1.5007780082987552,
+      "grad_norm": 0.6237315923312742,
+      "learning_rate": 5.8627900976894035e-06,
+      "loss": 0.434,
+      "step": 1929
+    },
+    {
+      "epoch": 1.5015560165975104,
+      "grad_norm": 0.5482413942891893,
+      "learning_rate": 5.858329573350184e-06,
+      "loss": 0.3793,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5023340248962656,
+      "grad_norm": 0.5261765393474149,
+      "learning_rate": 5.853868345055488e-06,
+      "loss": 0.3858,
+      "step": 1931
+    },
+    {
+      "epoch": 1.5031120331950207,
+      "grad_norm": 0.5082895159322758,
+      "learning_rate": 5.849406416464171e-06,
+      "loss": 0.328,
+      "step": 1932
+    },
+    {
+      "epoch": 1.503890041493776,
+      "grad_norm": 0.6289273585541298,
+      "learning_rate": 5.844943791235669e-06,
+      "loss": 0.4908,
+      "step": 1933
+    },
+    {
+      "epoch": 1.504668049792531,
+      "grad_norm": 0.5077294842883142,
+      "learning_rate": 5.8404804730299825e-06,
+      "loss": 0.3686,
+      "step": 1934
+    },
+    {
+      "epoch": 1.5054460580912863,
+      "grad_norm": 0.5521768562314273,
+      "learning_rate": 5.836016465507687e-06,
+      "loss": 0.362,
+      "step": 1935
+    },
+    {
+      "epoch": 1.5062240663900415,
+      "grad_norm": 0.5290324211121863,
+      "learning_rate": 5.831551772329918e-06,
+      "loss": 0.4042,
+      "step": 1936
+    },
+    {
+      "epoch": 1.5070020746887967,
+      "grad_norm": 0.5812645184018235,
+      "learning_rate": 5.827086397158375e-06,
+      "loss": 0.4115,
+      "step": 1937
+    },
+    {
+      "epoch": 1.5077800829875518,
+      "grad_norm": 0.5098984194867257,
+      "learning_rate": 5.822620343655319e-06,
+      "loss": 0.3572,
+      "step": 1938
+    },
+    {
+      "epoch": 1.508558091286307,
+      "grad_norm": 0.5696761305253618,
+      "learning_rate": 5.818153615483563e-06,
+      "loss": 0.4203,
+      "step": 1939
+    },
+    {
+      "epoch": 1.5093360995850622,
+      "grad_norm": 0.5997224977442267,
+      "learning_rate": 5.813686216306478e-06,
+      "loss": 0.4011,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5101141078838174,
+      "grad_norm": 0.5741958049752846,
+      "learning_rate": 5.809218149787981e-06,
+      "loss": 0.4074,
+      "step": 1941
+    },
+    {
+      "epoch": 1.5108921161825726,
+      "grad_norm": 0.5599921304501567,
+      "learning_rate": 5.8047494195925405e-06,
+      "loss": 0.4108,
+      "step": 1942
+    },
+    {
+      "epoch": 1.5116701244813278,
+      "grad_norm": 0.6144443661840264,
+      "learning_rate": 5.800280029385165e-06,
+      "loss": 0.4078,
+      "step": 1943
+    },
+    {
+      "epoch": 1.512448132780083,
+      "grad_norm": 0.6013291488947935,
+      "learning_rate": 5.795809982831408e-06,
+      "loss": 0.3911,
+      "step": 1944
+    },
+    {
+      "epoch": 1.5132261410788381,
+      "grad_norm": 0.46011079213566924,
+      "learning_rate": 5.79133928359736e-06,
+      "loss": 0.3761,
+      "step": 1945
+    },
+    {
+      "epoch": 1.5140041493775933,
+      "grad_norm": 0.5830247744824051,
+      "learning_rate": 5.7868679353496474e-06,
+      "loss": 0.4126,
+      "step": 1946
+    },
+    {
+      "epoch": 1.5147821576763485,
+      "grad_norm": 0.7036562312550645,
+      "learning_rate": 5.782395941755426e-06,
+      "loss": 0.4858,
+      "step": 1947
+    },
+    {
+      "epoch": 1.5155601659751037,
+      "grad_norm": 0.5689709828510874,
+      "learning_rate": 5.777923306482383e-06,
+      "loss": 0.3963,
+      "step": 1948
+    },
+    {
+      "epoch": 1.516338174273859,
+      "grad_norm": 0.6777262728806724,
+      "learning_rate": 5.7734500331987345e-06,
+      "loss": 0.4322,
+      "step": 1949
+    },
+    {
+      "epoch": 1.517116182572614,
+      "grad_norm": 0.5998425991896148,
+      "learning_rate": 5.768976125573213e-06,
+      "loss": 0.4307,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5178941908713695,
+      "grad_norm": 0.57258949627251,
+      "learning_rate": 5.7645015872750794e-06,
+      "loss": 0.3867,
+      "step": 1951
+    },
+    {
+      "epoch": 1.5186721991701244,
+      "grad_norm": 0.6101685819534187,
+      "learning_rate": 5.760026421974107e-06,
+      "loss": 0.413,
+      "step": 1952
+    },
+    {
+      "epoch": 1.5194502074688798,
+      "grad_norm": 0.5137690189791451,
+      "learning_rate": 5.755550633340584e-06,
+      "loss": 0.3876,
+      "step": 1953
+    },
+    {
+      "epoch": 1.5202282157676348,
+      "grad_norm": 0.5512831189975047,
+      "learning_rate": 5.75107422504531e-06,
+      "loss": 0.4076,
+      "step": 1954
+    },
+    {
+      "epoch": 1.5210062240663902,
+      "grad_norm": 0.6056277157540597,
+      "learning_rate": 5.746597200759591e-06,
+      "loss": 0.4338,
+      "step": 1955
+    },
+    {
+      "epoch": 1.5217842323651452,
+      "grad_norm": 0.4804215456005456,
+      "learning_rate": 5.742119564155243e-06,
+      "loss": 0.3203,
+      "step": 1956
+    },
+    {
+      "epoch": 1.5225622406639006,
+      "grad_norm": 0.6910987784092586,
+      "learning_rate": 5.737641318904581e-06,
+      "loss": 0.4564,
+      "step": 1957
+    },
+    {
+      "epoch": 1.5233402489626555,
+      "grad_norm": 0.5634671432854095,
+      "learning_rate": 5.733162468680417e-06,
+      "loss": 0.3914,
+      "step": 1958
+    },
+    {
+      "epoch": 1.524118257261411,
+      "grad_norm": 0.5265443255559377,
+      "learning_rate": 5.728683017156064e-06,
+      "loss": 0.3736,
+      "step": 1959
+    },
+    {
+      "epoch": 1.5248962655601659,
+      "grad_norm": 0.591494344975664,
+      "learning_rate": 5.724202968005324e-06,
+      "loss": 0.4265,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5256742738589213,
+      "grad_norm": 0.5519209219421536,
+      "learning_rate": 5.719722324902495e-06,
+      "loss": 0.3807,
+      "step": 1961
+    },
+    {
+      "epoch": 1.5264522821576763,
+      "grad_norm": 0.5541409323380365,
+      "learning_rate": 5.7152410915223525e-06,
+      "loss": 0.4429,
+      "step": 1962
+    },
+    {
+      "epoch": 1.5272302904564317,
+      "grad_norm": 0.585347184394982,
+      "learning_rate": 5.710759271540164e-06,
+      "loss": 0.3973,
+      "step": 1963
+    },
+    {
+      "epoch": 1.5280082987551866,
+      "grad_norm": 0.5523615167208807,
+      "learning_rate": 5.706276868631675e-06,
+      "loss": 0.4292,
+      "step": 1964
+    },
+    {
+      "epoch": 1.528786307053942,
+      "grad_norm": 0.5408633043874059,
+      "learning_rate": 5.701793886473114e-06,
+      "loss": 0.4292,
+      "step": 1965
+    },
+    {
+      "epoch": 1.529564315352697,
+      "grad_norm": 0.5311941394156889,
+      "learning_rate": 5.697310328741175e-06,
+      "loss": 0.3947,
+      "step": 1966
+    },
+    {
+      "epoch": 1.5303423236514524,
+      "grad_norm": 0.4838134413697601,
+      "learning_rate": 5.692826199113032e-06,
+      "loss": 0.3684,
+      "step": 1967
+    },
+    {
+      "epoch": 1.5311203319502074,
+      "grad_norm": 0.48910472646434655,
+      "learning_rate": 5.688341501266325e-06,
+      "loss": 0.3837,
+      "step": 1968
+    },
+    {
+      "epoch": 1.5318983402489628,
+      "grad_norm": 0.6025503759381499,
+      "learning_rate": 5.683856238879161e-06,
+      "loss": 0.4156,
+      "step": 1969
+    },
+    {
+      "epoch": 1.5326763485477177,
+      "grad_norm": 0.5716362273934151,
+      "learning_rate": 5.679370415630109e-06,
+      "loss": 0.4408,
+      "step": 1970
+    },
+    {
+      "epoch": 1.5334543568464731,
+      "grad_norm": 0.5860719045038633,
+      "learning_rate": 5.674884035198198e-06,
+      "loss": 0.3969,
+      "step": 1971
+    },
+    {
+      "epoch": 1.534232365145228,
+      "grad_norm": 0.5612013318644737,
+      "learning_rate": 5.670397101262917e-06,
+      "loss": 0.3912,
+      "step": 1972
+    },
+    {
+      "epoch": 1.5350103734439835,
+      "grad_norm": 0.5591199868144547,
+      "learning_rate": 5.665909617504204e-06,
+      "loss": 0.4248,
+      "step": 1973
+    },
+    {
+      "epoch": 1.5357883817427385,
+      "grad_norm": 0.5610410239671493,
+      "learning_rate": 5.6614215876024515e-06,
+      "loss": 0.3622,
+      "step": 1974
+    },
+    {
+      "epoch": 1.5365663900414939,
+      "grad_norm": 0.6092610824321522,
+      "learning_rate": 5.656933015238497e-06,
+      "loss": 0.4471,
+      "step": 1975
+    },
+    {
+      "epoch": 1.5373443983402488,
+      "grad_norm": 0.5142077251564491,
+      "learning_rate": 5.652443904093627e-06,
+      "loss": 0.3719,
+      "step": 1976
+    },
+    {
+      "epoch": 1.5381224066390042,
+      "grad_norm": 0.5575256608166222,
+      "learning_rate": 5.647954257849566e-06,
+      "loss": 0.4484,
+      "step": 1977
+    },
+    {
+      "epoch": 1.5389004149377592,
+      "grad_norm": 0.5764466466364058,
+      "learning_rate": 5.643464080188481e-06,
+      "loss": 0.4028,
+      "step": 1978
+    },
+    {
+      "epoch": 1.5396784232365146,
+      "grad_norm": 0.5393388055818888,
+      "learning_rate": 5.638973374792971e-06,
+      "loss": 0.3705,
+      "step": 1979
+    },
+    {
+      "epoch": 1.5404564315352696,
+      "grad_norm": 0.5481899007973284,
+      "learning_rate": 5.63448214534607e-06,
+      "loss": 0.4332,
+      "step": 1980
+    },
+    {
+      "epoch": 1.541234439834025,
+      "grad_norm": 0.5950443331633755,
+      "learning_rate": 5.629990395531241e-06,
+      "loss": 0.3989,
+      "step": 1981
+    },
+    {
+      "epoch": 1.54201244813278,
+      "grad_norm": 0.5567887859708718,
+      "learning_rate": 5.625498129032375e-06,
+      "loss": 0.3833,
+      "step": 1982
+    },
+    {
+      "epoch": 1.5427904564315353,
+      "grad_norm": 0.5625219683593582,
+      "learning_rate": 5.621005349533786e-06,
+      "loss": 0.3491,
+      "step": 1983
+    },
+    {
+      "epoch": 1.5435684647302903,
+      "grad_norm": 0.6265762841245444,
+      "learning_rate": 5.616512060720206e-06,
+      "loss": 0.4463,
+      "step": 1984
+    },
+    {
+      "epoch": 1.5443464730290457,
+      "grad_norm": 0.5325121702987277,
+      "learning_rate": 5.6120182662767915e-06,
+      "loss": 0.3594,
+      "step": 1985
+    },
+    {
+      "epoch": 1.5451244813278007,
+      "grad_norm": 0.514642222964822,
+      "learning_rate": 5.607523969889107e-06,
+      "loss": 0.3544,
+      "step": 1986
+    },
+    {
+      "epoch": 1.545902489626556,
+      "grad_norm": 0.5643536377434794,
+      "learning_rate": 5.603029175243133e-06,
+      "loss": 0.398,
+      "step": 1987
+    },
+    {
+      "epoch": 1.546680497925311,
+      "grad_norm": 0.5830063946034949,
+      "learning_rate": 5.5985338860252544e-06,
+      "loss": 0.3964,
+      "step": 1988
+    },
+    {
+      "epoch": 1.5474585062240664,
+      "grad_norm": 0.5254216102249979,
+      "learning_rate": 5.594038105922267e-06,
+      "loss": 0.3875,
+      "step": 1989
+    },
+    {
+      "epoch": 1.5482365145228216,
+      "grad_norm": 0.5627642438001997,
+      "learning_rate": 5.589541838621364e-06,
+      "loss": 0.392,
+      "step": 1990
+    },
+    {
+      "epoch": 1.5490145228215768,
+      "grad_norm": 0.5624495285963266,
+      "learning_rate": 5.585045087810142e-06,
+      "loss": 0.365,
+      "step": 1991
+    },
+    {
+      "epoch": 1.549792531120332,
+      "grad_norm": 0.6284335295588108,
+      "learning_rate": 5.580547857176592e-06,
+      "loss": 0.4506,
+      "step": 1992
+    },
+    {
+      "epoch": 1.5505705394190872,
+      "grad_norm": 0.5972847575220268,
+      "learning_rate": 5.576050150409098e-06,
+      "loss": 0.4196,
+      "step": 1993
+    },
+    {
+      "epoch": 1.5513485477178424,
+      "grad_norm": 0.5452503680041826,
+      "learning_rate": 5.571551971196439e-06,
+      "loss": 0.3843,
+      "step": 1994
+    },
+    {
+      "epoch": 1.5521265560165975,
+      "grad_norm": 0.6343585292830105,
+      "learning_rate": 5.5670533232277754e-06,
+      "loss": 0.4312,
+      "step": 1995
+    },
+    {
+      "epoch": 1.5529045643153527,
+      "grad_norm": 0.5111253341957419,
+      "learning_rate": 5.562554210192659e-06,
+      "loss": 0.3763,
+      "step": 1996
+    },
+    {
+      "epoch": 1.553682572614108,
+      "grad_norm": 0.5112381793219121,
+      "learning_rate": 5.558054635781013e-06,
+      "loss": 0.4033,
+      "step": 1997
+    },
+    {
+      "epoch": 1.554460580912863,
+      "grad_norm": 0.5619847389742358,
+      "learning_rate": 5.553554603683147e-06,
+      "loss": 0.4616,
+      "step": 1998
+    },
+    {
+      "epoch": 1.5552385892116183,
+      "grad_norm": 0.49246528646988497,
+      "learning_rate": 5.549054117589746e-06,
+      "loss": 0.3414,
+      "step": 1999
+    },
+    {
+      "epoch": 1.5560165975103735,
+      "grad_norm": 0.48851090999328967,
+      "learning_rate": 5.544553181191864e-06,
+      "loss": 0.3951,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5567946058091287,
+      "grad_norm": 0.8718460708261238,
+      "learning_rate": 5.540051798180926e-06,
+      "loss": 0.487,
+      "step": 2001
+    },
+    {
+      "epoch": 1.5575726141078838,
+      "grad_norm": 0.5090687869337716,
+      "learning_rate": 5.53554997224872e-06,
+      "loss": 0.3414,
+      "step": 2002
+    },
+    {
+      "epoch": 1.558350622406639,
+      "grad_norm": 0.5086070176188225,
+      "learning_rate": 5.531047707087403e-06,
+      "loss": 0.343,
+      "step": 2003
+    },
+    {
+      "epoch": 1.5591286307053942,
+      "grad_norm": 0.5057564322024687,
+      "learning_rate": 5.526545006389489e-06,
+      "loss": 0.4312,
+      "step": 2004
+    },
+    {
+      "epoch": 1.5599066390041494,
+      "grad_norm": 0.5462122572238138,
+      "learning_rate": 5.522041873847849e-06,
+      "loss": 0.3489,
+      "step": 2005
+    },
+    {
+      "epoch": 1.5606846473029046,
+      "grad_norm": 0.6406755747268688,
+      "learning_rate": 5.517538313155708e-06,
+      "loss": 0.4341,
+      "step": 2006
+    },
+    {
+      "epoch": 1.5614626556016598,
+      "grad_norm": 0.46597490009730924,
+      "learning_rate": 5.513034328006644e-06,
+      "loss": 0.366,
+      "step": 2007
+    },
+    {
+      "epoch": 1.562240663900415,
+      "grad_norm": 0.5280403166469179,
+      "learning_rate": 5.508529922094581e-06,
+      "loss": 0.4076,
+      "step": 2008
+    },
+    {
+      "epoch": 1.5630186721991701,
+      "grad_norm": 0.517931553215537,
+      "learning_rate": 5.50402509911379e-06,
+      "loss": 0.3515,
+      "step": 2009
+    },
+    {
+      "epoch": 1.5637966804979253,
+      "grad_norm": 0.6013170965367755,
+      "learning_rate": 5.499519862758882e-06,
+      "loss": 0.3799,
+      "step": 2010
+    },
+    {
+      "epoch": 1.5645746887966805,
+      "grad_norm": 0.5415014443209477,
+      "learning_rate": 5.4950142167248084e-06,
+      "loss": 0.3663,
+      "step": 2011
+    },
+    {
+      "epoch": 1.5653526970954357,
+      "grad_norm": 0.565467427025543,
+      "learning_rate": 5.4905081647068555e-06,
+      "loss": 0.4298,
+      "step": 2012
+    },
+    {
+      "epoch": 1.5661307053941909,
+      "grad_norm": 0.5062221630378054,
+      "learning_rate": 5.486001710400647e-06,
+      "loss": 0.3902,
+      "step": 2013
+    },
+    {
+      "epoch": 1.566908713692946,
+      "grad_norm": 0.5210934071830751,
+      "learning_rate": 5.481494857502129e-06,
+      "loss": 0.3701,
+      "step": 2014
+    },
+    {
+      "epoch": 1.5676867219917012,
+      "grad_norm": 0.5815972724685885,
+      "learning_rate": 5.476987609707579e-06,
+      "loss": 0.3938,
+      "step": 2015
+    },
+    {
+      "epoch": 1.5684647302904564,
+      "grad_norm": 0.5958807057927087,
+      "learning_rate": 5.472479970713599e-06,
+      "loss": 0.4071,
+      "step": 2016
+    },
+    {
+      "epoch": 1.5692427385892116,
+      "grad_norm": 0.5671962050796571,
+      "learning_rate": 5.46797194421711e-06,
+      "loss": 0.365,
+      "step": 2017
+    },
+    {
+      "epoch": 1.5700207468879668,
+      "grad_norm": 0.5572237085930942,
+      "learning_rate": 5.463463533915352e-06,
+      "loss": 0.4481,
+      "step": 2018
+    },
+    {
+      "epoch": 1.570798755186722,
+      "grad_norm": 0.5390902231502043,
+      "learning_rate": 5.458954743505877e-06,
+      "loss": 0.3892,
+      "step": 2019
+    },
+    {
+      "epoch": 1.5715767634854771,
+      "grad_norm": 0.5119900562108131,
+      "learning_rate": 5.454445576686555e-06,
+      "loss": 0.372,
+      "step": 2020
+    },
+    {
+      "epoch": 1.5723547717842323,
+      "grad_norm": 0.5448160846688016,
+      "learning_rate": 5.449936037155556e-06,
+      "loss": 0.3929,
+      "step": 2021
+    },
+    {
+      "epoch": 1.5731327800829875,
+      "grad_norm": 0.5768215881224731,
+      "learning_rate": 5.445426128611367e-06,
+      "loss": 0.4213,
+      "step": 2022
+    },
+    {
+      "epoch": 1.5739107883817427,
+      "grad_norm": 0.5350381999395692,
+      "learning_rate": 5.4409158547527616e-06,
+      "loss": 0.3986,
+      "step": 2023
+    },
+    {
+      "epoch": 1.5746887966804979,
+      "grad_norm": 0.5213615529701441,
+      "learning_rate": 5.436405219278831e-06,
+      "loss": 0.4168,
+      "step": 2024
+    },
+    {
+      "epoch": 1.575466804979253,
+      "grad_norm": 0.47445773672308517,
+      "learning_rate": 5.431894225888948e-06,
+      "loss": 0.3562,
+      "step": 2025
+    },
+    {
+      "epoch": 1.5762448132780082,
+      "grad_norm": 0.5132262197786499,
+      "learning_rate": 5.427382878282789e-06,
+      "loss": 0.3831,
+      "step": 2026
+    },
+    {
+      "epoch": 1.5770228215767634,
+      "grad_norm": 0.5314676552317412,
+      "learning_rate": 5.422871180160317e-06,
+      "loss": 0.3944,
+      "step": 2027
+    },
+    {
+      "epoch": 1.5778008298755186,
+      "grad_norm": 0.634514185317542,
+      "learning_rate": 5.418359135221782e-06,
+      "loss": 0.4586,
+      "step": 2028
+    },
+    {
+      "epoch": 1.5785788381742738,
+      "grad_norm": 0.5670133573240419,
+      "learning_rate": 5.413846747167719e-06,
+      "loss": 0.4185,
+      "step": 2029
+    },
+    {
+      "epoch": 1.5793568464730292,
+      "grad_norm": 0.5633025809202581,
+      "learning_rate": 5.409334019698945e-06,
+      "loss": 0.3537,
+      "step": 2030
+    },
+    {
+      "epoch": 1.5801348547717842,
+      "grad_norm": 0.5155689149222971,
+      "learning_rate": 5.404820956516555e-06,
+      "loss": 0.3818,
+      "step": 2031
+    },
+    {
+      "epoch": 1.5809128630705396,
+      "grad_norm": 0.5466102005968764,
+      "learning_rate": 5.400307561321918e-06,
+      "loss": 0.3659,
+      "step": 2032
+    },
+    {
+      "epoch": 1.5816908713692945,
+      "grad_norm": 0.5025594153087526,
+      "learning_rate": 5.3957938378166775e-06,
+      "loss": 0.3716,
+      "step": 2033
+    },
+    {
+      "epoch": 1.58246887966805,
+      "grad_norm": 0.5773544404414245,
+      "learning_rate": 5.391279789702747e-06,
+      "loss": 0.3809,
+      "step": 2034
+    },
+    {
+      "epoch": 1.583246887966805,
+      "grad_norm": 0.6494217279058303,
+      "learning_rate": 5.386765420682301e-06,
+      "loss": 0.4902,
+      "step": 2035
+    },
+    {
+      "epoch": 1.5840248962655603,
+      "grad_norm": 0.5450400867785532,
+      "learning_rate": 5.3822507344577835e-06,
+      "loss": 0.3753,
+      "step": 2036
+    },
+    {
+      "epoch": 1.5848029045643153,
+      "grad_norm": 0.572885909973732,
+      "learning_rate": 5.377735734731894e-06,
+      "loss": 0.4276,
+      "step": 2037
+    },
+    {
+      "epoch": 1.5855809128630707,
+      "grad_norm": 0.4947722913765326,
+      "learning_rate": 5.373220425207593e-06,
+      "loss": 0.3894,
+      "step": 2038
+    },
+    {
+      "epoch": 1.5863589211618256,
+      "grad_norm": 0.500728292740155,
+      "learning_rate": 5.36870480958809e-06,
+      "loss": 0.3745,
+      "step": 2039
+    },
+    {
+      "epoch": 1.587136929460581,
+      "grad_norm": 0.581801997857012,
+      "learning_rate": 5.364188891576852e-06,
+      "loss": 0.4235,
+      "step": 2040
+    },
+    {
+      "epoch": 1.587914937759336,
+      "grad_norm": 0.6422034189844766,
+      "learning_rate": 5.359672674877588e-06,
+      "loss": 0.4487,
+      "step": 2041
+    },
+    {
+      "epoch": 1.5886929460580914,
+      "grad_norm": 0.5701343145317651,
+      "learning_rate": 5.355156163194257e-06,
+      "loss": 0.3858,
+      "step": 2042
+    },
+    {
+      "epoch": 1.5894709543568464,
+      "grad_norm": 0.48343989676020743,
+      "learning_rate": 5.3506393602310584e-06,
+      "loss": 0.3511,
+      "step": 2043
+    },
+    {
+      "epoch": 1.5902489626556018,
+      "grad_norm": 0.5308367122291103,
+      "learning_rate": 5.346122269692425e-06,
+      "loss": 0.418,
+      "step": 2044
+    },
+    {
+      "epoch": 1.5910269709543567,
+      "grad_norm": 0.5531150969810927,
+      "learning_rate": 5.341604895283032e-06,
+      "loss": 0.4134,
+      "step": 2045
+    },
+    {
+      "epoch": 1.5918049792531122,
+      "grad_norm": 0.5407937389936823,
+      "learning_rate": 5.337087240707787e-06,
+      "loss": 0.392,
+      "step": 2046
+    },
+    {
+      "epoch": 1.5925829875518671,
+      "grad_norm": 0.5342789528578493,
+      "learning_rate": 5.332569309671826e-06,
+      "loss": 0.356,
+      "step": 2047
+    },
+    {
+      "epoch": 1.5933609958506225,
+      "grad_norm": 0.5974158292002194,
+      "learning_rate": 5.328051105880508e-06,
+      "loss": 0.4101,
+      "step": 2048
+    },
+    {
+      "epoch": 1.5941390041493775,
+      "grad_norm": 0.5171223566299462,
+      "learning_rate": 5.3235326330394235e-06,
+      "loss": 0.3548,
+      "step": 2049
+    },
+    {
+      "epoch": 1.5949170124481329,
+      "grad_norm": 0.5739805007573618,
+      "learning_rate": 5.319013894854378e-06,
+      "loss": 0.4007,
+      "step": 2050
+    },
+    {
+      "epoch": 1.5956950207468878,
+      "grad_norm": 0.5311109848971282,
+      "learning_rate": 5.314494895031395e-06,
+      "loss": 0.3526,
+      "step": 2051
+    },
+    {
+      "epoch": 1.5964730290456433,
+      "grad_norm": 0.6183273267288776,
+      "learning_rate": 5.309975637276717e-06,
+      "loss": 0.4369,
+      "step": 2052
+    },
+    {
+      "epoch": 1.5972510373443982,
+      "grad_norm": 0.5233267180517742,
+      "learning_rate": 5.305456125296789e-06,
+      "loss": 0.355,
+      "step": 2053
+    },
+    {
+      "epoch": 1.5980290456431536,
+      "grad_norm": 0.544480900561584,
+      "learning_rate": 5.300936362798273e-06,
+      "loss": 0.4239,
+      "step": 2054
+    },
+    {
+      "epoch": 1.5988070539419086,
+      "grad_norm": 0.5307295672103095,
+      "learning_rate": 5.296416353488037e-06,
+      "loss": 0.3991,
+      "step": 2055
+    },
+    {
+      "epoch": 1.599585062240664,
+      "grad_norm": 0.5707195297422504,
+      "learning_rate": 5.291896101073144e-06,
+      "loss": 0.3861,
+      "step": 2056
+    },
+    {
+      "epoch": 1.600363070539419,
+      "grad_norm": 0.581388093497825,
+      "learning_rate": 5.287375609260862e-06,
+      "loss": 0.4096,
+      "step": 2057
+    },
+    {
+      "epoch": 1.6011410788381744,
+      "grad_norm": 0.5962131309565865,
+      "learning_rate": 5.2828548817586535e-06,
+      "loss": 0.4287,
+      "step": 2058
+    },
+    {
+      "epoch": 1.6019190871369293,
+      "grad_norm": 0.542815062886424,
+      "learning_rate": 5.278333922274173e-06,
+      "loss": 0.3759,
+      "step": 2059
+    },
+    {
+      "epoch": 1.6026970954356847,
+      "grad_norm": 0.5417673610753667,
+      "learning_rate": 5.2738127345152705e-06,
+      "loss": 0.4097,
+      "step": 2060
+    },
+    {
+      "epoch": 1.6034751037344397,
+      "grad_norm": 0.5280286792050063,
+      "learning_rate": 5.269291322189975e-06,
+      "loss": 0.4121,
+      "step": 2061
+    },
+    {
+      "epoch": 1.604253112033195,
+      "grad_norm": 0.5203799778844658,
+      "learning_rate": 5.264769689006509e-06,
+      "loss": 0.4219,
+      "step": 2062
+    },
+    {
+      "epoch": 1.60503112033195,
+      "grad_norm": 0.49954491772179105,
+      "learning_rate": 5.260247838673269e-06,
+      "loss": 0.3966,
+      "step": 2063
+    },
+    {
+      "epoch": 1.6058091286307055,
+      "grad_norm": 0.5353510942834874,
+      "learning_rate": 5.255725774898831e-06,
+      "loss": 0.4051,
+      "step": 2064
+    },
+    {
+      "epoch": 1.6065871369294604,
+      "grad_norm": 0.6347880065208158,
+      "learning_rate": 5.251203501391951e-06,
+      "loss": 0.4813,
+      "step": 2065
+    },
+    {
+      "epoch": 1.6073651452282158,
+      "grad_norm": 0.5012016943595701,
+      "learning_rate": 5.246681021861546e-06,
+      "loss": 0.374,
+      "step": 2066
+    },
+    {
+      "epoch": 1.6081431535269708,
+      "grad_norm": 0.48166755425056446,
+      "learning_rate": 5.242158340016715e-06,
+      "loss": 0.3776,
+      "step": 2067
+    },
+    {
+      "epoch": 1.6089211618257262,
+      "grad_norm": 0.6493325414333341,
+      "learning_rate": 5.237635459566716e-06,
+      "loss": 0.4277,
+      "step": 2068
+    },
+    {
+      "epoch": 1.6096991701244814,
+      "grad_norm": 0.5889423116546074,
+      "learning_rate": 5.233112384220972e-06,
+      "loss": 0.4471,
+      "step": 2069
+    },
+    {
+      "epoch": 1.6104771784232366,
+      "grad_norm": 0.5003788660925224,
+      "learning_rate": 5.22858911768906e-06,
+      "loss": 0.3724,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6112551867219918,
+      "grad_norm": 0.570255365658206,
+      "learning_rate": 5.2240656636807216e-06,
+      "loss": 0.3708,
+      "step": 2071
+    },
+    {
+      "epoch": 1.612033195020747,
+      "grad_norm": 0.5663407609646942,
+      "learning_rate": 5.219542025905849e-06,
+      "loss": 0.3631,
+      "step": 2072
+    },
+    {
+      "epoch": 1.6128112033195021,
+      "grad_norm": 0.5181056504589183,
+      "learning_rate": 5.215018208074485e-06,
+      "loss": 0.3388,
+      "step": 2073
+    },
+    {
+      "epoch": 1.6135892116182573,
+      "grad_norm": 0.48963501523756564,
+      "learning_rate": 5.210494213896819e-06,
+      "loss": 0.3731,
+      "step": 2074
+    },
+    {
+      "epoch": 1.6143672199170125,
+      "grad_norm": 0.5232952649939704,
+      "learning_rate": 5.205970047083189e-06,
+      "loss": 0.4142,
+      "step": 2075
+    },
+    {
+      "epoch": 1.6151452282157677,
+      "grad_norm": 0.48038757698907136,
+      "learning_rate": 5.20144571134407e-06,
+      "loss": 0.3874,
+      "step": 2076
+    },
+    {
+      "epoch": 1.6159232365145229,
+      "grad_norm": 0.661222739343923,
+      "learning_rate": 5.196921210390077e-06,
+      "loss": 0.4628,
+      "step": 2077
+    },
+    {
+      "epoch": 1.616701244813278,
+      "grad_norm": 0.514810469178608,
+      "learning_rate": 5.1923965479319615e-06,
+      "loss": 0.4006,
+      "step": 2078
+    },
+    {
+      "epoch": 1.6174792531120332,
+      "grad_norm": 0.5954168702771858,
+      "learning_rate": 5.187871727680607e-06,
+      "loss": 0.4128,
+      "step": 2079
+    },
+    {
+      "epoch": 1.6182572614107884,
+      "grad_norm": 0.5694112466639059,
+      "learning_rate": 5.183346753347025e-06,
+      "loss": 0.3862,
+      "step": 2080
+    },
+    {
+      "epoch": 1.6190352697095436,
+      "grad_norm": 0.486061856821325,
+      "learning_rate": 5.178821628642356e-06,
+      "loss": 0.3386,
+      "step": 2081
+    },
+    {
+      "epoch": 1.6198132780082988,
+      "grad_norm": 0.5908904498137724,
+      "learning_rate": 5.174296357277865e-06,
+      "loss": 0.437,
+      "step": 2082
+    },
+    {
+      "epoch": 1.620591286307054,
+      "grad_norm": 0.6596321251393273,
+      "learning_rate": 5.169770942964929e-06,
+      "loss": 0.4117,
+      "step": 2083
+    },
+    {
+      "epoch": 1.6213692946058091,
+      "grad_norm": 0.5679874075326472,
+      "learning_rate": 5.1652453894150516e-06,
+      "loss": 0.386,
+      "step": 2084
+    },
+    {
+      "epoch": 1.6221473029045643,
+      "grad_norm": 0.5614337986123863,
+      "learning_rate": 5.160719700339846e-06,
+      "loss": 0.427,
+      "step": 2085
+    },
+    {
+      "epoch": 1.6229253112033195,
+      "grad_norm": 0.5068259771641748,
+      "learning_rate": 5.156193879451038e-06,
+      "loss": 0.3747,
+      "step": 2086
+    },
+    {
+      "epoch": 1.6237033195020747,
+      "grad_norm": 0.5342608245446453,
+      "learning_rate": 5.15166793046046e-06,
+      "loss": 0.3998,
+      "step": 2087
+    },
+    {
+      "epoch": 1.6244813278008299,
+      "grad_norm": 0.48453536365653094,
+      "learning_rate": 5.147141857080052e-06,
+      "loss": 0.3804,
+      "step": 2088
+    },
+    {
+      "epoch": 1.625259336099585,
+      "grad_norm": 0.5937087813973254,
+      "learning_rate": 5.142615663021854e-06,
+      "loss": 0.4312,
+      "step": 2089
+    },
+    {
+      "epoch": 1.6260373443983402,
+      "grad_norm": 0.5530083700516197,
+      "learning_rate": 5.1380893519980056e-06,
+      "loss": 0.429,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6268153526970954,
+      "grad_norm": 0.45068038873556165,
+      "learning_rate": 5.133562927720742e-06,
+      "loss": 0.3471,
+      "step": 2091
+    },
+    {
+      "epoch": 1.6275933609958506,
+      "grad_norm": 0.5077176430545055,
+      "learning_rate": 5.1290363939023905e-06,
+      "loss": 0.388,
+      "step": 2092
+    },
+    {
+      "epoch": 1.6283713692946058,
+      "grad_norm": 0.5201811255473193,
+      "learning_rate": 5.1245097542553705e-06,
+      "loss": 0.399,
+      "step": 2093
+    },
+    {
+      "epoch": 1.629149377593361,
+      "grad_norm": 0.4790930119601685,
+      "learning_rate": 5.1199830124921845e-06,
+      "loss": 0.374,
+      "step": 2094
+    },
+    {
+      "epoch": 1.6299273858921162,
+      "grad_norm": 0.6125456959653823,
+      "learning_rate": 5.115456172325427e-06,
+      "loss": 0.4131,
+      "step": 2095
+    },
+    {
+      "epoch": 1.6307053941908713,
+      "grad_norm": 0.47799319606805274,
+      "learning_rate": 5.110929237467761e-06,
+      "loss": 0.3531,
+      "step": 2096
+    },
+    {
+      "epoch": 1.6314834024896265,
+      "grad_norm": 0.5364779503968133,
+      "learning_rate": 5.106402211631937e-06,
+      "loss": 0.3828,
+      "step": 2097
+    },
+    {
+      "epoch": 1.6322614107883817,
+      "grad_norm": 0.5889384265820924,
+      "learning_rate": 5.101875098530775e-06,
+      "loss": 0.4848,
+      "step": 2098
+    },
+    {
+      "epoch": 1.633039419087137,
+      "grad_norm": 0.5146633048099957,
+      "learning_rate": 5.09734790187717e-06,
+      "loss": 0.3675,
+      "step": 2099
+    },
+    {
+      "epoch": 1.633817427385892,
+      "grad_norm": 0.5796372361495907,
+      "learning_rate": 5.092820625384083e-06,
+      "loss": 0.4249,
+      "step": 2100
+    },
+    {
+      "epoch": 1.6345954356846473,
+      "grad_norm": 0.6560679744712808,
+      "learning_rate": 5.088293272764541e-06,
+      "loss": 0.4623,
+      "step": 2101
+    },
+    {
+      "epoch": 1.6353734439834025,
+      "grad_norm": 0.46532361226284535,
+      "learning_rate": 5.083765847731632e-06,
+      "loss": 0.2975,
+      "step": 2102
+    },
+    {
+      "epoch": 1.6361514522821576,
+      "grad_norm": 0.5403109241011086,
+      "learning_rate": 5.079238353998508e-06,
+      "loss": 0.4102,
+      "step": 2103
+    },
+    {
+      "epoch": 1.6369294605809128,
+      "grad_norm": 0.5409071584866033,
+      "learning_rate": 5.074710795278375e-06,
+      "loss": 0.384,
+      "step": 2104
+    },
+    {
+      "epoch": 1.637707468879668,
+      "grad_norm": 0.5515644385579896,
+      "learning_rate": 5.070183175284488e-06,
+      "loss": 0.4144,
+      "step": 2105
+    },
+    {
+      "epoch": 1.6384854771784232,
+      "grad_norm": 0.5642917104016661,
+      "learning_rate": 5.065655497730158e-06,
+      "loss": 0.4192,
+      "step": 2106
+    },
+    {
+      "epoch": 1.6392634854771784,
+      "grad_norm": 0.5347023306377896,
+      "learning_rate": 5.061127766328738e-06,
+      "loss": 0.3791,
+      "step": 2107
+    },
+    {
+      "epoch": 1.6400414937759336,
+      "grad_norm": 0.5269459002798191,
+      "learning_rate": 5.056599984793635e-06,
+      "loss": 0.3761,
+      "step": 2108
+    },
+    {
+      "epoch": 1.640819502074689,
+      "grad_norm": 0.5480385564483365,
+      "learning_rate": 5.052072156838287e-06,
+      "loss": 0.4105,
+      "step": 2109
+    },
+    {
+      "epoch": 1.641597510373444,
+      "grad_norm": 0.6148023431697652,
+      "learning_rate": 5.047544286176171e-06,
+      "loss": 0.4301,
+      "step": 2110
+    },
+    {
+      "epoch": 1.6423755186721993,
+      "grad_norm": 0.550083097948547,
+      "learning_rate": 5.043016376520805e-06,
+      "loss": 0.3982,
+      "step": 2111
+    },
+    {
+      "epoch": 1.6431535269709543,
+      "grad_norm": 0.4566722510304721,
+      "learning_rate": 5.038488431585735e-06,
+      "loss": 0.3495,
+      "step": 2112
+    },
+    {
+      "epoch": 1.6439315352697097,
+      "grad_norm": 0.5955473947162069,
+      "learning_rate": 5.033960455084535e-06,
+      "loss": 0.4151,
+      "step": 2113
+    },
+    {
+      "epoch": 1.6447095435684647,
+      "grad_norm": 0.5815210771672532,
+      "learning_rate": 5.029432450730807e-06,
+      "loss": 0.4158,
+      "step": 2114
+    },
+    {
+      "epoch": 1.64548755186722,
+      "grad_norm": 0.5545184178202704,
+      "learning_rate": 5.024904422238176e-06,
+      "loss": 0.4468,
+      "step": 2115
+    },
+    {
+      "epoch": 1.646265560165975,
+      "grad_norm": 0.541284252886828,
+      "learning_rate": 5.020376373320285e-06,
+      "loss": 0.3919,
+      "step": 2116
+    },
+    {
+      "epoch": 1.6470435684647304,
+      "grad_norm": 0.528027251176678,
+      "learning_rate": 5.015848307690795e-06,
+      "loss": 0.3892,
+      "step": 2117
+    },
+    {
+      "epoch": 1.6478215767634854,
+      "grad_norm": 0.5171123479600179,
+      "learning_rate": 5.011320229063381e-06,
+      "loss": 0.3724,
+      "step": 2118
+    },
+    {
+      "epoch": 1.6485995850622408,
+      "grad_norm": 0.5644129110099912,
+      "learning_rate": 5.006792141151726e-06,
+      "loss": 0.4301,
+      "step": 2119
+    },
+    {
+      "epoch": 1.6493775933609958,
+      "grad_norm": 0.5302895420435279,
+      "learning_rate": 5.002264047669526e-06,
+      "loss": 0.3871,
+      "step": 2120
+    },
+    {
+      "epoch": 1.6501556016597512,
+      "grad_norm": 0.5134703939905793,
+      "learning_rate": 4.997735952330475e-06,
+      "loss": 0.437,
+      "step": 2121
+    },
+    {
+      "epoch": 1.6509336099585061,
+      "grad_norm": 0.45989345445711877,
+      "learning_rate": 4.993207858848275e-06,
+      "loss": 0.4078,
+      "step": 2122
+    },
+    {
+      "epoch": 1.6517116182572615,
+      "grad_norm": 0.5029684805889143,
+      "learning_rate": 4.988679770936622e-06,
+      "loss": 0.3509,
+      "step": 2123
+    },
+    {
+      "epoch": 1.6524896265560165,
+      "grad_norm": 0.7011728529449217,
+      "learning_rate": 4.984151692309206e-06,
+      "loss": 0.5236,
+      "step": 2124
+    },
+    {
+      "epoch": 1.653267634854772,
+      "grad_norm": 0.48034442951513306,
+      "learning_rate": 4.979623626679716e-06,
+      "loss": 0.314,
+      "step": 2125
+    },
+    {
+      "epoch": 1.6540456431535269,
+      "grad_norm": 0.5534171883788328,
+      "learning_rate": 4.975095577761826e-06,
+      "loss": 0.3654,
+      "step": 2126
+    },
+    {
+      "epoch": 1.6548236514522823,
+      "grad_norm": 0.5371396396496473,
+      "learning_rate": 4.970567549269193e-06,
+      "loss": 0.3947,
+      "step": 2127
+    },
+    {
+      "epoch": 1.6556016597510372,
+      "grad_norm": 0.47215217379829066,
+      "learning_rate": 4.966039544915466e-06,
+      "loss": 0.4154,
+      "step": 2128
+    },
+    {
+      "epoch": 1.6563796680497926,
+      "grad_norm": 0.5440824432296227,
+      "learning_rate": 4.961511568414266e-06,
+      "loss": 0.3491,
+      "step": 2129
+    },
+    {
+      "epoch": 1.6571576763485476,
+      "grad_norm": 0.5735866958071315,
+      "learning_rate": 4.956983623479197e-06,
+      "loss": 0.4237,
+      "step": 2130
+    },
+    {
+      "epoch": 1.657935684647303,
+      "grad_norm": 0.557184053810615,
+      "learning_rate": 4.95245571382383e-06,
+      "loss": 0.3852,
+      "step": 2131
+    },
+    {
+      "epoch": 1.658713692946058,
+      "grad_norm": 0.5385224919841279,
+      "learning_rate": 4.947927843161714e-06,
+      "loss": 0.4005,
+      "step": 2132
+    },
+    {
+      "epoch": 1.6594917012448134,
+      "grad_norm": 0.4964365859006009,
+      "learning_rate": 4.943400015206366e-06,
+      "loss": 0.3446,
+      "step": 2133
+    },
+    {
+      "epoch": 1.6602697095435683,
+      "grad_norm": 0.5900262249082231,
+      "learning_rate": 4.9388722336712616e-06,
+      "loss": 0.4221,
+      "step": 2134
+    },
+    {
+      "epoch": 1.6610477178423237,
+      "grad_norm": 0.5687414392870107,
+      "learning_rate": 4.934344502269845e-06,
+      "loss": 0.4064,
+      "step": 2135
+    },
+    {
+      "epoch": 1.6618257261410787,
+      "grad_norm": 0.484227851108785,
+      "learning_rate": 4.929816824715514e-06,
+      "loss": 0.3713,
+      "step": 2136
+    },
+    {
+      "epoch": 1.6626037344398341,
+      "grad_norm": 0.5136813605346126,
+      "learning_rate": 4.925289204721628e-06,
+      "loss": 0.3809,
+      "step": 2137
+    },
+    {
+      "epoch": 1.663381742738589,
+      "grad_norm": 0.5423219894955067,
+      "learning_rate": 4.920761646001494e-06,
+      "loss": 0.3999,
+      "step": 2138
+    },
+    {
+      "epoch": 1.6641597510373445,
+      "grad_norm": 0.6143401547801955,
+      "learning_rate": 4.916234152268368e-06,
+      "loss": 0.4147,
+      "step": 2139
+    },
+    {
+      "epoch": 1.6649377593360994,
+      "grad_norm": 0.5984278053057214,
+      "learning_rate": 4.911706727235461e-06,
+      "loss": 0.366,
+      "step": 2140
+    },
+    {
+      "epoch": 1.6657157676348548,
+      "grad_norm": 0.556572823841576,
+      "learning_rate": 4.907179374615918e-06,
+      "loss": 0.418,
+      "step": 2141
+    },
+    {
+      "epoch": 1.6664937759336098,
+      "grad_norm": 0.5918690264426519,
+      "learning_rate": 4.902652098122831e-06,
+      "loss": 0.3949,
+      "step": 2142
+    },
+    {
+      "epoch": 1.6672717842323652,
+      "grad_norm": 0.6160409648459483,
+      "learning_rate": 4.898124901469226e-06,
+      "loss": 0.4567,
+      "step": 2143
+    },
+    {
+      "epoch": 1.6680497925311202,
+      "grad_norm": 0.6450149115744205,
+      "learning_rate": 4.8935977883680655e-06,
+      "loss": 0.4481,
+      "step": 2144
+    },
+    {
+      "epoch": 1.6688278008298756,
+      "grad_norm": 0.5777977810075752,
+      "learning_rate": 4.889070762532241e-06,
+      "loss": 0.4221,
+      "step": 2145
+    },
+    {
+      "epoch": 1.6696058091286305,
+      "grad_norm": 0.5830743244487808,
+      "learning_rate": 4.884543827674575e-06,
+      "loss": 0.4415,
+      "step": 2146
+    },
+    {
+      "epoch": 1.670383817427386,
+      "grad_norm": 0.4686162612487805,
+      "learning_rate": 4.880016987507816e-06,
+      "loss": 0.3585,
+      "step": 2147
+    },
+    {
+      "epoch": 1.671161825726141,
+      "grad_norm": 0.5030093323283157,
+      "learning_rate": 4.875490245744631e-06,
+      "loss": 0.4403,
+      "step": 2148
+    },
+    {
+      "epoch": 1.6719398340248963,
+      "grad_norm": 0.567476119884263,
+      "learning_rate": 4.870963606097612e-06,
+      "loss": 0.3978,
+      "step": 2149
+    },
+    {
+      "epoch": 1.6727178423236515,
+      "grad_norm": 0.5769378444250371,
+      "learning_rate": 4.8664370722792595e-06,
+      "loss": 0.3864,
+      "step": 2150
+    },
+    {
+      "epoch": 1.6734958506224067,
+      "grad_norm": 0.5731228545369228,
+      "learning_rate": 4.861910648001996e-06,
+      "loss": 0.3921,
+      "step": 2151
+    },
+    {
+      "epoch": 1.6742738589211619,
+      "grad_norm": 0.44653527193110076,
+      "learning_rate": 4.857384336978147e-06,
+      "loss": 0.3691,
+      "step": 2152
+    },
+    {
+      "epoch": 1.675051867219917,
+      "grad_norm": 0.5603552002006846,
+      "learning_rate": 4.852858142919948e-06,
+      "loss": 0.3879,
+      "step": 2153
+    },
+    {
+      "epoch": 1.6758298755186722,
+      "grad_norm": 0.596058953067879,
+      "learning_rate": 4.8483320695395405e-06,
+      "loss": 0.4534,
+      "step": 2154
+    },
+    {
+      "epoch": 1.6766078838174274,
+      "grad_norm": 0.572014988834668,
+      "learning_rate": 4.843806120548963e-06,
+      "loss": 0.4015,
+      "step": 2155
+    },
+    {
+      "epoch": 1.6773858921161826,
+      "grad_norm": 0.5519278671813699,
+      "learning_rate": 4.8392802996601565e-06,
+      "loss": 0.4136,
+      "step": 2156
+    },
+    {
+      "epoch": 1.6781639004149378,
+      "grad_norm": 0.5439941800406779,
+      "learning_rate": 4.83475461058495e-06,
+      "loss": 0.3716,
+      "step": 2157
+    },
+    {
+      "epoch": 1.678941908713693,
+      "grad_norm": 0.5289981306564617,
+      "learning_rate": 4.830229057035074e-06,
+      "loss": 0.3723,
+      "step": 2158
+    },
+    {
+      "epoch": 1.6797199170124482,
+      "grad_norm": 0.5453559692179399,
+      "learning_rate": 4.825703642722137e-06,
+      "loss": 0.4451,
+      "step": 2159
+    },
+    {
+      "epoch": 1.6804979253112033,
+      "grad_norm": 0.5573703150937594,
+      "learning_rate": 4.821178371357644e-06,
+      "loss": 0.3985,
+      "step": 2160
+    },
+    {
+      "epoch": 1.6812759336099585,
+      "grad_norm": 0.6012337893481218,
+      "learning_rate": 4.816653246652976e-06,
+      "loss": 0.4467,
+      "step": 2161
+    },
+    {
+      "epoch": 1.6820539419087137,
+      "grad_norm": 0.4953313193073563,
+      "learning_rate": 4.812128272319394e-06,
+      "loss": 0.3221,
+      "step": 2162
+    },
+    {
+      "epoch": 1.682831950207469,
+      "grad_norm": 0.5707820839429782,
+      "learning_rate": 4.80760345206804e-06,
+      "loss": 0.3951,
+      "step": 2163
+    },
+    {
+      "epoch": 1.683609958506224,
+      "grad_norm": 0.5098643643009366,
+      "learning_rate": 4.803078789609925e-06,
+      "loss": 0.3953,
+      "step": 2164
+    },
+    {
+      "epoch": 1.6843879668049793,
+      "grad_norm": 0.4897942825723685,
+      "learning_rate": 4.798554288655931e-06,
+      "loss": 0.3661,
+      "step": 2165
+    },
+    {
+      "epoch": 1.6851659751037344,
+      "grad_norm": 0.508256924256849,
+      "learning_rate": 4.794029952916812e-06,
+      "loss": 0.407,
+      "step": 2166
+    },
+    {
+      "epoch": 1.6859439834024896,
+      "grad_norm": 0.5779278918078747,
+      "learning_rate": 4.7895057861031805e-06,
+      "loss": 0.4128,
+      "step": 2167
+    },
+    {
+      "epoch": 1.6867219917012448,
+      "grad_norm": 0.5430074776758674,
+      "learning_rate": 4.784981791925516e-06,
+      "loss": 0.4267,
+      "step": 2168
+    },
+    {
+      "epoch": 1.6875,
+      "grad_norm": 0.5329469914509679,
+      "learning_rate": 4.780457974094152e-06,
+      "loss": 0.3397,
+      "step": 2169
+    },
+    {
+      "epoch": 1.6882780082987552,
+      "grad_norm": 0.5363082084339857,
+      "learning_rate": 4.775934336319281e-06,
+      "loss": 0.4017,
+      "step": 2170
+    },
+    {
+      "epoch": 1.6890560165975104,
+      "grad_norm": 0.4505395078807037,
+      "learning_rate": 4.771410882310942e-06,
+      "loss": 0.3465,
+      "step": 2171
+    },
+    {
+      "epoch": 1.6898340248962656,
+      "grad_norm": 0.6072200813904377,
+      "learning_rate": 4.76688761577903e-06,
+      "loss": 0.4195,
+      "step": 2172
+    },
+    {
+      "epoch": 1.6906120331950207,
+      "grad_norm": 0.5421866374616311,
+      "learning_rate": 4.762364540433285e-06,
+      "loss": 0.3662,
+      "step": 2173
+    },
+    {
+      "epoch": 1.691390041493776,
+      "grad_norm": 0.5242749438726794,
+      "learning_rate": 4.757841659983284e-06,
+      "loss": 0.39,
+      "step": 2174
+    },
+    {
+      "epoch": 1.692168049792531,
+      "grad_norm": 0.6291793218030014,
+      "learning_rate": 4.753318978138455e-06,
+      "loss": 0.4526,
+      "step": 2175
+    },
+    {
+      "epoch": 1.6929460580912863,
+      "grad_norm": 0.5145845968592745,
+      "learning_rate": 4.748796498608051e-06,
+      "loss": 0.3832,
+      "step": 2176
+    },
+    {
+      "epoch": 1.6937240663900415,
+      "grad_norm": 0.5550112134750425,
+      "learning_rate": 4.74427422510117e-06,
+      "loss": 0.3683,
+      "step": 2177
+    },
+    {
+      "epoch": 1.6945020746887967,
+      "grad_norm": 0.5073048213974801,
+      "learning_rate": 4.739752161326733e-06,
+      "loss": 0.3853,
+      "step": 2178
+    },
+    {
+      "epoch": 1.6952800829875518,
+      "grad_norm": 0.5666629877235039,
+      "learning_rate": 4.735230310993491e-06,
+      "loss": 0.3829,
+      "step": 2179
+    },
+    {
+      "epoch": 1.696058091286307,
+      "grad_norm": 0.636352542305011,
+      "learning_rate": 4.7307086778100256e-06,
+      "loss": 0.3888,
+      "step": 2180
+    },
+    {
+      "epoch": 1.6968360995850622,
+      "grad_norm": 0.5629476163977468,
+      "learning_rate": 4.726187265484731e-06,
+      "loss": 0.4066,
+      "step": 2181
+    },
+    {
+      "epoch": 1.6976141078838174,
+      "grad_norm": 0.6475724525026995,
+      "learning_rate": 4.721666077725829e-06,
+      "loss": 0.4989,
+      "step": 2182
+    },
+    {
+      "epoch": 1.6983921161825726,
+      "grad_norm": 0.4998648413555906,
+      "learning_rate": 4.717145118241348e-06,
+      "loss": 0.4092,
+      "step": 2183
+    },
+    {
+      "epoch": 1.6991701244813278,
+      "grad_norm": 0.5432990360289148,
+      "learning_rate": 4.712624390739141e-06,
+      "loss": 0.4122,
+      "step": 2184
+    },
+    {
+      "epoch": 1.699948132780083,
+      "grad_norm": 0.5332800024914512,
+      "learning_rate": 4.708103898926857e-06,
+      "loss": 0.3878,
+      "step": 2185
+    },
+    {
+      "epoch": 1.7007261410788381,
+      "grad_norm": 0.5498007885733288,
+      "learning_rate": 4.703583646511964e-06,
+      "loss": 0.4152,
+      "step": 2186
+    },
+    {
+      "epoch": 1.7015041493775933,
+      "grad_norm": 0.6129623617257118,
+      "learning_rate": 4.699063637201728e-06,
+      "loss": 0.4088,
+      "step": 2187
+    },
+    {
+      "epoch": 1.7022821576763485,
+      "grad_norm": 0.6186385178471406,
+      "learning_rate": 4.694543874703212e-06,
+      "loss": 0.4448,
+      "step": 2188
+    },
+    {
+      "epoch": 1.7030601659751037,
+      "grad_norm": 0.4953257417876297,
+      "learning_rate": 4.6900243627232866e-06,
+      "loss": 0.3574,
+      "step": 2189
+    },
+    {
+      "epoch": 1.703838174273859,
+      "grad_norm": 0.5572899286802421,
+      "learning_rate": 4.685505104968606e-06,
+      "loss": 0.4279,
+      "step": 2190
+    },
+    {
+      "epoch": 1.704616182572614,
+      "grad_norm": 0.5326528380329922,
+      "learning_rate": 4.680986105145625e-06,
+      "loss": 0.4026,
+      "step": 2191
+    },
+    {
+      "epoch": 1.7053941908713695,
+      "grad_norm": 0.5505707310911756,
+      "learning_rate": 4.676467366960577e-06,
+      "loss": 0.3982,
+      "step": 2192
+    },
+    {
+      "epoch": 1.7061721991701244,
+      "grad_norm": 0.49667861098704164,
+      "learning_rate": 4.671948894119492e-06,
+      "loss": 0.3917,
+      "step": 2193
+    },
+    {
+      "epoch": 1.7069502074688798,
+      "grad_norm": 0.5230673510042897,
+      "learning_rate": 4.667430690328177e-06,
+      "loss": 0.3854,
+      "step": 2194
+    },
+    {
+      "epoch": 1.7077282157676348,
+      "grad_norm": 0.5632897016394673,
+      "learning_rate": 4.662912759292214e-06,
+      "loss": 0.4372,
+      "step": 2195
+    },
+    {
+      "epoch": 1.7085062240663902,
+      "grad_norm": 0.5031235319000656,
+      "learning_rate": 4.65839510471697e-06,
+      "loss": 0.3623,
+      "step": 2196
+    },
+    {
+      "epoch": 1.7092842323651452,
+      "grad_norm": 0.49639028578236394,
+      "learning_rate": 4.653877730307576e-06,
+      "loss": 0.3609,
+      "step": 2197
+    },
+    {
+      "epoch": 1.7100622406639006,
+      "grad_norm": 0.5715836289016729,
+      "learning_rate": 4.649360639768944e-06,
+      "loss": 0.4389,
+      "step": 2198
+    },
+    {
+      "epoch": 1.7108402489626555,
+      "grad_norm": 0.5608572959162271,
+      "learning_rate": 4.644843836805744e-06,
+      "loss": 0.4076,
+      "step": 2199
+    },
+    {
+      "epoch": 1.711618257261411,
+      "grad_norm": 0.5727085137777886,
+      "learning_rate": 4.640327325122411e-06,
+      "loss": 0.3784,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7123962655601659,
+      "grad_norm": 0.5172519751311586,
+      "learning_rate": 4.6358111084231495e-06,
+      "loss": 0.4068,
+      "step": 2201
+    },
+    {
+      "epoch": 1.7131742738589213,
+      "grad_norm": 0.5854834959383782,
+      "learning_rate": 4.63129519041191e-06,
+      "loss": 0.4002,
+      "step": 2202
+    },
+    {
+      "epoch": 1.7139522821576763,
+      "grad_norm": 0.5431493253772754,
+      "learning_rate": 4.62677957479241e-06,
+      "loss": 0.3769,
+      "step": 2203
+    },
+    {
+      "epoch": 1.7147302904564317,
+      "grad_norm": 0.596678914952974,
+      "learning_rate": 4.622264265268107e-06,
+      "loss": 0.396,
+      "step": 2204
+    },
+    {
+      "epoch": 1.7155082987551866,
+      "grad_norm": 0.49249679820113956,
+      "learning_rate": 4.61774926554222e-06,
+      "loss": 0.3812,
+      "step": 2205
+    },
+    {
+      "epoch": 1.716286307053942,
+      "grad_norm": 0.5515894982350646,
+      "learning_rate": 4.6132345793177e-06,
+      "loss": 0.4203,
+      "step": 2206
+    },
+    {
+      "epoch": 1.717064315352697,
+      "grad_norm": 0.583325476747495,
+      "learning_rate": 4.608720210297254e-06,
+      "loss": 0.3783,
+      "step": 2207
+    },
+    {
+      "epoch": 1.7178423236514524,
+      "grad_norm": 0.517830767405429,
+      "learning_rate": 4.604206162183323e-06,
+      "loss": 0.4085,
+      "step": 2208
+    },
+    {
+      "epoch": 1.7186203319502074,
+      "grad_norm": 0.5554428099016708,
+      "learning_rate": 4.599692438678082e-06,
+      "loss": 0.4113,
+      "step": 2209
+    },
+    {
+      "epoch": 1.7193983402489628,
+      "grad_norm": 0.574077269297451,
+      "learning_rate": 4.5951790434834466e-06,
+      "loss": 0.3993,
+      "step": 2210
+    },
+    {
+      "epoch": 1.7201763485477177,
+      "grad_norm": 0.521567190980388,
+      "learning_rate": 4.5906659803010565e-06,
+      "loss": 0.3706,
+      "step": 2211
+    },
+    {
+      "epoch": 1.7209543568464731,
+      "grad_norm": 0.5433536991423498,
+      "learning_rate": 4.5861532528322815e-06,
+      "loss": 0.4033,
+      "step": 2212
+    },
+    {
+      "epoch": 1.721732365145228,
+      "grad_norm": 0.5467254562075335,
+      "learning_rate": 4.581640864778219e-06,
+      "loss": 0.413,
+      "step": 2213
+    },
+    {
+      "epoch": 1.7225103734439835,
+      "grad_norm": 0.5978865527801327,
+      "learning_rate": 4.577128819839683e-06,
+      "loss": 0.4392,
+      "step": 2214
+    },
+    {
+      "epoch": 1.7232883817427385,
+      "grad_norm": 0.5626800844569438,
+      "learning_rate": 4.572617121717211e-06,
+      "loss": 0.3659,
+      "step": 2215
+    },
+    {
+      "epoch": 1.7240663900414939,
+      "grad_norm": 0.5260077489514698,
+      "learning_rate": 4.568105774111053e-06,
+      "loss": 0.3996,
+      "step": 2216
+    },
+    {
+      "epoch": 1.7248443983402488,
+      "grad_norm": 0.5113003828253708,
+      "learning_rate": 4.563594780721172e-06,
+      "loss": 0.3813,
+      "step": 2217
+    },
+    {
+      "epoch": 1.7256224066390042,
+      "grad_norm": 0.6061087788159928,
+      "learning_rate": 4.559084145247239e-06,
+      "loss": 0.4691,
+      "step": 2218
+    },
+    {
+      "epoch": 1.7264004149377592,
+      "grad_norm": 0.5389428129675444,
+      "learning_rate": 4.554573871388637e-06,
+      "loss": 0.4172,
+      "step": 2219
+    },
+    {
+      "epoch": 1.7271784232365146,
+      "grad_norm": 0.5881816063550038,
+      "learning_rate": 4.550063962844445e-06,
+      "loss": 0.3933,
+      "step": 2220
+    },
+    {
+      "epoch": 1.7279564315352696,
+      "grad_norm": 0.5745824213099524,
+      "learning_rate": 4.545554423313446e-06,
+      "loss": 0.4226,
+      "step": 2221
+    },
+    {
+      "epoch": 1.728734439834025,
+      "grad_norm": 0.5146106560226387,
+      "learning_rate": 4.5410452564941235e-06,
+      "loss": 0.3645,
+      "step": 2222
+    },
+    {
+      "epoch": 1.72951244813278,
+      "grad_norm": 0.5572731663364356,
+      "learning_rate": 4.536536466084649e-06,
+      "loss": 0.4211,
+      "step": 2223
+    },
+    {
+      "epoch": 1.7302904564315353,
+      "grad_norm": 0.515307393331826,
+      "learning_rate": 4.532028055782892e-06,
+      "loss": 0.3885,
+      "step": 2224
+    },
+    {
+      "epoch": 1.7310684647302903,
+      "grad_norm": 0.5376655914727702,
+      "learning_rate": 4.527520029286402e-06,
+      "loss": 0.4226,
+      "step": 2225
+    },
+    {
+      "epoch": 1.7318464730290457,
+      "grad_norm": 0.5571654575398322,
+      "learning_rate": 4.523012390292424e-06,
+      "loss": 0.3711,
+      "step": 2226
+    },
+    {
+      "epoch": 1.7326244813278007,
+      "grad_norm": 0.5081922124814116,
+      "learning_rate": 4.5185051424978735e-06,
+      "loss": 0.3654,
+      "step": 2227
+    },
+    {
+      "epoch": 1.733402489626556,
+      "grad_norm": 0.6043645437083652,
+      "learning_rate": 4.513998289599355e-06,
+      "loss": 0.4614,
+      "step": 2228
+    },
+    {
+      "epoch": 1.734180497925311,
+      "grad_norm": 0.5413212230946557,
+      "learning_rate": 4.509491835293146e-06,
+      "loss": 0.4195,
+      "step": 2229
+    },
+    {
+      "epoch": 1.7349585062240664,
+      "grad_norm": 0.5019547771130523,
+      "learning_rate": 4.504985783275193e-06,
+      "loss": 0.3575,
+      "step": 2230
+    },
+    {
+      "epoch": 1.7357365145228216,
+      "grad_norm": 0.5470247993517471,
+      "learning_rate": 4.50048013724112e-06,
+      "loss": 0.3757,
+      "step": 2231
+    },
+    {
+      "epoch": 1.7365145228215768,
+      "grad_norm": 0.574308189217535,
+      "learning_rate": 4.4959749008862114e-06,
+      "loss": 0.4158,
+      "step": 2232
+    },
+    {
+      "epoch": 1.737292531120332,
+      "grad_norm": 0.5585431270877467,
+      "learning_rate": 4.491470077905421e-06,
+      "loss": 0.326,
+      "step": 2233
+    },
+    {
+      "epoch": 1.7380705394190872,
+      "grad_norm": 0.5940925564328631,
+      "learning_rate": 4.486965671993358e-06,
+      "loss": 0.4148,
+      "step": 2234
+    },
+    {
+      "epoch": 1.7388485477178424,
+      "grad_norm": 0.5623959337595166,
+      "learning_rate": 4.482461686844293e-06,
+      "loss": 0.3992,
+      "step": 2235
+    },
+    {
+      "epoch": 1.7396265560165975,
+      "grad_norm": 0.558044809369712,
+      "learning_rate": 4.4779581261521535e-06,
+      "loss": 0.39,
+      "step": 2236
+    },
+    {
+      "epoch": 1.7404045643153527,
+      "grad_norm": 0.5371928720803721,
+      "learning_rate": 4.473454993610513e-06,
+      "loss": 0.4029,
+      "step": 2237
+    },
+    {
+      "epoch": 1.741182572614108,
+      "grad_norm": 0.5528334264134458,
+      "learning_rate": 4.468952292912599e-06,
+      "loss": 0.4059,
+      "step": 2238
+    },
+    {
+      "epoch": 1.741960580912863,
+      "grad_norm": 0.5275029801240765,
+      "learning_rate": 4.464450027751281e-06,
+      "loss": 0.3769,
+      "step": 2239
+    },
+    {
+      "epoch": 1.7427385892116183,
+      "grad_norm": 0.5775700083856213,
+      "learning_rate": 4.459948201819077e-06,
+      "loss": 0.4002,
+      "step": 2240
+    },
+    {
+      "epoch": 1.7435165975103735,
+      "grad_norm": 0.5402179510227739,
+      "learning_rate": 4.455446818808137e-06,
+      "loss": 0.371,
+      "step": 2241
+    },
+    {
+      "epoch": 1.7442946058091287,
+      "grad_norm": 0.4876091576726739,
+      "learning_rate": 4.450945882410255e-06,
+      "loss": 0.3881,
+      "step": 2242
+    },
+    {
+      "epoch": 1.7450726141078838,
+      "grad_norm": 0.5466324038240191,
+      "learning_rate": 4.446445396316855e-06,
+      "loss": 0.3967,
+      "step": 2243
+    },
+    {
+      "epoch": 1.745850622406639,
+      "grad_norm": 0.577943353127805,
+      "learning_rate": 4.441945364218989e-06,
+      "loss": 0.4307,
+      "step": 2244
+    },
+    {
+      "epoch": 1.7466286307053942,
+      "grad_norm": 0.4897708720028192,
+      "learning_rate": 4.437445789807344e-06,
+      "loss": 0.3383,
+      "step": 2245
+    },
+    {
+      "epoch": 1.7474066390041494,
+      "grad_norm": 0.5005877428621953,
+      "learning_rate": 4.432946676772225e-06,
+      "loss": 0.4553,
+      "step": 2246
+    },
+    {
+      "epoch": 1.7481846473029046,
+      "grad_norm": 0.5462631101245773,
+      "learning_rate": 4.428448028803564e-06,
+      "loss": 0.44,
+      "step": 2247
+    },
+    {
+      "epoch": 1.7489626556016598,
+      "grad_norm": 0.5349076101332182,
+      "learning_rate": 4.423949849590903e-06,
+      "loss": 0.4154,
+      "step": 2248
+    },
+    {
+      "epoch": 1.749740663900415,
+      "grad_norm": 0.49989793708225994,
+      "learning_rate": 4.419452142823409e-06,
+      "loss": 0.4126,
+      "step": 2249
+    },
+    {
+      "epoch": 1.7505186721991701,
+      "grad_norm": 0.4748845165336068,
+      "learning_rate": 4.41495491218986e-06,
+      "loss": 0.346,
+      "step": 2250
+    },
+    {
+      "epoch": 1.7512966804979253,
+      "grad_norm": 0.5963588970710443,
+      "learning_rate": 4.410458161378638e-06,
+      "loss": 0.3998,
+      "step": 2251
+    },
+    {
+      "epoch": 1.7520746887966805,
+      "grad_norm": 0.63061346066848,
+      "learning_rate": 4.405961894077736e-06,
+      "loss": 0.4439,
+      "step": 2252
+    },
+    {
+      "epoch": 1.7528526970954357,
+      "grad_norm": 0.5114625103569146,
+      "learning_rate": 4.401466113974746e-06,
+      "loss": 0.3418,
+      "step": 2253
+    },
+    {
+      "epoch": 1.7536307053941909,
+      "grad_norm": 0.6182495769340287,
+      "learning_rate": 4.3969708247568684e-06,
+      "loss": 0.4431,
+      "step": 2254
+    },
+    {
+      "epoch": 1.754408713692946,
+      "grad_norm": 0.5323981884972765,
+      "learning_rate": 4.392476030110894e-06,
+      "loss": 0.3913,
+      "step": 2255
+    },
+    {
+      "epoch": 1.7551867219917012,
+      "grad_norm": 0.5906976868296338,
+      "learning_rate": 4.387981733723209e-06,
+      "loss": 0.3948,
+      "step": 2256
+    },
+    {
+      "epoch": 1.7559647302904564,
+      "grad_norm": 0.5057521166564412,
+      "learning_rate": 4.3834879392797945e-06,
+      "loss": 0.3195,
+      "step": 2257
+    },
+    {
+      "epoch": 1.7567427385892116,
+      "grad_norm": 0.5902515389188909,
+      "learning_rate": 4.378994650466215e-06,
+      "loss": 0.4603,
+      "step": 2258
+    },
+    {
+      "epoch": 1.7575207468879668,
+      "grad_norm": 0.538891918451021,
+      "learning_rate": 4.374501870967627e-06,
+      "loss": 0.3841,
+      "step": 2259
+    },
+    {
+      "epoch": 1.758298755186722,
+      "grad_norm": 0.5203287782929158,
+      "learning_rate": 4.37000960446876e-06,
+      "loss": 0.3582,
+      "step": 2260
+    },
+    {
+      "epoch": 1.7590767634854771,
+      "grad_norm": 0.5468413556172445,
+      "learning_rate": 4.36551785465393e-06,
+      "loss": 0.4023,
+      "step": 2261
+    },
+    {
+      "epoch": 1.7598547717842323,
+      "grad_norm": 0.47734379573947844,
+      "learning_rate": 4.361026625207031e-06,
+      "loss": 0.3541,
+      "step": 2262
+    },
+    {
+      "epoch": 1.7606327800829875,
+      "grad_norm": 0.46624562681076925,
+      "learning_rate": 4.3565359198115195e-06,
+      "loss": 0.3316,
+      "step": 2263
+    },
+    {
+      "epoch": 1.7614107883817427,
+      "grad_norm": 0.604252565851173,
+      "learning_rate": 4.352045742150435e-06,
+      "loss": 0.4052,
+      "step": 2264
+    },
+    {
+      "epoch": 1.7621887966804979,
+      "grad_norm": 0.5195695829894028,
+      "learning_rate": 4.347556095906374e-06,
+      "loss": 0.4168,
+      "step": 2265
+    },
+    {
+      "epoch": 1.762966804979253,
+      "grad_norm": 0.5651302600974065,
+      "learning_rate": 4.343066984761505e-06,
+      "loss": 0.4143,
+      "step": 2266
+    },
+    {
+      "epoch": 1.7637448132780082,
+      "grad_norm": 0.49011257783141626,
+      "learning_rate": 4.33857841239755e-06,
+      "loss": 0.3503,
+      "step": 2267
+    },
+    {
+      "epoch": 1.7645228215767634,
+      "grad_norm": 0.5647849825743663,
+      "learning_rate": 4.334090382495797e-06,
+      "loss": 0.4011,
+      "step": 2268
+    },
+    {
+      "epoch": 1.7653008298755186,
+      "grad_norm": 0.4980126266066467,
+      "learning_rate": 4.329602898737086e-06,
+      "loss": 0.4073,
+      "step": 2269
+    },
+    {
+      "epoch": 1.7660788381742738,
+      "grad_norm": 0.5676719987868803,
+      "learning_rate": 4.3251159648018024e-06,
+      "loss": 0.3937,
+      "step": 2270
+    },
+    {
+      "epoch": 1.7668568464730292,
+      "grad_norm": 0.6087512407454029,
+      "learning_rate": 4.320629584369893e-06,
+      "loss": 0.3728,
+      "step": 2271
+    },
+    {
+      "epoch": 1.7676348547717842,
+      "grad_norm": 0.5426026110074803,
+      "learning_rate": 4.316143761120841e-06,
+      "loss": 0.4158,
+      "step": 2272
+    },
+    {
+      "epoch": 1.7684128630705396,
+      "grad_norm": 0.5342423582210432,
+      "learning_rate": 4.311658498733678e-06,
+      "loss": 0.3639,
+      "step": 2273
+    },
+    {
+      "epoch": 1.7691908713692945,
+      "grad_norm": 0.5115516943705564,
+      "learning_rate": 4.3071738008869696e-06,
+      "loss": 0.3843,
+      "step": 2274
+    },
+    {
+      "epoch": 1.76996887966805,
+      "grad_norm": 0.5848636973356229,
+      "learning_rate": 4.302689671258826e-06,
+      "loss": 0.4566,
+      "step": 2275
+    },
+    {
+      "epoch": 1.770746887966805,
+      "grad_norm": 0.48606180582528424,
+      "learning_rate": 4.298206113526889e-06,
+      "loss": 0.372,
+      "step": 2276
+    },
+    {
+      "epoch": 1.7715248962655603,
+      "grad_norm": 0.4244589686053297,
+      "learning_rate": 4.293723131368325e-06,
+      "loss": 0.3344,
+      "step": 2277
+    },
+    {
+      "epoch": 1.7723029045643153,
+      "grad_norm": 0.5205349276859615,
+      "learning_rate": 4.289240728459838e-06,
+      "loss": 0.4106,
+      "step": 2278
+    },
+    {
+      "epoch": 1.7730809128630707,
+      "grad_norm": 0.5085234455471266,
+      "learning_rate": 4.284758908477649e-06,
+      "loss": 0.4564,
+      "step": 2279
+    },
+    {
+      "epoch": 1.7738589211618256,
+      "grad_norm": 0.41782740370059324,
+      "learning_rate": 4.280277675097509e-06,
+      "loss": 0.3199,
+      "step": 2280
+    },
+    {
+      "epoch": 1.774636929460581,
+      "grad_norm": 0.5142519680936156,
+      "learning_rate": 4.2757970319946764e-06,
+      "loss": 0.4129,
+      "step": 2281
+    },
+    {
+      "epoch": 1.775414937759336,
+      "grad_norm": 0.534737840326655,
+      "learning_rate": 4.271316982843936e-06,
+      "loss": 0.3918,
+      "step": 2282
+    },
+    {
+      "epoch": 1.7761929460580914,
+      "grad_norm": 0.47113423963377726,
+      "learning_rate": 4.266837531319584e-06,
+      "loss": 0.3249,
+      "step": 2283
+    },
+    {
+      "epoch": 1.7769709543568464,
+      "grad_norm": 0.5414478110092273,
+      "learning_rate": 4.2623586810954195e-06,
+      "loss": 0.4428,
+      "step": 2284
+    },
+    {
+      "epoch": 1.7777489626556018,
+      "grad_norm": 0.49701463246343175,
+      "learning_rate": 4.257880435844758e-06,
+      "loss": 0.3619,
+      "step": 2285
+    },
+    {
+      "epoch": 1.7785269709543567,
+      "grad_norm": 0.5032513157033761,
+      "learning_rate": 4.2534027992404094e-06,
+      "loss": 0.3848,
+      "step": 2286
+    },
+    {
+      "epoch": 1.7793049792531122,
+      "grad_norm": 0.5403558286477843,
+      "learning_rate": 4.248925774954693e-06,
+      "loss": 0.3721,
+      "step": 2287
+    },
+    {
+      "epoch": 1.7800829875518671,
+      "grad_norm": 0.5666898945834319,
+      "learning_rate": 4.244449366659417e-06,
+      "loss": 0.4152,
+      "step": 2288
+    },
+    {
+      "epoch": 1.7808609958506225,
+      "grad_norm": 0.6097381590154699,
+      "learning_rate": 4.2399735780258935e-06,
+      "loss": 0.4292,
+      "step": 2289
+    },
+    {
+      "epoch": 1.7816390041493775,
+      "grad_norm": 0.520864109387451,
+      "learning_rate": 4.235498412724922e-06,
+      "loss": 0.3551,
+      "step": 2290
+    },
+    {
+      "epoch": 1.7824170124481329,
+      "grad_norm": 0.6571306970791058,
+      "learning_rate": 4.2310238744267875e-06,
+      "loss": 0.3932,
+      "step": 2291
+    },
+    {
+      "epoch": 1.7831950207468878,
+      "grad_norm": 0.6032550447134705,
+      "learning_rate": 4.226549966801268e-06,
+      "loss": 0.4156,
+      "step": 2292
+    },
+    {
+      "epoch": 1.7839730290456433,
+      "grad_norm": 0.5255016721376868,
+      "learning_rate": 4.2220766935176174e-06,
+      "loss": 0.4254,
+      "step": 2293
+    },
+    {
+      "epoch": 1.7847510373443982,
+      "grad_norm": 0.5516234524389213,
+      "learning_rate": 4.217604058244577e-06,
+      "loss": 0.4478,
+      "step": 2294
+    },
+    {
+      "epoch": 1.7855290456431536,
+      "grad_norm": 0.44041293826862404,
+      "learning_rate": 4.213132064650355e-06,
+      "loss": 0.3332,
+      "step": 2295
+    },
+    {
+      "epoch": 1.7863070539419086,
+      "grad_norm": 0.5157755032915077,
+      "learning_rate": 4.20866071640264e-06,
+      "loss": 0.369,
+      "step": 2296
+    },
+    {
+      "epoch": 1.787085062240664,
+      "grad_norm": 0.6575400540613355,
+      "learning_rate": 4.204190017168593e-06,
+      "loss": 0.4172,
+      "step": 2297
+    },
+    {
+      "epoch": 1.787863070539419,
+      "grad_norm": 0.5396017512671248,
+      "learning_rate": 4.199719970614836e-06,
+      "loss": 0.414,
+      "step": 2298
+    },
+    {
+      "epoch": 1.7886410788381744,
+      "grad_norm": 0.5692934818455175,
+      "learning_rate": 4.195250580407463e-06,
+      "loss": 0.4239,
+      "step": 2299
+    },
+    {
+      "epoch": 1.7894190871369293,
+      "grad_norm": 0.6249278903918376,
+      "learning_rate": 4.19078185021202e-06,
+      "loss": 0.3954,
+      "step": 2300
+    },
+    {
+      "epoch": 1.7901970954356847,
+      "grad_norm": 0.46457596760442693,
+      "learning_rate": 4.186313783693525e-06,
+      "loss": 0.3664,
+      "step": 2301
+    },
+    {
+      "epoch": 1.7909751037344397,
+      "grad_norm": 0.6055029796726691,
+      "learning_rate": 4.181846384516439e-06,
+      "loss": 0.4191,
+      "step": 2302
+    },
+    {
+      "epoch": 1.791753112033195,
+      "grad_norm": 0.5601663987155778,
+      "learning_rate": 4.177379656344682e-06,
+      "loss": 0.4061,
+      "step": 2303
+    },
+    {
+      "epoch": 1.79253112033195,
+      "grad_norm": 0.468091294876262,
+      "learning_rate": 4.172913602841627e-06,
+      "loss": 0.3383,
+      "step": 2304
+    },
+    {
+      "epoch": 1.7933091286307055,
+      "grad_norm": 0.5130263302208022,
+      "learning_rate": 4.168448227670083e-06,
+      "loss": 0.4024,
+      "step": 2305
+    },
+    {
+      "epoch": 1.7940871369294604,
+      "grad_norm": 0.6185459884380178,
+      "learning_rate": 4.163983534492314e-06,
+      "loss": 0.384,
+      "step": 2306
+    },
+    {
+      "epoch": 1.7948651452282158,
+      "grad_norm": 0.5565623545478996,
+      "learning_rate": 4.159519526970018e-06,
+      "loss": 0.4063,
+      "step": 2307
+    },
+    {
+      "epoch": 1.7956431535269708,
+      "grad_norm": 0.5598631111207472,
+      "learning_rate": 4.155056208764334e-06,
+      "loss": 0.4106,
+      "step": 2308
+    },
+    {
+      "epoch": 1.7964211618257262,
+      "grad_norm": 0.5424598650605104,
+      "learning_rate": 4.15059358353583e-06,
+      "loss": 0.3882,
+      "step": 2309
+    },
+    {
+      "epoch": 1.7971991701244814,
+      "grad_norm": 0.5149326709743707,
+      "learning_rate": 4.146131654944514e-06,
+      "loss": 0.3853,
+      "step": 2310
+    },
+    {
+      "epoch": 1.7979771784232366,
+      "grad_norm": 0.5886300178790226,
+      "learning_rate": 4.141670426649817e-06,
+      "loss": 0.4057,
+      "step": 2311
+    },
+    {
+      "epoch": 1.7987551867219918,
+      "grad_norm": 0.4691253420351814,
+      "learning_rate": 4.137209902310596e-06,
+      "loss": 0.3469,
+      "step": 2312
+    },
+    {
+      "epoch": 1.799533195020747,
+      "grad_norm": 0.5225666539124797,
+      "learning_rate": 4.132750085585134e-06,
+      "loss": 0.3918,
+      "step": 2313
+    },
+    {
+      "epoch": 1.8003112033195021,
+      "grad_norm": 0.5473467034843277,
+      "learning_rate": 4.128290980131129e-06,
+      "loss": 0.4187,
+      "step": 2314
+    },
+    {
+      "epoch": 1.8010892116182573,
+      "grad_norm": 0.5312703530767225,
+      "learning_rate": 4.123832589605701e-06,
+      "loss": 0.3473,
+      "step": 2315
+    },
+    {
+      "epoch": 1.8018672199170125,
+      "grad_norm": 0.5537495168053346,
+      "learning_rate": 4.119374917665377e-06,
+      "loss": 0.3953,
+      "step": 2316
+    },
+    {
+      "epoch": 1.8026452282157677,
+      "grad_norm": 0.5432405101951062,
+      "learning_rate": 4.114917967966099e-06,
+      "loss": 0.391,
+      "step": 2317
+    },
+    {
+      "epoch": 1.8034232365145229,
+      "grad_norm": 0.575384278028712,
+      "learning_rate": 4.11046174416322e-06,
+      "loss": 0.4082,
+      "step": 2318
+    },
+    {
+      "epoch": 1.804201244813278,
+      "grad_norm": 0.5526157757947535,
+      "learning_rate": 4.106006249911489e-06,
+      "loss": 0.3991,
+      "step": 2319
+    },
+    {
+      "epoch": 1.8049792531120332,
+      "grad_norm": 0.564621353804366,
+      "learning_rate": 4.1015514888650655e-06,
+      "loss": 0.4054,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8057572614107884,
+      "grad_norm": 0.5387993953813537,
+      "learning_rate": 4.097097464677499e-06,
+      "loss": 0.3351,
+      "step": 2321
+    },
+    {
+      "epoch": 1.8065352697095436,
+      "grad_norm": 0.6523820727467549,
+      "learning_rate": 4.092644181001742e-06,
+      "loss": 0.4087,
+      "step": 2322
+    },
+    {
+      "epoch": 1.8073132780082988,
+      "grad_norm": 0.5862589605632227,
+      "learning_rate": 4.0881916414901345e-06,
+      "loss": 0.4472,
+      "step": 2323
+    },
+    {
+      "epoch": 1.808091286307054,
+      "grad_norm": 0.5450859206838072,
+      "learning_rate": 4.083739849794412e-06,
+      "loss": 0.363,
+      "step": 2324
+    },
+    {
+      "epoch": 1.8088692946058091,
+      "grad_norm": 0.551956340637267,
+      "learning_rate": 4.079288809565692e-06,
+      "loss": 0.3786,
+      "step": 2325
+    },
+    {
+      "epoch": 1.8096473029045643,
+      "grad_norm": 0.5380197821615292,
+      "learning_rate": 4.074838524454474e-06,
+      "loss": 0.3694,
+      "step": 2326
+    },
+    {
+      "epoch": 1.8104253112033195,
+      "grad_norm": 0.5348170221434071,
+      "learning_rate": 4.0703889981106455e-06,
+      "loss": 0.39,
+      "step": 2327
+    },
+    {
+      "epoch": 1.8112033195020747,
+      "grad_norm": 0.5540329291232371,
+      "learning_rate": 4.0659402341834635e-06,
+      "loss": 0.3939,
+      "step": 2328
+    },
+    {
+      "epoch": 1.8119813278008299,
+      "grad_norm": 0.5936682460265076,
+      "learning_rate": 4.061492236321566e-06,
+      "loss": 0.4489,
+      "step": 2329
+    },
+    {
+      "epoch": 1.812759336099585,
+      "grad_norm": 0.5656814655785755,
+      "learning_rate": 4.057045008172958e-06,
+      "loss": 0.377,
+      "step": 2330
+    },
+    {
+      "epoch": 1.8135373443983402,
+      "grad_norm": 0.5550516969565056,
+      "learning_rate": 4.052598553385016e-06,
+      "loss": 0.4239,
+      "step": 2331
+    },
+    {
+      "epoch": 1.8143153526970954,
+      "grad_norm": 0.558464835239899,
+      "learning_rate": 4.048152875604485e-06,
+      "loss": 0.3955,
+      "step": 2332
+    },
+    {
+      "epoch": 1.8150933609958506,
+      "grad_norm": 0.47402703995124423,
+      "learning_rate": 4.0437079784774656e-06,
+      "loss": 0.3841,
+      "step": 2333
+    },
+    {
+      "epoch": 1.8158713692946058,
+      "grad_norm": 0.5100326254843693,
+      "learning_rate": 4.039263865649424e-06,
+      "loss": 0.3559,
+      "step": 2334
+    },
+    {
+      "epoch": 1.816649377593361,
+      "grad_norm": 0.565201340667645,
+      "learning_rate": 4.034820540765178e-06,
+      "loss": 0.3943,
+      "step": 2335
+    },
+    {
+      "epoch": 1.8174273858921162,
+      "grad_norm": 0.568203490538433,
+      "learning_rate": 4.0303780074689055e-06,
+      "loss": 0.4097,
+      "step": 2336
+    },
+    {
+      "epoch": 1.8182053941908713,
+      "grad_norm": 0.5076445314517091,
+      "learning_rate": 4.0259362694041305e-06,
+      "loss": 0.4204,
+      "step": 2337
+    },
+    {
+      "epoch": 1.8189834024896265,
+      "grad_norm": 0.5298447282325358,
+      "learning_rate": 4.021495330213725e-06,
+      "loss": 0.4127,
+      "step": 2338
+    },
+    {
+      "epoch": 1.8197614107883817,
+      "grad_norm": 0.5169838786333888,
+      "learning_rate": 4.017055193539912e-06,
+      "loss": 0.3987,
+      "step": 2339
+    },
+    {
+      "epoch": 1.820539419087137,
+      "grad_norm": 0.5612765632439634,
+      "learning_rate": 4.0126158630242454e-06,
+      "loss": 0.397,
+      "step": 2340
+    },
+    {
+      "epoch": 1.821317427385892,
+      "grad_norm": 0.4905628650846357,
+      "learning_rate": 4.008177342307627e-06,
+      "loss": 0.3374,
+      "step": 2341
+    },
+    {
+      "epoch": 1.8220954356846473,
+      "grad_norm": 0.6083178526239025,
+      "learning_rate": 4.003739635030292e-06,
+      "loss": 0.4733,
+      "step": 2342
+    },
+    {
+      "epoch": 1.8228734439834025,
+      "grad_norm": 0.5640353430773815,
+      "learning_rate": 3.999302744831805e-06,
+      "loss": 0.3807,
+      "step": 2343
+    },
+    {
+      "epoch": 1.8236514522821576,
+      "grad_norm": 0.4538145328483939,
+      "learning_rate": 3.994866675351066e-06,
+      "loss": 0.3452,
+      "step": 2344
+    },
+    {
+      "epoch": 1.8244294605809128,
+      "grad_norm": 0.5601168075962589,
+      "learning_rate": 3.990431430226298e-06,
+      "loss": 0.4259,
+      "step": 2345
+    },
+    {
+      "epoch": 1.825207468879668,
+      "grad_norm": 0.5606207998360712,
+      "learning_rate": 3.98599701309505e-06,
+      "loss": 0.4176,
+      "step": 2346
+    },
+    {
+      "epoch": 1.8259854771784232,
+      "grad_norm": 0.5670620531511132,
+      "learning_rate": 3.981563427594189e-06,
+      "loss": 0.4017,
+      "step": 2347
+    },
+    {
+      "epoch": 1.8267634854771784,
+      "grad_norm": 0.5331544014500127,
+      "learning_rate": 3.977130677359904e-06,
+      "loss": 0.3986,
+      "step": 2348
+    },
+    {
+      "epoch": 1.8275414937759336,
+      "grad_norm": 0.5548877600622228,
+      "learning_rate": 3.972698766027695e-06,
+      "loss": 0.3877,
+      "step": 2349
+    },
+    {
+      "epoch": 1.828319502074689,
+      "grad_norm": 0.5763195423570318,
+      "learning_rate": 3.968267697232377e-06,
+      "loss": 0.4158,
+      "step": 2350
+    },
+    {
+      "epoch": 1.829097510373444,
+      "grad_norm": 0.5032725839438968,
+      "learning_rate": 3.963837474608075e-06,
+      "loss": 0.3594,
+      "step": 2351
+    },
+    {
+      "epoch": 1.8298755186721993,
+      "grad_norm": 0.583118758731178,
+      "learning_rate": 3.959408101788213e-06,
+      "loss": 0.448,
+      "step": 2352
+    },
+    {
+      "epoch": 1.8306535269709543,
+      "grad_norm": 0.49370234810382146,
+      "learning_rate": 3.954979582405528e-06,
+      "loss": 0.3429,
+      "step": 2353
+    },
+    {
+      "epoch": 1.8314315352697097,
+      "grad_norm": 0.5224561320470869,
+      "learning_rate": 3.950551920092049e-06,
+      "loss": 0.375,
+      "step": 2354
+    },
+    {
+      "epoch": 1.8322095435684647,
+      "grad_norm": 0.588644047523422,
+      "learning_rate": 3.9461251184791065e-06,
+      "loss": 0.4255,
+      "step": 2355
+    },
+    {
+      "epoch": 1.83298755186722,
+      "grad_norm": 0.5322642422085027,
+      "learning_rate": 3.941699181197321e-06,
+      "loss": 0.4016,
+      "step": 2356
+    },
+    {
+      "epoch": 1.833765560165975,
+      "grad_norm": 0.49897353781566667,
+      "learning_rate": 3.937274111876608e-06,
+      "loss": 0.3847,
+      "step": 2357
+    },
+    {
+      "epoch": 1.8345435684647304,
+      "grad_norm": 0.5219337819849348,
+      "learning_rate": 3.932849914146173e-06,
+      "loss": 0.378,
+      "step": 2358
+    },
+    {
+      "epoch": 1.8353215767634854,
+      "grad_norm": 0.5870882483908679,
+      "learning_rate": 3.9284265916345005e-06,
+      "loss": 0.402,
+      "step": 2359
+    },
+    {
+      "epoch": 1.8360995850622408,
+      "grad_norm": 0.501896380960453,
+      "learning_rate": 3.924004147969362e-06,
+      "loss": 0.3639,
+      "step": 2360
+    },
+    {
+      "epoch": 1.8368775933609958,
+      "grad_norm": 0.5457870332352799,
+      "learning_rate": 3.919582586777804e-06,
+      "loss": 0.3986,
+      "step": 2361
+    },
+    {
+      "epoch": 1.8376556016597512,
+      "grad_norm": 0.5886313147212028,
+      "learning_rate": 3.915161911686156e-06,
+      "loss": 0.4278,
+      "step": 2362
+    },
+    {
+      "epoch": 1.8384336099585061,
+      "grad_norm": 0.4791025143261753,
+      "learning_rate": 3.910742126320012e-06,
+      "loss": 0.3787,
+      "step": 2363
+    },
+    {
+      "epoch": 1.8392116182572615,
+      "grad_norm": 0.5155409464153237,
+      "learning_rate": 3.906323234304241e-06,
+      "loss": 0.3822,
+      "step": 2364
+    },
+    {
+      "epoch": 1.8399896265560165,
+      "grad_norm": 0.5642266814236212,
+      "learning_rate": 3.901905239262984e-06,
+      "loss": 0.4019,
+      "step": 2365
+    },
+    {
+      "epoch": 1.840767634854772,
+      "grad_norm": 0.4910592023008419,
+      "learning_rate": 3.897488144819637e-06,
+      "loss": 0.3398,
+      "step": 2366
+    },
+    {
+      "epoch": 1.8415456431535269,
+      "grad_norm": 0.554518111752704,
+      "learning_rate": 3.893071954596864e-06,
+      "loss": 0.4499,
+      "step": 2367
+    },
+    {
+      "epoch": 1.8423236514522823,
+      "grad_norm": 0.5418525430279066,
+      "learning_rate": 3.888656672216584e-06,
+      "loss": 0.3377,
+      "step": 2368
+    },
+    {
+      "epoch": 1.8431016597510372,
+      "grad_norm": 0.6888952554957641,
+      "learning_rate": 3.884242301299975e-06,
+      "loss": 0.479,
+      "step": 2369
+    },
+    {
+      "epoch": 1.8438796680497926,
+      "grad_norm": 0.5965256792775628,
+      "learning_rate": 3.8798288454674606e-06,
+      "loss": 0.4339,
+      "step": 2370
+    },
+    {
+      "epoch": 1.8446576763485476,
+      "grad_norm": 0.48934992388607995,
+      "learning_rate": 3.875416308338723e-06,
+      "loss": 0.357,
+      "step": 2371
+    },
+    {
+      "epoch": 1.845435684647303,
+      "grad_norm": 0.5556177335510989,
+      "learning_rate": 3.871004693532685e-06,
+      "loss": 0.3904,
+      "step": 2372
+    },
+    {
+      "epoch": 1.846213692946058,
+      "grad_norm": 0.556183650328025,
+      "learning_rate": 3.866594004667513e-06,
+      "loss": 0.4246,
+      "step": 2373
+    },
+    {
+      "epoch": 1.8469917012448134,
+      "grad_norm": 0.49985245458059463,
+      "learning_rate": 3.862184245360619e-06,
+      "loss": 0.3402,
+      "step": 2374
+    },
+    {
+      "epoch": 1.8477697095435683,
+      "grad_norm": 0.5614950863752375,
+      "learning_rate": 3.8577754192286434e-06,
+      "loss": 0.3898,
+      "step": 2375
+    },
+    {
+      "epoch": 1.8485477178423237,
+      "grad_norm": 0.5284084336663669,
+      "learning_rate": 3.8533675298874725e-06,
+      "loss": 0.3864,
+      "step": 2376
+    },
+    {
+      "epoch": 1.8493257261410787,
+      "grad_norm": 0.4950477682333912,
+      "learning_rate": 3.8489605809522154e-06,
+      "loss": 0.3593,
+      "step": 2377
+    },
+    {
+      "epoch": 1.8501037344398341,
+      "grad_norm": 0.5730071222715888,
+      "learning_rate": 3.844554576037212e-06,
+      "loss": 0.4062,
+      "step": 2378
+    },
+    {
+      "epoch": 1.850881742738589,
+      "grad_norm": 0.4828419488764784,
+      "learning_rate": 3.840149518756034e-06,
+      "loss": 0.3581,
+      "step": 2379
+    },
+    {
+      "epoch": 1.8516597510373445,
+      "grad_norm": 0.5375620875423266,
+      "learning_rate": 3.8357454127214675e-06,
+      "loss": 0.4437,
+      "step": 2380
+    },
+    {
+      "epoch": 1.8524377593360994,
+      "grad_norm": 0.48058133032180655,
+      "learning_rate": 3.831342261545522e-06,
+      "loss": 0.3629,
+      "step": 2381
+    },
+    {
+      "epoch": 1.8532157676348548,
+      "grad_norm": 0.5085076902889752,
+      "learning_rate": 3.826940068839424e-06,
+      "loss": 0.41,
+      "step": 2382
+    },
+    {
+      "epoch": 1.8539937759336098,
+      "grad_norm": 0.5019615579477165,
+      "learning_rate": 3.822538838213616e-06,
+      "loss": 0.3752,
+      "step": 2383
+    },
+    {
+      "epoch": 1.8547717842323652,
+      "grad_norm": 0.5966458770092311,
+      "learning_rate": 3.8181385732777455e-06,
+      "loss": 0.3788,
+      "step": 2384
+    },
+    {
+      "epoch": 1.8555497925311202,
+      "grad_norm": 0.5300898378627517,
+      "learning_rate": 3.8137392776406738e-06,
+      "loss": 0.3612,
+      "step": 2385
+    },
+    {
+      "epoch": 1.8563278008298756,
+      "grad_norm": 0.5907994448389025,
+      "learning_rate": 3.8093409549104675e-06,
+      "loss": 0.4468,
+      "step": 2386
+    },
+    {
+      "epoch": 1.8571058091286305,
+      "grad_norm": 0.5111151396692519,
+      "learning_rate": 3.8049436086943874e-06,
+      "loss": 0.3876,
+      "step": 2387
+    },
+    {
+      "epoch": 1.857883817427386,
+      "grad_norm": 0.4828434810829184,
+      "learning_rate": 3.8005472425989047e-06,
+      "loss": 0.4174,
+      "step": 2388
+    },
+    {
+      "epoch": 1.858661825726141,
+      "grad_norm": 0.4844299250205084,
+      "learning_rate": 3.7961518602296777e-06,
+      "loss": 0.3788,
+      "step": 2389
+    },
+    {
+      "epoch": 1.8594398340248963,
+      "grad_norm": 0.44218654177995786,
+      "learning_rate": 3.7917574651915625e-06,
+      "loss": 0.3189,
+      "step": 2390
+    },
+    {
+      "epoch": 1.8602178423236515,
+      "grad_norm": 0.5179258560443935,
+      "learning_rate": 3.7873640610886012e-06,
+      "loss": 0.3864,
+      "step": 2391
+    },
+    {
+      "epoch": 1.8609958506224067,
+      "grad_norm": 0.5304809772906861,
+      "learning_rate": 3.78297165152403e-06,
+      "loss": 0.3885,
+      "step": 2392
+    },
+    {
+      "epoch": 1.8617738589211619,
+      "grad_norm": 0.500140307995282,
+      "learning_rate": 3.7785802401002654e-06,
+      "loss": 0.3656,
+      "step": 2393
+    },
+    {
+      "epoch": 1.862551867219917,
+      "grad_norm": 0.5639610305197998,
+      "learning_rate": 3.774189830418902e-06,
+      "loss": 0.3741,
+      "step": 2394
+    },
+    {
+      "epoch": 1.8633298755186722,
+      "grad_norm": 0.5529748553180547,
+      "learning_rate": 3.7698004260807197e-06,
+      "loss": 0.3855,
+      "step": 2395
+    },
+    {
+      "epoch": 1.8641078838174274,
+      "grad_norm": 0.48423102657171724,
+      "learning_rate": 3.765412030685668e-06,
+      "loss": 0.3729,
+      "step": 2396
+    },
+    {
+      "epoch": 1.8648858921161826,
+      "grad_norm": 0.5738460484784457,
+      "learning_rate": 3.7610246478328737e-06,
+      "loss": 0.4425,
+      "step": 2397
+    },
+    {
+      "epoch": 1.8656639004149378,
+      "grad_norm": 0.5073288949477371,
+      "learning_rate": 3.7566382811206294e-06,
+      "loss": 0.3853,
+      "step": 2398
+    },
+    {
+      "epoch": 1.866441908713693,
+      "grad_norm": 0.6024234049909809,
+      "learning_rate": 3.7522529341463946e-06,
+      "loss": 0.4547,
+      "step": 2399
+    },
+    {
+      "epoch": 1.8672199170124482,
+      "grad_norm": 0.5036803150493163,
+      "learning_rate": 3.747868610506798e-06,
+      "loss": 0.4188,
+      "step": 2400
+    },
+    {
+      "epoch": 1.8679979253112033,
+      "grad_norm": 0.546520255968725,
+      "learning_rate": 3.7434853137976202e-06,
+      "loss": 0.3713,
+      "step": 2401
+    },
+    {
+      "epoch": 1.8687759336099585,
+      "grad_norm": 0.5629128106142152,
+      "learning_rate": 3.7391030476138063e-06,
+      "loss": 0.3859,
+      "step": 2402
+    },
+    {
+      "epoch": 1.8695539419087137,
+      "grad_norm": 0.5549446713672614,
+      "learning_rate": 3.7347218155494513e-06,
+      "loss": 0.3623,
+      "step": 2403
+    },
+    {
+      "epoch": 1.870331950207469,
+      "grad_norm": 0.5020766542901953,
+      "learning_rate": 3.7303416211978082e-06,
+      "loss": 0.4284,
+      "step": 2404
+    },
+    {
+      "epoch": 1.871109958506224,
+      "grad_norm": 0.5674728033748256,
+      "learning_rate": 3.7259624681512722e-06,
+      "loss": 0.3798,
+      "step": 2405
+    },
+    {
+      "epoch": 1.8718879668049793,
+      "grad_norm": 0.6330755020937452,
+      "learning_rate": 3.72158436000139e-06,
+      "loss": 0.4514,
+      "step": 2406
+    },
+    {
+      "epoch": 1.8726659751037344,
+      "grad_norm": 0.5312098208931773,
+      "learning_rate": 3.71720730033885e-06,
+      "loss": 0.352,
+      "step": 2407
+    },
+    {
+      "epoch": 1.8734439834024896,
+      "grad_norm": 0.48644783724736196,
+      "learning_rate": 3.7128312927534755e-06,
+      "loss": 0.3772,
+      "step": 2408
+    },
+    {
+      "epoch": 1.8742219917012448,
+      "grad_norm": 0.5158505736165422,
+      "learning_rate": 3.7084563408342357e-06,
+      "loss": 0.3906,
+      "step": 2409
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.5603210911401574,
+      "learning_rate": 3.7040824481692263e-06,
+      "loss": 0.4297,
+      "step": 2410
+    },
+    {
+      "epoch": 1.8757780082987552,
+      "grad_norm": 0.5612154950579507,
+      "learning_rate": 3.69970961834568e-06,
+      "loss": 0.4036,
+      "step": 2411
+    },
+    {
+      "epoch": 1.8765560165975104,
+      "grad_norm": 0.5344168996368373,
+      "learning_rate": 3.695337854949951e-06,
+      "loss": 0.4112,
+      "step": 2412
+    },
+    {
+      "epoch": 1.8773340248962656,
+      "grad_norm": 0.5799934195734067,
+      "learning_rate": 3.690967161567527e-06,
+      "loss": 0.4057,
+      "step": 2413
+    },
+    {
+      "epoch": 1.8781120331950207,
+      "grad_norm": 0.5287275792461299,
+      "learning_rate": 3.686597541783014e-06,
+      "loss": 0.393,
+      "step": 2414
+    },
+    {
+      "epoch": 1.878890041493776,
+      "grad_norm": 0.5279063568113497,
+      "learning_rate": 3.6822289991801363e-06,
+      "loss": 0.3809,
+      "step": 2415
+    },
+    {
+      "epoch": 1.879668049792531,
+      "grad_norm": 0.5668287985532174,
+      "learning_rate": 3.677861537341738e-06,
+      "loss": 0.3763,
+      "step": 2416
+    },
+    {
+      "epoch": 1.8804460580912863,
+      "grad_norm": 0.49879589625267545,
+      "learning_rate": 3.673495159849772e-06,
+      "loss": 0.3832,
+      "step": 2417
+    },
+    {
+      "epoch": 1.8812240663900415,
+      "grad_norm": 0.535052972480242,
+      "learning_rate": 3.6691298702853085e-06,
+      "loss": 0.4247,
+      "step": 2418
+    },
+    {
+      "epoch": 1.8820020746887967,
+      "grad_norm": 0.4982705002613858,
+      "learning_rate": 3.6647656722285185e-06,
+      "loss": 0.3609,
+      "step": 2419
+    },
+    {
+      "epoch": 1.8827800829875518,
+      "grad_norm": 0.5838789005678213,
+      "learning_rate": 3.6604025692586818e-06,
+      "loss": 0.4368,
+      "step": 2420
+    },
+    {
+      "epoch": 1.883558091286307,
+      "grad_norm": 0.511056481561275,
+      "learning_rate": 3.6560405649541826e-06,
+      "loss": 0.3962,
+      "step": 2421
+    },
+    {
+      "epoch": 1.8843360995850622,
+      "grad_norm": 0.48299372165887466,
+      "learning_rate": 3.6516796628924956e-06,
+      "loss": 0.3825,
+      "step": 2422
+    },
+    {
+      "epoch": 1.8851141078838174,
+      "grad_norm": 0.4956001952447112,
+      "learning_rate": 3.6473198666502025e-06,
+      "loss": 0.3778,
+      "step": 2423
+    },
+    {
+      "epoch": 1.8858921161825726,
+      "grad_norm": 0.5272800881549655,
+      "learning_rate": 3.6429611798029673e-06,
+      "loss": 0.3929,
+      "step": 2424
+    },
+    {
+      "epoch": 1.8866701244813278,
+      "grad_norm": 0.5330148291808138,
+      "learning_rate": 3.6386036059255512e-06,
+      "loss": 0.3788,
+      "step": 2425
+    },
+    {
+      "epoch": 1.887448132780083,
+      "grad_norm": 0.5055304452919412,
+      "learning_rate": 3.634247148591802e-06,
+      "loss": 0.3634,
+      "step": 2426
+    },
+    {
+      "epoch": 1.8882261410788381,
+      "grad_norm": 0.5744775492733354,
+      "learning_rate": 3.629891811374648e-06,
+      "loss": 0.4338,
+      "step": 2427
+    },
+    {
+      "epoch": 1.8890041493775933,
+      "grad_norm": 0.5145785832831375,
+      "learning_rate": 3.6255375978461027e-06,
+      "loss": 0.3876,
+      "step": 2428
+    },
+    {
+      "epoch": 1.8897821576763485,
+      "grad_norm": 0.548825563135965,
+      "learning_rate": 3.6211845115772538e-06,
+      "loss": 0.3726,
+      "step": 2429
+    },
+    {
+      "epoch": 1.8905601659751037,
+      "grad_norm": 0.4676142240018973,
+      "learning_rate": 3.616832556138271e-06,
+      "loss": 0.3193,
+      "step": 2430
+    },
+    {
+      "epoch": 1.891338174273859,
+      "grad_norm": 0.5596198875428283,
+      "learning_rate": 3.612481735098389e-06,
+      "loss": 0.4435,
+      "step": 2431
+    },
+    {
+      "epoch": 1.892116182572614,
+      "grad_norm": 0.4771075668791299,
+      "learning_rate": 3.608132052025918e-06,
+      "loss": 0.3654,
+      "step": 2432
+    },
+    {
+      "epoch": 1.8928941908713695,
+      "grad_norm": 0.49269879382212933,
+      "learning_rate": 3.6037835104882334e-06,
+      "loss": 0.3842,
+      "step": 2433
+    },
+    {
+      "epoch": 1.8936721991701244,
+      "grad_norm": 0.5338151382266385,
+      "learning_rate": 3.599436114051771e-06,
+      "loss": 0.4168,
+      "step": 2434
+    },
+    {
+      "epoch": 1.8944502074688798,
+      "grad_norm": 0.5136836224272903,
+      "learning_rate": 3.595089866282034e-06,
+      "loss": 0.3814,
+      "step": 2435
+    },
+    {
+      "epoch": 1.8952282157676348,
+      "grad_norm": 0.48715344136834215,
+      "learning_rate": 3.590744770743576e-06,
+      "loss": 0.4142,
+      "step": 2436
+    },
+    {
+      "epoch": 1.8960062240663902,
+      "grad_norm": 0.5615181927817199,
+      "learning_rate": 3.5864008310000113e-06,
+      "loss": 0.4215,
+      "step": 2437
+    },
+    {
+      "epoch": 1.8967842323651452,
+      "grad_norm": 0.517706057361045,
+      "learning_rate": 3.582058050614002e-06,
+      "loss": 0.373,
+      "step": 2438
+    },
+    {
+      "epoch": 1.8975622406639006,
+      "grad_norm": 0.5263208035928584,
+      "learning_rate": 3.577716433147263e-06,
+      "loss": 0.421,
+      "step": 2439
+    },
+    {
+      "epoch": 1.8983402489626555,
+      "grad_norm": 0.5208281496922639,
+      "learning_rate": 3.5733759821605562e-06,
+      "loss": 0.3662,
+      "step": 2440
+    },
+    {
+      "epoch": 1.899118257261411,
+      "grad_norm": 0.5213576248024293,
+      "learning_rate": 3.5690367012136824e-06,
+      "loss": 0.3875,
+      "step": 2441
+    },
+    {
+      "epoch": 1.8998962655601659,
+      "grad_norm": 0.5365597639672628,
+      "learning_rate": 3.5646985938654854e-06,
+      "loss": 0.3898,
+      "step": 2442
+    },
+    {
+      "epoch": 1.9006742738589213,
+      "grad_norm": 0.629812047464079,
+      "learning_rate": 3.5603616636738446e-06,
+      "loss": 0.448,
+      "step": 2443
+    },
+    {
+      "epoch": 1.9014522821576763,
+      "grad_norm": 0.5170729190127444,
+      "learning_rate": 3.55602591419568e-06,
+      "loss": 0.3824,
+      "step": 2444
+    },
+    {
+      "epoch": 1.9022302904564317,
+      "grad_norm": 0.5786767774994407,
+      "learning_rate": 3.5516913489869333e-06,
+      "loss": 0.4489,
+      "step": 2445
+    },
+    {
+      "epoch": 1.9030082987551866,
+      "grad_norm": 0.5362227651959804,
+      "learning_rate": 3.5473579716025824e-06,
+      "loss": 0.4125,
+      "step": 2446
+    },
+    {
+      "epoch": 1.903786307053942,
+      "grad_norm": 0.542048057540201,
+      "learning_rate": 3.5430257855966314e-06,
+      "loss": 0.3864,
+      "step": 2447
+    },
+    {
+      "epoch": 1.904564315352697,
+      "grad_norm": 0.594995367506249,
+      "learning_rate": 3.538694794522101e-06,
+      "loss": 0.4179,
+      "step": 2448
+    },
+    {
+      "epoch": 1.9053423236514524,
+      "grad_norm": 0.5670166152029226,
+      "learning_rate": 3.5343650019310395e-06,
+      "loss": 0.3922,
+      "step": 2449
+    },
+    {
+      "epoch": 1.9061203319502074,
+      "grad_norm": 0.48903643100305183,
+      "learning_rate": 3.530036411374506e-06,
+      "loss": 0.3798,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9068983402489628,
+      "grad_norm": 0.5621720871930199,
+      "learning_rate": 3.5257090264025777e-06,
+      "loss": 0.4177,
+      "step": 2451
+    },
+    {
+      "epoch": 1.9076763485477177,
+      "grad_norm": 0.5509345480823824,
+      "learning_rate": 3.5213828505643388e-06,
+      "loss": 0.3757,
+      "step": 2452
+    },
+    {
+      "epoch": 1.9084543568464731,
+      "grad_norm": 0.7314509191475778,
+      "learning_rate": 3.5170578874078874e-06,
+      "loss": 0.4681,
+      "step": 2453
+    },
+    {
+      "epoch": 1.909232365145228,
+      "grad_norm": 0.5342385318910443,
+      "learning_rate": 3.5127341404803257e-06,
+      "loss": 0.3659,
+      "step": 2454
+    },
+    {
+      "epoch": 1.9100103734439835,
+      "grad_norm": 0.5462591792812815,
+      "learning_rate": 3.5084116133277524e-06,
+      "loss": 0.3927,
+      "step": 2455
+    },
+    {
+      "epoch": 1.9107883817427385,
+      "grad_norm": 0.5415848189672374,
+      "learning_rate": 3.504090309495275e-06,
+      "loss": 0.3851,
+      "step": 2456
+    },
+    {
+      "epoch": 1.9115663900414939,
+      "grad_norm": 0.5889778823825181,
+      "learning_rate": 3.4997702325269883e-06,
+      "loss": 0.4264,
+      "step": 2457
+    },
+    {
+      "epoch": 1.9123443983402488,
+      "grad_norm": 0.5820322078180551,
+      "learning_rate": 3.495451385965991e-06,
+      "loss": 0.4345,
+      "step": 2458
+    },
+    {
+      "epoch": 1.9131224066390042,
+      "grad_norm": 0.45941635469206576,
+      "learning_rate": 3.4911337733543638e-06,
+      "loss": 0.343,
+      "step": 2459
+    },
+    {
+      "epoch": 1.9139004149377592,
+      "grad_norm": 0.5915049242569727,
+      "learning_rate": 3.486817398233179e-06,
+      "loss": 0.4585,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9146784232365146,
+      "grad_norm": 0.46989189191362746,
+      "learning_rate": 3.4825022641424966e-06,
+      "loss": 0.3726,
+      "step": 2461
+    },
+    {
+      "epoch": 1.9154564315352696,
+      "grad_norm": 0.5239063309617389,
+      "learning_rate": 3.478188374621354e-06,
+      "loss": 0.3767,
+      "step": 2462
+    },
+    {
+      "epoch": 1.916234439834025,
+      "grad_norm": 0.5328455017189653,
+      "learning_rate": 3.473875733207771e-06,
+      "loss": 0.3737,
+      "step": 2463
+    },
+    {
+      "epoch": 1.91701244813278,
+      "grad_norm": 0.5818411000328347,
+      "learning_rate": 3.4695643434387403e-06,
+      "loss": 0.3851,
+      "step": 2464
+    },
+    {
+      "epoch": 1.9177904564315353,
+      "grad_norm": 0.5427098930204252,
+      "learning_rate": 3.465254208850234e-06,
+      "loss": 0.4194,
+      "step": 2465
+    },
+    {
+      "epoch": 1.9185684647302903,
+      "grad_norm": 0.5170143428116506,
+      "learning_rate": 3.460945332977188e-06,
+      "loss": 0.3771,
+      "step": 2466
+    },
+    {
+      "epoch": 1.9193464730290457,
+      "grad_norm": 0.6159064913731166,
+      "learning_rate": 3.4566377193535115e-06,
+      "loss": 0.4243,
+      "step": 2467
+    },
+    {
+      "epoch": 1.9201244813278007,
+      "grad_norm": 0.5888181290066158,
+      "learning_rate": 3.4523313715120775e-06,
+      "loss": 0.3673,
+      "step": 2468
+    },
+    {
+      "epoch": 1.920902489626556,
+      "grad_norm": 0.5659907461324543,
+      "learning_rate": 3.4480262929847147e-06,
+      "loss": 0.4206,
+      "step": 2469
+    },
+    {
+      "epoch": 1.921680497925311,
+      "grad_norm": 0.5897493557065933,
+      "learning_rate": 3.44372248730222e-06,
+      "loss": 0.3893,
+      "step": 2470
+    },
+    {
+      "epoch": 1.9224585062240664,
+      "grad_norm": 0.5659996781510208,
+      "learning_rate": 3.4394199579943404e-06,
+      "loss": 0.3959,
+      "step": 2471
+    },
+    {
+      "epoch": 1.9232365145228216,
+      "grad_norm": 0.5428360173706658,
+      "learning_rate": 3.435118708589778e-06,
+      "loss": 0.3411,
+      "step": 2472
+    },
+    {
+      "epoch": 1.9240145228215768,
+      "grad_norm": 0.5880989686732919,
+      "learning_rate": 3.430818742616182e-06,
+      "loss": 0.46,
+      "step": 2473
+    },
+    {
+      "epoch": 1.924792531120332,
+      "grad_norm": 0.5245632052190078,
+      "learning_rate": 3.426520063600156e-06,
+      "loss": 0.3581,
+      "step": 2474
+    },
+    {
+      "epoch": 1.9255705394190872,
+      "grad_norm": 0.5336038999579295,
+      "learning_rate": 3.4222226750672428e-06,
+      "loss": 0.3776,
+      "step": 2475
+    },
+    {
+      "epoch": 1.9263485477178424,
+      "grad_norm": 0.5780961522752593,
+      "learning_rate": 3.417926580541926e-06,
+      "loss": 0.4328,
+      "step": 2476
+    },
+    {
+      "epoch": 1.9271265560165975,
+      "grad_norm": 0.5678776359242258,
+      "learning_rate": 3.413631783547633e-06,
+      "loss": 0.3849,
+      "step": 2477
+    },
+    {
+      "epoch": 1.9279045643153527,
+      "grad_norm": 0.5075140688649143,
+      "learning_rate": 3.4093382876067206e-06,
+      "loss": 0.3793,
+      "step": 2478
+    },
+    {
+      "epoch": 1.928682572614108,
+      "grad_norm": 0.5617846783011771,
+      "learning_rate": 3.4050460962404853e-06,
+      "loss": 0.3616,
+      "step": 2479
+    },
+    {
+      "epoch": 1.929460580912863,
+      "grad_norm": 0.5366753532539731,
+      "learning_rate": 3.4007552129691466e-06,
+      "loss": 0.3759,
+      "step": 2480
+    },
+    {
+      "epoch": 1.9302385892116183,
+      "grad_norm": 0.627270213632953,
+      "learning_rate": 3.396465641311857e-06,
+      "loss": 0.4388,
+      "step": 2481
+    },
+    {
+      "epoch": 1.9310165975103735,
+      "grad_norm": 0.6005131795126498,
+      "learning_rate": 3.3921773847866916e-06,
+      "loss": 0.3647,
+      "step": 2482
+    },
+    {
+      "epoch": 1.9317946058091287,
+      "grad_norm": 0.5625347860858282,
+      "learning_rate": 3.3878904469106448e-06,
+      "loss": 0.4101,
+      "step": 2483
+    },
+    {
+      "epoch": 1.9325726141078838,
+      "grad_norm": 0.5595715157037091,
+      "learning_rate": 3.3836048311996333e-06,
+      "loss": 0.3996,
+      "step": 2484
+    },
+    {
+      "epoch": 1.933350622406639,
+      "grad_norm": 0.44674920755694725,
+      "learning_rate": 3.3793205411684836e-06,
+      "loss": 0.3292,
+      "step": 2485
+    },
+    {
+      "epoch": 1.9341286307053942,
+      "grad_norm": 0.5373344533557178,
+      "learning_rate": 3.375037580330943e-06,
+      "loss": 0.3717,
+      "step": 2486
+    },
+    {
+      "epoch": 1.9349066390041494,
+      "grad_norm": 0.5943119429454425,
+      "learning_rate": 3.370755952199659e-06,
+      "loss": 0.403,
+      "step": 2487
+    },
+    {
+      "epoch": 1.9356846473029046,
+      "grad_norm": 0.5314601217984761,
+      "learning_rate": 3.3664756602861964e-06,
+      "loss": 0.3797,
+      "step": 2488
+    },
+    {
+      "epoch": 1.9364626556016598,
+      "grad_norm": 0.571578768128056,
+      "learning_rate": 3.3621967081010177e-06,
+      "loss": 0.4027,
+      "step": 2489
+    },
+    {
+      "epoch": 1.937240663900415,
+      "grad_norm": 0.5481632523628887,
+      "learning_rate": 3.357919099153485e-06,
+      "loss": 0.4294,
+      "step": 2490
+    },
+    {
+      "epoch": 1.9380186721991701,
+      "grad_norm": 0.5834721158542985,
+      "learning_rate": 3.3536428369518663e-06,
+      "loss": 0.3934,
+      "step": 2491
+    },
+    {
+      "epoch": 1.9387966804979253,
+      "grad_norm": 0.5394534272249718,
+      "learning_rate": 3.3493679250033175e-06,
+      "loss": 0.4061,
+      "step": 2492
+    },
+    {
+      "epoch": 1.9395746887966805,
+      "grad_norm": 0.48960651424469515,
+      "learning_rate": 3.345094366813891e-06,
+      "loss": 0.4072,
+      "step": 2493
+    },
+    {
+      "epoch": 1.9403526970954357,
+      "grad_norm": 0.51960616363486,
+      "learning_rate": 3.3408221658885266e-06,
+      "loss": 0.366,
+      "step": 2494
+    },
+    {
+      "epoch": 1.9411307053941909,
+      "grad_norm": 0.4984235567720628,
+      "learning_rate": 3.336551325731053e-06,
+      "loss": 0.331,
+      "step": 2495
+    },
+    {
+      "epoch": 1.941908713692946,
+      "grad_norm": 0.5407785186314209,
+      "learning_rate": 3.3322818498441843e-06,
+      "loss": 0.4469,
+      "step": 2496
+    },
+    {
+      "epoch": 1.9426867219917012,
+      "grad_norm": 0.4879432500965268,
+      "learning_rate": 3.328013741729512e-06,
+      "loss": 0.3743,
+      "step": 2497
+    },
+    {
+      "epoch": 1.9434647302904564,
+      "grad_norm": 0.502899932268725,
+      "learning_rate": 3.3237470048875087e-06,
+      "loss": 0.4218,
+      "step": 2498
+    },
+    {
+      "epoch": 1.9442427385892116,
+      "grad_norm": 0.5175189847951854,
+      "learning_rate": 3.319481642817517e-06,
+      "loss": 0.3503,
+      "step": 2499
+    },
+    {
+      "epoch": 1.9450207468879668,
+      "grad_norm": 0.5136499478853049,
+      "learning_rate": 3.315217659017762e-06,
+      "loss": 0.41,
+      "step": 2500
+    },
+    {
+      "epoch": 1.945798755186722,
+      "grad_norm": 0.5816147788335556,
+      "learning_rate": 3.3109550569853278e-06,
+      "loss": 0.3606,
+      "step": 2501
+    },
+    {
+      "epoch": 1.9465767634854771,
+      "grad_norm": 0.5717113792032724,
+      "learning_rate": 3.3066938402161714e-06,
+      "loss": 0.4236,
+      "step": 2502
+    },
+    {
+      "epoch": 1.9473547717842323,
+      "grad_norm": 0.5877222937108594,
+      "learning_rate": 3.3024340122051146e-06,
+      "loss": 0.4107,
+      "step": 2503
+    },
+    {
+      "epoch": 1.9481327800829875,
+      "grad_norm": 0.5206315010982968,
+      "learning_rate": 3.2981755764458336e-06,
+      "loss": 0.3904,
+      "step": 2504
+    },
+    {
+      "epoch": 1.9489107883817427,
+      "grad_norm": 0.5016812407982975,
+      "learning_rate": 3.2939185364308707e-06,
+      "loss": 0.3894,
+      "step": 2505
+    },
+    {
+      "epoch": 1.9496887966804979,
+      "grad_norm": 0.5291742600527443,
+      "learning_rate": 3.289662895651618e-06,
+      "loss": 0.3971,
+      "step": 2506
+    },
+    {
+      "epoch": 1.950466804979253,
+      "grad_norm": 0.6282164598053888,
+      "learning_rate": 3.2854086575983228e-06,
+      "loss": 0.3858,
+      "step": 2507
+    },
+    {
+      "epoch": 1.9512448132780082,
+      "grad_norm": 0.5635397260462525,
+      "learning_rate": 3.2811558257600784e-06,
+      "loss": 0.4387,
+      "step": 2508
+    },
+    {
+      "epoch": 1.9520228215767634,
+      "grad_norm": 0.5096501316378109,
+      "learning_rate": 3.276904403624831e-06,
+      "loss": 0.3501,
+      "step": 2509
+    },
+    {
+      "epoch": 1.9528008298755186,
+      "grad_norm": 0.6111168512269416,
+      "learning_rate": 3.272654394679366e-06,
+      "loss": 0.4326,
+      "step": 2510
+    },
+    {
+      "epoch": 1.9535788381742738,
+      "grad_norm": 0.5418842076907006,
+      "learning_rate": 3.268405802409308e-06,
+      "loss": 0.3693,
+      "step": 2511
+    },
+    {
+      "epoch": 1.9543568464730292,
+      "grad_norm": 0.5404932797353206,
+      "learning_rate": 3.264158630299127e-06,
+      "loss": 0.4213,
+      "step": 2512
+    },
+    {
+      "epoch": 1.9551348547717842,
+      "grad_norm": 0.5861354019110064,
+      "learning_rate": 3.259912881832119e-06,
+      "loss": 0.4029,
+      "step": 2513
+    },
+    {
+      "epoch": 1.9559128630705396,
+      "grad_norm": 0.5356362675913765,
+      "learning_rate": 3.2556685604904205e-06,
+      "loss": 0.3867,
+      "step": 2514
+    },
+    {
+      "epoch": 1.9566908713692945,
+      "grad_norm": 0.54223116029722,
+      "learning_rate": 3.2514256697549945e-06,
+      "loss": 0.4496,
+      "step": 2515
+    },
+    {
+      "epoch": 1.95746887966805,
+      "grad_norm": 0.5435091387313036,
+      "learning_rate": 3.2471842131056274e-06,
+      "loss": 0.418,
+      "step": 2516
+    },
+    {
+      "epoch": 1.958246887966805,
+      "grad_norm": 0.41570246860113086,
+      "learning_rate": 3.242944194020936e-06,
+      "loss": 0.3407,
+      "step": 2517
+    },
+    {
+      "epoch": 1.9590248962655603,
+      "grad_norm": 0.5273282001656338,
+      "learning_rate": 3.2387056159783524e-06,
+      "loss": 0.4066,
+      "step": 2518
+    },
+    {
+      "epoch": 1.9598029045643153,
+      "grad_norm": 0.5761857734147758,
+      "learning_rate": 3.2344684824541307e-06,
+      "loss": 0.443,
+      "step": 2519
+    },
+    {
+      "epoch": 1.9605809128630707,
+      "grad_norm": 0.49329106587138516,
+      "learning_rate": 3.2302327969233345e-06,
+      "loss": 0.391,
+      "step": 2520
+    },
+    {
+      "epoch": 1.9613589211618256,
+      "grad_norm": 0.5266802785933788,
+      "learning_rate": 3.2259985628598466e-06,
+      "loss": 0.435,
+      "step": 2521
+    },
+    {
+      "epoch": 1.962136929460581,
+      "grad_norm": 0.4855291141340067,
+      "learning_rate": 3.22176578373636e-06,
+      "loss": 0.3954,
+      "step": 2522
+    },
+    {
+      "epoch": 1.962914937759336,
+      "grad_norm": 0.5138281491569104,
+      "learning_rate": 3.217534463024366e-06,
+      "loss": 0.3802,
+      "step": 2523
+    },
+    {
+      "epoch": 1.9636929460580914,
+      "grad_norm": 0.578874549934224,
+      "learning_rate": 3.2133046041941683e-06,
+      "loss": 0.427,
+      "step": 2524
+    },
+    {
+      "epoch": 1.9644709543568464,
+      "grad_norm": 0.4614820766484502,
+      "learning_rate": 3.2090762107148647e-06,
+      "loss": 0.3557,
+      "step": 2525
+    },
+    {
+      "epoch": 1.9652489626556018,
+      "grad_norm": 0.5735337805436925,
+      "learning_rate": 3.204849286054359e-06,
+      "loss": 0.4427,
+      "step": 2526
+    },
+    {
+      "epoch": 1.9660269709543567,
+      "grad_norm": 0.48042309491733587,
+      "learning_rate": 3.2006238336793427e-06,
+      "loss": 0.3549,
+      "step": 2527
+    },
+    {
+      "epoch": 1.9668049792531122,
+      "grad_norm": 0.5831956599513581,
+      "learning_rate": 3.196399857055304e-06,
+      "loss": 0.3922,
+      "step": 2528
+    },
+    {
+      "epoch": 1.9675829875518671,
+      "grad_norm": 0.49683045326097897,
+      "learning_rate": 3.1921773596465223e-06,
+      "loss": 0.3869,
+      "step": 2529
+    },
+    {
+      "epoch": 1.9683609958506225,
+      "grad_norm": 0.45279709130823576,
+      "learning_rate": 3.187956344916058e-06,
+      "loss": 0.3896,
+      "step": 2530
+    },
+    {
+      "epoch": 1.9691390041493775,
+      "grad_norm": 0.4898993958637394,
+      "learning_rate": 3.183736816325762e-06,
+      "loss": 0.4156,
+      "step": 2531
+    },
+    {
+      "epoch": 1.9699170124481329,
+      "grad_norm": 0.5075582548930789,
+      "learning_rate": 3.1795187773362624e-06,
+      "loss": 0.382,
+      "step": 2532
+    },
+    {
+      "epoch": 1.9706950207468878,
+      "grad_norm": 0.4979793522446796,
+      "learning_rate": 3.175302231406966e-06,
+      "loss": 0.3566,
+      "step": 2533
+    },
+    {
+      "epoch": 1.9714730290456433,
+      "grad_norm": 0.5324096316532091,
+      "learning_rate": 3.1710871819960544e-06,
+      "loss": 0.4206,
+      "step": 2534
+    },
+    {
+      "epoch": 1.9722510373443982,
+      "grad_norm": 0.5349733708942778,
+      "learning_rate": 3.1668736325604853e-06,
+      "loss": 0.3904,
+      "step": 2535
+    },
+    {
+      "epoch": 1.9730290456431536,
+      "grad_norm": 0.5181986229175758,
+      "learning_rate": 3.1626615865559852e-06,
+      "loss": 0.3635,
+      "step": 2536
+    },
+    {
+      "epoch": 1.9738070539419086,
+      "grad_norm": 0.5604847106495198,
+      "learning_rate": 3.158451047437041e-06,
+      "loss": 0.4489,
+      "step": 2537
+    },
+    {
+      "epoch": 1.974585062240664,
+      "grad_norm": 0.4941908032684207,
+      "learning_rate": 3.1542420186569155e-06,
+      "loss": 0.3906,
+      "step": 2538
+    },
+    {
+      "epoch": 1.975363070539419,
+      "grad_norm": 0.5206530409441004,
+      "learning_rate": 3.1500345036676214e-06,
+      "loss": 0.3972,
+      "step": 2539
+    },
+    {
+      "epoch": 1.9761410788381744,
+      "grad_norm": 0.5752879929594736,
+      "learning_rate": 3.1458285059199387e-06,
+      "loss": 0.3556,
+      "step": 2540
+    },
+    {
+      "epoch": 1.9769190871369293,
+      "grad_norm": 0.5113254168725803,
+      "learning_rate": 3.141624028863396e-06,
+      "loss": 0.3781,
+      "step": 2541
+    },
+    {
+      "epoch": 1.9776970954356847,
+      "grad_norm": 0.4676391950528534,
+      "learning_rate": 3.1374210759462787e-06,
+      "loss": 0.3447,
+      "step": 2542
+    },
+    {
+      "epoch": 1.9784751037344397,
+      "grad_norm": 0.5478924014946802,
+      "learning_rate": 3.133219650615623e-06,
+      "loss": 0.4354,
+      "step": 2543
+    },
+    {
+      "epoch": 1.979253112033195,
+      "grad_norm": 0.5156655177904961,
+      "learning_rate": 3.1290197563172097e-06,
+      "loss": 0.4107,
+      "step": 2544
+    },
+    {
+      "epoch": 1.98003112033195,
+      "grad_norm": 0.47845594423466914,
+      "learning_rate": 3.1248213964955653e-06,
+      "loss": 0.3595,
+      "step": 2545
+    },
+    {
+      "epoch": 1.9808091286307055,
+      "grad_norm": 0.49897937124340214,
+      "learning_rate": 3.1206245745939546e-06,
+      "loss": 0.4089,
+      "step": 2546
+    },
+    {
+      "epoch": 1.9815871369294604,
+      "grad_norm": 0.4313655200489229,
+      "learning_rate": 3.1164292940543885e-06,
+      "loss": 0.3596,
+      "step": 2547
+    },
+    {
+      "epoch": 1.9823651452282158,
+      "grad_norm": 0.656311620148762,
+      "learning_rate": 3.112235558317604e-06,
+      "loss": 0.4442,
+      "step": 2548
+    },
+    {
+      "epoch": 1.9831431535269708,
+      "grad_norm": 0.5777356274312242,
+      "learning_rate": 3.1080433708230803e-06,
+      "loss": 0.4123,
+      "step": 2549
+    },
+    {
+      "epoch": 1.9839211618257262,
+      "grad_norm": 0.49057166434244737,
+      "learning_rate": 3.103852735009022e-06,
+      "loss": 0.3733,
+      "step": 2550
+    },
+    {
+      "epoch": 1.9846991701244814,
+      "grad_norm": 0.5416497486288832,
+      "learning_rate": 3.0996636543123597e-06,
+      "loss": 0.3786,
+      "step": 2551
+    },
+    {
+      "epoch": 1.9854771784232366,
+      "grad_norm": 0.5487353583783505,
+      "learning_rate": 3.0954761321687545e-06,
+      "loss": 0.462,
+      "step": 2552
+    },
+    {
+      "epoch": 1.9862551867219918,
+      "grad_norm": 0.5103217755134126,
+      "learning_rate": 3.091290172012582e-06,
+      "loss": 0.361,
+      "step": 2553
+    },
+    {
+      "epoch": 1.987033195020747,
+      "grad_norm": 0.49198253079552295,
+      "learning_rate": 3.087105777276942e-06,
+      "loss": 0.3907,
+      "step": 2554
+    },
+    {
+      "epoch": 1.9878112033195021,
+      "grad_norm": 0.5501782725335493,
+      "learning_rate": 3.0829229513936472e-06,
+      "loss": 0.3936,
+      "step": 2555
+    },
+    {
+      "epoch": 1.9885892116182573,
+      "grad_norm": 0.5894571694669006,
+      "learning_rate": 3.078741697793225e-06,
+      "loss": 0.4538,
+      "step": 2556
+    },
+    {
+      "epoch": 1.9893672199170125,
+      "grad_norm": 0.537227487641102,
+      "learning_rate": 3.0745620199049176e-06,
+      "loss": 0.3818,
+      "step": 2557
+    },
+    {
+      "epoch": 1.9901452282157677,
+      "grad_norm": 0.562969104086542,
+      "learning_rate": 3.0703839211566654e-06,
+      "loss": 0.387,
+      "step": 2558
+    },
+    {
+      "epoch": 1.9909232365145229,
+      "grad_norm": 0.5791464487197951,
+      "learning_rate": 3.0662074049751235e-06,
+      "loss": 0.4109,
+      "step": 2559
+    },
+    {
+      "epoch": 1.991701244813278,
+      "grad_norm": 0.5500132047484221,
+      "learning_rate": 3.062032474785639e-06,
+      "loss": 0.4142,
+      "step": 2560
+    },
+    {
+      "epoch": 1.9924792531120332,
+      "grad_norm": 0.5279718607568004,
+      "learning_rate": 3.0578591340122695e-06,
+      "loss": 0.4166,
+      "step": 2561
+    },
+    {
+      "epoch": 1.9932572614107884,
+      "grad_norm": 0.6144423433574359,
+      "learning_rate": 3.053687386077759e-06,
+      "loss": 0.4689,
+      "step": 2562
+    },
+    {
+      "epoch": 1.9940352697095436,
+      "grad_norm": 0.5343288179760488,
+      "learning_rate": 3.0495172344035505e-06,
+      "loss": 0.422,
+      "step": 2563
+    },
+    {
+      "epoch": 1.9948132780082988,
+      "grad_norm": 0.5972634942724403,
+      "learning_rate": 3.0453486824097794e-06,
+      "loss": 0.4246,
+      "step": 2564
+    },
+    {
+      "epoch": 1.995591286307054,
+      "grad_norm": 0.5705997323624956,
+      "learning_rate": 3.041181733515263e-06,
+      "loss": 0.3538,
+      "step": 2565
+    },
+    {
+      "epoch": 1.9963692946058091,
+      "grad_norm": 0.5167048713052862,
+      "learning_rate": 3.0370163911375096e-06,
+      "loss": 0.401,
+      "step": 2566
+    },
+    {
+      "epoch": 1.9971473029045643,
+      "grad_norm": 0.5207107226449875,
+      "learning_rate": 3.0328526586927063e-06,
+      "loss": 0.4113,
+      "step": 2567
+    },
+    {
+      "epoch": 1.9979253112033195,
+      "grad_norm": 0.49877886832391677,
+      "learning_rate": 3.0286905395957233e-06,
+      "loss": 0.3661,
+      "step": 2568
+    },
+    {
+      "epoch": 1.9987033195020747,
+      "grad_norm": 0.5656859910339247,
+      "learning_rate": 3.024530037260102e-06,
+      "loss": 0.3529,
+      "step": 2569
+    },
+    {
+      "epoch": 1.9994813278008299,
+      "grad_norm": 0.5695313718398513,
+      "learning_rate": 3.0203711550980635e-06,
+      "loss": 0.4491,
+      "step": 2570
+    },
+    {
+      "epoch": 2.0002593360995853,
+      "grad_norm": 1.6352052063241584,
+      "learning_rate": 3.016213896520499e-06,
+      "loss": 0.6215,
+      "step": 2571
+    },
+    {
+      "epoch": 2.0010373443983402,
+      "grad_norm": 0.5389752169459832,
+      "learning_rate": 3.012058264936964e-06,
+      "loss": 0.3565,
+      "step": 2572
+    },
+    {
+      "epoch": 2.0018153526970957,
+      "grad_norm": 0.5399139873851488,
+      "learning_rate": 3.0079042637556867e-06,
+      "loss": 0.3596,
+      "step": 2573
+    },
+    {
+      "epoch": 2.0025933609958506,
+      "grad_norm": 0.4894333369120865,
+      "learning_rate": 3.00375189638355e-06,
+      "loss": 0.3487,
+      "step": 2574
+    },
+    {
+      "epoch": 2.003371369294606,
+      "grad_norm": 0.5553880462343063,
+      "learning_rate": 2.9996011662261047e-06,
+      "loss": 0.3509,
+      "step": 2575
+    },
+    {
+      "epoch": 2.004149377593361,
+      "grad_norm": 0.536326597982821,
+      "learning_rate": 2.9954520766875504e-06,
+      "loss": 0.3235,
+      "step": 2576
+    },
+    {
+      "epoch": 2.0049273858921164,
+      "grad_norm": 0.5344774729648599,
+      "learning_rate": 2.9913046311707496e-06,
+      "loss": 0.3312,
+      "step": 2577
+    },
+    {
+      "epoch": 2.0057053941908713,
+      "grad_norm": 0.521414737026206,
+      "learning_rate": 2.987158833077214e-06,
+      "loss": 0.2989,
+      "step": 2578
+    },
+    {
+      "epoch": 2.0064834024896268,
+      "grad_norm": 0.5998353456668626,
+      "learning_rate": 2.9830146858070986e-06,
+      "loss": 0.3564,
+      "step": 2579
+    },
+    {
+      "epoch": 2.0072614107883817,
+      "grad_norm": 0.5282151618448686,
+      "learning_rate": 2.9788721927592125e-06,
+      "loss": 0.3513,
+      "step": 2580
+    },
+    {
+      "epoch": 2.008039419087137,
+      "grad_norm": 0.5611772864729125,
+      "learning_rate": 2.9747313573310012e-06,
+      "loss": 0.3837,
+      "step": 2581
+    },
+    {
+      "epoch": 2.008817427385892,
+      "grad_norm": 0.5157477795341966,
+      "learning_rate": 2.970592182918557e-06,
+      "loss": 0.3375,
+      "step": 2582
+    },
+    {
+      "epoch": 2.0095954356846475,
+      "grad_norm": 0.5871812235033396,
+      "learning_rate": 2.9664546729166038e-06,
+      "loss": 0.3759,
+      "step": 2583
+    },
+    {
+      "epoch": 2.0103734439834025,
+      "grad_norm": 0.5175045661849054,
+      "learning_rate": 2.9623188307185045e-06,
+      "loss": 0.3493,
+      "step": 2584
+    },
+    {
+      "epoch": 2.011151452282158,
+      "grad_norm": 0.5339702846181945,
+      "learning_rate": 2.958184659716255e-06,
+      "loss": 0.3033,
+      "step": 2585
+    },
+    {
+      "epoch": 2.011929460580913,
+      "grad_norm": 0.6374743223435375,
+      "learning_rate": 2.954052163300476e-06,
+      "loss": 0.3904,
+      "step": 2586
+    },
+    {
+      "epoch": 2.0127074688796682,
+      "grad_norm": 0.4886751168640502,
+      "learning_rate": 2.9499213448604197e-06,
+      "loss": 0.3487,
+      "step": 2587
+    },
+    {
+      "epoch": 2.013485477178423,
+      "grad_norm": 0.5231754372750812,
+      "learning_rate": 2.9457922077839587e-06,
+      "loss": 0.3813,
+      "step": 2588
+    },
+    {
+      "epoch": 2.0142634854771786,
+      "grad_norm": 0.5721187318535549,
+      "learning_rate": 2.9416647554575883e-06,
+      "loss": 0.3216,
+      "step": 2589
+    },
+    {
+      "epoch": 2.0150414937759336,
+      "grad_norm": 0.5167419625337567,
+      "learning_rate": 2.9375389912664197e-06,
+      "loss": 0.3303,
+      "step": 2590
+    },
+    {
+      "epoch": 2.015819502074689,
+      "grad_norm": 0.5333034061479656,
+      "learning_rate": 2.933414918594184e-06,
+      "loss": 0.3539,
+      "step": 2591
+    },
+    {
+      "epoch": 2.016597510373444,
+      "grad_norm": 0.5404777369240766,
+      "learning_rate": 2.929292540823222e-06,
+      "loss": 0.3497,
+      "step": 2592
+    },
+    {
+      "epoch": 2.0173755186721993,
+      "grad_norm": 0.5484805908618148,
+      "learning_rate": 2.925171861334483e-06,
+      "loss": 0.3595,
+      "step": 2593
+    },
+    {
+      "epoch": 2.0181535269709543,
+      "grad_norm": 0.5608966736702051,
+      "learning_rate": 2.9210528835075287e-06,
+      "loss": 0.356,
+      "step": 2594
+    },
+    {
+      "epoch": 2.0189315352697097,
+      "grad_norm": 0.5755177523074986,
+      "learning_rate": 2.916935610720518e-06,
+      "loss": 0.3084,
+      "step": 2595
+    },
+    {
+      "epoch": 2.0197095435684647,
+      "grad_norm": 0.555853327129972,
+      "learning_rate": 2.9128200463502197e-06,
+      "loss": 0.3457,
+      "step": 2596
+    },
+    {
+      "epoch": 2.02048755186722,
+      "grad_norm": 0.5236066015796267,
+      "learning_rate": 2.9087061937719915e-06,
+      "loss": 0.3281,
+      "step": 2597
+    },
+    {
+      "epoch": 2.021265560165975,
+      "grad_norm": 0.49947973058857686,
+      "learning_rate": 2.904594056359797e-06,
+      "loss": 0.3665,
+      "step": 2598
+    },
+    {
+      "epoch": 2.0220435684647304,
+      "grad_norm": 0.5544006468912936,
+      "learning_rate": 2.90048363748619e-06,
+      "loss": 0.3763,
+      "step": 2599
+    },
+    {
+      "epoch": 2.0228215767634854,
+      "grad_norm": 0.503767736296507,
+      "learning_rate": 2.896374940522307e-06,
+      "loss": 0.3199,
+      "step": 2600
+    },
+    {
+      "epoch": 2.023599585062241,
+      "grad_norm": 0.46231984963529804,
+      "learning_rate": 2.892267968837887e-06,
+      "loss": 0.3394,
+      "step": 2601
+    },
+    {
+      "epoch": 2.0243775933609958,
+      "grad_norm": 0.47149070397071335,
+      "learning_rate": 2.888162725801239e-06,
+      "loss": 0.3649,
+      "step": 2602
+    },
+    {
+      "epoch": 2.025155601659751,
+      "grad_norm": 0.45189520535501254,
+      "learning_rate": 2.8840592147792658e-06,
+      "loss": 0.2799,
+      "step": 2603
+    },
+    {
+      "epoch": 2.025933609958506,
+      "grad_norm": 0.464121191246612,
+      "learning_rate": 2.8799574391374453e-06,
+      "loss": 0.3213,
+      "step": 2604
+    },
+    {
+      "epoch": 2.0267116182572615,
+      "grad_norm": 0.5193312825450764,
+      "learning_rate": 2.8758574022398273e-06,
+      "loss": 0.3391,
+      "step": 2605
+    },
+    {
+      "epoch": 2.0274896265560165,
+      "grad_norm": 0.5137549765227485,
+      "learning_rate": 2.871759107449047e-06,
+      "loss": 0.3114,
+      "step": 2606
+    },
+    {
+      "epoch": 2.028267634854772,
+      "grad_norm": 0.48598393354024266,
+      "learning_rate": 2.8676625581262985e-06,
+      "loss": 0.3443,
+      "step": 2607
+    },
+    {
+      "epoch": 2.029045643153527,
+      "grad_norm": 0.5021265001151275,
+      "learning_rate": 2.8635677576313515e-06,
+      "loss": 0.3466,
+      "step": 2608
+    },
+    {
+      "epoch": 2.0298236514522823,
+      "grad_norm": 0.5262434980508413,
+      "learning_rate": 2.8594747093225404e-06,
+      "loss": 0.391,
+      "step": 2609
+    },
+    {
+      "epoch": 2.0306016597510372,
+      "grad_norm": 0.4748068778782379,
+      "learning_rate": 2.8553834165567613e-06,
+      "loss": 0.3299,
+      "step": 2610
+    },
+    {
+      "epoch": 2.0313796680497926,
+      "grad_norm": 0.5446640950961169,
+      "learning_rate": 2.851293882689472e-06,
+      "loss": 0.3226,
+      "step": 2611
+    },
+    {
+      "epoch": 2.0321576763485476,
+      "grad_norm": 0.5617157550905628,
+      "learning_rate": 2.847206111074686e-06,
+      "loss": 0.394,
+      "step": 2612
+    },
+    {
+      "epoch": 2.032935684647303,
+      "grad_norm": 0.47393241284920096,
+      "learning_rate": 2.8431201050649737e-06,
+      "loss": 0.2983,
+      "step": 2613
+    },
+    {
+      "epoch": 2.033713692946058,
+      "grad_norm": 0.5024587345469693,
+      "learning_rate": 2.8390358680114548e-06,
+      "loss": 0.354,
+      "step": 2614
+    },
+    {
+      "epoch": 2.0344917012448134,
+      "grad_norm": 0.5058476373016942,
+      "learning_rate": 2.8349534032638e-06,
+      "loss": 0.3151,
+      "step": 2615
+    },
+    {
+      "epoch": 2.0352697095435683,
+      "grad_norm": 0.5384483367692212,
+      "learning_rate": 2.8308727141702263e-06,
+      "loss": 0.3531,
+      "step": 2616
+    },
+    {
+      "epoch": 2.0360477178423237,
+      "grad_norm": 0.5068351738494636,
+      "learning_rate": 2.826793804077494e-06,
+      "loss": 0.3295,
+      "step": 2617
+    },
+    {
+      "epoch": 2.0368257261410787,
+      "grad_norm": 0.5460974763251636,
+      "learning_rate": 2.8227166763309045e-06,
+      "loss": 0.38,
+      "step": 2618
+    },
+    {
+      "epoch": 2.037603734439834,
+      "grad_norm": 0.48366765104704507,
+      "learning_rate": 2.818641334274297e-06,
+      "loss": 0.3014,
+      "step": 2619
+    },
+    {
+      "epoch": 2.038381742738589,
+      "grad_norm": 0.49944045091447115,
+      "learning_rate": 2.814567781250049e-06,
+      "loss": 0.3497,
+      "step": 2620
+    },
+    {
+      "epoch": 2.0391597510373445,
+      "grad_norm": 0.5224714027276477,
+      "learning_rate": 2.8104960205990615e-06,
+      "loss": 0.369,
+      "step": 2621
+    },
+    {
+      "epoch": 2.0399377593360994,
+      "grad_norm": 0.5063534715432086,
+      "learning_rate": 2.806426055660782e-06,
+      "loss": 0.3116,
+      "step": 2622
+    },
+    {
+      "epoch": 2.040715767634855,
+      "grad_norm": 0.5567830265930057,
+      "learning_rate": 2.8023578897731662e-06,
+      "loss": 0.4056,
+      "step": 2623
+    },
+    {
+      "epoch": 2.04149377593361,
+      "grad_norm": 0.44728952983638326,
+      "learning_rate": 2.7982915262727094e-06,
+      "loss": 0.3011,
+      "step": 2624
+    },
+    {
+      "epoch": 2.042271784232365,
+      "grad_norm": 0.6243601584097669,
+      "learning_rate": 2.7942269684944244e-06,
+      "loss": 0.3574,
+      "step": 2625
+    },
+    {
+      "epoch": 2.04304979253112,
+      "grad_norm": 0.5204122962296256,
+      "learning_rate": 2.7901642197718337e-06,
+      "loss": 0.3318,
+      "step": 2626
+    },
+    {
+      "epoch": 2.0438278008298756,
+      "grad_norm": 0.5963474532205301,
+      "learning_rate": 2.786103283436994e-06,
+      "loss": 0.3852,
+      "step": 2627
+    },
+    {
+      "epoch": 2.0446058091286305,
+      "grad_norm": 0.532403241593858,
+      "learning_rate": 2.7820441628204574e-06,
+      "loss": 0.3381,
+      "step": 2628
+    },
+    {
+      "epoch": 2.045383817427386,
+      "grad_norm": 0.546639532024546,
+      "learning_rate": 2.777986861251298e-06,
+      "loss": 0.3157,
+      "step": 2629
+    },
+    {
+      "epoch": 2.046161825726141,
+      "grad_norm": 0.4866681962384701,
+      "learning_rate": 2.7739313820570967e-06,
+      "loss": 0.2732,
+      "step": 2630
+    },
+    {
+      "epoch": 2.0469398340248963,
+      "grad_norm": 0.577932860077872,
+      "learning_rate": 2.7698777285639334e-06,
+      "loss": 0.3782,
+      "step": 2631
+    },
+    {
+      "epoch": 2.0477178423236513,
+      "grad_norm": 0.5349442060295214,
+      "learning_rate": 2.765825904096403e-06,
+      "loss": 0.36,
+      "step": 2632
+    },
+    {
+      "epoch": 2.0484958506224067,
+      "grad_norm": 0.52354830009273,
+      "learning_rate": 2.7617759119775856e-06,
+      "loss": 0.3838,
+      "step": 2633
+    },
+    {
+      "epoch": 2.0492738589211617,
+      "grad_norm": 0.43194016455493656,
+      "learning_rate": 2.7577277555290693e-06,
+      "loss": 0.3244,
+      "step": 2634
+    },
+    {
+      "epoch": 2.050051867219917,
+      "grad_norm": 0.5257337967562634,
+      "learning_rate": 2.7536814380709308e-06,
+      "loss": 0.3308,
+      "step": 2635
+    },
+    {
+      "epoch": 2.050829875518672,
+      "grad_norm": 0.4947545597776326,
+      "learning_rate": 2.749636962921743e-06,
+      "loss": 0.3479,
+      "step": 2636
+    },
+    {
+      "epoch": 2.0516078838174274,
+      "grad_norm": 0.48858755925931086,
+      "learning_rate": 2.7455943333985643e-06,
+      "loss": 0.3278,
+      "step": 2637
+    },
+    {
+      "epoch": 2.0523858921161824,
+      "grad_norm": 0.5625659862290451,
+      "learning_rate": 2.74155355281694e-06,
+      "loss": 0.4042,
+      "step": 2638
+    },
+    {
+      "epoch": 2.053163900414938,
+      "grad_norm": 0.4606844148744298,
+      "learning_rate": 2.737514624490902e-06,
+      "loss": 0.2704,
+      "step": 2639
+    },
+    {
+      "epoch": 2.0539419087136928,
+      "grad_norm": 0.4928782622989064,
+      "learning_rate": 2.7334775517329585e-06,
+      "loss": 0.3001,
+      "step": 2640
+    },
+    {
+      "epoch": 2.054719917012448,
+      "grad_norm": 0.5726244688774949,
+      "learning_rate": 2.7294423378540996e-06,
+      "loss": 0.4198,
+      "step": 2641
+    },
+    {
+      "epoch": 2.055497925311203,
+      "grad_norm": 0.47208038870562663,
+      "learning_rate": 2.725408986163789e-06,
+      "loss": 0.3411,
+      "step": 2642
+    },
+    {
+      "epoch": 2.0562759336099585,
+      "grad_norm": 0.47712621913645975,
+      "learning_rate": 2.721377499969965e-06,
+      "loss": 0.3182,
+      "step": 2643
+    },
+    {
+      "epoch": 2.0570539419087135,
+      "grad_norm": 0.5092951737552278,
+      "learning_rate": 2.717347882579029e-06,
+      "loss": 0.3049,
+      "step": 2644
+    },
+    {
+      "epoch": 2.057831950207469,
+      "grad_norm": 0.5651425763311481,
+      "learning_rate": 2.7133201372958606e-06,
+      "loss": 0.46,
+      "step": 2645
+    },
+    {
+      "epoch": 2.058609958506224,
+      "grad_norm": 0.509964321942029,
+      "learning_rate": 2.7092942674238e-06,
+      "loss": 0.3239,
+      "step": 2646
+    },
+    {
+      "epoch": 2.0593879668049793,
+      "grad_norm": 0.5434795864403968,
+      "learning_rate": 2.7052702762646397e-06,
+      "loss": 0.316,
+      "step": 2647
+    },
+    {
+      "epoch": 2.0601659751037342,
+      "grad_norm": 0.5219988739924837,
+      "learning_rate": 2.70124816711865e-06,
+      "loss": 0.3607,
+      "step": 2648
+    },
+    {
+      "epoch": 2.0609439834024896,
+      "grad_norm": 0.5104008856927058,
+      "learning_rate": 2.697227943284537e-06,
+      "loss": 0.3487,
+      "step": 2649
+    },
+    {
+      "epoch": 2.061721991701245,
+      "grad_norm": 0.5245651462352366,
+      "learning_rate": 2.6932096080594804e-06,
+      "loss": 0.3364,
+      "step": 2650
+    },
+    {
+      "epoch": 2.0625,
+      "grad_norm": 0.6173504233989373,
+      "learning_rate": 2.689193164739094e-06,
+      "loss": 0.3775,
+      "step": 2651
+    },
+    {
+      "epoch": 2.063278008298755,
+      "grad_norm": 0.5667726193171813,
+      "learning_rate": 2.685178616617448e-06,
+      "loss": 0.3583,
+      "step": 2652
+    },
+    {
+      "epoch": 2.0640560165975104,
+      "grad_norm": 0.5822246071934846,
+      "learning_rate": 2.681165966987064e-06,
+      "loss": 0.3115,
+      "step": 2653
+    },
+    {
+      "epoch": 2.0648340248962658,
+      "grad_norm": 0.5361766276979218,
+      "learning_rate": 2.6771552191388933e-06,
+      "loss": 0.3429,
+      "step": 2654
+    },
+    {
+      "epoch": 2.0656120331950207,
+      "grad_norm": 0.5105067133676949,
+      "learning_rate": 2.6731463763623365e-06,
+      "loss": 0.3127,
+      "step": 2655
+    },
+    {
+      "epoch": 2.066390041493776,
+      "grad_norm": 0.5411578574992859,
+      "learning_rate": 2.6691394419452297e-06,
+      "loss": 0.3787,
+      "step": 2656
+    },
+    {
+      "epoch": 2.067168049792531,
+      "grad_norm": 0.4819853030166084,
+      "learning_rate": 2.665134419173844e-06,
+      "loss": 0.3207,
+      "step": 2657
+    },
+    {
+      "epoch": 2.0679460580912865,
+      "grad_norm": 0.5285705295358947,
+      "learning_rate": 2.661131311332882e-06,
+      "loss": 0.3277,
+      "step": 2658
+    },
+    {
+      "epoch": 2.0687240663900415,
+      "grad_norm": 0.490396467556533,
+      "learning_rate": 2.6571301217054758e-06,
+      "loss": 0.3269,
+      "step": 2659
+    },
+    {
+      "epoch": 2.069502074688797,
+      "grad_norm": 0.5046679370247246,
+      "learning_rate": 2.653130853573186e-06,
+      "loss": 0.321,
+      "step": 2660
+    },
+    {
+      "epoch": 2.070280082987552,
+      "grad_norm": 0.5440720801465362,
+      "learning_rate": 2.6491335102159942e-06,
+      "loss": 0.3451,
+      "step": 2661
+    },
+    {
+      "epoch": 2.0710580912863072,
+      "grad_norm": 0.5216033458445741,
+      "learning_rate": 2.6451380949123064e-06,
+      "loss": 0.3307,
+      "step": 2662
+    },
+    {
+      "epoch": 2.071836099585062,
+      "grad_norm": 0.550844166661634,
+      "learning_rate": 2.6411446109389456e-06,
+      "loss": 0.3081,
+      "step": 2663
+    },
+    {
+      "epoch": 2.0726141078838176,
+      "grad_norm": 0.5525395670662854,
+      "learning_rate": 2.6371530615711533e-06,
+      "loss": 0.3363,
+      "step": 2664
+    },
+    {
+      "epoch": 2.0733921161825726,
+      "grad_norm": 0.578294332815424,
+      "learning_rate": 2.6331634500825766e-06,
+      "loss": 0.3709,
+      "step": 2665
+    },
+    {
+      "epoch": 2.074170124481328,
+      "grad_norm": 0.5812860548137052,
+      "learning_rate": 2.6291757797452845e-06,
+      "loss": 0.3179,
+      "step": 2666
+    },
+    {
+      "epoch": 2.074948132780083,
+      "grad_norm": 0.5398262682480355,
+      "learning_rate": 2.6251900538297464e-06,
+      "loss": 0.3791,
+      "step": 2667
+    },
+    {
+      "epoch": 2.0757261410788383,
+      "grad_norm": 0.47307920151632366,
+      "learning_rate": 2.62120627560484e-06,
+      "loss": 0.3162,
+      "step": 2668
+    },
+    {
+      "epoch": 2.0765041493775933,
+      "grad_norm": 0.44688557054797395,
+      "learning_rate": 2.6172244483378457e-06,
+      "loss": 0.3044,
+      "step": 2669
+    },
+    {
+      "epoch": 2.0772821576763487,
+      "grad_norm": 0.5288613640652648,
+      "learning_rate": 2.6132445752944374e-06,
+      "loss": 0.3787,
+      "step": 2670
+    },
+    {
+      "epoch": 2.0780601659751037,
+      "grad_norm": 0.5087738538696289,
+      "learning_rate": 2.6092666597386996e-06,
+      "loss": 0.3444,
+      "step": 2671
+    },
+    {
+      "epoch": 2.078838174273859,
+      "grad_norm": 0.49059779993810976,
+      "learning_rate": 2.6052907049330966e-06,
+      "loss": 0.3347,
+      "step": 2672
+    },
+    {
+      "epoch": 2.079616182572614,
+      "grad_norm": 0.5487015792056315,
+      "learning_rate": 2.6013167141384923e-06,
+      "loss": 0.3633,
+      "step": 2673
+    },
+    {
+      "epoch": 2.0803941908713695,
+      "grad_norm": 0.5665265233855062,
+      "learning_rate": 2.597344690614144e-06,
+      "loss": 0.3871,
+      "step": 2674
+    },
+    {
+      "epoch": 2.0811721991701244,
+      "grad_norm": 0.5745152927732928,
+      "learning_rate": 2.593374637617684e-06,
+      "loss": 0.3595,
+      "step": 2675
+    },
+    {
+      "epoch": 2.08195020746888,
+      "grad_norm": 0.4937773285369367,
+      "learning_rate": 2.589406558405142e-06,
+      "loss": 0.3416,
+      "step": 2676
+    },
+    {
+      "epoch": 2.082728215767635,
+      "grad_norm": 0.5528210407296126,
+      "learning_rate": 2.585440456230916e-06,
+      "loss": 0.4024,
+      "step": 2677
+    },
+    {
+      "epoch": 2.08350622406639,
+      "grad_norm": 0.47208743314072193,
+      "learning_rate": 2.58147633434779e-06,
+      "loss": 0.2775,
+      "step": 2678
+    },
+    {
+      "epoch": 2.084284232365145,
+      "grad_norm": 0.498783996595592,
+      "learning_rate": 2.5775141960069234e-06,
+      "loss": 0.3708,
+      "step": 2679
+    },
+    {
+      "epoch": 2.0850622406639006,
+      "grad_norm": 0.4918408230365275,
+      "learning_rate": 2.5735540444578467e-06,
+      "loss": 0.3378,
+      "step": 2680
+    },
+    {
+      "epoch": 2.0858402489626555,
+      "grad_norm": 0.5561077859152129,
+      "learning_rate": 2.569595882948462e-06,
+      "loss": 0.4026,
+      "step": 2681
+    },
+    {
+      "epoch": 2.086618257261411,
+      "grad_norm": 0.49283053368876256,
+      "learning_rate": 2.56563971472504e-06,
+      "loss": 0.3441,
+      "step": 2682
+    },
+    {
+      "epoch": 2.087396265560166,
+      "grad_norm": 0.4883938311046349,
+      "learning_rate": 2.5616855430322154e-06,
+      "loss": 0.3494,
+      "step": 2683
+    },
+    {
+      "epoch": 2.0881742738589213,
+      "grad_norm": 0.48658609867416414,
+      "learning_rate": 2.557733371112985e-06,
+      "loss": 0.3441,
+      "step": 2684
+    },
+    {
+      "epoch": 2.0889522821576763,
+      "grad_norm": 0.4387672733339187,
+      "learning_rate": 2.553783202208707e-06,
+      "loss": 0.3427,
+      "step": 2685
+    },
+    {
+      "epoch": 2.0897302904564317,
+      "grad_norm": 0.5391911487275327,
+      "learning_rate": 2.5498350395590975e-06,
+      "loss": 0.3534,
+      "step": 2686
+    },
+    {
+      "epoch": 2.0905082987551866,
+      "grad_norm": 0.4182473482552586,
+      "learning_rate": 2.545888886402224e-06,
+      "loss": 0.3165,
+      "step": 2687
+    },
+    {
+      "epoch": 2.091286307053942,
+      "grad_norm": 0.4878068544908817,
+      "learning_rate": 2.5419447459745095e-06,
+      "loss": 0.3667,
+      "step": 2688
+    },
+    {
+      "epoch": 2.092064315352697,
+      "grad_norm": 0.5344373089308796,
+      "learning_rate": 2.5380026215107225e-06,
+      "loss": 0.3125,
+      "step": 2689
+    },
+    {
+      "epoch": 2.0928423236514524,
+      "grad_norm": 0.5891915228063328,
+      "learning_rate": 2.5340625162439836e-06,
+      "loss": 0.3687,
+      "step": 2690
+    },
+    {
+      "epoch": 2.0936203319502074,
+      "grad_norm": 0.5617097983563056,
+      "learning_rate": 2.5301244334057484e-06,
+      "loss": 0.3378,
+      "step": 2691
+    },
+    {
+      "epoch": 2.0943983402489628,
+      "grad_norm": 0.48363234134750943,
+      "learning_rate": 2.5261883762258243e-06,
+      "loss": 0.352,
+      "step": 2692
+    },
+    {
+      "epoch": 2.0951763485477177,
+      "grad_norm": 0.5771224731632358,
+      "learning_rate": 2.5222543479323513e-06,
+      "loss": 0.3515,
+      "step": 2693
+    },
+    {
+      "epoch": 2.095954356846473,
+      "grad_norm": 0.4946067793355006,
+      "learning_rate": 2.5183223517518057e-06,
+      "loss": 0.3228,
+      "step": 2694
+    },
+    {
+      "epoch": 2.096732365145228,
+      "grad_norm": 0.5061231916801904,
+      "learning_rate": 2.5143923909090008e-06,
+      "loss": 0.333,
+      "step": 2695
+    },
+    {
+      "epoch": 2.0975103734439835,
+      "grad_norm": 0.49945109221655254,
+      "learning_rate": 2.5104644686270706e-06,
+      "loss": 0.3497,
+      "step": 2696
+    },
+    {
+      "epoch": 2.0982883817427385,
+      "grad_norm": 0.5574993782581119,
+      "learning_rate": 2.506538588127494e-06,
+      "loss": 0.3721,
+      "step": 2697
+    },
+    {
+      "epoch": 2.099066390041494,
+      "grad_norm": 0.5700818280175377,
+      "learning_rate": 2.5026147526300592e-06,
+      "loss": 0.4163,
+      "step": 2698
+    },
+    {
+      "epoch": 2.099844398340249,
+      "grad_norm": 0.5511165239920374,
+      "learning_rate": 2.498692965352884e-06,
+      "loss": 0.3398,
+      "step": 2699
+    },
+    {
+      "epoch": 2.1006224066390042,
+      "grad_norm": 0.4973784457578659,
+      "learning_rate": 2.4947732295124117e-06,
+      "loss": 0.2968,
+      "step": 2700
+    },
+    {
+      "epoch": 2.101400414937759,
+      "grad_norm": 0.614396063818474,
+      "learning_rate": 2.490855548323391e-06,
+      "loss": 0.3707,
+      "step": 2701
+    },
+    {
+      "epoch": 2.1021784232365146,
+      "grad_norm": 0.5092308288313256,
+      "learning_rate": 2.4869399249988956e-06,
+      "loss": 0.2907,
+      "step": 2702
+    },
+    {
+      "epoch": 2.1029564315352696,
+      "grad_norm": 0.5441124467863855,
+      "learning_rate": 2.4830263627503063e-06,
+      "loss": 0.3373,
+      "step": 2703
+    },
+    {
+      "epoch": 2.103734439834025,
+      "grad_norm": 0.5774306776057851,
+      "learning_rate": 2.4791148647873165e-06,
+      "loss": 0.3321,
+      "step": 2704
+    },
+    {
+      "epoch": 2.10451244813278,
+      "grad_norm": 0.5319441211159899,
+      "learning_rate": 2.475205434317925e-06,
+      "loss": 0.3831,
+      "step": 2705
+    },
+    {
+      "epoch": 2.1052904564315353,
+      "grad_norm": 0.4471985506708388,
+      "learning_rate": 2.471298074548434e-06,
+      "loss": 0.3191,
+      "step": 2706
+    },
+    {
+      "epoch": 2.1060684647302903,
+      "grad_norm": 0.5370458841941612,
+      "learning_rate": 2.4673927886834493e-06,
+      "loss": 0.4101,
+      "step": 2707
+    },
+    {
+      "epoch": 2.1068464730290457,
+      "grad_norm": 0.5646695425010381,
+      "learning_rate": 2.463489579925875e-06,
+      "loss": 0.3621,
+      "step": 2708
+    },
+    {
+      "epoch": 2.1076244813278007,
+      "grad_norm": 0.46570037407295617,
+      "learning_rate": 2.4595884514769124e-06,
+      "loss": 0.2971,
+      "step": 2709
+    },
+    {
+      "epoch": 2.108402489626556,
+      "grad_norm": 0.5876151668373774,
+      "learning_rate": 2.4556894065360553e-06,
+      "loss": 0.4134,
+      "step": 2710
+    },
+    {
+      "epoch": 2.109180497925311,
+      "grad_norm": 0.5716831478863226,
+      "learning_rate": 2.4517924483010915e-06,
+      "loss": 0.3017,
+      "step": 2711
+    },
+    {
+      "epoch": 2.1099585062240664,
+      "grad_norm": 0.43826223264155284,
+      "learning_rate": 2.447897579968089e-06,
+      "loss": 0.3237,
+      "step": 2712
+    },
+    {
+      "epoch": 2.1107365145228214,
+      "grad_norm": 0.5062397982546614,
+      "learning_rate": 2.4440048047314145e-06,
+      "loss": 0.4017,
+      "step": 2713
+    },
+    {
+      "epoch": 2.111514522821577,
+      "grad_norm": 0.47766815890085457,
+      "learning_rate": 2.4401141257837096e-06,
+      "loss": 0.3568,
+      "step": 2714
+    },
+    {
+      "epoch": 2.1122925311203318,
+      "grad_norm": 0.5682886323707971,
+      "learning_rate": 2.436225546315899e-06,
+      "loss": 0.3217,
+      "step": 2715
+    },
+    {
+      "epoch": 2.113070539419087,
+      "grad_norm": 0.548148043491232,
+      "learning_rate": 2.432339069517186e-06,
+      "loss": 0.3126,
+      "step": 2716
+    },
+    {
+      "epoch": 2.113848547717842,
+      "grad_norm": 0.511501486789409,
+      "learning_rate": 2.4284546985750442e-06,
+      "loss": 0.337,
+      "step": 2717
+    },
+    {
+      "epoch": 2.1146265560165975,
+      "grad_norm": 0.5321024367611105,
+      "learning_rate": 2.4245724366752314e-06,
+      "loss": 0.3418,
+      "step": 2718
+    },
+    {
+      "epoch": 2.1154045643153525,
+      "grad_norm": 0.4527513269393408,
+      "learning_rate": 2.420692287001763e-06,
+      "loss": 0.294,
+      "step": 2719
+    },
+    {
+      "epoch": 2.116182572614108,
+      "grad_norm": 0.5093570668294939,
+      "learning_rate": 2.4168142527369283e-06,
+      "loss": 0.3463,
+      "step": 2720
+    },
+    {
+      "epoch": 2.116960580912863,
+      "grad_norm": 0.44778385529240544,
+      "learning_rate": 2.4129383370612865e-06,
+      "loss": 0.2871,
+      "step": 2721
+    },
+    {
+      "epoch": 2.1177385892116183,
+      "grad_norm": 0.5875627536032844,
+      "learning_rate": 2.4090645431536456e-06,
+      "loss": 0.3966,
+      "step": 2722
+    },
+    {
+      "epoch": 2.1185165975103732,
+      "grad_norm": 0.47457112386339756,
+      "learning_rate": 2.405192874191091e-06,
+      "loss": 0.2828,
+      "step": 2723
+    },
+    {
+      "epoch": 2.1192946058091287,
+      "grad_norm": 0.5687038285517749,
+      "learning_rate": 2.4013233333489492e-06,
+      "loss": 0.3535,
+      "step": 2724
+    },
+    {
+      "epoch": 2.1200726141078836,
+      "grad_norm": 0.543742736194438,
+      "learning_rate": 2.397455923800811e-06,
+      "loss": 0.333,
+      "step": 2725
+    },
+    {
+      "epoch": 2.120850622406639,
+      "grad_norm": 0.5730722236727601,
+      "learning_rate": 2.393590648718516e-06,
+      "loss": 0.4006,
+      "step": 2726
+    },
+    {
+      "epoch": 2.121628630705394,
+      "grad_norm": 0.5698422413122737,
+      "learning_rate": 2.389727511272153e-06,
+      "loss": 0.3861,
+      "step": 2727
+    },
+    {
+      "epoch": 2.1224066390041494,
+      "grad_norm": 0.49869263479628423,
+      "learning_rate": 2.3858665146300597e-06,
+      "loss": 0.3353,
+      "step": 2728
+    },
+    {
+      "epoch": 2.123184647302905,
+      "grad_norm": 0.6136487014342571,
+      "learning_rate": 2.3820076619588154e-06,
+      "loss": 0.3515,
+      "step": 2729
+    },
+    {
+      "epoch": 2.1239626556016598,
+      "grad_norm": 0.4932662315713693,
+      "learning_rate": 2.3781509564232436e-06,
+      "loss": 0.3442,
+      "step": 2730
+    },
+    {
+      "epoch": 2.1247406639004147,
+      "grad_norm": 0.6164527404854443,
+      "learning_rate": 2.3742964011864047e-06,
+      "loss": 0.3446,
+      "step": 2731
+    },
+    {
+      "epoch": 2.12551867219917,
+      "grad_norm": 0.5442389182906625,
+      "learning_rate": 2.3704439994095957e-06,
+      "loss": 0.3134,
+      "step": 2732
+    },
+    {
+      "epoch": 2.1262966804979255,
+      "grad_norm": 0.5124637181060794,
+      "learning_rate": 2.3665937542523493e-06,
+      "loss": 0.3333,
+      "step": 2733
+    },
+    {
+      "epoch": 2.1270746887966805,
+      "grad_norm": 0.5089726271328416,
+      "learning_rate": 2.3627456688724277e-06,
+      "loss": 0.3564,
+      "step": 2734
+    },
+    {
+      "epoch": 2.1278526970954355,
+      "grad_norm": 0.5627496407231042,
+      "learning_rate": 2.3588997464258222e-06,
+      "loss": 0.3454,
+      "step": 2735
+    },
+    {
+      "epoch": 2.128630705394191,
+      "grad_norm": 0.5450445800686692,
+      "learning_rate": 2.3550559900667508e-06,
+      "loss": 0.3338,
+      "step": 2736
+    },
+    {
+      "epoch": 2.1294087136929463,
+      "grad_norm": 0.5560274503241461,
+      "learning_rate": 2.351214402947656e-06,
+      "loss": 0.3311,
+      "step": 2737
+    },
+    {
+      "epoch": 2.1301867219917012,
+      "grad_norm": 0.6146042587441861,
+      "learning_rate": 2.347374988219195e-06,
+      "loss": 0.3924,
+      "step": 2738
+    },
+    {
+      "epoch": 2.1309647302904566,
+      "grad_norm": 0.5596677883918503,
+      "learning_rate": 2.3435377490302536e-06,
+      "loss": 0.3482,
+      "step": 2739
+    },
+    {
+      "epoch": 2.1317427385892116,
+      "grad_norm": 0.5975464994502611,
+      "learning_rate": 2.3397026885279224e-06,
+      "loss": 0.3822,
+      "step": 2740
+    },
+    {
+      "epoch": 2.132520746887967,
+      "grad_norm": 0.5618437806435863,
+      "learning_rate": 2.3358698098575146e-06,
+      "loss": 0.3804,
+      "step": 2741
+    },
+    {
+      "epoch": 2.133298755186722,
+      "grad_norm": 0.4963614852080694,
+      "learning_rate": 2.3320391161625506e-06,
+      "loss": 0.305,
+      "step": 2742
+    },
+    {
+      "epoch": 2.1340767634854774,
+      "grad_norm": 0.5230331336057213,
+      "learning_rate": 2.328210610584752e-06,
+      "loss": 0.3347,
+      "step": 2743
+    },
+    {
+      "epoch": 2.1348547717842323,
+      "grad_norm": 0.6016105643810923,
+      "learning_rate": 2.32438429626406e-06,
+      "loss": 0.4037,
+      "step": 2744
+    },
+    {
+      "epoch": 2.1356327800829877,
+      "grad_norm": 0.5266929691284897,
+      "learning_rate": 2.320560176338605e-06,
+      "loss": 0.3151,
+      "step": 2745
+    },
+    {
+      "epoch": 2.1364107883817427,
+      "grad_norm": 0.48942851550720723,
+      "learning_rate": 2.3167382539447246e-06,
+      "loss": 0.353,
+      "step": 2746
+    },
+    {
+      "epoch": 2.137188796680498,
+      "grad_norm": 0.5386384626974129,
+      "learning_rate": 2.3129185322169533e-06,
+      "loss": 0.3947,
+      "step": 2747
+    },
+    {
+      "epoch": 2.137966804979253,
+      "grad_norm": 0.5417651659281724,
+      "learning_rate": 2.3091010142880184e-06,
+      "loss": 0.3804,
+      "step": 2748
+    },
+    {
+      "epoch": 2.1387448132780085,
+      "grad_norm": 0.5166387009103842,
+      "learning_rate": 2.3052857032888464e-06,
+      "loss": 0.3102,
+      "step": 2749
+    },
+    {
+      "epoch": 2.1395228215767634,
+      "grad_norm": 0.562199320785755,
+      "learning_rate": 2.301472602348544e-06,
+      "loss": 0.3701,
+      "step": 2750
+    },
+    {
+      "epoch": 2.140300829875519,
+      "grad_norm": 0.4965916906214273,
+      "learning_rate": 2.297661714594412e-06,
+      "loss": 0.3265,
+      "step": 2751
+    },
+    {
+      "epoch": 2.141078838174274,
+      "grad_norm": 0.4909217350708387,
+      "learning_rate": 2.293853043151934e-06,
+      "loss": 0.3686,
+      "step": 2752
+    },
+    {
+      "epoch": 2.141856846473029,
+      "grad_norm": 0.44442751315509066,
+      "learning_rate": 2.290046591144777e-06,
+      "loss": 0.2802,
+      "step": 2753
+    },
+    {
+      "epoch": 2.142634854771784,
+      "grad_norm": 0.5238988272387529,
+      "learning_rate": 2.286242361694786e-06,
+      "loss": 0.3897,
+      "step": 2754
+    },
+    {
+      "epoch": 2.1434128630705396,
+      "grad_norm": 0.546910892348062,
+      "learning_rate": 2.2824403579219857e-06,
+      "loss": 0.3111,
+      "step": 2755
+    },
+    {
+      "epoch": 2.1441908713692945,
+      "grad_norm": 0.49043141999881396,
+      "learning_rate": 2.278640582944572e-06,
+      "loss": 0.3298,
+      "step": 2756
+    },
+    {
+      "epoch": 2.14496887966805,
+      "grad_norm": 0.5345200803579407,
+      "learning_rate": 2.2748430398789163e-06,
+      "loss": 0.3569,
+      "step": 2757
+    },
+    {
+      "epoch": 2.145746887966805,
+      "grad_norm": 0.5797677474231967,
+      "learning_rate": 2.271047731839557e-06,
+      "loss": 0.3445,
+      "step": 2758
+    },
+    {
+      "epoch": 2.1465248962655603,
+      "grad_norm": 0.5636437042894075,
+      "learning_rate": 2.2672546619392003e-06,
+      "loss": 0.3365,
+      "step": 2759
+    },
+    {
+      "epoch": 2.1473029045643153,
+      "grad_norm": 0.45888206259969627,
+      "learning_rate": 2.2634638332887205e-06,
+      "loss": 0.3138,
+      "step": 2760
+    },
+    {
+      "epoch": 2.1480809128630707,
+      "grad_norm": 0.4900764971416145,
+      "learning_rate": 2.2596752489971426e-06,
+      "loss": 0.3357,
+      "step": 2761
+    },
+    {
+      "epoch": 2.1488589211618256,
+      "grad_norm": 0.48313793649887893,
+      "learning_rate": 2.2558889121716642e-06,
+      "loss": 0.3438,
+      "step": 2762
+    },
+    {
+      "epoch": 2.149636929460581,
+      "grad_norm": 0.502351521244763,
+      "learning_rate": 2.2521048259176353e-06,
+      "loss": 0.3678,
+      "step": 2763
+    },
+    {
+      "epoch": 2.150414937759336,
+      "grad_norm": 0.434189978924552,
+      "learning_rate": 2.248322993338552e-06,
+      "loss": 0.2881,
+      "step": 2764
+    },
+    {
+      "epoch": 2.1511929460580914,
+      "grad_norm": 0.5256994061051214,
+      "learning_rate": 2.2445434175360763e-06,
+      "loss": 0.3779,
+      "step": 2765
+    },
+    {
+      "epoch": 2.1519709543568464,
+      "grad_norm": 0.5136126283911633,
+      "learning_rate": 2.240766101610005e-06,
+      "loss": 0.3297,
+      "step": 2766
+    },
+    {
+      "epoch": 2.152748962655602,
+      "grad_norm": 0.45601334298791607,
+      "learning_rate": 2.236991048658295e-06,
+      "loss": 0.2841,
+      "step": 2767
+    },
+    {
+      "epoch": 2.1535269709543567,
+      "grad_norm": 0.45662335615834665,
+      "learning_rate": 2.2332182617770355e-06,
+      "loss": 0.2929,
+      "step": 2768
+    },
+    {
+      "epoch": 2.154304979253112,
+      "grad_norm": 0.5619791667739492,
+      "learning_rate": 2.229447744060461e-06,
+      "loss": 0.3387,
+      "step": 2769
+    },
+    {
+      "epoch": 2.155082987551867,
+      "grad_norm": 0.5196774684271482,
+      "learning_rate": 2.2256794986009535e-06,
+      "loss": 0.3432,
+      "step": 2770
+    },
+    {
+      "epoch": 2.1558609958506225,
+      "grad_norm": 0.4933504209381443,
+      "learning_rate": 2.2219135284890165e-06,
+      "loss": 0.3027,
+      "step": 2771
+    },
+    {
+      "epoch": 2.1566390041493775,
+      "grad_norm": 0.5882655332988618,
+      "learning_rate": 2.218149836813298e-06,
+      "loss": 0.3728,
+      "step": 2772
+    },
+    {
+      "epoch": 2.157417012448133,
+      "grad_norm": 0.6143959093235075,
+      "learning_rate": 2.214388426660575e-06,
+      "loss": 0.3867,
+      "step": 2773
+    },
+    {
+      "epoch": 2.158195020746888,
+      "grad_norm": 0.5955767553687161,
+      "learning_rate": 2.2106293011157487e-06,
+      "loss": 0.3708,
+      "step": 2774
+    },
+    {
+      "epoch": 2.1589730290456433,
+      "grad_norm": 0.5231376210123042,
+      "learning_rate": 2.206872463261858e-06,
+      "loss": 0.3138,
+      "step": 2775
+    },
+    {
+      "epoch": 2.159751037344398,
+      "grad_norm": 0.5641473456727385,
+      "learning_rate": 2.2031179161800507e-06,
+      "loss": 0.3681,
+      "step": 2776
+    },
+    {
+      "epoch": 2.1605290456431536,
+      "grad_norm": 0.5356789582256931,
+      "learning_rate": 2.1993656629496068e-06,
+      "loss": 0.3597,
+      "step": 2777
+    },
+    {
+      "epoch": 2.1613070539419086,
+      "grad_norm": 0.5691974172811576,
+      "learning_rate": 2.1956157066479207e-06,
+      "loss": 0.3709,
+      "step": 2778
+    },
+    {
+      "epoch": 2.162085062240664,
+      "grad_norm": 0.5066054595163624,
+      "learning_rate": 2.191868050350503e-06,
+      "loss": 0.3558,
+      "step": 2779
+    },
+    {
+      "epoch": 2.162863070539419,
+      "grad_norm": 0.5523594243965524,
+      "learning_rate": 2.188122697130979e-06,
+      "loss": 0.3826,
+      "step": 2780
+    },
+    {
+      "epoch": 2.1636410788381744,
+      "grad_norm": 0.5339601633455411,
+      "learning_rate": 2.1843796500610858e-06,
+      "loss": 0.3094,
+      "step": 2781
+    },
+    {
+      "epoch": 2.1644190871369293,
+      "grad_norm": 0.5429939806329077,
+      "learning_rate": 2.180638912210667e-06,
+      "loss": 0.3821,
+      "step": 2782
+    },
+    {
+      "epoch": 2.1651970954356847,
+      "grad_norm": 0.5311059143637099,
+      "learning_rate": 2.176900486647673e-06,
+      "loss": 0.3381,
+      "step": 2783
+    },
+    {
+      "epoch": 2.1659751037344397,
+      "grad_norm": 0.5846295890587605,
+      "learning_rate": 2.1731643764381593e-06,
+      "loss": 0.2979,
+      "step": 2784
+    },
+    {
+      "epoch": 2.166753112033195,
+      "grad_norm": 0.5207934871845048,
+      "learning_rate": 2.169430584646281e-06,
+      "loss": 0.3062,
+      "step": 2785
+    },
+    {
+      "epoch": 2.16753112033195,
+      "grad_norm": 0.5137448809457972,
+      "learning_rate": 2.1656991143342937e-06,
+      "loss": 0.3457,
+      "step": 2786
+    },
+    {
+      "epoch": 2.1683091286307055,
+      "grad_norm": 0.535910169613968,
+      "learning_rate": 2.161969968562542e-06,
+      "loss": 0.3522,
+      "step": 2787
+    },
+    {
+      "epoch": 2.1690871369294604,
+      "grad_norm": 0.4895676625893986,
+      "learning_rate": 2.1582431503894746e-06,
+      "loss": 0.3187,
+      "step": 2788
+    },
+    {
+      "epoch": 2.169865145228216,
+      "grad_norm": 0.5373403347458946,
+      "learning_rate": 2.1545186628716274e-06,
+      "loss": 0.3559,
+      "step": 2789
+    },
+    {
+      "epoch": 2.170643153526971,
+      "grad_norm": 0.4798647253278432,
+      "learning_rate": 2.150796509063617e-06,
+      "loss": 0.3035,
+      "step": 2790
+    },
+    {
+      "epoch": 2.171421161825726,
+      "grad_norm": 0.512376746861512,
+      "learning_rate": 2.1470766920181603e-06,
+      "loss": 0.3204,
+      "step": 2791
+    },
+    {
+      "epoch": 2.172199170124481,
+      "grad_norm": 0.5265798142599396,
+      "learning_rate": 2.1433592147860438e-06,
+      "loss": 0.394,
+      "step": 2792
+    },
+    {
+      "epoch": 2.1729771784232366,
+      "grad_norm": 0.5625939555018764,
+      "learning_rate": 2.1396440804161444e-06,
+      "loss": 0.3994,
+      "step": 2793
+    },
+    {
+      "epoch": 2.1737551867219915,
+      "grad_norm": 0.4968399431642992,
+      "learning_rate": 2.135931291955413e-06,
+      "loss": 0.293,
+      "step": 2794
+    },
+    {
+      "epoch": 2.174533195020747,
+      "grad_norm": 0.5488272473029706,
+      "learning_rate": 2.132220852448876e-06,
+      "loss": 0.391,
+      "step": 2795
+    },
+    {
+      "epoch": 2.175311203319502,
+      "grad_norm": 0.5136748198547881,
+      "learning_rate": 2.1285127649396418e-06,
+      "loss": 0.3386,
+      "step": 2796
+    },
+    {
+      "epoch": 2.1760892116182573,
+      "grad_norm": 0.4608058300146933,
+      "learning_rate": 2.1248070324688773e-06,
+      "loss": 0.3056,
+      "step": 2797
+    },
+    {
+      "epoch": 2.1768672199170123,
+      "grad_norm": 0.5164249791830957,
+      "learning_rate": 2.121103658075825e-06,
+      "loss": 0.3302,
+      "step": 2798
+    },
+    {
+      "epoch": 2.1776452282157677,
+      "grad_norm": 0.45939478094107083,
+      "learning_rate": 2.117402644797792e-06,
+      "loss": 0.3035,
+      "step": 2799
+    },
+    {
+      "epoch": 2.1784232365145226,
+      "grad_norm": 0.5421460777652679,
+      "learning_rate": 2.11370399567015e-06,
+      "loss": 0.3701,
+      "step": 2800
+    },
+    {
+      "epoch": 2.179201244813278,
+      "grad_norm": 0.48126236427357966,
+      "learning_rate": 2.1100077137263304e-06,
+      "loss": 0.3053,
+      "step": 2801
+    },
+    {
+      "epoch": 2.179979253112033,
+      "grad_norm": 0.5168083672854645,
+      "learning_rate": 2.106313801997824e-06,
+      "loss": 0.3457,
+      "step": 2802
+    },
+    {
+      "epoch": 2.1807572614107884,
+      "grad_norm": 0.5463609047442576,
+      "learning_rate": 2.102622263514177e-06,
+      "loss": 0.3436,
+      "step": 2803
+    },
+    {
+      "epoch": 2.1815352697095434,
+      "grad_norm": 0.4816421255282539,
+      "learning_rate": 2.0989331013029897e-06,
+      "loss": 0.311,
+      "step": 2804
+    },
+    {
+      "epoch": 2.1823132780082988,
+      "grad_norm": 0.5403565780664573,
+      "learning_rate": 2.0952463183899134e-06,
+      "loss": 0.3902,
+      "step": 2805
+    },
+    {
+      "epoch": 2.1830912863070537,
+      "grad_norm": 0.49413006546771443,
+      "learning_rate": 2.0915619177986477e-06,
+      "loss": 0.3459,
+      "step": 2806
+    },
+    {
+      "epoch": 2.183869294605809,
+      "grad_norm": 0.4690524611460225,
+      "learning_rate": 2.0878799025509412e-06,
+      "loss": 0.3495,
+      "step": 2807
+    },
+    {
+      "epoch": 2.1846473029045645,
+      "grad_norm": 0.519413095098871,
+      "learning_rate": 2.084200275666577e-06,
+      "loss": 0.3438,
+      "step": 2808
+    },
+    {
+      "epoch": 2.1854253112033195,
+      "grad_norm": 0.4573826225702993,
+      "learning_rate": 2.0805230401633926e-06,
+      "loss": 0.3129,
+      "step": 2809
+    },
+    {
+      "epoch": 2.1862033195020745,
+      "grad_norm": 0.6045822896800577,
+      "learning_rate": 2.0768481990572562e-06,
+      "loss": 0.3765,
+      "step": 2810
+    },
+    {
+      "epoch": 2.18698132780083,
+      "grad_norm": 0.5379338065609193,
+      "learning_rate": 2.0731757553620695e-06,
+      "loss": 0.3786,
+      "step": 2811
+    },
+    {
+      "epoch": 2.1877593360995853,
+      "grad_norm": 0.5663782351701511,
+      "learning_rate": 2.0695057120897794e-06,
+      "loss": 0.3697,
+      "step": 2812
+    },
+    {
+      "epoch": 2.1885373443983402,
+      "grad_norm": 0.4769231645084492,
+      "learning_rate": 2.0658380722503484e-06,
+      "loss": 0.2904,
+      "step": 2813
+    },
+    {
+      "epoch": 2.189315352697095,
+      "grad_norm": 0.5109286088907352,
+      "learning_rate": 2.0621728388517842e-06,
+      "loss": 0.3072,
+      "step": 2814
+    },
+    {
+      "epoch": 2.1900933609958506,
+      "grad_norm": 0.6843691819179663,
+      "learning_rate": 2.058510014900108e-06,
+      "loss": 0.3969,
+      "step": 2815
+    },
+    {
+      "epoch": 2.190871369294606,
+      "grad_norm": 0.5787212506943372,
+      "learning_rate": 2.054849603399369e-06,
+      "loss": 0.39,
+      "step": 2816
+    },
+    {
+      "epoch": 2.191649377593361,
+      "grad_norm": 0.548485124497276,
+      "learning_rate": 2.0511916073516446e-06,
+      "loss": 0.3157,
+      "step": 2817
+    },
+    {
+      "epoch": 2.1924273858921164,
+      "grad_norm": 0.537004272181865,
+      "learning_rate": 2.0475360297570196e-06,
+      "loss": 0.3773,
+      "step": 2818
+    },
+    {
+      "epoch": 2.1932053941908713,
+      "grad_norm": 0.5074428653285438,
+      "learning_rate": 2.043882873613603e-06,
+      "loss": 0.3349,
+      "step": 2819
+    },
+    {
+      "epoch": 2.1939834024896268,
+      "grad_norm": 0.535496158717886,
+      "learning_rate": 2.0402321419175154e-06,
+      "loss": 0.3415,
+      "step": 2820
+    },
+    {
+      "epoch": 2.1947614107883817,
+      "grad_norm": 0.5527749396907288,
+      "learning_rate": 2.0365838376628895e-06,
+      "loss": 0.37,
+      "step": 2821
+    },
+    {
+      "epoch": 2.195539419087137,
+      "grad_norm": 0.4678660395258337,
+      "learning_rate": 2.032937963841868e-06,
+      "loss": 0.3294,
+      "step": 2822
+    },
+    {
+      "epoch": 2.196317427385892,
+      "grad_norm": 0.533640939287966,
+      "learning_rate": 2.0292945234445978e-06,
+      "loss": 0.3721,
+      "step": 2823
+    },
+    {
+      "epoch": 2.1970954356846475,
+      "grad_norm": 0.5287997620056658,
+      "learning_rate": 2.0256535194592325e-06,
+      "loss": 0.3323,
+      "step": 2824
+    },
+    {
+      "epoch": 2.1978734439834025,
+      "grad_norm": 0.5524280475489568,
+      "learning_rate": 2.0220149548719265e-06,
+      "loss": 0.3686,
+      "step": 2825
+    },
+    {
+      "epoch": 2.198651452282158,
+      "grad_norm": 0.5450922710241994,
+      "learning_rate": 2.0183788326668334e-06,
+      "loss": 0.3123,
+      "step": 2826
+    },
+    {
+      "epoch": 2.199429460580913,
+      "grad_norm": 0.6070685773993206,
+      "learning_rate": 2.0147451558261044e-06,
+      "loss": 0.3659,
+      "step": 2827
+    },
+    {
+      "epoch": 2.2002074688796682,
+      "grad_norm": 0.5353869566076621,
+      "learning_rate": 2.0111139273298858e-06,
+      "loss": 0.3293,
+      "step": 2828
+    },
+    {
+      "epoch": 2.200985477178423,
+      "grad_norm": 0.5719806548220162,
+      "learning_rate": 2.0074851501563097e-06,
+      "loss": 0.3071,
+      "step": 2829
+    },
+    {
+      "epoch": 2.2017634854771786,
+      "grad_norm": 0.5807668494353859,
+      "learning_rate": 2.0038588272815087e-06,
+      "loss": 0.3412,
+      "step": 2830
+    },
+    {
+      "epoch": 2.2025414937759336,
+      "grad_norm": 0.5012125394670285,
+      "learning_rate": 2.0002349616795936e-06,
+      "loss": 0.3138,
+      "step": 2831
+    },
+    {
+      "epoch": 2.203319502074689,
+      "grad_norm": 0.5055218732843547,
+      "learning_rate": 1.996613556322664e-06,
+      "loss": 0.3432,
+      "step": 2832
+    },
+    {
+      "epoch": 2.204097510373444,
+      "grad_norm": 0.454915749853858,
+      "learning_rate": 1.9929946141808017e-06,
+      "loss": 0.3196,
+      "step": 2833
+    },
+    {
+      "epoch": 2.2048755186721993,
+      "grad_norm": 0.5624599630640468,
+      "learning_rate": 1.9893781382220625e-06,
+      "loss": 0.359,
+      "step": 2834
+    },
+    {
+      "epoch": 2.2056535269709543,
+      "grad_norm": 0.5364675934120889,
+      "learning_rate": 1.9857641314124908e-06,
+      "loss": 0.3308,
+      "step": 2835
+    },
+    {
+      "epoch": 2.2064315352697097,
+      "grad_norm": 0.5583255251605639,
+      "learning_rate": 1.982152596716095e-06,
+      "loss": 0.3562,
+      "step": 2836
+    },
+    {
+      "epoch": 2.2072095435684647,
+      "grad_norm": 0.49907634649615285,
+      "learning_rate": 1.9785435370948597e-06,
+      "loss": 0.3444,
+      "step": 2837
+    },
+    {
+      "epoch": 2.20798755186722,
+      "grad_norm": 0.4992457675095568,
+      "learning_rate": 1.9749369555087456e-06,
+      "loss": 0.2795,
+      "step": 2838
+    },
+    {
+      "epoch": 2.208765560165975,
+      "grad_norm": 0.49355390606868993,
+      "learning_rate": 1.971332854915669e-06,
+      "loss": 0.3538,
+      "step": 2839
+    },
+    {
+      "epoch": 2.2095435684647304,
+      "grad_norm": 0.4764439422990594,
+      "learning_rate": 1.967731238271526e-06,
+      "loss": 0.3358,
+      "step": 2840
+    },
+    {
+      "epoch": 2.2103215767634854,
+      "grad_norm": 0.480144435268386,
+      "learning_rate": 1.964132108530161e-06,
+      "loss": 0.3545,
+      "step": 2841
+    },
+    {
+      "epoch": 2.211099585062241,
+      "grad_norm": 0.44752540940589614,
+      "learning_rate": 1.9605354686433876e-06,
+      "loss": 0.3088,
+      "step": 2842
+    },
+    {
+      "epoch": 2.2118775933609958,
+      "grad_norm": 0.5317709388862399,
+      "learning_rate": 1.956941321560976e-06,
+      "loss": 0.3413,
+      "step": 2843
+    },
+    {
+      "epoch": 2.212655601659751,
+      "grad_norm": 0.5250896363336793,
+      "learning_rate": 1.953349670230651e-06,
+      "loss": 0.3104,
+      "step": 2844
+    },
+    {
+      "epoch": 2.213433609958506,
+      "grad_norm": 0.5582520758265934,
+      "learning_rate": 1.9497605175980897e-06,
+      "loss": 0.3555,
+      "step": 2845
+    },
+    {
+      "epoch": 2.2142116182572615,
+      "grad_norm": 0.5614123788773931,
+      "learning_rate": 1.9461738666069224e-06,
+      "loss": 0.3315,
+      "step": 2846
+    },
+    {
+      "epoch": 2.2149896265560165,
+      "grad_norm": 0.5431950868527019,
+      "learning_rate": 1.942589720198725e-06,
+      "loss": 0.4078,
+      "step": 2847
+    },
+    {
+      "epoch": 2.215767634854772,
+      "grad_norm": 0.5329635080931919,
+      "learning_rate": 1.9390080813130214e-06,
+      "loss": 0.3309,
+      "step": 2848
+    },
+    {
+      "epoch": 2.216545643153527,
+      "grad_norm": 0.4518780396783634,
+      "learning_rate": 1.9354289528872773e-06,
+      "loss": 0.3283,
+      "step": 2849
+    },
+    {
+      "epoch": 2.2173236514522823,
+      "grad_norm": 0.5063905035886941,
+      "learning_rate": 1.9318523378569003e-06,
+      "loss": 0.2959,
+      "step": 2850
+    },
+    {
+      "epoch": 2.2181016597510372,
+      "grad_norm": 0.5371015362708798,
+      "learning_rate": 1.9282782391552368e-06,
+      "loss": 0.3521,
+      "step": 2851
+    },
+    {
+      "epoch": 2.2188796680497926,
+      "grad_norm": 0.4888189475697543,
+      "learning_rate": 1.9247066597135694e-06,
+      "loss": 0.3411,
+      "step": 2852
+    },
+    {
+      "epoch": 2.2196576763485476,
+      "grad_norm": 0.4919797701314707,
+      "learning_rate": 1.9211376024611132e-06,
+      "loss": 0.3372,
+      "step": 2853
+    },
+    {
+      "epoch": 2.220435684647303,
+      "grad_norm": 0.5122959611255326,
+      "learning_rate": 1.9175710703250188e-06,
+      "loss": 0.3206,
+      "step": 2854
+    },
+    {
+      "epoch": 2.221213692946058,
+      "grad_norm": 0.5913852487421755,
+      "learning_rate": 1.914007066230357e-06,
+      "loss": 0.3961,
+      "step": 2855
+    },
+    {
+      "epoch": 2.2219917012448134,
+      "grad_norm": 0.4743815031593825,
+      "learning_rate": 1.910445593100139e-06,
+      "loss": 0.331,
+      "step": 2856
+    },
+    {
+      "epoch": 2.2227697095435683,
+      "grad_norm": 0.5008079856431371,
+      "learning_rate": 1.9068866538552844e-06,
+      "loss": 0.3338,
+      "step": 2857
+    },
+    {
+      "epoch": 2.2235477178423237,
+      "grad_norm": 0.6303398739470664,
+      "learning_rate": 1.9033302514146478e-06,
+      "loss": 0.3578,
+      "step": 2858
+    },
+    {
+      "epoch": 2.2243257261410787,
+      "grad_norm": 0.5111734211345424,
+      "learning_rate": 1.899776388694999e-06,
+      "loss": 0.3481,
+      "step": 2859
+    },
+    {
+      "epoch": 2.225103734439834,
+      "grad_norm": 0.5608444391391786,
+      "learning_rate": 1.8962250686110178e-06,
+      "loss": 0.3295,
+      "step": 2860
+    },
+    {
+      "epoch": 2.225881742738589,
+      "grad_norm": 0.5004032159223387,
+      "learning_rate": 1.892676294075312e-06,
+      "loss": 0.3509,
+      "step": 2861
+    },
+    {
+      "epoch": 2.2266597510373445,
+      "grad_norm": 0.49306368053557514,
+      "learning_rate": 1.8891300679983899e-06,
+      "loss": 0.3435,
+      "step": 2862
+    },
+    {
+      "epoch": 2.2274377593360994,
+      "grad_norm": 0.5705984143983756,
+      "learning_rate": 1.8855863932886731e-06,
+      "loss": 0.3722,
+      "step": 2863
+    },
+    {
+      "epoch": 2.228215767634855,
+      "grad_norm": 0.4905565964135863,
+      "learning_rate": 1.8820452728524978e-06,
+      "loss": 0.3448,
+      "step": 2864
+    },
+    {
+      "epoch": 2.22899377593361,
+      "grad_norm": 0.5257120766639164,
+      "learning_rate": 1.8785067095940928e-06,
+      "loss": 0.3435,
+      "step": 2865
+    },
+    {
+      "epoch": 2.229771784232365,
+      "grad_norm": 0.623129476889233,
+      "learning_rate": 1.8749707064156025e-06,
+      "loss": 0.4253,
+      "step": 2866
+    },
+    {
+      "epoch": 2.23054979253112,
+      "grad_norm": 0.5318172190375682,
+      "learning_rate": 1.8714372662170605e-06,
+      "loss": 0.3289,
+      "step": 2867
+    },
+    {
+      "epoch": 2.2313278008298756,
+      "grad_norm": 0.5693644860726017,
+      "learning_rate": 1.8679063918964047e-06,
+      "loss": 0.3684,
+      "step": 2868
+    },
+    {
+      "epoch": 2.2321058091286305,
+      "grad_norm": 0.5042762570462336,
+      "learning_rate": 1.8643780863494676e-06,
+      "loss": 0.2937,
+      "step": 2869
+    },
+    {
+      "epoch": 2.232883817427386,
+      "grad_norm": 0.5034119284383248,
+      "learning_rate": 1.860852352469974e-06,
+      "loss": 0.3933,
+      "step": 2870
+    },
+    {
+      "epoch": 2.233661825726141,
+      "grad_norm": 0.533289178170175,
+      "learning_rate": 1.8573291931495408e-06,
+      "loss": 0.3423,
+      "step": 2871
+    },
+    {
+      "epoch": 2.2344398340248963,
+      "grad_norm": 0.5670150595157059,
+      "learning_rate": 1.8538086112776722e-06,
+      "loss": 0.3317,
+      "step": 2872
+    },
+    {
+      "epoch": 2.2352178423236513,
+      "grad_norm": 0.5145889679546656,
+      "learning_rate": 1.8502906097417588e-06,
+      "loss": 0.3699,
+      "step": 2873
+    },
+    {
+      "epoch": 2.2359958506224067,
+      "grad_norm": 0.48593320428659686,
+      "learning_rate": 1.8467751914270754e-06,
+      "loss": 0.3104,
+      "step": 2874
+    },
+    {
+      "epoch": 2.2367738589211617,
+      "grad_norm": 0.5252689277773418,
+      "learning_rate": 1.8432623592167781e-06,
+      "loss": 0.3391,
+      "step": 2875
+    },
+    {
+      "epoch": 2.237551867219917,
+      "grad_norm": 0.47190527825405904,
+      "learning_rate": 1.8397521159919012e-06,
+      "loss": 0.312,
+      "step": 2876
+    },
+    {
+      "epoch": 2.238329875518672,
+      "grad_norm": 0.4843142401288107,
+      "learning_rate": 1.836244464631357e-06,
+      "loss": 0.3625,
+      "step": 2877
+    },
+    {
+      "epoch": 2.2391078838174274,
+      "grad_norm": 0.5459963977712915,
+      "learning_rate": 1.832739408011931e-06,
+      "loss": 0.3803,
+      "step": 2878
+    },
+    {
+      "epoch": 2.2398858921161824,
+      "grad_norm": 0.5026701271297953,
+      "learning_rate": 1.8292369490082811e-06,
+      "loss": 0.3157,
+      "step": 2879
+    },
+    {
+      "epoch": 2.240663900414938,
+      "grad_norm": 0.4994024808870511,
+      "learning_rate": 1.8257370904929361e-06,
+      "loss": 0.3313,
+      "step": 2880
+    },
+    {
+      "epoch": 2.2414419087136928,
+      "grad_norm": 0.5642753568976328,
+      "learning_rate": 1.8222398353362857e-06,
+      "loss": 0.3573,
+      "step": 2881
+    },
+    {
+      "epoch": 2.242219917012448,
+      "grad_norm": 0.5360524287588306,
+      "learning_rate": 1.8187451864065964e-06,
+      "loss": 0.3862,
+      "step": 2882
+    },
+    {
+      "epoch": 2.242997925311203,
+      "grad_norm": 0.5149713090844903,
+      "learning_rate": 1.8152531465699847e-06,
+      "loss": 0.3324,
+      "step": 2883
+    },
+    {
+      "epoch": 2.2437759336099585,
+      "grad_norm": 0.5175046738215061,
+      "learning_rate": 1.8117637186904325e-06,
+      "loss": 0.3543,
+      "step": 2884
+    },
+    {
+      "epoch": 2.2445539419087135,
+      "grad_norm": 0.5488300842671736,
+      "learning_rate": 1.8082769056297856e-06,
+      "loss": 0.3956,
+      "step": 2885
+    },
+    {
+      "epoch": 2.245331950207469,
+      "grad_norm": 0.53432931255888,
+      "learning_rate": 1.804792710247732e-06,
+      "loss": 0.3176,
+      "step": 2886
+    },
+    {
+      "epoch": 2.2461099585062243,
+      "grad_norm": 0.5265044430921535,
+      "learning_rate": 1.8013111354018276e-06,
+      "loss": 0.38,
+      "step": 2887
+    },
+    {
+      "epoch": 2.2468879668049793,
+      "grad_norm": 0.5060099534317481,
+      "learning_rate": 1.797832183947466e-06,
+      "loss": 0.3439,
+      "step": 2888
+    },
+    {
+      "epoch": 2.2476659751037342,
+      "grad_norm": 0.4716729447185719,
+      "learning_rate": 1.7943558587378973e-06,
+      "loss": 0.3281,
+      "step": 2889
+    },
+    {
+      "epoch": 2.2484439834024896,
+      "grad_norm": 0.4791587603638165,
+      "learning_rate": 1.7908821626242152e-06,
+      "loss": 0.2944,
+      "step": 2890
+    },
+    {
+      "epoch": 2.249221991701245,
+      "grad_norm": 0.4677772313513279,
+      "learning_rate": 1.7874110984553567e-06,
+      "loss": 0.3194,
+      "step": 2891
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.572500227639187,
+      "learning_rate": 1.7839426690781014e-06,
+      "loss": 0.3885,
+      "step": 2892
+    },
+    {
+      "epoch": 2.250778008298755,
+      "grad_norm": 0.5455348190044512,
+      "learning_rate": 1.7804768773370673e-06,
+      "loss": 0.3626,
+      "step": 2893
+    },
+    {
+      "epoch": 2.2515560165975104,
+      "grad_norm": 0.4954271928099586,
+      "learning_rate": 1.7770137260747095e-06,
+      "loss": 0.3202,
+      "step": 2894
+    },
+    {
+      "epoch": 2.2523340248962658,
+      "grad_norm": 0.49289325080780677,
+      "learning_rate": 1.773553218131317e-06,
+      "loss": 0.3178,
+      "step": 2895
+    },
+    {
+      "epoch": 2.2531120331950207,
+      "grad_norm": 0.5559409411593644,
+      "learning_rate": 1.7700953563450108e-06,
+      "loss": 0.3795,
+      "step": 2896
+    },
+    {
+      "epoch": 2.2538900414937757,
+      "grad_norm": 0.5457850773240672,
+      "learning_rate": 1.7666401435517433e-06,
+      "loss": 0.3096,
+      "step": 2897
+    },
+    {
+      "epoch": 2.254668049792531,
+      "grad_norm": 0.596855527356718,
+      "learning_rate": 1.7631875825852917e-06,
+      "loss": 0.2994,
+      "step": 2898
+    },
+    {
+      "epoch": 2.2554460580912865,
+      "grad_norm": 0.47177198100787565,
+      "learning_rate": 1.7597376762772605e-06,
+      "loss": 0.3305,
+      "step": 2899
+    },
+    {
+      "epoch": 2.2562240663900415,
+      "grad_norm": 0.5518331101914562,
+      "learning_rate": 1.7562904274570758e-06,
+      "loss": 0.3327,
+      "step": 2900
+    },
+    {
+      "epoch": 2.2570020746887964,
+      "grad_norm": 0.5711784440065772,
+      "learning_rate": 1.7528458389519865e-06,
+      "loss": 0.3585,
+      "step": 2901
+    },
+    {
+      "epoch": 2.257780082987552,
+      "grad_norm": 0.49768797645112844,
+      "learning_rate": 1.7494039135870527e-06,
+      "loss": 0.2951,
+      "step": 2902
+    },
+    {
+      "epoch": 2.2585580912863072,
+      "grad_norm": 0.5581190143678272,
+      "learning_rate": 1.7459646541851622e-06,
+      "loss": 0.355,
+      "step": 2903
+    },
+    {
+      "epoch": 2.259336099585062,
+      "grad_norm": 0.506714262140307,
+      "learning_rate": 1.7425280635670023e-06,
+      "loss": 0.3315,
+      "step": 2904
+    },
+    {
+      "epoch": 2.2601141078838176,
+      "grad_norm": 0.6861684229340813,
+      "learning_rate": 1.739094144551084e-06,
+      "loss": 0.4504,
+      "step": 2905
+    },
+    {
+      "epoch": 2.2608921161825726,
+      "grad_norm": 0.5970383340101914,
+      "learning_rate": 1.7356628999537222e-06,
+      "loss": 0.3624,
+      "step": 2906
+    },
+    {
+      "epoch": 2.261670124481328,
+      "grad_norm": 0.5226497235339852,
+      "learning_rate": 1.7322343325890328e-06,
+      "loss": 0.3023,
+      "step": 2907
+    },
+    {
+      "epoch": 2.262448132780083,
+      "grad_norm": 0.5426434914255696,
+      "learning_rate": 1.7288084452689475e-06,
+      "loss": 0.3765,
+      "step": 2908
+    },
+    {
+      "epoch": 2.2632261410788383,
+      "grad_norm": 0.5134937648991882,
+      "learning_rate": 1.7253852408031895e-06,
+      "loss": 0.3376,
+      "step": 2909
+    },
+    {
+      "epoch": 2.2640041493775933,
+      "grad_norm": 0.5252982395777775,
+      "learning_rate": 1.721964721999287e-06,
+      "loss": 0.3323,
+      "step": 2910
+    },
+    {
+      "epoch": 2.2647821576763487,
+      "grad_norm": 0.49586532834365005,
+      "learning_rate": 1.7185468916625647e-06,
+      "loss": 0.2932,
+      "step": 2911
+    },
+    {
+      "epoch": 2.2655601659751037,
+      "grad_norm": 0.5334889192684478,
+      "learning_rate": 1.7151317525961402e-06,
+      "loss": 0.4249,
+      "step": 2912
+    },
+    {
+      "epoch": 2.266338174273859,
+      "grad_norm": 0.47178878138510955,
+      "learning_rate": 1.7117193076009308e-06,
+      "loss": 0.3082,
+      "step": 2913
+    },
+    {
+      "epoch": 2.267116182572614,
+      "grad_norm": 0.5157495465402331,
+      "learning_rate": 1.708309559475635e-06,
+      "loss": 0.3474,
+      "step": 2914
+    },
+    {
+      "epoch": 2.2678941908713695,
+      "grad_norm": 0.5258122410858934,
+      "learning_rate": 1.7049025110167444e-06,
+      "loss": 0.3448,
+      "step": 2915
+    },
+    {
+      "epoch": 2.2686721991701244,
+      "grad_norm": 0.4287343041399282,
+      "learning_rate": 1.7014981650185368e-06,
+      "loss": 0.2973,
+      "step": 2916
+    },
+    {
+      "epoch": 2.26945020746888,
+      "grad_norm": 0.5230697720929522,
+      "learning_rate": 1.6980965242730713e-06,
+      "loss": 0.3898,
+      "step": 2917
+    },
+    {
+      "epoch": 2.270228215767635,
+      "grad_norm": 0.6212598518489162,
+      "learning_rate": 1.6946975915701907e-06,
+      "loss": 0.375,
+      "step": 2918
+    },
+    {
+      "epoch": 2.27100622406639,
+      "grad_norm": 0.5330936217214772,
+      "learning_rate": 1.6913013696975155e-06,
+      "loss": 0.3427,
+      "step": 2919
+    },
+    {
+      "epoch": 2.271784232365145,
+      "grad_norm": 0.4898764990278097,
+      "learning_rate": 1.6879078614404432e-06,
+      "loss": 0.3279,
+      "step": 2920
+    },
+    {
+      "epoch": 2.2725622406639006,
+      "grad_norm": 0.5491894252365416,
+      "learning_rate": 1.684517069582145e-06,
+      "loss": 0.377,
+      "step": 2921
+    },
+    {
+      "epoch": 2.2733402489626555,
+      "grad_norm": 0.5839792897825692,
+      "learning_rate": 1.6811289969035654e-06,
+      "loss": 0.3634,
+      "step": 2922
+    },
+    {
+      "epoch": 2.274118257261411,
+      "grad_norm": 0.5980515690458624,
+      "learning_rate": 1.6777436461834184e-06,
+      "loss": 0.3331,
+      "step": 2923
+    },
+    {
+      "epoch": 2.274896265560166,
+      "grad_norm": 0.5468955962950518,
+      "learning_rate": 1.674361020198187e-06,
+      "loss": 0.3727,
+      "step": 2924
+    },
+    {
+      "epoch": 2.2756742738589213,
+      "grad_norm": 0.5396812482614491,
+      "learning_rate": 1.6709811217221127e-06,
+      "loss": 0.4139,
+      "step": 2925
+    },
+    {
+      "epoch": 2.2764522821576763,
+      "grad_norm": 0.4613414649116614,
+      "learning_rate": 1.6676039535272115e-06,
+      "loss": 0.3034,
+      "step": 2926
+    },
+    {
+      "epoch": 2.2772302904564317,
+      "grad_norm": 0.5244086803146649,
+      "learning_rate": 1.664229518383253e-06,
+      "loss": 0.362,
+      "step": 2927
+    },
+    {
+      "epoch": 2.2780082987551866,
+      "grad_norm": 0.6091113320780822,
+      "learning_rate": 1.6608578190577611e-06,
+      "loss": 0.3576,
+      "step": 2928
+    },
+    {
+      "epoch": 2.278786307053942,
+      "grad_norm": 0.5467589313410434,
+      "learning_rate": 1.6574888583160287e-06,
+      "loss": 0.322,
+      "step": 2929
+    },
+    {
+      "epoch": 2.279564315352697,
+      "grad_norm": 0.5033509169559522,
+      "learning_rate": 1.654122638921088e-06,
+      "loss": 0.3402,
+      "step": 2930
+    },
+    {
+      "epoch": 2.2803423236514524,
+      "grad_norm": 0.5266554062616822,
+      "learning_rate": 1.6507591636337373e-06,
+      "loss": 0.3645,
+      "step": 2931
+    },
+    {
+      "epoch": 2.2811203319502074,
+      "grad_norm": 0.5016233980921003,
+      "learning_rate": 1.647398435212511e-06,
+      "loss": 0.3304,
+      "step": 2932
+    },
+    {
+      "epoch": 2.2818983402489628,
+      "grad_norm": 0.5102035908880437,
+      "learning_rate": 1.6440404564136969e-06,
+      "loss": 0.2953,
+      "step": 2933
+    },
+    {
+      "epoch": 2.2826763485477177,
+      "grad_norm": 0.5381287182798608,
+      "learning_rate": 1.6406852299913334e-06,
+      "loss": 0.3778,
+      "step": 2934
+    },
+    {
+      "epoch": 2.283454356846473,
+      "grad_norm": 0.5122452409375381,
+      "learning_rate": 1.6373327586971905e-06,
+      "loss": 0.3478,
+      "step": 2935
+    },
+    {
+      "epoch": 2.284232365145228,
+      "grad_norm": 0.5277556261943144,
+      "learning_rate": 1.6339830452807842e-06,
+      "loss": 0.3109,
+      "step": 2936
+    },
+    {
+      "epoch": 2.2850103734439835,
+      "grad_norm": 0.538256099765555,
+      "learning_rate": 1.6306360924893699e-06,
+      "loss": 0.3551,
+      "step": 2937
+    },
+    {
+      "epoch": 2.2857883817427385,
+      "grad_norm": 0.4831608670577681,
+      "learning_rate": 1.6272919030679363e-06,
+      "loss": 0.2976,
+      "step": 2938
+    },
+    {
+      "epoch": 2.286566390041494,
+      "grad_norm": 0.50457913570934,
+      "learning_rate": 1.6239504797592065e-06,
+      "loss": 0.3685,
+      "step": 2939
+    },
+    {
+      "epoch": 2.287344398340249,
+      "grad_norm": 0.45132759715721543,
+      "learning_rate": 1.620611825303635e-06,
+      "loss": 0.3326,
+      "step": 2940
+    },
+    {
+      "epoch": 2.2881224066390042,
+      "grad_norm": 0.5277583730098235,
+      "learning_rate": 1.617275942439407e-06,
+      "loss": 0.3795,
+      "step": 2941
+    },
+    {
+      "epoch": 2.288900414937759,
+      "grad_norm": 0.5383322640968198,
+      "learning_rate": 1.613942833902431e-06,
+      "loss": 0.3211,
+      "step": 2942
+    },
+    {
+      "epoch": 2.2896784232365146,
+      "grad_norm": 0.6072907246816951,
+      "learning_rate": 1.6106125024263436e-06,
+      "loss": 0.3693,
+      "step": 2943
+    },
+    {
+      "epoch": 2.2904564315352696,
+      "grad_norm": 0.5508931909203599,
+      "learning_rate": 1.6072849507425026e-06,
+      "loss": 0.358,
+      "step": 2944
+    },
+    {
+      "epoch": 2.291234439834025,
+      "grad_norm": 0.5419036805273529,
+      "learning_rate": 1.603960181579985e-06,
+      "loss": 0.3207,
+      "step": 2945
+    },
+    {
+      "epoch": 2.29201244813278,
+      "grad_norm": 0.49090569833764114,
+      "learning_rate": 1.6006381976655872e-06,
+      "loss": 0.3242,
+      "step": 2946
+    },
+    {
+      "epoch": 2.2927904564315353,
+      "grad_norm": 0.5061293185295338,
+      "learning_rate": 1.5973190017238211e-06,
+      "loss": 0.3625,
+      "step": 2947
+    },
+    {
+      "epoch": 2.2935684647302903,
+      "grad_norm": 0.4934128632057655,
+      "learning_rate": 1.59400259647691e-06,
+      "loss": 0.3164,
+      "step": 2948
+    },
+    {
+      "epoch": 2.2943464730290457,
+      "grad_norm": 0.5061938308890133,
+      "learning_rate": 1.5906889846447921e-06,
+      "loss": 0.3853,
+      "step": 2949
+    },
+    {
+      "epoch": 2.2951244813278007,
+      "grad_norm": 0.5758697167850283,
+      "learning_rate": 1.5873781689451128e-06,
+      "loss": 0.3562,
+      "step": 2950
+    },
+    {
+      "epoch": 2.295902489626556,
+      "grad_norm": 0.5050010156522317,
+      "learning_rate": 1.5840701520932195e-06,
+      "loss": 0.3363,
+      "step": 2951
+    },
+    {
+      "epoch": 2.296680497925311,
+      "grad_norm": 0.4892281855510828,
+      "learning_rate": 1.5807649368021733e-06,
+      "loss": 0.3437,
+      "step": 2952
+    },
+    {
+      "epoch": 2.2974585062240664,
+      "grad_norm": 0.4559171411992778,
+      "learning_rate": 1.5774625257827337e-06,
+      "loss": 0.3199,
+      "step": 2953
+    },
+    {
+      "epoch": 2.2982365145228214,
+      "grad_norm": 0.6113647859234117,
+      "learning_rate": 1.574162921743354e-06,
+      "loss": 0.3814,
+      "step": 2954
+    },
+    {
+      "epoch": 2.299014522821577,
+      "grad_norm": 0.517045689018187,
+      "learning_rate": 1.5708661273901977e-06,
+      "loss": 0.2991,
+      "step": 2955
+    },
+    {
+      "epoch": 2.2997925311203318,
+      "grad_norm": 0.5098761467245114,
+      "learning_rate": 1.5675721454271097e-06,
+      "loss": 0.3828,
+      "step": 2956
+    },
+    {
+      "epoch": 2.300570539419087,
+      "grad_norm": 0.4807281037479323,
+      "learning_rate": 1.564280978555644e-06,
+      "loss": 0.3203,
+      "step": 2957
+    },
+    {
+      "epoch": 2.301348547717842,
+      "grad_norm": 0.5354798416139216,
+      "learning_rate": 1.5609926294750315e-06,
+      "loss": 0.3885,
+      "step": 2958
+    },
+    {
+      "epoch": 2.3021265560165975,
+      "grad_norm": 0.49294582130916015,
+      "learning_rate": 1.5577071008821993e-06,
+      "loss": 0.3369,
+      "step": 2959
+    },
+    {
+      "epoch": 2.3029045643153525,
+      "grad_norm": 0.5378389180029753,
+      "learning_rate": 1.5544243954717637e-06,
+      "loss": 0.3471,
+      "step": 2960
+    },
+    {
+      "epoch": 2.303682572614108,
+      "grad_norm": 0.446658455463621,
+      "learning_rate": 1.5511445159360179e-06,
+      "loss": 0.3128,
+      "step": 2961
+    },
+    {
+      "epoch": 2.304460580912863,
+      "grad_norm": 0.4850483863577994,
+      "learning_rate": 1.5478674649649429e-06,
+      "loss": 0.3591,
+      "step": 2962
+    },
+    {
+      "epoch": 2.3052385892116183,
+      "grad_norm": 0.5470572300100535,
+      "learning_rate": 1.5445932452461992e-06,
+      "loss": 0.3625,
+      "step": 2963
+    },
+    {
+      "epoch": 2.3060165975103732,
+      "grad_norm": 0.5493224406790418,
+      "learning_rate": 1.5413218594651242e-06,
+      "loss": 0.3252,
+      "step": 2964
+    },
+    {
+      "epoch": 2.3067946058091287,
+      "grad_norm": 0.550630845658632,
+      "learning_rate": 1.5380533103047317e-06,
+      "loss": 0.3505,
+      "step": 2965
+    },
+    {
+      "epoch": 2.307572614107884,
+      "grad_norm": 0.6298107024456089,
+      "learning_rate": 1.534787600445709e-06,
+      "loss": 0.3243,
+      "step": 2966
+    },
+    {
+      "epoch": 2.308350622406639,
+      "grad_norm": 0.49365994753063064,
+      "learning_rate": 1.5315247325664145e-06,
+      "loss": 0.3325,
+      "step": 2967
+    },
+    {
+      "epoch": 2.309128630705394,
+      "grad_norm": 0.4853473213821622,
+      "learning_rate": 1.528264709342876e-06,
+      "loss": 0.3555,
+      "step": 2968
+    },
+    {
+      "epoch": 2.3099066390041494,
+      "grad_norm": 0.4534099887825028,
+      "learning_rate": 1.5250075334487874e-06,
+      "loss": 0.318,
+      "step": 2969
+    },
+    {
+      "epoch": 2.310684647302905,
+      "grad_norm": 0.4909480341693995,
+      "learning_rate": 1.5217532075555098e-06,
+      "loss": 0.3423,
+      "step": 2970
+    },
+    {
+      "epoch": 2.3114626556016598,
+      "grad_norm": 0.5413151641931779,
+      "learning_rate": 1.5185017343320652e-06,
+      "loss": 0.3205,
+      "step": 2971
+    },
+    {
+      "epoch": 2.3122406639004147,
+      "grad_norm": 0.48184717371016483,
+      "learning_rate": 1.5152531164451318e-06,
+      "loss": 0.3503,
+      "step": 2972
+    },
+    {
+      "epoch": 2.31301867219917,
+      "grad_norm": 0.5840539568645444,
+      "learning_rate": 1.5120073565590554e-06,
+      "loss": 0.3825,
+      "step": 2973
+    },
+    {
+      "epoch": 2.3137966804979255,
+      "grad_norm": 0.5608889820687402,
+      "learning_rate": 1.5087644573358322e-06,
+      "loss": 0.348,
+      "step": 2974
+    },
+    {
+      "epoch": 2.3145746887966805,
+      "grad_norm": 0.5803702995041913,
+      "learning_rate": 1.505524421435108e-06,
+      "loss": 0.3614,
+      "step": 2975
+    },
+    {
+      "epoch": 2.3153526970954355,
+      "grad_norm": 0.50360254186434,
+      "learning_rate": 1.5022872515141912e-06,
+      "loss": 0.3073,
+      "step": 2976
+    },
+    {
+      "epoch": 2.316130705394191,
+      "grad_norm": 0.5173237097651724,
+      "learning_rate": 1.4990529502280276e-06,
+      "loss": 0.3346,
+      "step": 2977
+    },
+    {
+      "epoch": 2.3169087136929463,
+      "grad_norm": 0.5358935214509994,
+      "learning_rate": 1.495821520229222e-06,
+      "loss": 0.3221,
+      "step": 2978
+    },
+    {
+      "epoch": 2.3176867219917012,
+      "grad_norm": 0.5346209257641478,
+      "learning_rate": 1.4925929641680132e-06,
+      "loss": 0.3553,
+      "step": 2979
+    },
+    {
+      "epoch": 2.318464730290456,
+      "grad_norm": 0.450674801716226,
+      "learning_rate": 1.4893672846922896e-06,
+      "loss": 0.3045,
+      "step": 2980
+    },
+    {
+      "epoch": 2.3192427385892116,
+      "grad_norm": 0.4917336707058322,
+      "learning_rate": 1.4861444844475824e-06,
+      "loss": 0.4154,
+      "step": 2981
+    },
+    {
+      "epoch": 2.320020746887967,
+      "grad_norm": 0.49867086098247126,
+      "learning_rate": 1.4829245660770542e-06,
+      "loss": 0.332,
+      "step": 2982
+    },
+    {
+      "epoch": 2.320798755186722,
+      "grad_norm": 0.4869043936829677,
+      "learning_rate": 1.4797075322215082e-06,
+      "loss": 0.3573,
+      "step": 2983
+    },
+    {
+      "epoch": 2.3215767634854774,
+      "grad_norm": 0.5166448380563277,
+      "learning_rate": 1.476493385519383e-06,
+      "loss": 0.3306,
+      "step": 2984
+    },
+    {
+      "epoch": 2.3223547717842323,
+      "grad_norm": 0.5139799627834639,
+      "learning_rate": 1.4732821286067472e-06,
+      "loss": 0.3252,
+      "step": 2985
+    },
+    {
+      "epoch": 2.3231327800829877,
+      "grad_norm": 0.5439126798103188,
+      "learning_rate": 1.4700737641172997e-06,
+      "loss": 0.3716,
+      "step": 2986
+    },
+    {
+      "epoch": 2.3239107883817427,
+      "grad_norm": 0.5026298334949774,
+      "learning_rate": 1.4668682946823675e-06,
+      "loss": 0.3076,
+      "step": 2987
+    },
+    {
+      "epoch": 2.324688796680498,
+      "grad_norm": 0.5211145567478304,
+      "learning_rate": 1.4636657229309036e-06,
+      "loss": 0.358,
+      "step": 2988
+    },
+    {
+      "epoch": 2.325466804979253,
+      "grad_norm": 0.47195381818035415,
+      "learning_rate": 1.4604660514894836e-06,
+      "loss": 0.3078,
+      "step": 2989
+    },
+    {
+      "epoch": 2.3262448132780085,
+      "grad_norm": 0.4822120490191706,
+      "learning_rate": 1.4572692829823053e-06,
+      "loss": 0.3476,
+      "step": 2990
+    },
+    {
+      "epoch": 2.3270228215767634,
+      "grad_norm": 0.576983776877411,
+      "learning_rate": 1.4540754200311852e-06,
+      "loss": 0.3696,
+      "step": 2991
+    },
+    {
+      "epoch": 2.327800829875519,
+      "grad_norm": 0.5429473408553198,
+      "learning_rate": 1.4508844652555581e-06,
+      "loss": 0.3392,
+      "step": 2992
+    },
+    {
+      "epoch": 2.328578838174274,
+      "grad_norm": 0.4972106941188701,
+      "learning_rate": 1.4476964212724686e-06,
+      "loss": 0.3258,
+      "step": 2993
+    },
+    {
+      "epoch": 2.329356846473029,
+      "grad_norm": 0.5265377193863773,
+      "learning_rate": 1.4445112906965819e-06,
+      "loss": 0.3343,
+      "step": 2994
+    },
+    {
+      "epoch": 2.330134854771784,
+      "grad_norm": 0.5198398974699232,
+      "learning_rate": 1.4413290761401677e-06,
+      "loss": 0.3498,
+      "step": 2995
+    },
+    {
+      "epoch": 2.3309128630705396,
+      "grad_norm": 0.5324519748304007,
+      "learning_rate": 1.4381497802131078e-06,
+      "loss": 0.322,
+      "step": 2996
+    },
+    {
+      "epoch": 2.3316908713692945,
+      "grad_norm": 0.5144930635076752,
+      "learning_rate": 1.4349734055228887e-06,
+      "loss": 0.3593,
+      "step": 2997
+    },
+    {
+      "epoch": 2.33246887966805,
+      "grad_norm": 0.48612793010482996,
+      "learning_rate": 1.431799954674597e-06,
+      "loss": 0.2878,
+      "step": 2998
+    },
+    {
+      "epoch": 2.333246887966805,
+      "grad_norm": 0.5108829517984143,
+      "learning_rate": 1.428629430270932e-06,
+      "loss": 0.3702,
+      "step": 2999
+    },
+    {
+      "epoch": 2.3340248962655603,
+      "grad_norm": 0.4819826689383623,
+      "learning_rate": 1.4254618349121808e-06,
+      "loss": 0.3778,
+      "step": 3000
+    },
+    {
+      "epoch": 2.3348029045643153,
+      "grad_norm": 0.44386606442061377,
+      "learning_rate": 1.4222971711962346e-06,
+      "loss": 0.2919,
+      "step": 3001
+    },
+    {
+      "epoch": 2.3355809128630707,
+      "grad_norm": 0.45589748873367897,
+      "learning_rate": 1.4191354417185832e-06,
+      "loss": 0.3617,
+      "step": 3002
+    },
+    {
+      "epoch": 2.3363589211618256,
+      "grad_norm": 0.45038827243239254,
+      "learning_rate": 1.4159766490723004e-06,
+      "loss": 0.3188,
+      "step": 3003
+    },
+    {
+      "epoch": 2.337136929460581,
+      "grad_norm": 0.468764184221201,
+      "learning_rate": 1.4128207958480634e-06,
+      "loss": 0.305,
+      "step": 3004
+    },
+    {
+      "epoch": 2.337914937759336,
+      "grad_norm": 0.5055078589720686,
+      "learning_rate": 1.4096678846341266e-06,
+      "loss": 0.3527,
+      "step": 3005
+    },
+    {
+      "epoch": 2.3386929460580914,
+      "grad_norm": 0.5255604422372142,
+      "learning_rate": 1.4065179180163397e-06,
+      "loss": 0.3146,
+      "step": 3006
+    },
+    {
+      "epoch": 2.3394709543568464,
+      "grad_norm": 0.48738871536417383,
+      "learning_rate": 1.4033708985781346e-06,
+      "loss": 0.3171,
+      "step": 3007
+    },
+    {
+      "epoch": 2.340248962655602,
+      "grad_norm": 0.5076115223489897,
+      "learning_rate": 1.4002268289005265e-06,
+      "loss": 0.41,
+      "step": 3008
+    },
+    {
+      "epoch": 2.3410269709543567,
+      "grad_norm": 0.5338977558772645,
+      "learning_rate": 1.397085711562111e-06,
+      "loss": 0.3039,
+      "step": 3009
+    },
+    {
+      "epoch": 2.341804979253112,
+      "grad_norm": 0.5362882631099625,
+      "learning_rate": 1.393947549139063e-06,
+      "loss": 0.3456,
+      "step": 3010
+    },
+    {
+      "epoch": 2.342582987551867,
+      "grad_norm": 0.5948132949332909,
+      "learning_rate": 1.3908123442051336e-06,
+      "loss": 0.3433,
+      "step": 3011
+    },
+    {
+      "epoch": 2.3433609958506225,
+      "grad_norm": 0.5703336308985073,
+      "learning_rate": 1.3876800993316475e-06,
+      "loss": 0.3614,
+      "step": 3012
+    },
+    {
+      "epoch": 2.3441390041493775,
+      "grad_norm": 0.5069500451639604,
+      "learning_rate": 1.3845508170875028e-06,
+      "loss": 0.3654,
+      "step": 3013
+    },
+    {
+      "epoch": 2.344917012448133,
+      "grad_norm": 0.5528757240024685,
+      "learning_rate": 1.3814245000391674e-06,
+      "loss": 0.3588,
+      "step": 3014
+    },
+    {
+      "epoch": 2.345695020746888,
+      "grad_norm": 0.5457721140795525,
+      "learning_rate": 1.3783011507506778e-06,
+      "loss": 0.308,
+      "step": 3015
+    },
+    {
+      "epoch": 2.3464730290456433,
+      "grad_norm": 0.6044949391398314,
+      "learning_rate": 1.3751807717836363e-06,
+      "loss": 0.3558,
+      "step": 3016
+    },
+    {
+      "epoch": 2.347251037344398,
+      "grad_norm": 0.5168155910521238,
+      "learning_rate": 1.3720633656972076e-06,
+      "loss": 0.347,
+      "step": 3017
+    },
+    {
+      "epoch": 2.3480290456431536,
+      "grad_norm": 0.5039029416096547,
+      "learning_rate": 1.3689489350481228e-06,
+      "loss": 0.2869,
+      "step": 3018
+    },
+    {
+      "epoch": 2.3488070539419086,
+      "grad_norm": 0.6083846660160064,
+      "learning_rate": 1.3658374823906634e-06,
+      "loss": 0.3988,
+      "step": 3019
+    },
+    {
+      "epoch": 2.349585062240664,
+      "grad_norm": 0.583765977216566,
+      "learning_rate": 1.3627290102766815e-06,
+      "loss": 0.3453,
+      "step": 3020
+    },
+    {
+      "epoch": 2.350363070539419,
+      "grad_norm": 0.4646435659850748,
+      "learning_rate": 1.3596235212555714e-06,
+      "loss": 0.294,
+      "step": 3021
+    },
+    {
+      "epoch": 2.3511410788381744,
+      "grad_norm": 0.5525868523683777,
+      "learning_rate": 1.3565210178742915e-06,
+      "loss": 0.427,
+      "step": 3022
+    },
+    {
+      "epoch": 2.3519190871369293,
+      "grad_norm": 0.46918135643486375,
+      "learning_rate": 1.3534215026773478e-06,
+      "loss": 0.3315,
+      "step": 3023
+    },
+    {
+      "epoch": 2.3526970954356847,
+      "grad_norm": 0.477117401617014,
+      "learning_rate": 1.3503249782067895e-06,
+      "loss": 0.3197,
+      "step": 3024
+    },
+    {
+      "epoch": 2.3534751037344397,
+      "grad_norm": 0.5788798129096405,
+      "learning_rate": 1.3472314470022252e-06,
+      "loss": 0.3575,
+      "step": 3025
+    },
+    {
+      "epoch": 2.354253112033195,
+      "grad_norm": 0.4778224724931448,
+      "learning_rate": 1.344140911600797e-06,
+      "loss": 0.2681,
+      "step": 3026
+    },
+    {
+      "epoch": 2.35503112033195,
+      "grad_norm": 0.5331021538618675,
+      "learning_rate": 1.3410533745371968e-06,
+      "loss": 0.3152,
+      "step": 3027
+    },
+    {
+      "epoch": 2.3558091286307055,
+      "grad_norm": 0.5497025497197451,
+      "learning_rate": 1.3379688383436546e-06,
+      "loss": 0.3503,
+      "step": 3028
+    },
+    {
+      "epoch": 2.3565871369294604,
+      "grad_norm": 0.5905598383876354,
+      "learning_rate": 1.3348873055499395e-06,
+      "loss": 0.3357,
+      "step": 3029
+    },
+    {
+      "epoch": 2.357365145228216,
+      "grad_norm": 0.467431479023594,
+      "learning_rate": 1.3318087786833616e-06,
+      "loss": 0.2876,
+      "step": 3030
+    },
+    {
+      "epoch": 2.358143153526971,
+      "grad_norm": 0.5015377614611838,
+      "learning_rate": 1.3287332602687585e-06,
+      "loss": 0.3693,
+      "step": 3031
+    },
+    {
+      "epoch": 2.358921161825726,
+      "grad_norm": 0.5387948628273856,
+      "learning_rate": 1.3256607528285047e-06,
+      "loss": 0.345,
+      "step": 3032
+    },
+    {
+      "epoch": 2.359699170124481,
+      "grad_norm": 0.5721723954951033,
+      "learning_rate": 1.322591258882505e-06,
+      "loss": 0.3068,
+      "step": 3033
+    },
+    {
+      "epoch": 2.3604771784232366,
+      "grad_norm": 0.522671361102078,
+      "learning_rate": 1.3195247809481925e-06,
+      "loss": 0.3444,
+      "step": 3034
+    },
+    {
+      "epoch": 2.3612551867219915,
+      "grad_norm": 0.5057801887400056,
+      "learning_rate": 1.3164613215405258e-06,
+      "loss": 0.3573,
+      "step": 3035
+    },
+    {
+      "epoch": 2.362033195020747,
+      "grad_norm": 0.5570482869999064,
+      "learning_rate": 1.3134008831719891e-06,
+      "loss": 0.3387,
+      "step": 3036
+    },
+    {
+      "epoch": 2.362811203319502,
+      "grad_norm": 0.5941033744629296,
+      "learning_rate": 1.310343468352588e-06,
+      "loss": 0.3598,
+      "step": 3037
+    },
+    {
+      "epoch": 2.3635892116182573,
+      "grad_norm": 0.5031018976061788,
+      "learning_rate": 1.307289079589849e-06,
+      "loss": 0.3651,
+      "step": 3038
+    },
+    {
+      "epoch": 2.3643672199170123,
+      "grad_norm": 0.4304120675374911,
+      "learning_rate": 1.3042377193888168e-06,
+      "loss": 0.3246,
+      "step": 3039
+    },
+    {
+      "epoch": 2.3651452282157677,
+      "grad_norm": 0.5435285841424613,
+      "learning_rate": 1.3011893902520511e-06,
+      "loss": 0.3786,
+      "step": 3040
+    },
+    {
+      "epoch": 2.3659232365145226,
+      "grad_norm": 0.48751769955336255,
+      "learning_rate": 1.298144094679627e-06,
+      "loss": 0.3582,
+      "step": 3041
+    },
+    {
+      "epoch": 2.366701244813278,
+      "grad_norm": 0.4913295091157043,
+      "learning_rate": 1.2951018351691314e-06,
+      "loss": 0.3093,
+      "step": 3042
+    },
+    {
+      "epoch": 2.367479253112033,
+      "grad_norm": 0.5278093240339529,
+      "learning_rate": 1.2920626142156605e-06,
+      "loss": 0.3917,
+      "step": 3043
+    },
+    {
+      "epoch": 2.3682572614107884,
+      "grad_norm": 0.5665975987810312,
+      "learning_rate": 1.2890264343118208e-06,
+      "loss": 0.3571,
+      "step": 3044
+    },
+    {
+      "epoch": 2.369035269709544,
+      "grad_norm": 0.48777219190156945,
+      "learning_rate": 1.285993297947718e-06,
+      "loss": 0.3423,
+      "step": 3045
+    },
+    {
+      "epoch": 2.3698132780082988,
+      "grad_norm": 0.5368897479145567,
+      "learning_rate": 1.2829632076109722e-06,
+      "loss": 0.343,
+      "step": 3046
+    },
+    {
+      "epoch": 2.3705912863070537,
+      "grad_norm": 0.6365638846161286,
+      "learning_rate": 1.2799361657866944e-06,
+      "loss": 0.3257,
+      "step": 3047
+    },
+    {
+      "epoch": 2.371369294605809,
+      "grad_norm": 0.528468053821051,
+      "learning_rate": 1.2769121749575042e-06,
+      "loss": 0.3678,
+      "step": 3048
+    },
+    {
+      "epoch": 2.3721473029045645,
+      "grad_norm": 0.4819976830026667,
+      "learning_rate": 1.2738912376035157e-06,
+      "loss": 0.3834,
+      "step": 3049
+    },
+    {
+      "epoch": 2.3729253112033195,
+      "grad_norm": 0.4410613189794332,
+      "learning_rate": 1.270873356202334e-06,
+      "loss": 0.3022,
+      "step": 3050
+    },
+    {
+      "epoch": 2.3737033195020745,
+      "grad_norm": 0.48384685430263574,
+      "learning_rate": 1.2678585332290671e-06,
+      "loss": 0.3587,
+      "step": 3051
+    },
+    {
+      "epoch": 2.37448132780083,
+      "grad_norm": 0.5424578763986233,
+      "learning_rate": 1.2648467711563063e-06,
+      "loss": 0.3555,
+      "step": 3052
+    },
+    {
+      "epoch": 2.3752593360995853,
+      "grad_norm": 0.6254421354016114,
+      "learning_rate": 1.261838072454137e-06,
+      "loss": 0.3529,
+      "step": 3053
+    },
+    {
+      "epoch": 2.3760373443983402,
+      "grad_norm": 0.5586494466459092,
+      "learning_rate": 1.2588324395901308e-06,
+      "loss": 0.3896,
+      "step": 3054
+    },
+    {
+      "epoch": 2.376815352697095,
+      "grad_norm": 0.5092825818408565,
+      "learning_rate": 1.2558298750293456e-06,
+      "loss": 0.3319,
+      "step": 3055
+    },
+    {
+      "epoch": 2.3775933609958506,
+      "grad_norm": 0.5301689551978039,
+      "learning_rate": 1.2528303812343229e-06,
+      "loss": 0.3721,
+      "step": 3056
+    },
+    {
+      "epoch": 2.378371369294606,
+      "grad_norm": 0.5319758517844274,
+      "learning_rate": 1.249833960665084e-06,
+      "loss": 0.3721,
+      "step": 3057
+    },
+    {
+      "epoch": 2.379149377593361,
+      "grad_norm": 0.4717987654142552,
+      "learning_rate": 1.2468406157791313e-06,
+      "loss": 0.3379,
+      "step": 3058
+    },
+    {
+      "epoch": 2.379927385892116,
+      "grad_norm": 0.48111745519825455,
+      "learning_rate": 1.2438503490314447e-06,
+      "loss": 0.3554,
+      "step": 3059
+    },
+    {
+      "epoch": 2.3807053941908713,
+      "grad_norm": 0.5444430138394475,
+      "learning_rate": 1.2408631628744794e-06,
+      "loss": 0.3446,
+      "step": 3060
+    },
+    {
+      "epoch": 2.3814834024896268,
+      "grad_norm": 0.4684428105393605,
+      "learning_rate": 1.237879059758163e-06,
+      "loss": 0.3325,
+      "step": 3061
+    },
+    {
+      "epoch": 2.3822614107883817,
+      "grad_norm": 0.4242690714548586,
+      "learning_rate": 1.2348980421298962e-06,
+      "loss": 0.3111,
+      "step": 3062
+    },
+    {
+      "epoch": 2.383039419087137,
+      "grad_norm": 0.48717586867384405,
+      "learning_rate": 1.2319201124345481e-06,
+      "loss": 0.316,
+      "step": 3063
+    },
+    {
+      "epoch": 2.383817427385892,
+      "grad_norm": 0.4977298263005144,
+      "learning_rate": 1.2289452731144551e-06,
+      "loss": 0.3745,
+      "step": 3064
+    },
+    {
+      "epoch": 2.3845954356846475,
+      "grad_norm": 0.4415815722969752,
+      "learning_rate": 1.2259735266094219e-06,
+      "loss": 0.3037,
+      "step": 3065
+    },
+    {
+      "epoch": 2.3853734439834025,
+      "grad_norm": 0.4979654310024448,
+      "learning_rate": 1.2230048753567093e-06,
+      "loss": 0.3284,
+      "step": 3066
+    },
+    {
+      "epoch": 2.386151452282158,
+      "grad_norm": 0.5339054329047498,
+      "learning_rate": 1.2200393217910493e-06,
+      "loss": 0.3768,
+      "step": 3067
+    },
+    {
+      "epoch": 2.386929460580913,
+      "grad_norm": 0.5127094079200619,
+      "learning_rate": 1.217076868344625e-06,
+      "loss": 0.3103,
+      "step": 3068
+    },
+    {
+      "epoch": 2.3877074688796682,
+      "grad_norm": 0.4864238581161296,
+      "learning_rate": 1.214117517447082e-06,
+      "loss": 0.3754,
+      "step": 3069
+    },
+    {
+      "epoch": 2.388485477178423,
+      "grad_norm": 0.504491077764483,
+      "learning_rate": 1.2111612715255223e-06,
+      "loss": 0.3214,
+      "step": 3070
+    },
+    {
+      "epoch": 2.3892634854771786,
+      "grad_norm": 0.3886455983075693,
+      "learning_rate": 1.2082081330044926e-06,
+      "loss": 0.2582,
+      "step": 3071
+    },
+    {
+      "epoch": 2.3900414937759336,
+      "grad_norm": 0.5397876567128261,
+      "learning_rate": 1.2052581043060046e-06,
+      "loss": 0.3503,
+      "step": 3072
+    },
+    {
+      "epoch": 2.390819502074689,
+      "grad_norm": 0.5237847996271592,
+      "learning_rate": 1.2023111878495069e-06,
+      "loss": 0.3395,
+      "step": 3073
+    },
+    {
+      "epoch": 2.391597510373444,
+      "grad_norm": 0.554676336375452,
+      "learning_rate": 1.1993673860519023e-06,
+      "loss": 0.3777,
+      "step": 3074
+    },
+    {
+      "epoch": 2.3923755186721993,
+      "grad_norm": 0.5347689712427608,
+      "learning_rate": 1.1964267013275388e-06,
+      "loss": 0.3422,
+      "step": 3075
+    },
+    {
+      "epoch": 2.3931535269709543,
+      "grad_norm": 0.4360561877320133,
+      "learning_rate": 1.1934891360882046e-06,
+      "loss": 0.3007,
+      "step": 3076
+    },
+    {
+      "epoch": 2.3939315352697097,
+      "grad_norm": 0.5596369386007711,
+      "learning_rate": 1.1905546927431362e-06,
+      "loss": 0.393,
+      "step": 3077
+    },
+    {
+      "epoch": 2.3947095435684647,
+      "grad_norm": 0.5700535401222714,
+      "learning_rate": 1.187623373699001e-06,
+      "loss": 0.3618,
+      "step": 3078
+    },
+    {
+      "epoch": 2.39548755186722,
+      "grad_norm": 0.45495220344979437,
+      "learning_rate": 1.18469518135991e-06,
+      "loss": 0.2743,
+      "step": 3079
+    },
+    {
+      "epoch": 2.396265560165975,
+      "grad_norm": 0.5370021449756419,
+      "learning_rate": 1.1817701181274078e-06,
+      "loss": 0.3619,
+      "step": 3080
+    },
+    {
+      "epoch": 2.3970435684647304,
+      "grad_norm": 0.5214687732579755,
+      "learning_rate": 1.178848186400473e-06,
+      "loss": 0.3234,
+      "step": 3081
+    },
+    {
+      "epoch": 2.3978215767634854,
+      "grad_norm": 0.5065194954689153,
+      "learning_rate": 1.1759293885755162e-06,
+      "loss": 0.3291,
+      "step": 3082
+    },
+    {
+      "epoch": 2.398599585062241,
+      "grad_norm": 0.5401307606228358,
+      "learning_rate": 1.1730137270463776e-06,
+      "loss": 0.3346,
+      "step": 3083
+    },
+    {
+      "epoch": 2.3993775933609958,
+      "grad_norm": 0.4817510896589676,
+      "learning_rate": 1.1701012042043246e-06,
+      "loss": 0.3026,
+      "step": 3084
+    },
+    {
+      "epoch": 2.400155601659751,
+      "grad_norm": 0.5575460909311463,
+      "learning_rate": 1.1671918224380507e-06,
+      "loss": 0.3737,
+      "step": 3085
+    },
+    {
+      "epoch": 2.400933609958506,
+      "grad_norm": 0.5689417613319715,
+      "learning_rate": 1.1642855841336741e-06,
+      "loss": 0.3548,
+      "step": 3086
+    },
+    {
+      "epoch": 2.4017116182572615,
+      "grad_norm": 0.4739932763426165,
+      "learning_rate": 1.1613824916747334e-06,
+      "loss": 0.303,
+      "step": 3087
+    },
+    {
+      "epoch": 2.4024896265560165,
+      "grad_norm": 0.5193155454672156,
+      "learning_rate": 1.15848254744219e-06,
+      "loss": 0.3442,
+      "step": 3088
+    },
+    {
+      "epoch": 2.403267634854772,
+      "grad_norm": 0.4938502080180903,
+      "learning_rate": 1.1555857538144172e-06,
+      "loss": 0.3867,
+      "step": 3089
+    },
+    {
+      "epoch": 2.404045643153527,
+      "grad_norm": 0.49474423825255537,
+      "learning_rate": 1.152692113167212e-06,
+      "loss": 0.3316,
+      "step": 3090
+    },
+    {
+      "epoch": 2.4048236514522823,
+      "grad_norm": 0.5471003984079785,
+      "learning_rate": 1.1498016278737817e-06,
+      "loss": 0.3286,
+      "step": 3091
+    },
+    {
+      "epoch": 2.4056016597510372,
+      "grad_norm": 0.5388616296729767,
+      "learning_rate": 1.1469143003047427e-06,
+      "loss": 0.3016,
+      "step": 3092
+    },
+    {
+      "epoch": 2.4063796680497926,
+      "grad_norm": 0.5751872748521937,
+      "learning_rate": 1.14403013282813e-06,
+      "loss": 0.4013,
+      "step": 3093
+    },
+    {
+      "epoch": 2.4071576763485476,
+      "grad_norm": 0.5097016694948893,
+      "learning_rate": 1.1411491278093755e-06,
+      "loss": 0.36,
+      "step": 3094
+    },
+    {
+      "epoch": 2.407935684647303,
+      "grad_norm": 0.4832696477732091,
+      "learning_rate": 1.1382712876113305e-06,
+      "loss": 0.2809,
+      "step": 3095
+    },
+    {
+      "epoch": 2.408713692946058,
+      "grad_norm": 0.5827264610745134,
+      "learning_rate": 1.135396614594239e-06,
+      "loss": 0.4201,
+      "step": 3096
+    },
+    {
+      "epoch": 2.4094917012448134,
+      "grad_norm": 0.5175154002989621,
+      "learning_rate": 1.132525111115752e-06,
+      "loss": 0.3168,
+      "step": 3097
+    },
+    {
+      "epoch": 2.4102697095435683,
+      "grad_norm": 0.5048115843713185,
+      "learning_rate": 1.1296567795309255e-06,
+      "loss": 0.3838,
+      "step": 3098
+    },
+    {
+      "epoch": 2.4110477178423237,
+      "grad_norm": 0.4385767483371883,
+      "learning_rate": 1.126791622192206e-06,
+      "loss": 0.3094,
+      "step": 3099
+    },
+    {
+      "epoch": 2.4118257261410787,
+      "grad_norm": 0.41661624301419037,
+      "learning_rate": 1.1239296414494422e-06,
+      "loss": 0.3067,
+      "step": 3100
+    },
+    {
+      "epoch": 2.412603734439834,
+      "grad_norm": 0.5454495643211161,
+      "learning_rate": 1.1210708396498754e-06,
+      "loss": 0.3521,
+      "step": 3101
+    },
+    {
+      "epoch": 2.413381742738589,
+      "grad_norm": 0.5060999948263826,
+      "learning_rate": 1.11821521913814e-06,
+      "loss": 0.3486,
+      "step": 3102
+    },
+    {
+      "epoch": 2.4141597510373445,
+      "grad_norm": 0.48740766182404377,
+      "learning_rate": 1.1153627822562618e-06,
+      "loss": 0.3487,
+      "step": 3103
+    },
+    {
+      "epoch": 2.4149377593360994,
+      "grad_norm": 0.4875414283353957,
+      "learning_rate": 1.1125135313436552e-06,
+      "loss": 0.3353,
+      "step": 3104
+    },
+    {
+      "epoch": 2.415715767634855,
+      "grad_norm": 0.43696156567116196,
+      "learning_rate": 1.1096674687371212e-06,
+      "loss": 0.326,
+      "step": 3105
+    },
+    {
+      "epoch": 2.41649377593361,
+      "grad_norm": 0.5146355618724683,
+      "learning_rate": 1.1068245967708474e-06,
+      "loss": 0.351,
+      "step": 3106
+    },
+    {
+      "epoch": 2.417271784232365,
+      "grad_norm": 0.5361804957452366,
+      "learning_rate": 1.1039849177764022e-06,
+      "loss": 0.3178,
+      "step": 3107
+    },
+    {
+      "epoch": 2.41804979253112,
+      "grad_norm": 0.5264804682998299,
+      "learning_rate": 1.1011484340827378e-06,
+      "loss": 0.3133,
+      "step": 3108
+    },
+    {
+      "epoch": 2.4188278008298756,
+      "grad_norm": 0.5454389022811766,
+      "learning_rate": 1.0983151480161851e-06,
+      "loss": 0.3201,
+      "step": 3109
+    },
+    {
+      "epoch": 2.4196058091286305,
+      "grad_norm": 0.5137835128502314,
+      "learning_rate": 1.0954850619004492e-06,
+      "loss": 0.3375,
+      "step": 3110
+    },
+    {
+      "epoch": 2.420383817427386,
+      "grad_norm": 0.5953407134415154,
+      "learning_rate": 1.0926581780566164e-06,
+      "loss": 0.337,
+      "step": 3111
+    },
+    {
+      "epoch": 2.421161825726141,
+      "grad_norm": 0.4554826239088497,
+      "learning_rate": 1.0898344988031434e-06,
+      "loss": 0.3027,
+      "step": 3112
+    },
+    {
+      "epoch": 2.4219398340248963,
+      "grad_norm": 0.5167710553259773,
+      "learning_rate": 1.087014026455858e-06,
+      "loss": 0.3899,
+      "step": 3113
+    },
+    {
+      "epoch": 2.4227178423236513,
+      "grad_norm": 0.5353937024859269,
+      "learning_rate": 1.084196763327962e-06,
+      "loss": 0.327,
+      "step": 3114
+    },
+    {
+      "epoch": 2.4234958506224067,
+      "grad_norm": 0.5744379973047649,
+      "learning_rate": 1.0813827117300175e-06,
+      "loss": 0.3625,
+      "step": 3115
+    },
+    {
+      "epoch": 2.4242738589211617,
+      "grad_norm": 0.560699129454184,
+      "learning_rate": 1.0785718739699619e-06,
+      "loss": 0.3564,
+      "step": 3116
+    },
+    {
+      "epoch": 2.425051867219917,
+      "grad_norm": 0.565561145233325,
+      "learning_rate": 1.0757642523530891e-06,
+      "loss": 0.3536,
+      "step": 3117
+    },
+    {
+      "epoch": 2.425829875518672,
+      "grad_norm": 0.5062074727948518,
+      "learning_rate": 1.0729598491820586e-06,
+      "loss": 0.3755,
+      "step": 3118
+    },
+    {
+      "epoch": 2.4266078838174274,
+      "grad_norm": 0.5104020891370028,
+      "learning_rate": 1.0701586667568935e-06,
+      "loss": 0.299,
+      "step": 3119
+    },
+    {
+      "epoch": 2.4273858921161824,
+      "grad_norm": 0.571596808538653,
+      "learning_rate": 1.067360707374967e-06,
+      "loss": 0.364,
+      "step": 3120
+    },
+    {
+      "epoch": 2.428163900414938,
+      "grad_norm": 0.60424045757905,
+      "learning_rate": 1.06456597333102e-06,
+      "loss": 0.3657,
+      "step": 3121
+    },
+    {
+      "epoch": 2.4289419087136928,
+      "grad_norm": 0.5184059772481233,
+      "learning_rate": 1.061774466917137e-06,
+      "loss": 0.3354,
+      "step": 3122
+    },
+    {
+      "epoch": 2.429719917012448,
+      "grad_norm": 0.4572622296992252,
+      "learning_rate": 1.0589861904227612e-06,
+      "loss": 0.3122,
+      "step": 3123
+    },
+    {
+      "epoch": 2.4304979253112036,
+      "grad_norm": 0.5102758381110947,
+      "learning_rate": 1.0562011461346893e-06,
+      "loss": 0.3634,
+      "step": 3124
+    },
+    {
+      "epoch": 2.4312759336099585,
+      "grad_norm": 0.49852575449192943,
+      "learning_rate": 1.0534193363370615e-06,
+      "loss": 0.3052,
+      "step": 3125
+    },
+    {
+      "epoch": 2.4320539419087135,
+      "grad_norm": 0.48676794076916496,
+      "learning_rate": 1.050640763311368e-06,
+      "loss": 0.3417,
+      "step": 3126
+    },
+    {
+      "epoch": 2.432831950207469,
+      "grad_norm": 0.5712441136147515,
+      "learning_rate": 1.0478654293364443e-06,
+      "loss": 0.3594,
+      "step": 3127
+    },
+    {
+      "epoch": 2.4336099585062243,
+      "grad_norm": 0.5125676954675802,
+      "learning_rate": 1.0450933366884691e-06,
+      "loss": 0.304,
+      "step": 3128
+    },
+    {
+      "epoch": 2.4343879668049793,
+      "grad_norm": 0.5402171123957263,
+      "learning_rate": 1.042324487640962e-06,
+      "loss": 0.3608,
+      "step": 3129
+    },
+    {
+      "epoch": 2.4351659751037342,
+      "grad_norm": 0.5024534917721193,
+      "learning_rate": 1.0395588844647853e-06,
+      "loss": 0.3335,
+      "step": 3130
+    },
+    {
+      "epoch": 2.4359439834024896,
+      "grad_norm": 0.4813634138887595,
+      "learning_rate": 1.0367965294281351e-06,
+      "loss": 0.3581,
+      "step": 3131
+    },
+    {
+      "epoch": 2.436721991701245,
+      "grad_norm": 0.4949673191394014,
+      "learning_rate": 1.0340374247965462e-06,
+      "loss": 0.3442,
+      "step": 3132
+    },
+    {
+      "epoch": 2.4375,
+      "grad_norm": 0.47248109437793817,
+      "learning_rate": 1.0312815728328867e-06,
+      "loss": 0.3451,
+      "step": 3133
+    },
+    {
+      "epoch": 2.438278008298755,
+      "grad_norm": 0.46281889889822025,
+      "learning_rate": 1.0285289757973582e-06,
+      "loss": 0.292,
+      "step": 3134
+    },
+    {
+      "epoch": 2.4390560165975104,
+      "grad_norm": 0.5594951853068969,
+      "learning_rate": 1.025779635947492e-06,
+      "loss": 0.3631,
+      "step": 3135
+    },
+    {
+      "epoch": 2.4398340248962658,
+      "grad_norm": 0.5201324982179255,
+      "learning_rate": 1.0230335555381449e-06,
+      "loss": 0.3489,
+      "step": 3136
+    },
+    {
+      "epoch": 2.4406120331950207,
+      "grad_norm": 0.5208722824188714,
+      "learning_rate": 1.0202907368215064e-06,
+      "loss": 0.2923,
+      "step": 3137
+    },
+    {
+      "epoch": 2.4413900414937757,
+      "grad_norm": 0.565810052490398,
+      "learning_rate": 1.017551182047088e-06,
+      "loss": 0.3853,
+      "step": 3138
+    },
+    {
+      "epoch": 2.442168049792531,
+      "grad_norm": 0.5616129226393695,
+      "learning_rate": 1.014814893461723e-06,
+      "loss": 0.384,
+      "step": 3139
+    },
+    {
+      "epoch": 2.4429460580912865,
+      "grad_norm": 0.5597273426158481,
+      "learning_rate": 1.0120818733095694e-06,
+      "loss": 0.3532,
+      "step": 3140
+    },
+    {
+      "epoch": 2.4437240663900415,
+      "grad_norm": 0.5520519225102601,
+      "learning_rate": 1.0093521238320986e-06,
+      "loss": 0.3106,
+      "step": 3141
+    },
+    {
+      "epoch": 2.4445020746887964,
+      "grad_norm": 0.527533454850931,
+      "learning_rate": 1.006625647268108e-06,
+      "loss": 0.333,
+      "step": 3142
+    },
+    {
+      "epoch": 2.445280082987552,
+      "grad_norm": 0.5216939527706409,
+      "learning_rate": 1.0039024458537022e-06,
+      "loss": 0.3388,
+      "step": 3143
+    },
+    {
+      "epoch": 2.4460580912863072,
+      "grad_norm": 0.5040208865929308,
+      "learning_rate": 1.0011825218223047e-06,
+      "loss": 0.3473,
+      "step": 3144
+    },
+    {
+      "epoch": 2.446836099585062,
+      "grad_norm": 0.5788944795176509,
+      "learning_rate": 9.984658774046535e-07,
+      "loss": 0.3605,
+      "step": 3145
+    },
+    {
+      "epoch": 2.4476141078838176,
+      "grad_norm": 0.5438186858400587,
+      "learning_rate": 9.957525148287889e-07,
+      "loss": 0.3688,
+      "step": 3146
+    },
+    {
+      "epoch": 2.4483921161825726,
+      "grad_norm": 0.5366861751040062,
+      "learning_rate": 9.930424363200659e-07,
+      "loss": 0.3433,
+      "step": 3147
+    },
+    {
+      "epoch": 2.449170124481328,
+      "grad_norm": 0.5403505614265545,
+      "learning_rate": 9.903356441011436e-07,
+      "loss": 0.3608,
+      "step": 3148
+    },
+    {
+      "epoch": 2.449948132780083,
+      "grad_norm": 0.488769668323972,
+      "learning_rate": 9.876321403919876e-07,
+      "loss": 0.318,
+      "step": 3149
+    },
+    {
+      "epoch": 2.4507261410788383,
+      "grad_norm": 0.5033648974466448,
+      "learning_rate": 9.849319274098646e-07,
+      "loss": 0.3155,
+      "step": 3150
+    },
+    {
+      "epoch": 2.4515041493775933,
+      "grad_norm": 0.5533021531735421,
+      "learning_rate": 9.822350073693438e-07,
+      "loss": 0.3944,
+      "step": 3151
+    },
+    {
+      "epoch": 2.4522821576763487,
+      "grad_norm": 0.5324118757200348,
+      "learning_rate": 9.79541382482292e-07,
+      "loss": 0.3056,
+      "step": 3152
+    },
+    {
+      "epoch": 2.4530601659751037,
+      "grad_norm": 0.5951195923913314,
+      "learning_rate": 9.768510549578757e-07,
+      "loss": 0.3569,
+      "step": 3153
+    },
+    {
+      "epoch": 2.453838174273859,
+      "grad_norm": 0.5513507345605964,
+      "learning_rate": 9.74164027002556e-07,
+      "loss": 0.3927,
+      "step": 3154
+    },
+    {
+      "epoch": 2.454616182572614,
+      "grad_norm": 0.4827411787362634,
+      "learning_rate": 9.714803008200868e-07,
+      "loss": 0.333,
+      "step": 3155
+    },
+    {
+      "epoch": 2.4553941908713695,
+      "grad_norm": 0.5426120529206822,
+      "learning_rate": 9.68799878611517e-07,
+      "loss": 0.3293,
+      "step": 3156
+    },
+    {
+      "epoch": 2.4561721991701244,
+      "grad_norm": 0.459242183948806,
+      "learning_rate": 9.661227625751806e-07,
+      "loss": 0.3557,
+      "step": 3157
+    },
+    {
+      "epoch": 2.45695020746888,
+      "grad_norm": 0.5112048099903694,
+      "learning_rate": 9.634489549067067e-07,
+      "loss": 0.326,
+      "step": 3158
+    },
+    {
+      "epoch": 2.457728215767635,
+      "grad_norm": 0.532436914267851,
+      "learning_rate": 9.607784577990058e-07,
+      "loss": 0.3502,
+      "step": 3159
+    },
+    {
+      "epoch": 2.45850622406639,
+      "grad_norm": 0.5549467094237107,
+      "learning_rate": 9.58111273442276e-07,
+      "loss": 0.3554,
+      "step": 3160
+    },
+    {
+      "epoch": 2.459284232365145,
+      "grad_norm": 0.46760366341006937,
+      "learning_rate": 9.554474040239986e-07,
+      "loss": 0.3242,
+      "step": 3161
+    },
+    {
+      "epoch": 2.4600622406639006,
+      "grad_norm": 0.5431311254299672,
+      "learning_rate": 9.527868517289313e-07,
+      "loss": 0.3337,
+      "step": 3162
+    },
+    {
+      "epoch": 2.4608402489626555,
+      "grad_norm": 0.5154190266675683,
+      "learning_rate": 9.501296187391201e-07,
+      "loss": 0.3602,
+      "step": 3163
+    },
+    {
+      "epoch": 2.461618257261411,
+      "grad_norm": 0.5246154359140993,
+      "learning_rate": 9.474757072338803e-07,
+      "loss": 0.303,
+      "step": 3164
+    },
+    {
+      "epoch": 2.462396265560166,
+      "grad_norm": 0.5530449119916304,
+      "learning_rate": 9.448251193898062e-07,
+      "loss": 0.2966,
+      "step": 3165
+    },
+    {
+      "epoch": 2.4631742738589213,
+      "grad_norm": 0.5417029625162313,
+      "learning_rate": 9.421778573807705e-07,
+      "loss": 0.3742,
+      "step": 3166
+    },
+    {
+      "epoch": 2.4639522821576763,
+      "grad_norm": 0.47789632428720763,
+      "learning_rate": 9.395339233779093e-07,
+      "loss": 0.3555,
+      "step": 3167
+    },
+    {
+      "epoch": 2.4647302904564317,
+      "grad_norm": 0.47605369333663666,
+      "learning_rate": 9.368933195496388e-07,
+      "loss": 0.2919,
+      "step": 3168
+    },
+    {
+      "epoch": 2.4655082987551866,
+      "grad_norm": 0.4630381710178329,
+      "learning_rate": 9.342560480616364e-07,
+      "loss": 0.3345,
+      "step": 3169
+    },
+    {
+      "epoch": 2.466286307053942,
+      "grad_norm": 0.5275281009712566,
+      "learning_rate": 9.316221110768498e-07,
+      "loss": 0.3762,
+      "step": 3170
+    },
+    {
+      "epoch": 2.467064315352697,
+      "grad_norm": 0.5315391816147098,
+      "learning_rate": 9.289915107554925e-07,
+      "loss": 0.339,
+      "step": 3171
+    },
+    {
+      "epoch": 2.4678423236514524,
+      "grad_norm": 0.48314261368278943,
+      "learning_rate": 9.263642492550401e-07,
+      "loss": 0.32,
+      "step": 3172
+    },
+    {
+      "epoch": 2.4686203319502074,
+      "grad_norm": 0.5452045224407734,
+      "learning_rate": 9.237403287302299e-07,
+      "loss": 0.346,
+      "step": 3173
+    },
+    {
+      "epoch": 2.4693983402489628,
+      "grad_norm": 0.5519305520844204,
+      "learning_rate": 9.211197513330605e-07,
+      "loss": 0.3329,
+      "step": 3174
+    },
+    {
+      "epoch": 2.4701763485477177,
+      "grad_norm": 0.47912314253906996,
+      "learning_rate": 9.185025192127872e-07,
+      "loss": 0.3266,
+      "step": 3175
+    },
+    {
+      "epoch": 2.470954356846473,
+      "grad_norm": 0.5385912921219231,
+      "learning_rate": 9.158886345159223e-07,
+      "loss": 0.348,
+      "step": 3176
+    },
+    {
+      "epoch": 2.471732365145228,
+      "grad_norm": 0.5110030728180253,
+      "learning_rate": 9.132780993862323e-07,
+      "loss": 0.3553,
+      "step": 3177
+    },
+    {
+      "epoch": 2.4725103734439835,
+      "grad_norm": 0.5030685666205549,
+      "learning_rate": 9.10670915964737e-07,
+      "loss": 0.2938,
+      "step": 3178
+    },
+    {
+      "epoch": 2.4732883817427385,
+      "grad_norm": 0.5153063651868162,
+      "learning_rate": 9.080670863897079e-07,
+      "loss": 0.3172,
+      "step": 3179
+    },
+    {
+      "epoch": 2.474066390041494,
+      "grad_norm": 0.5815101522802505,
+      "learning_rate": 9.054666127966638e-07,
+      "loss": 0.358,
+      "step": 3180
+    },
+    {
+      "epoch": 2.474844398340249,
+      "grad_norm": 0.47627757796554937,
+      "learning_rate": 9.028694973183738e-07,
+      "loss": 0.3039,
+      "step": 3181
+    },
+    {
+      "epoch": 2.4756224066390042,
+      "grad_norm": 0.5104243580531602,
+      "learning_rate": 9.002757420848518e-07,
+      "loss": 0.3247,
+      "step": 3182
+    },
+    {
+      "epoch": 2.476400414937759,
+      "grad_norm": 0.5695673719085329,
+      "learning_rate": 8.976853492233517e-07,
+      "loss": 0.3447,
+      "step": 3183
+    },
+    {
+      "epoch": 2.4771784232365146,
+      "grad_norm": 0.4599434183487293,
+      "learning_rate": 8.950983208583797e-07,
+      "loss": 0.2562,
+      "step": 3184
+    },
+    {
+      "epoch": 2.4779564315352696,
+      "grad_norm": 0.5893005680859899,
+      "learning_rate": 8.925146591116701e-07,
+      "loss": 0.4249,
+      "step": 3185
+    },
+    {
+      "epoch": 2.478734439834025,
+      "grad_norm": 0.47513713081842096,
+      "learning_rate": 8.899343661022075e-07,
+      "loss": 0.3254,
+      "step": 3186
+    },
+    {
+      "epoch": 2.47951244813278,
+      "grad_norm": 0.5153188056818723,
+      "learning_rate": 8.87357443946208e-07,
+      "loss": 0.3997,
+      "step": 3187
+    },
+    {
+      "epoch": 2.4802904564315353,
+      "grad_norm": 0.5061566726174315,
+      "learning_rate": 8.847838947571191e-07,
+      "loss": 0.3333,
+      "step": 3188
+    },
+    {
+      "epoch": 2.4810684647302903,
+      "grad_norm": 0.5001649734512056,
+      "learning_rate": 8.822137206456333e-07,
+      "loss": 0.3135,
+      "step": 3189
+    },
+    {
+      "epoch": 2.4818464730290457,
+      "grad_norm": 0.5290803260918499,
+      "learning_rate": 8.796469237196636e-07,
+      "loss": 0.3348,
+      "step": 3190
+    },
+    {
+      "epoch": 2.4826244813278007,
+      "grad_norm": 0.483444244129546,
+      "learning_rate": 8.770835060843596e-07,
+      "loss": 0.3992,
+      "step": 3191
+    },
+    {
+      "epoch": 2.483402489626556,
+      "grad_norm": 0.4984002004259697,
+      "learning_rate": 8.745234698420973e-07,
+      "loss": 0.3495,
+      "step": 3192
+    },
+    {
+      "epoch": 2.484180497925311,
+      "grad_norm": 0.5100333261950786,
+      "learning_rate": 8.719668170924788e-07,
+      "loss": 0.3109,
+      "step": 3193
+    },
+    {
+      "epoch": 2.4849585062240664,
+      "grad_norm": 0.5188032051131133,
+      "learning_rate": 8.694135499323364e-07,
+      "loss": 0.3242,
+      "step": 3194
+    },
+    {
+      "epoch": 2.4857365145228214,
+      "grad_norm": 0.5492410910713144,
+      "learning_rate": 8.668636704557176e-07,
+      "loss": 0.3732,
+      "step": 3195
+    },
+    {
+      "epoch": 2.486514522821577,
+      "grad_norm": 0.4956485994615074,
+      "learning_rate": 8.643171807538974e-07,
+      "loss": 0.3577,
+      "step": 3196
+    },
+    {
+      "epoch": 2.4872925311203318,
+      "grad_norm": 0.4683983350400083,
+      "learning_rate": 8.617740829153681e-07,
+      "loss": 0.2816,
+      "step": 3197
+    },
+    {
+      "epoch": 2.488070539419087,
+      "grad_norm": 0.4776377125289487,
+      "learning_rate": 8.592343790258418e-07,
+      "loss": 0.3766,
+      "step": 3198
+    },
+    {
+      "epoch": 2.488848547717842,
+      "grad_norm": 0.5723900623216642,
+      "learning_rate": 8.566980711682454e-07,
+      "loss": 0.3562,
+      "step": 3199
+    },
+    {
+      "epoch": 2.4896265560165975,
+      "grad_norm": 0.5143521292636568,
+      "learning_rate": 8.541651614227215e-07,
+      "loss": 0.3597,
+      "step": 3200
+    },
+    {
+      "epoch": 2.4904045643153525,
+      "grad_norm": 0.5752136756446573,
+      "learning_rate": 8.516356518666264e-07,
+      "loss": 0.3402,
+      "step": 3201
+    },
+    {
+      "epoch": 2.491182572614108,
+      "grad_norm": 0.5135280125302544,
+      "learning_rate": 8.491095445745262e-07,
+      "loss": 0.3615,
+      "step": 3202
+    },
+    {
+      "epoch": 2.491960580912863,
+      "grad_norm": 0.4795504840532863,
+      "learning_rate": 8.465868416181977e-07,
+      "loss": 0.2999,
+      "step": 3203
+    },
+    {
+      "epoch": 2.4927385892116183,
+      "grad_norm": 0.5078129726468947,
+      "learning_rate": 8.440675450666253e-07,
+      "loss": 0.3227,
+      "step": 3204
+    },
+    {
+      "epoch": 2.4935165975103732,
+      "grad_norm": 0.5621434139265608,
+      "learning_rate": 8.415516569859999e-07,
+      "loss": 0.3564,
+      "step": 3205
+    },
+    {
+      "epoch": 2.4942946058091287,
+      "grad_norm": 0.5017078600492831,
+      "learning_rate": 8.390391794397168e-07,
+      "loss": 0.3898,
+      "step": 3206
+    },
+    {
+      "epoch": 2.495072614107884,
+      "grad_norm": 0.44258123092190704,
+      "learning_rate": 8.365301144883736e-07,
+      "loss": 0.3104,
+      "step": 3207
+    },
+    {
+      "epoch": 2.495850622406639,
+      "grad_norm": 0.4518831412232179,
+      "learning_rate": 8.340244641897716e-07,
+      "loss": 0.3249,
+      "step": 3208
+    },
+    {
+      "epoch": 2.496628630705394,
+      "grad_norm": 0.46029241860461284,
+      "learning_rate": 8.315222305989057e-07,
+      "loss": 0.3665,
+      "step": 3209
+    },
+    {
+      "epoch": 2.4974066390041494,
+      "grad_norm": 0.4981559064044681,
+      "learning_rate": 8.290234157679767e-07,
+      "loss": 0.3313,
+      "step": 3210
+    },
+    {
+      "epoch": 2.498184647302905,
+      "grad_norm": 0.4332523444622387,
+      "learning_rate": 8.265280217463734e-07,
+      "loss": 0.3106,
+      "step": 3211
+    },
+    {
+      "epoch": 2.4989626556016598,
+      "grad_norm": 0.5049255465189482,
+      "learning_rate": 8.240360505806866e-07,
+      "loss": 0.3573,
+      "step": 3212
+    },
+    {
+      "epoch": 2.4997406639004147,
+      "grad_norm": 0.48753390647434636,
+      "learning_rate": 8.215475043146959e-07,
+      "loss": 0.3557,
+      "step": 3213
+    },
+    {
+      "epoch": 2.50051867219917,
+      "grad_norm": 0.506445398684548,
+      "learning_rate": 8.190623849893681e-07,
+      "loss": 0.3383,
+      "step": 3214
+    },
+    {
+      "epoch": 2.5012966804979255,
+      "grad_norm": 0.5777673894837403,
+      "learning_rate": 8.165806946428695e-07,
+      "loss": 0.3526,
+      "step": 3215
+    },
+    {
+      "epoch": 2.5020746887966805,
+      "grad_norm": 0.5475802313208992,
+      "learning_rate": 8.141024353105437e-07,
+      "loss": 0.3707,
+      "step": 3216
+    },
+    {
+      "epoch": 2.5028526970954355,
+      "grad_norm": 0.5050261406363964,
+      "learning_rate": 8.11627609024927e-07,
+      "loss": 0.3184,
+      "step": 3217
+    },
+    {
+      "epoch": 2.503630705394191,
+      "grad_norm": 0.534419838090669,
+      "learning_rate": 8.091562178157375e-07,
+      "loss": 0.3699,
+      "step": 3218
+    },
+    {
+      "epoch": 2.5044087136929463,
+      "grad_norm": 0.46631296136539624,
+      "learning_rate": 8.066882637098755e-07,
+      "loss": 0.2996,
+      "step": 3219
+    },
+    {
+      "epoch": 2.5051867219917012,
+      "grad_norm": 0.5019149066353601,
+      "learning_rate": 8.04223748731427e-07,
+      "loss": 0.3541,
+      "step": 3220
+    },
+    {
+      "epoch": 2.505964730290456,
+      "grad_norm": 0.5194864214704392,
+      "learning_rate": 8.0176267490165e-07,
+      "loss": 0.3895,
+      "step": 3221
+    },
+    {
+      "epoch": 2.5067427385892116,
+      "grad_norm": 0.46692786367104405,
+      "learning_rate": 7.993050442389855e-07,
+      "loss": 0.3353,
+      "step": 3222
+    },
+    {
+      "epoch": 2.507520746887967,
+      "grad_norm": 0.444045419960878,
+      "learning_rate": 7.968508587590495e-07,
+      "loss": 0.311,
+      "step": 3223
+    },
+    {
+      "epoch": 2.508298755186722,
+      "grad_norm": 0.5313563517486721,
+      "learning_rate": 7.944001204746321e-07,
+      "loss": 0.3682,
+      "step": 3224
+    },
+    {
+      "epoch": 2.509076763485477,
+      "grad_norm": 0.5356068616299807,
+      "learning_rate": 7.919528313956954e-07,
+      "loss": 0.3782,
+      "step": 3225
+    },
+    {
+      "epoch": 2.5098547717842323,
+      "grad_norm": 0.4837275416365924,
+      "learning_rate": 7.895089935293742e-07,
+      "loss": 0.3069,
+      "step": 3226
+    },
+    {
+      "epoch": 2.5106327800829877,
+      "grad_norm": 0.4738171413085141,
+      "learning_rate": 7.870686088799728e-07,
+      "loss": 0.3025,
+      "step": 3227
+    },
+    {
+      "epoch": 2.5114107883817427,
+      "grad_norm": 0.615626405077517,
+      "learning_rate": 7.846316794489611e-07,
+      "loss": 0.3606,
+      "step": 3228
+    },
+    {
+      "epoch": 2.5121887966804977,
+      "grad_norm": 0.6051409151205197,
+      "learning_rate": 7.821982072349776e-07,
+      "loss": 0.3345,
+      "step": 3229
+    },
+    {
+      "epoch": 2.512966804979253,
+      "grad_norm": 0.5711184712623798,
+      "learning_rate": 7.797681942338248e-07,
+      "loss": 0.3679,
+      "step": 3230
+    },
+    {
+      "epoch": 2.5137448132780085,
+      "grad_norm": 0.5625268531837001,
+      "learning_rate": 7.773416424384689e-07,
+      "loss": 0.3828,
+      "step": 3231
+    },
+    {
+      "epoch": 2.5145228215767634,
+      "grad_norm": 0.49990029460566027,
+      "learning_rate": 7.749185538390325e-07,
+      "loss": 0.3513,
+      "step": 3232
+    },
+    {
+      "epoch": 2.5153008298755184,
+      "grad_norm": 0.5098426588395253,
+      "learning_rate": 7.724989304228053e-07,
+      "loss": 0.3699,
+      "step": 3233
+    },
+    {
+      "epoch": 2.516078838174274,
+      "grad_norm": 0.4622273629052422,
+      "learning_rate": 7.700827741742323e-07,
+      "loss": 0.3207,
+      "step": 3234
+    },
+    {
+      "epoch": 2.516856846473029,
+      "grad_norm": 0.45900890321990595,
+      "learning_rate": 7.676700870749094e-07,
+      "loss": 0.3542,
+      "step": 3235
+    },
+    {
+      "epoch": 2.517634854771784,
+      "grad_norm": 0.5883985088429243,
+      "learning_rate": 7.65260871103598e-07,
+      "loss": 0.3454,
+      "step": 3236
+    },
+    {
+      "epoch": 2.5184128630705396,
+      "grad_norm": 0.47956204671089253,
+      "learning_rate": 7.628551282362018e-07,
+      "loss": 0.3725,
+      "step": 3237
+    },
+    {
+      "epoch": 2.5191908713692945,
+      "grad_norm": 0.5103559297499981,
+      "learning_rate": 7.604528604457833e-07,
+      "loss": 0.3091,
+      "step": 3238
+    },
+    {
+      "epoch": 2.51996887966805,
+      "grad_norm": 0.5037231689527538,
+      "learning_rate": 7.580540697025523e-07,
+      "loss": 0.361,
+      "step": 3239
+    },
+    {
+      "epoch": 2.520746887966805,
+      "grad_norm": 0.4955115366213704,
+      "learning_rate": 7.556587579738656e-07,
+      "loss": 0.3124,
+      "step": 3240
+    },
+    {
+      "epoch": 2.5215248962655603,
+      "grad_norm": 0.6183109874586269,
+      "learning_rate": 7.532669272242321e-07,
+      "loss": 0.3364,
+      "step": 3241
+    },
+    {
+      "epoch": 2.5223029045643153,
+      "grad_norm": 0.5307953590054815,
+      "learning_rate": 7.508785794152984e-07,
+      "loss": 0.3692,
+      "step": 3242
+    },
+    {
+      "epoch": 2.5230809128630707,
+      "grad_norm": 0.49152179950758873,
+      "learning_rate": 7.484937165058592e-07,
+      "loss": 0.3113,
+      "step": 3243
+    },
+    {
+      "epoch": 2.5238589211618256,
+      "grad_norm": 0.5027960325075005,
+      "learning_rate": 7.461123404518506e-07,
+      "loss": 0.3243,
+      "step": 3244
+    },
+    {
+      "epoch": 2.524636929460581,
+      "grad_norm": 0.5523536445967931,
+      "learning_rate": 7.437344532063473e-07,
+      "loss": 0.3637,
+      "step": 3245
+    },
+    {
+      "epoch": 2.525414937759336,
+      "grad_norm": 0.48900392034211876,
+      "learning_rate": 7.41360056719565e-07,
+      "loss": 0.302,
+      "step": 3246
+    },
+    {
+      "epoch": 2.5261929460580914,
+      "grad_norm": 0.44551858948116446,
+      "learning_rate": 7.389891529388544e-07,
+      "loss": 0.3117,
+      "step": 3247
+    },
+    {
+      "epoch": 2.5269709543568464,
+      "grad_norm": 0.5981595719600599,
+      "learning_rate": 7.366217438087025e-07,
+      "loss": 0.3594,
+      "step": 3248
+    },
+    {
+      "epoch": 2.527748962655602,
+      "grad_norm": 0.49398475639905176,
+      "learning_rate": 7.342578312707305e-07,
+      "loss": 0.3321,
+      "step": 3249
+    },
+    {
+      "epoch": 2.5285269709543567,
+      "grad_norm": 0.49451101994029345,
+      "learning_rate": 7.318974172636917e-07,
+      "loss": 0.3427,
+      "step": 3250
+    },
+    {
+      "epoch": 2.529304979253112,
+      "grad_norm": 0.48169715610190483,
+      "learning_rate": 7.295405037234693e-07,
+      "loss": 0.3599,
+      "step": 3251
+    },
+    {
+      "epoch": 2.530082987551867,
+      "grad_norm": 0.5188483024694391,
+      "learning_rate": 7.271870925830787e-07,
+      "loss": 0.3183,
+      "step": 3252
+    },
+    {
+      "epoch": 2.5308609958506225,
+      "grad_norm": 0.4629485943015748,
+      "learning_rate": 7.248371857726555e-07,
+      "loss": 0.2776,
+      "step": 3253
+    },
+    {
+      "epoch": 2.5316390041493775,
+      "grad_norm": 0.5810137889454958,
+      "learning_rate": 7.224907852194701e-07,
+      "loss": 0.4117,
+      "step": 3254
+    },
+    {
+      "epoch": 2.532417012448133,
+      "grad_norm": 0.5264356732889829,
+      "learning_rate": 7.201478928479138e-07,
+      "loss": 0.34,
+      "step": 3255
+    },
+    {
+      "epoch": 2.533195020746888,
+      "grad_norm": 0.5146742817977837,
+      "learning_rate": 7.178085105794963e-07,
+      "loss": 0.3332,
+      "step": 3256
+    },
+    {
+      "epoch": 2.5339730290456433,
+      "grad_norm": 0.49424870822331646,
+      "learning_rate": 7.154726403328566e-07,
+      "loss": 0.351,
+      "step": 3257
+    },
+    {
+      "epoch": 2.534751037344398,
+      "grad_norm": 0.5286568413328394,
+      "learning_rate": 7.13140284023745e-07,
+      "loss": 0.346,
+      "step": 3258
+    },
+    {
+      "epoch": 2.5355290456431536,
+      "grad_norm": 0.4943568217080096,
+      "learning_rate": 7.108114435650381e-07,
+      "loss": 0.2808,
+      "step": 3259
+    },
+    {
+      "epoch": 2.5363070539419086,
+      "grad_norm": 0.5354996999124659,
+      "learning_rate": 7.08486120866722e-07,
+      "loss": 0.3469,
+      "step": 3260
+    },
+    {
+      "epoch": 2.537085062240664,
+      "grad_norm": 0.4932158858760802,
+      "learning_rate": 7.061643178359007e-07,
+      "loss": 0.3281,
+      "step": 3261
+    },
+    {
+      "epoch": 2.537863070539419,
+      "grad_norm": 0.5532418796957336,
+      "learning_rate": 7.038460363767941e-07,
+      "loss": 0.3478,
+      "step": 3262
+    },
+    {
+      "epoch": 2.5386410788381744,
+      "grad_norm": 0.5233040953709623,
+      "learning_rate": 7.01531278390728e-07,
+      "loss": 0.3297,
+      "step": 3263
+    },
+    {
+      "epoch": 2.5394190871369293,
+      "grad_norm": 0.5574519225943647,
+      "learning_rate": 6.992200457761427e-07,
+      "loss": 0.3486,
+      "step": 3264
+    },
+    {
+      "epoch": 2.5401970954356847,
+      "grad_norm": 0.5628195878191264,
+      "learning_rate": 6.969123404285865e-07,
+      "loss": 0.3381,
+      "step": 3265
+    },
+    {
+      "epoch": 2.5409751037344397,
+      "grad_norm": 0.51864056004475,
+      "learning_rate": 6.946081642407149e-07,
+      "loss": 0.3598,
+      "step": 3266
+    },
+    {
+      "epoch": 2.541753112033195,
+      "grad_norm": 0.4709716860018638,
+      "learning_rate": 6.923075191022877e-07,
+      "loss": 0.3517,
+      "step": 3267
+    },
+    {
+      "epoch": 2.54253112033195,
+      "grad_norm": 0.596167520845791,
+      "learning_rate": 6.900104069001696e-07,
+      "loss": 0.3637,
+      "step": 3268
+    },
+    {
+      "epoch": 2.5433091286307055,
+      "grad_norm": 0.4824917470181129,
+      "learning_rate": 6.877168295183279e-07,
+      "loss": 0.3346,
+      "step": 3269
+    },
+    {
+      "epoch": 2.5440871369294604,
+      "grad_norm": 0.45764692218164027,
+      "learning_rate": 6.854267888378313e-07,
+      "loss": 0.378,
+      "step": 3270
+    },
+    {
+      "epoch": 2.544865145228216,
+      "grad_norm": 0.5031072778974947,
+      "learning_rate": 6.831402867368464e-07,
+      "loss": 0.3183,
+      "step": 3271
+    },
+    {
+      "epoch": 2.545643153526971,
+      "grad_norm": 0.595965546841046,
+      "learning_rate": 6.808573250906387e-07,
+      "loss": 0.3981,
+      "step": 3272
+    },
+    {
+      "epoch": 2.546421161825726,
+      "grad_norm": 0.49820853434746387,
+      "learning_rate": 6.785779057715714e-07,
+      "loss": 0.3427,
+      "step": 3273
+    },
+    {
+      "epoch": 2.5471991701244816,
+      "grad_norm": 0.44024707410923725,
+      "learning_rate": 6.76302030649097e-07,
+      "loss": 0.3086,
+      "step": 3274
+    },
+    {
+      "epoch": 2.5479771784232366,
+      "grad_norm": 0.5153685917770826,
+      "learning_rate": 6.74029701589769e-07,
+      "loss": 0.3789,
+      "step": 3275
+    },
+    {
+      "epoch": 2.5487551867219915,
+      "grad_norm": 0.4521256687230434,
+      "learning_rate": 6.717609204572273e-07,
+      "loss": 0.2982,
+      "step": 3276
+    },
+    {
+      "epoch": 2.549533195020747,
+      "grad_norm": 0.5781581718917221,
+      "learning_rate": 6.694956891122034e-07,
+      "loss": 0.3648,
+      "step": 3277
+    },
+    {
+      "epoch": 2.5503112033195023,
+      "grad_norm": 0.5218079418684556,
+      "learning_rate": 6.672340094125191e-07,
+      "loss": 0.3706,
+      "step": 3278
+    },
+    {
+      "epoch": 2.5510892116182573,
+      "grad_norm": 0.5076444445185639,
+      "learning_rate": 6.649758832130776e-07,
+      "loss": 0.3352,
+      "step": 3279
+    },
+    {
+      "epoch": 2.5518672199170123,
+      "grad_norm": 0.5896566869534183,
+      "learning_rate": 6.627213123658766e-07,
+      "loss": 0.3793,
+      "step": 3280
+    },
+    {
+      "epoch": 2.5526452282157677,
+      "grad_norm": 0.49034683357420644,
+      "learning_rate": 6.604702987199901e-07,
+      "loss": 0.3092,
+      "step": 3281
+    },
+    {
+      "epoch": 2.553423236514523,
+      "grad_norm": 0.47811139295184085,
+      "learning_rate": 6.582228441215777e-07,
+      "loss": 0.32,
+      "step": 3282
+    },
+    {
+      "epoch": 2.554201244813278,
+      "grad_norm": 0.5188604175862692,
+      "learning_rate": 6.559789504138831e-07,
+      "loss": 0.3263,
+      "step": 3283
+    },
+    {
+      "epoch": 2.554979253112033,
+      "grad_norm": 0.5421269281950594,
+      "learning_rate": 6.53738619437222e-07,
+      "loss": 0.3698,
+      "step": 3284
+    },
+    {
+      "epoch": 2.5557572614107884,
+      "grad_norm": 0.5218955771998076,
+      "learning_rate": 6.515018530289968e-07,
+      "loss": 0.3101,
+      "step": 3285
+    },
+    {
+      "epoch": 2.556535269709544,
+      "grad_norm": 0.5465702385302932,
+      "learning_rate": 6.492686530236803e-07,
+      "loss": 0.3352,
+      "step": 3286
+    },
+    {
+      "epoch": 2.5573132780082988,
+      "grad_norm": 0.5974452849752158,
+      "learning_rate": 6.470390212528215e-07,
+      "loss": 0.3521,
+      "step": 3287
+    },
+    {
+      "epoch": 2.5580912863070537,
+      "grad_norm": 0.5084415218024398,
+      "learning_rate": 6.448129595450459e-07,
+      "loss": 0.3199,
+      "step": 3288
+    },
+    {
+      "epoch": 2.558869294605809,
+      "grad_norm": 0.5462254652283819,
+      "learning_rate": 6.425904697260471e-07,
+      "loss": 0.3605,
+      "step": 3289
+    },
+    {
+      "epoch": 2.5596473029045645,
+      "grad_norm": 0.5158414235735261,
+      "learning_rate": 6.403715536185917e-07,
+      "loss": 0.3509,
+      "step": 3290
+    },
+    {
+      "epoch": 2.5604253112033195,
+      "grad_norm": 0.49419011238638405,
+      "learning_rate": 6.381562130425146e-07,
+      "loss": 0.3381,
+      "step": 3291
+    },
+    {
+      "epoch": 2.5612033195020745,
+      "grad_norm": 0.5613893625488073,
+      "learning_rate": 6.359444498147189e-07,
+      "loss": 0.3689,
+      "step": 3292
+    },
+    {
+      "epoch": 2.56198132780083,
+      "grad_norm": 0.47988005354211666,
+      "learning_rate": 6.337362657491719e-07,
+      "loss": 0.3381,
+      "step": 3293
+    },
+    {
+      "epoch": 2.5627593360995853,
+      "grad_norm": 0.5289628017842056,
+      "learning_rate": 6.315316626569074e-07,
+      "loss": 0.3681,
+      "step": 3294
+    },
+    {
+      "epoch": 2.5635373443983402,
+      "grad_norm": 0.5382397504321745,
+      "learning_rate": 6.29330642346021e-07,
+      "loss": 0.3199,
+      "step": 3295
+    },
+    {
+      "epoch": 2.564315352697095,
+      "grad_norm": 0.4969471810906602,
+      "learning_rate": 6.271332066216717e-07,
+      "loss": 0.2929,
+      "step": 3296
+    },
+    {
+      "epoch": 2.5650933609958506,
+      "grad_norm": 0.5056165341590063,
+      "learning_rate": 6.249393572860768e-07,
+      "loss": 0.3434,
+      "step": 3297
+    },
+    {
+      "epoch": 2.565871369294606,
+      "grad_norm": 0.5417365021662949,
+      "learning_rate": 6.227490961385119e-07,
+      "loss": 0.3806,
+      "step": 3298
+    },
+    {
+      "epoch": 2.566649377593361,
+      "grad_norm": 0.5304116964193522,
+      "learning_rate": 6.205624249753133e-07,
+      "loss": 0.3381,
+      "step": 3299
+    },
+    {
+      "epoch": 2.567427385892116,
+      "grad_norm": 0.49523511038721624,
+      "learning_rate": 6.183793455898662e-07,
+      "loss": 0.3415,
+      "step": 3300
+    },
+    {
+      "epoch": 2.5682053941908713,
+      "grad_norm": 0.5632534441839724,
+      "learning_rate": 6.161998597726171e-07,
+      "loss": 0.3255,
+      "step": 3301
+    },
+    {
+      "epoch": 2.5689834024896268,
+      "grad_norm": 0.5376588748709361,
+      "learning_rate": 6.140239693110622e-07,
+      "loss": 0.3496,
+      "step": 3302
+    },
+    {
+      "epoch": 2.5697614107883817,
+      "grad_norm": 0.5205880278261968,
+      "learning_rate": 6.118516759897492e-07,
+      "loss": 0.3871,
+      "step": 3303
+    },
+    {
+      "epoch": 2.5705394190871367,
+      "grad_norm": 0.5245156283771348,
+      "learning_rate": 6.096829815902755e-07,
+      "loss": 0.3453,
+      "step": 3304
+    },
+    {
+      "epoch": 2.571317427385892,
+      "grad_norm": 0.5088870068772565,
+      "learning_rate": 6.075178878912846e-07,
+      "loss": 0.326,
+      "step": 3305
+    },
+    {
+      "epoch": 2.5720954356846475,
+      "grad_norm": 0.5038707751828274,
+      "learning_rate": 6.053563966684734e-07,
+      "loss": 0.3869,
+      "step": 3306
+    },
+    {
+      "epoch": 2.5728734439834025,
+      "grad_norm": 0.5181790005194951,
+      "learning_rate": 6.031985096945764e-07,
+      "loss": 0.3586,
+      "step": 3307
+    },
+    {
+      "epoch": 2.5736514522821574,
+      "grad_norm": 0.5086601933283552,
+      "learning_rate": 6.010442287393764e-07,
+      "loss": 0.348,
+      "step": 3308
+    },
+    {
+      "epoch": 2.574429460580913,
+      "grad_norm": 0.7483888755999825,
+      "learning_rate": 5.988935555697006e-07,
+      "loss": 0.3621,
+      "step": 3309
+    },
+    {
+      "epoch": 2.5752074688796682,
+      "grad_norm": 0.48378365850608623,
+      "learning_rate": 5.967464919494115e-07,
+      "loss": 0.3588,
+      "step": 3310
+    },
+    {
+      "epoch": 2.575985477178423,
+      "grad_norm": 0.5057368610144527,
+      "learning_rate": 5.946030396394176e-07,
+      "loss": 0.29,
+      "step": 3311
+    },
+    {
+      "epoch": 2.576763485477178,
+      "grad_norm": 0.5553275544628083,
+      "learning_rate": 5.924632003976594e-07,
+      "loss": 0.336,
+      "step": 3312
+    },
+    {
+      "epoch": 2.5775414937759336,
+      "grad_norm": 0.5550654269993699,
+      "learning_rate": 5.903269759791197e-07,
+      "loss": 0.3935,
+      "step": 3313
+    },
+    {
+      "epoch": 2.578319502074689,
+      "grad_norm": 0.4994983951650039,
+      "learning_rate": 5.881943681358127e-07,
+      "loss": 0.363,
+      "step": 3314
+    },
+    {
+      "epoch": 2.579097510373444,
+      "grad_norm": 0.503177669928531,
+      "learning_rate": 5.860653786167886e-07,
+      "loss": 0.3258,
+      "step": 3315
+    },
+    {
+      "epoch": 2.5798755186721993,
+      "grad_norm": 0.4873785518325095,
+      "learning_rate": 5.839400091681291e-07,
+      "loss": 0.296,
+      "step": 3316
+    },
+    {
+      "epoch": 2.5806535269709543,
+      "grad_norm": 0.5297269551989893,
+      "learning_rate": 5.818182615329482e-07,
+      "loss": 0.3517,
+      "step": 3317
+    },
+    {
+      "epoch": 2.5814315352697097,
+      "grad_norm": 0.5135248050912038,
+      "learning_rate": 5.797001374513872e-07,
+      "loss": 0.3495,
+      "step": 3318
+    },
+    {
+      "epoch": 2.5822095435684647,
+      "grad_norm": 0.6130031099745724,
+      "learning_rate": 5.775856386606177e-07,
+      "loss": 0.3345,
+      "step": 3319
+    },
+    {
+      "epoch": 2.58298755186722,
+      "grad_norm": 0.5563873515518287,
+      "learning_rate": 5.754747668948368e-07,
+      "loss": 0.3683,
+      "step": 3320
+    },
+    {
+      "epoch": 2.583765560165975,
+      "grad_norm": 0.5074153720848996,
+      "learning_rate": 5.733675238852676e-07,
+      "loss": 0.3835,
+      "step": 3321
+    },
+    {
+      "epoch": 2.5845435684647304,
+      "grad_norm": 0.5264397590675893,
+      "learning_rate": 5.71263911360157e-07,
+      "loss": 0.3342,
+      "step": 3322
+    },
+    {
+      "epoch": 2.5853215767634854,
+      "grad_norm": 0.45686832576537534,
+      "learning_rate": 5.691639310447739e-07,
+      "loss": 0.3054,
+      "step": 3323
+    },
+    {
+      "epoch": 2.586099585062241,
+      "grad_norm": 0.5353439564401214,
+      "learning_rate": 5.670675846614082e-07,
+      "loss": 0.3704,
+      "step": 3324
+    },
+    {
+      "epoch": 2.5868775933609958,
+      "grad_norm": 0.5644419537138687,
+      "learning_rate": 5.649748739293709e-07,
+      "loss": 0.3715,
+      "step": 3325
+    },
+    {
+      "epoch": 2.587655601659751,
+      "grad_norm": 0.4893716208077568,
+      "learning_rate": 5.628858005649879e-07,
+      "loss": 0.2986,
+      "step": 3326
+    },
+    {
+      "epoch": 2.588433609958506,
+      "grad_norm": 0.5165861285538652,
+      "learning_rate": 5.60800366281607e-07,
+      "loss": 0.3113,
+      "step": 3327
+    },
+    {
+      "epoch": 2.5892116182572615,
+      "grad_norm": 0.4662294670847626,
+      "learning_rate": 5.587185727895866e-07,
+      "loss": 0.3711,
+      "step": 3328
+    },
+    {
+      "epoch": 2.5899896265560165,
+      "grad_norm": 0.6036852853628438,
+      "learning_rate": 5.566404217963001e-07,
+      "loss": 0.3948,
+      "step": 3329
+    },
+    {
+      "epoch": 2.590767634854772,
+      "grad_norm": 0.5056481270666889,
+      "learning_rate": 5.545659150061383e-07,
+      "loss": 0.3077,
+      "step": 3330
+    },
+    {
+      "epoch": 2.591545643153527,
+      "grad_norm": 0.4663584485186856,
+      "learning_rate": 5.524950541204954e-07,
+      "loss": 0.3682,
+      "step": 3331
+    },
+    {
+      "epoch": 2.5923236514522823,
+      "grad_norm": 0.5232465953187401,
+      "learning_rate": 5.504278408377828e-07,
+      "loss": 0.2964,
+      "step": 3332
+    },
+    {
+      "epoch": 2.5931016597510372,
+      "grad_norm": 0.5057476063725741,
+      "learning_rate": 5.483642768534147e-07,
+      "loss": 0.3824,
+      "step": 3333
+    },
+    {
+      "epoch": 2.5938796680497926,
+      "grad_norm": 0.4299999257856115,
+      "learning_rate": 5.463043638598153e-07,
+      "loss": 0.3153,
+      "step": 3334
+    },
+    {
+      "epoch": 2.5946576763485476,
+      "grad_norm": 0.48387095347985126,
+      "learning_rate": 5.442481035464137e-07,
+      "loss": 0.3382,
+      "step": 3335
+    },
+    {
+      "epoch": 2.595435684647303,
+      "grad_norm": 0.4825324490471688,
+      "learning_rate": 5.421954975996441e-07,
+      "loss": 0.3363,
+      "step": 3336
+    },
+    {
+      "epoch": 2.596213692946058,
+      "grad_norm": 0.5592698773181579,
+      "learning_rate": 5.401465477029416e-07,
+      "loss": 0.379,
+      "step": 3337
+    },
+    {
+      "epoch": 2.5969917012448134,
+      "grad_norm": 0.4565778627368813,
+      "learning_rate": 5.381012555367449e-07,
+      "loss": 0.2954,
+      "step": 3338
+    },
+    {
+      "epoch": 2.5977697095435683,
+      "grad_norm": 0.4596807870003037,
+      "learning_rate": 5.360596227784915e-07,
+      "loss": 0.3079,
+      "step": 3339
+    },
+    {
+      "epoch": 2.5985477178423237,
+      "grad_norm": 0.5677811949063425,
+      "learning_rate": 5.340216511026186e-07,
+      "loss": 0.3583,
+      "step": 3340
+    },
+    {
+      "epoch": 2.5993257261410787,
+      "grad_norm": 0.5042624753699725,
+      "learning_rate": 5.319873421805599e-07,
+      "loss": 0.3311,
+      "step": 3341
+    },
+    {
+      "epoch": 2.600103734439834,
+      "grad_norm": 0.4822515712170292,
+      "learning_rate": 5.299566976807457e-07,
+      "loss": 0.342,
+      "step": 3342
+    },
+    {
+      "epoch": 2.600881742738589,
+      "grad_norm": 0.5834897247883266,
+      "learning_rate": 5.279297192686011e-07,
+      "loss": 0.4078,
+      "step": 3343
+    },
+    {
+      "epoch": 2.6016597510373445,
+      "grad_norm": 0.5167083326863519,
+      "learning_rate": 5.259064086065435e-07,
+      "loss": 0.3484,
+      "step": 3344
+    },
+    {
+      "epoch": 2.6024377593360994,
+      "grad_norm": 0.4795467584593253,
+      "learning_rate": 5.238867673539833e-07,
+      "loss": 0.3569,
+      "step": 3345
+    },
+    {
+      "epoch": 2.603215767634855,
+      "grad_norm": 0.49506384098542094,
+      "learning_rate": 5.218707971673215e-07,
+      "loss": 0.355,
+      "step": 3346
+    },
+    {
+      "epoch": 2.60399377593361,
+      "grad_norm": 0.4931197474514569,
+      "learning_rate": 5.19858499699945e-07,
+      "loss": 0.3197,
+      "step": 3347
+    },
+    {
+      "epoch": 2.604771784232365,
+      "grad_norm": 0.5144182096023574,
+      "learning_rate": 5.178498766022361e-07,
+      "loss": 0.3778,
+      "step": 3348
+    },
+    {
+      "epoch": 2.60554979253112,
+      "grad_norm": 0.5002679770089425,
+      "learning_rate": 5.158449295215534e-07,
+      "loss": 0.3085,
+      "step": 3349
+    },
+    {
+      "epoch": 2.6063278008298756,
+      "grad_norm": 0.5486548087506213,
+      "learning_rate": 5.1384366010225e-07,
+      "loss": 0.325,
+      "step": 3350
+    },
+    {
+      "epoch": 2.6071058091286305,
+      "grad_norm": 0.5793159638526021,
+      "learning_rate": 5.118460699856581e-07,
+      "loss": 0.3866,
+      "step": 3351
+    },
+    {
+      "epoch": 2.607883817427386,
+      "grad_norm": 0.4948043597800364,
+      "learning_rate": 5.09852160810091e-07,
+      "loss": 0.3352,
+      "step": 3352
+    },
+    {
+      "epoch": 2.608661825726141,
+      "grad_norm": 0.5413788881774823,
+      "learning_rate": 5.078619342108481e-07,
+      "loss": 0.3985,
+      "step": 3353
+    },
+    {
+      "epoch": 2.6094398340248963,
+      "grad_norm": 0.48497437147611705,
+      "learning_rate": 5.058753918202031e-07,
+      "loss": 0.2831,
+      "step": 3354
+    },
+    {
+      "epoch": 2.6102178423236513,
+      "grad_norm": 0.5668295966481246,
+      "learning_rate": 5.038925352674118e-07,
+      "loss": 0.3881,
+      "step": 3355
+    },
+    {
+      "epoch": 2.6109958506224067,
+      "grad_norm": 0.458608054845604,
+      "learning_rate": 5.019133661787056e-07,
+      "loss": 0.287,
+      "step": 3356
+    },
+    {
+      "epoch": 2.611773858921162,
+      "grad_norm": 0.4848101063292013,
+      "learning_rate": 4.999378861772907e-07,
+      "loss": 0.3788,
+      "step": 3357
+    },
+    {
+      "epoch": 2.612551867219917,
+      "grad_norm": 0.5155752159642913,
+      "learning_rate": 4.979660968833522e-07,
+      "loss": 0.2849,
+      "step": 3358
+    },
+    {
+      "epoch": 2.613329875518672,
+      "grad_norm": 0.4859347345472739,
+      "learning_rate": 4.959979999140419e-07,
+      "loss": 0.3518,
+      "step": 3359
+    },
+    {
+      "epoch": 2.6141078838174274,
+      "grad_norm": 0.45277050035462085,
+      "learning_rate": 4.940335968834875e-07,
+      "loss": 0.291,
+      "step": 3360
+    },
+    {
+      "epoch": 2.614885892116183,
+      "grad_norm": 0.48336743055255776,
+      "learning_rate": 4.92072889402786e-07,
+      "loss": 0.368,
+      "step": 3361
+    },
+    {
+      "epoch": 2.615663900414938,
+      "grad_norm": 0.4699418425636682,
+      "learning_rate": 4.901158790800037e-07,
+      "loss": 0.3006,
+      "step": 3362
+    },
+    {
+      "epoch": 2.6164419087136928,
+      "grad_norm": 0.5603395706613405,
+      "learning_rate": 4.88162567520174e-07,
+      "loss": 0.3486,
+      "step": 3363
+    },
+    {
+      "epoch": 2.617219917012448,
+      "grad_norm": 0.5563644946026781,
+      "learning_rate": 4.862129563252988e-07,
+      "loss": 0.3574,
+      "step": 3364
+    },
+    {
+      "epoch": 2.6179979253112036,
+      "grad_norm": 0.49869937033261297,
+      "learning_rate": 4.842670470943423e-07,
+      "loss": 0.3267,
+      "step": 3365
+    },
+    {
+      "epoch": 2.6187759336099585,
+      "grad_norm": 0.531839553962176,
+      "learning_rate": 4.82324841423234e-07,
+      "loss": 0.3617,
+      "step": 3366
+    },
+    {
+      "epoch": 2.6195539419087135,
+      "grad_norm": 0.5704308247512896,
+      "learning_rate": 4.803863409048671e-07,
+      "loss": 0.3844,
+      "step": 3367
+    },
+    {
+      "epoch": 2.620331950207469,
+      "grad_norm": 0.4587955138173075,
+      "learning_rate": 4.78451547129094e-07,
+      "loss": 0.3124,
+      "step": 3368
+    },
+    {
+      "epoch": 2.6211099585062243,
+      "grad_norm": 0.4465984664465274,
+      "learning_rate": 4.7652046168272847e-07,
+      "loss": 0.2928,
+      "step": 3369
+    },
+    {
+      "epoch": 2.6218879668049793,
+      "grad_norm": 0.6004421679075467,
+      "learning_rate": 4.745930861495407e-07,
+      "loss": 0.4082,
+      "step": 3370
+    },
+    {
+      "epoch": 2.6226659751037342,
+      "grad_norm": 0.5747000872818623,
+      "learning_rate": 4.726694221102612e-07,
+      "loss": 0.3376,
+      "step": 3371
+    },
+    {
+      "epoch": 2.6234439834024896,
+      "grad_norm": 0.5103490770692138,
+      "learning_rate": 4.707494711425764e-07,
+      "loss": 0.3801,
+      "step": 3372
+    },
+    {
+      "epoch": 2.624221991701245,
+      "grad_norm": 0.4831174643989958,
+      "learning_rate": 4.6883323482112277e-07,
+      "loss": 0.3291,
+      "step": 3373
+    },
+    {
+      "epoch": 2.625,
+      "grad_norm": 0.5227551862691285,
+      "learning_rate": 4.66920714717497e-07,
+      "loss": 0.3371,
+      "step": 3374
+    },
+    {
+      "epoch": 2.625778008298755,
+      "grad_norm": 0.5432997004700547,
+      "learning_rate": 4.650119124002417e-07,
+      "loss": 0.3362,
+      "step": 3375
+    },
+    {
+      "epoch": 2.6265560165975104,
+      "grad_norm": 0.586991558277173,
+      "learning_rate": 4.631068294348567e-07,
+      "loss": 0.3409,
+      "step": 3376
+    },
+    {
+      "epoch": 2.6273340248962658,
+      "grad_norm": 0.5512422415452808,
+      "learning_rate": 4.6120546738378513e-07,
+      "loss": 0.3537,
+      "step": 3377
+    },
+    {
+      "epoch": 2.6281120331950207,
+      "grad_norm": 0.49190112443097833,
+      "learning_rate": 4.593078278064211e-07,
+      "loss": 0.3791,
+      "step": 3378
+    },
+    {
+      "epoch": 2.6288900414937757,
+      "grad_norm": 0.5395109036077609,
+      "learning_rate": 4.5741391225910925e-07,
+      "loss": 0.3281,
+      "step": 3379
+    },
+    {
+      "epoch": 2.629668049792531,
+      "grad_norm": 0.5230919410739344,
+      "learning_rate": 4.5552372229513306e-07,
+      "loss": 0.2902,
+      "step": 3380
+    },
+    {
+      "epoch": 2.6304460580912865,
+      "grad_norm": 0.5183170845437576,
+      "learning_rate": 4.5363725946472536e-07,
+      "loss": 0.386,
+      "step": 3381
+    },
+    {
+      "epoch": 2.6312240663900415,
+      "grad_norm": 0.5409342359197947,
+      "learning_rate": 4.5175452531506127e-07,
+      "loss": 0.3675,
+      "step": 3382
+    },
+    {
+      "epoch": 2.6320020746887964,
+      "grad_norm": 0.5331751008286417,
+      "learning_rate": 4.4987552139025614e-07,
+      "loss": 0.292,
+      "step": 3383
+    },
+    {
+      "epoch": 2.632780082987552,
+      "grad_norm": 0.5662066374958181,
+      "learning_rate": 4.480002492313701e-07,
+      "loss": 0.3789,
+      "step": 3384
+    },
+    {
+      "epoch": 2.6335580912863072,
+      "grad_norm": 0.5943773707047357,
+      "learning_rate": 4.4612871037639674e-07,
+      "loss": 0.3851,
+      "step": 3385
+    },
+    {
+      "epoch": 2.634336099585062,
+      "grad_norm": 0.5312868437997001,
+      "learning_rate": 4.4426090636027317e-07,
+      "loss": 0.343,
+      "step": 3386
+    },
+    {
+      "epoch": 2.635114107883817,
+      "grad_norm": 0.4874810083542142,
+      "learning_rate": 4.4239683871487025e-07,
+      "loss": 0.334,
+      "step": 3387
+    },
+    {
+      "epoch": 2.6358921161825726,
+      "grad_norm": 0.49272799455495797,
+      "learning_rate": 4.4053650896899504e-07,
+      "loss": 0.3515,
+      "step": 3388
+    },
+    {
+      "epoch": 2.636670124481328,
+      "grad_norm": 0.4491960693710848,
+      "learning_rate": 4.386799186483903e-07,
+      "loss": 0.3162,
+      "step": 3389
+    },
+    {
+      "epoch": 2.637448132780083,
+      "grad_norm": 0.5166364747900952,
+      "learning_rate": 4.368270692757309e-07,
+      "loss": 0.3495,
+      "step": 3390
+    },
+    {
+      "epoch": 2.638226141078838,
+      "grad_norm": 0.46243518522076266,
+      "learning_rate": 4.3497796237062293e-07,
+      "loss": 0.3305,
+      "step": 3391
+    },
+    {
+      "epoch": 2.6390041493775933,
+      "grad_norm": 0.4677345275063639,
+      "learning_rate": 4.3313259944960384e-07,
+      "loss": 0.3147,
+      "step": 3392
+    },
+    {
+      "epoch": 2.6397821576763487,
+      "grad_norm": 0.49974506331484847,
+      "learning_rate": 4.312909820261413e-07,
+      "loss": 0.3415,
+      "step": 3393
+    },
+    {
+      "epoch": 2.6405601659751037,
+      "grad_norm": 0.527077211952503,
+      "learning_rate": 4.2945311161062984e-07,
+      "loss": 0.3279,
+      "step": 3394
+    },
+    {
+      "epoch": 2.641338174273859,
+      "grad_norm": 0.5350475329610251,
+      "learning_rate": 4.276189897103916e-07,
+      "loss": 0.3383,
+      "step": 3395
+    },
+    {
+      "epoch": 2.642116182572614,
+      "grad_norm": 0.5441094897151925,
+      "learning_rate": 4.2578861782967276e-07,
+      "loss": 0.3739,
+      "step": 3396
+    },
+    {
+      "epoch": 2.6428941908713695,
+      "grad_norm": 0.4426855124230752,
+      "learning_rate": 4.239619974696468e-07,
+      "loss": 0.32,
+      "step": 3397
+    },
+    {
+      "epoch": 2.6436721991701244,
+      "grad_norm": 0.5423516469116146,
+      "learning_rate": 4.2213913012840947e-07,
+      "loss": 0.3693,
+      "step": 3398
+    },
+    {
+      "epoch": 2.64445020746888,
+      "grad_norm": 0.5424332148683576,
+      "learning_rate": 4.2032001730097474e-07,
+      "loss": 0.3611,
+      "step": 3399
+    },
+    {
+      "epoch": 2.645228215767635,
+      "grad_norm": 0.5198326163774357,
+      "learning_rate": 4.185046604792847e-07,
+      "loss": 0.299,
+      "step": 3400
+    },
+    {
+      "epoch": 2.64600622406639,
+      "grad_norm": 0.5495027939890326,
+      "learning_rate": 4.1669306115219355e-07,
+      "loss": 0.3566,
+      "step": 3401
+    },
+    {
+      "epoch": 2.646784232365145,
+      "grad_norm": 0.5529333576601743,
+      "learning_rate": 4.148852208054799e-07,
+      "loss": 0.3464,
+      "step": 3402
+    },
+    {
+      "epoch": 2.6475622406639006,
+      "grad_norm": 0.5283614246313405,
+      "learning_rate": 4.1308114092183384e-07,
+      "loss": 0.3232,
+      "step": 3403
+    },
+    {
+      "epoch": 2.6483402489626555,
+      "grad_norm": 0.5164344048159928,
+      "learning_rate": 4.1128082298086455e-07,
+      "loss": 0.344,
+      "step": 3404
+    },
+    {
+      "epoch": 2.649118257261411,
+      "grad_norm": 0.4750318668317547,
+      "learning_rate": 4.0948426845909774e-07,
+      "loss": 0.3083,
+      "step": 3405
+    },
+    {
+      "epoch": 2.649896265560166,
+      "grad_norm": 0.5627520377670958,
+      "learning_rate": 4.0769147882996797e-07,
+      "loss": 0.3622,
+      "step": 3406
+    },
+    {
+      "epoch": 2.6506742738589213,
+      "grad_norm": 0.5002444324930331,
+      "learning_rate": 4.059024555638247e-07,
+      "loss": 0.314,
+      "step": 3407
+    },
+    {
+      "epoch": 2.6514522821576763,
+      "grad_norm": 0.5165376785670365,
+      "learning_rate": 4.041172001279281e-07,
+      "loss": 0.3531,
+      "step": 3408
+    },
+    {
+      "epoch": 2.6522302904564317,
+      "grad_norm": 0.533900821600652,
+      "learning_rate": 4.0233571398644864e-07,
+      "loss": 0.3773,
+      "step": 3409
+    },
+    {
+      "epoch": 2.6530082987551866,
+      "grad_norm": 0.529978663161268,
+      "learning_rate": 4.0055799860046375e-07,
+      "loss": 0.3667,
+      "step": 3410
+    },
+    {
+      "epoch": 2.653786307053942,
+      "grad_norm": 0.5217265405911854,
+      "learning_rate": 3.9878405542796093e-07,
+      "loss": 0.3344,
+      "step": 3411
+    },
+    {
+      "epoch": 2.654564315352697,
+      "grad_norm": 0.5260692735945439,
+      "learning_rate": 3.9701388592383093e-07,
+      "loss": 0.3497,
+      "step": 3412
+    },
+    {
+      "epoch": 2.6553423236514524,
+      "grad_norm": 0.5559857965754573,
+      "learning_rate": 3.9524749153987185e-07,
+      "loss": 0.368,
+      "step": 3413
+    },
+    {
+      "epoch": 2.6561203319502074,
+      "grad_norm": 0.5799268513823591,
+      "learning_rate": 3.934848737247848e-07,
+      "loss": 0.3331,
+      "step": 3414
+    },
+    {
+      "epoch": 2.6568983402489628,
+      "grad_norm": 0.5326283236744563,
+      "learning_rate": 3.9172603392417355e-07,
+      "loss": 0.3681,
+      "step": 3415
+    },
+    {
+      "epoch": 2.6576763485477177,
+      "grad_norm": 0.554861100082353,
+      "learning_rate": 3.899709735805435e-07,
+      "loss": 0.316,
+      "step": 3416
+    },
+    {
+      "epoch": 2.658454356846473,
+      "grad_norm": 0.5370288971452704,
+      "learning_rate": 3.882196941332983e-07,
+      "loss": 0.3237,
+      "step": 3417
+    },
+    {
+      "epoch": 2.659232365145228,
+      "grad_norm": 0.559197557290659,
+      "learning_rate": 3.864721970187446e-07,
+      "loss": 0.3698,
+      "step": 3418
+    },
+    {
+      "epoch": 2.6600103734439835,
+      "grad_norm": 0.5076713267690927,
+      "learning_rate": 3.84728483670086e-07,
+      "loss": 0.3157,
+      "step": 3419
+    },
+    {
+      "epoch": 2.6607883817427385,
+      "grad_norm": 0.5197034571600956,
+      "learning_rate": 3.829885555174184e-07,
+      "loss": 0.3907,
+      "step": 3420
+    },
+    {
+      "epoch": 2.661566390041494,
+      "grad_norm": 0.5120375390658074,
+      "learning_rate": 3.812524139877405e-07,
+      "loss": 0.3568,
+      "step": 3421
+    },
+    {
+      "epoch": 2.662344398340249,
+      "grad_norm": 0.5565116290355612,
+      "learning_rate": 3.7952006050493763e-07,
+      "loss": 0.3486,
+      "step": 3422
+    },
+    {
+      "epoch": 2.6631224066390042,
+      "grad_norm": 0.5627226538375619,
+      "learning_rate": 3.777914964897966e-07,
+      "loss": 0.3567,
+      "step": 3423
+    },
+    {
+      "epoch": 2.663900414937759,
+      "grad_norm": 0.48615905819088606,
+      "learning_rate": 3.760667233599896e-07,
+      "loss": 0.3303,
+      "step": 3424
+    },
+    {
+      "epoch": 2.6646784232365146,
+      "grad_norm": 0.5119869398023774,
+      "learning_rate": 3.743457425300817e-07,
+      "loss": 0.3389,
+      "step": 3425
+    },
+    {
+      "epoch": 2.6654564315352696,
+      "grad_norm": 0.5545349631084645,
+      "learning_rate": 3.7262855541153056e-07,
+      "loss": 0.324,
+      "step": 3426
+    },
+    {
+      "epoch": 2.666234439834025,
+      "grad_norm": 0.5250077758838492,
+      "learning_rate": 3.709151634126784e-07,
+      "loss": 0.3814,
+      "step": 3427
+    },
+    {
+      "epoch": 2.66701244813278,
+      "grad_norm": 0.48749542006494356,
+      "learning_rate": 3.6920556793875737e-07,
+      "loss": 0.3168,
+      "step": 3428
+    },
+    {
+      "epoch": 2.6677904564315353,
+      "grad_norm": 0.6188026898711676,
+      "learning_rate": 3.6749977039188423e-07,
+      "loss": 0.3604,
+      "step": 3429
+    },
+    {
+      "epoch": 2.6685684647302903,
+      "grad_norm": 0.5134255815768569,
+      "learning_rate": 3.6579777217106315e-07,
+      "loss": 0.3794,
+      "step": 3430
+    },
+    {
+      "epoch": 2.6693464730290457,
+      "grad_norm": 0.45665516360156594,
+      "learning_rate": 3.640995746721798e-07,
+      "loss": 0.2765,
+      "step": 3431
+    },
+    {
+      "epoch": 2.6701244813278007,
+      "grad_norm": 0.48759496629896854,
+      "learning_rate": 3.6240517928800513e-07,
+      "loss": 0.3462,
+      "step": 3432
+    },
+    {
+      "epoch": 2.670902489626556,
+      "grad_norm": 0.4729497841774127,
+      "learning_rate": 3.6071458740818965e-07,
+      "loss": 0.3363,
+      "step": 3433
+    },
+    {
+      "epoch": 2.671680497925311,
+      "grad_norm": 0.45250046479373396,
+      "learning_rate": 3.5902780041926545e-07,
+      "loss": 0.3252,
+      "step": 3434
+    },
+    {
+      "epoch": 2.6724585062240664,
+      "grad_norm": 0.5224093056556541,
+      "learning_rate": 3.573448197046442e-07,
+      "loss": 0.3467,
+      "step": 3435
+    },
+    {
+      "epoch": 2.673236514522822,
+      "grad_norm": 0.5093563096151861,
+      "learning_rate": 3.556656466446151e-07,
+      "loss": 0.3161,
+      "step": 3436
+    },
+    {
+      "epoch": 2.674014522821577,
+      "grad_norm": 0.5491304175450111,
+      "learning_rate": 3.5399028261634715e-07,
+      "loss": 0.3914,
+      "step": 3437
+    },
+    {
+      "epoch": 2.6747925311203318,
+      "grad_norm": 0.5046625079520122,
+      "learning_rate": 3.52318728993879e-07,
+      "loss": 0.3088,
+      "step": 3438
+    },
+    {
+      "epoch": 2.675570539419087,
+      "grad_norm": 0.554402140000515,
+      "learning_rate": 3.5065098714813284e-07,
+      "loss": 0.3636,
+      "step": 3439
+    },
+    {
+      "epoch": 2.6763485477178426,
+      "grad_norm": 0.570650152120489,
+      "learning_rate": 3.4898705844689794e-07,
+      "loss": 0.3418,
+      "step": 3440
+    },
+    {
+      "epoch": 2.6771265560165975,
+      "grad_norm": 0.47622647405145924,
+      "learning_rate": 3.473269442548394e-07,
+      "loss": 0.3236,
+      "step": 3441
+    },
+    {
+      "epoch": 2.6779045643153525,
+      "grad_norm": 0.5214121190815084,
+      "learning_rate": 3.456706459334935e-07,
+      "loss": 0.3369,
+      "step": 3442
+    },
+    {
+      "epoch": 2.678682572614108,
+      "grad_norm": 0.5382303450267175,
+      "learning_rate": 3.440181648412644e-07,
+      "loss": 0.3629,
+      "step": 3443
+    },
+    {
+      "epoch": 2.6794605809128633,
+      "grad_norm": 0.5117419369853605,
+      "learning_rate": 3.4236950233343114e-07,
+      "loss": 0.3606,
+      "step": 3444
+    },
+    {
+      "epoch": 2.6802385892116183,
+      "grad_norm": 0.44585205241424863,
+      "learning_rate": 3.407246597621339e-07,
+      "loss": 0.3125,
+      "step": 3445
+    },
+    {
+      "epoch": 2.6810165975103732,
+      "grad_norm": 0.5428427569018188,
+      "learning_rate": 3.3908363847638515e-07,
+      "loss": 0.3473,
+      "step": 3446
+    },
+    {
+      "epoch": 2.6817946058091287,
+      "grad_norm": 0.6445557666056658,
+      "learning_rate": 3.3744643982206336e-07,
+      "loss": 0.3401,
+      "step": 3447
+    },
+    {
+      "epoch": 2.682572614107884,
+      "grad_norm": 0.5518694505140711,
+      "learning_rate": 3.358130651419067e-07,
+      "loss": 0.3916,
+      "step": 3448
+    },
+    {
+      "epoch": 2.683350622406639,
+      "grad_norm": 0.6043233810064366,
+      "learning_rate": 3.3418351577552466e-07,
+      "loss": 0.3009,
+      "step": 3449
+    },
+    {
+      "epoch": 2.684128630705394,
+      "grad_norm": 0.49447271888530747,
+      "learning_rate": 3.325577930593826e-07,
+      "loss": 0.2937,
+      "step": 3450
+    },
+    {
+      "epoch": 2.6849066390041494,
+      "grad_norm": 0.5475333896122111,
+      "learning_rate": 3.3093589832681127e-07,
+      "loss": 0.3625,
+      "step": 3451
+    },
+    {
+      "epoch": 2.685684647302905,
+      "grad_norm": 0.5210455738945173,
+      "learning_rate": 3.29317832908001e-07,
+      "loss": 0.3564,
+      "step": 3452
+    },
+    {
+      "epoch": 2.6864626556016598,
+      "grad_norm": 0.4589319434554803,
+      "learning_rate": 3.2770359813000196e-07,
+      "loss": 0.2953,
+      "step": 3453
+    },
+    {
+      "epoch": 2.6872406639004147,
+      "grad_norm": 0.5742192602085746,
+      "learning_rate": 3.260931953167212e-07,
+      "loss": 0.3757,
+      "step": 3454
+    },
+    {
+      "epoch": 2.68801867219917,
+      "grad_norm": 0.4885294989835661,
+      "learning_rate": 3.2448662578892453e-07,
+      "loss": 0.3326,
+      "step": 3455
+    },
+    {
+      "epoch": 2.6887966804979255,
+      "grad_norm": 0.5051254724309303,
+      "learning_rate": 3.2288389086423345e-07,
+      "loss": 0.3907,
+      "step": 3456
+    },
+    {
+      "epoch": 2.6895746887966805,
+      "grad_norm": 0.48979436709830765,
+      "learning_rate": 3.212849918571248e-07,
+      "loss": 0.335,
+      "step": 3457
+    },
+    {
+      "epoch": 2.6903526970954355,
+      "grad_norm": 0.5472128264184987,
+      "learning_rate": 3.19689930078928e-07,
+      "loss": 0.3584,
+      "step": 3458
+    },
+    {
+      "epoch": 2.691130705394191,
+      "grad_norm": 0.4811260996840474,
+      "learning_rate": 3.1809870683782664e-07,
+      "loss": 0.3474,
+      "step": 3459
+    },
+    {
+      "epoch": 2.6919087136929463,
+      "grad_norm": 0.511925237427149,
+      "learning_rate": 3.165113234388567e-07,
+      "loss": 0.307,
+      "step": 3460
+    },
+    {
+      "epoch": 2.6926867219917012,
+      "grad_norm": 0.5726112511629228,
+      "learning_rate": 3.149277811839041e-07,
+      "loss": 0.4212,
+      "step": 3461
+    },
+    {
+      "epoch": 2.693464730290456,
+      "grad_norm": 0.4534788500504679,
+      "learning_rate": 3.1334808137170327e-07,
+      "loss": 0.3085,
+      "step": 3462
+    },
+    {
+      "epoch": 2.6942427385892116,
+      "grad_norm": 0.5292427988419659,
+      "learning_rate": 3.1177222529784057e-07,
+      "loss": 0.3502,
+      "step": 3463
+    },
+    {
+      "epoch": 2.695020746887967,
+      "grad_norm": 0.5685847573092989,
+      "learning_rate": 3.102002142547456e-07,
+      "loss": 0.3783,
+      "step": 3464
+    },
+    {
+      "epoch": 2.695798755186722,
+      "grad_norm": 0.5385756619673522,
+      "learning_rate": 3.08632049531698e-07,
+      "loss": 0.3525,
+      "step": 3465
+    },
+    {
+      "epoch": 2.696576763485477,
+      "grad_norm": 0.5500596721810548,
+      "learning_rate": 3.070677324148219e-07,
+      "loss": 0.3538,
+      "step": 3466
+    },
+    {
+      "epoch": 2.6973547717842323,
+      "grad_norm": 0.48077435862461537,
+      "learning_rate": 3.0550726418708487e-07,
+      "loss": 0.3285,
+      "step": 3467
+    },
+    {
+      "epoch": 2.6981327800829877,
+      "grad_norm": 0.4737285066664222,
+      "learning_rate": 3.0395064612829985e-07,
+      "loss": 0.3207,
+      "step": 3468
+    },
+    {
+      "epoch": 2.6989107883817427,
+      "grad_norm": 0.5406259486243598,
+      "learning_rate": 3.0239787951511837e-07,
+      "loss": 0.3741,
+      "step": 3469
+    },
+    {
+      "epoch": 2.6996887966804977,
+      "grad_norm": 0.4673420876421153,
+      "learning_rate": 3.0084896562103784e-07,
+      "loss": 0.3018,
+      "step": 3470
+    },
+    {
+      "epoch": 2.700466804979253,
+      "grad_norm": 0.522490775027069,
+      "learning_rate": 2.993039057163927e-07,
+      "loss": 0.3302,
+      "step": 3471
+    },
+    {
+      "epoch": 2.7012448132780085,
+      "grad_norm": 0.524280035772461,
+      "learning_rate": 2.9776270106835645e-07,
+      "loss": 0.3149,
+      "step": 3472
+    },
+    {
+      "epoch": 2.7020228215767634,
+      "grad_norm": 0.549831357100294,
+      "learning_rate": 2.9622535294094426e-07,
+      "loss": 0.3618,
+      "step": 3473
+    },
+    {
+      "epoch": 2.7028008298755184,
+      "grad_norm": 0.5204058747304708,
+      "learning_rate": 2.9469186259500293e-07,
+      "loss": 0.3628,
+      "step": 3474
+    },
+    {
+      "epoch": 2.703578838174274,
+      "grad_norm": 0.5207303617107185,
+      "learning_rate": 2.9316223128822155e-07,
+      "loss": 0.3475,
+      "step": 3475
+    },
+    {
+      "epoch": 2.704356846473029,
+      "grad_norm": 0.5602402534675832,
+      "learning_rate": 2.916364602751187e-07,
+      "loss": 0.3871,
+      "step": 3476
+    },
+    {
+      "epoch": 2.705134854771784,
+      "grad_norm": 0.5260195637419884,
+      "learning_rate": 2.901145508070491e-07,
+      "loss": 0.3419,
+      "step": 3477
+    },
+    {
+      "epoch": 2.7059128630705396,
+      "grad_norm": 0.5055678939589567,
+      "learning_rate": 2.885965041322014e-07,
+      "loss": 0.3625,
+      "step": 3478
+    },
+    {
+      "epoch": 2.7066908713692945,
+      "grad_norm": 0.5683844680205359,
+      "learning_rate": 2.8708232149559533e-07,
+      "loss": 0.3352,
+      "step": 3479
+    },
+    {
+      "epoch": 2.70746887966805,
+      "grad_norm": 0.5223666763504834,
+      "learning_rate": 2.855720041390814e-07,
+      "loss": 0.2835,
+      "step": 3480
+    },
+    {
+      "epoch": 2.708246887966805,
+      "grad_norm": 0.5397320210180953,
+      "learning_rate": 2.840655533013403e-07,
+      "loss": 0.3721,
+      "step": 3481
+    },
+    {
+      "epoch": 2.7090248962655603,
+      "grad_norm": 0.513895200421175,
+      "learning_rate": 2.8256297021788116e-07,
+      "loss": 0.3022,
+      "step": 3482
+    },
+    {
+      "epoch": 2.7098029045643153,
+      "grad_norm": 0.511314541824273,
+      "learning_rate": 2.8106425612104136e-07,
+      "loss": 0.3301,
+      "step": 3483
+    },
+    {
+      "epoch": 2.7105809128630707,
+      "grad_norm": 0.527190667569253,
+      "learning_rate": 2.7956941223998555e-07,
+      "loss": 0.3622,
+      "step": 3484
+    },
+    {
+      "epoch": 2.7113589211618256,
+      "grad_norm": 0.5441849348793651,
+      "learning_rate": 2.780784398007036e-07,
+      "loss": 0.3266,
+      "step": 3485
+    },
+    {
+      "epoch": 2.712136929460581,
+      "grad_norm": 0.5334726455991334,
+      "learning_rate": 2.7659134002601064e-07,
+      "loss": 0.3718,
+      "step": 3486
+    },
+    {
+      "epoch": 2.712914937759336,
+      "grad_norm": 0.5148653406257049,
+      "learning_rate": 2.7510811413554493e-07,
+      "loss": 0.368,
+      "step": 3487
+    },
+    {
+      "epoch": 2.7136929460580914,
+      "grad_norm": 0.5335878813033207,
+      "learning_rate": 2.736287633457679e-07,
+      "loss": 0.3298,
+      "step": 3488
+    },
+    {
+      "epoch": 2.7144709543568464,
+      "grad_norm": 0.4998220548892212,
+      "learning_rate": 2.721532888699646e-07,
+      "loss": 0.3396,
+      "step": 3489
+    },
+    {
+      "epoch": 2.715248962655602,
+      "grad_norm": 0.4699061059051531,
+      "learning_rate": 2.70681691918237e-07,
+      "loss": 0.3658,
+      "step": 3490
+    },
+    {
+      "epoch": 2.7160269709543567,
+      "grad_norm": 0.4629573671078091,
+      "learning_rate": 2.692139736975119e-07,
+      "loss": 0.3489,
+      "step": 3491
+    },
+    {
+      "epoch": 2.716804979253112,
+      "grad_norm": 0.49239873985780913,
+      "learning_rate": 2.677501354115297e-07,
+      "loss": 0.3375,
+      "step": 3492
+    },
+    {
+      "epoch": 2.717582987551867,
+      "grad_norm": 0.5251018565582074,
+      "learning_rate": 2.662901782608529e-07,
+      "loss": 0.3404,
+      "step": 3493
+    },
+    {
+      "epoch": 2.7183609958506225,
+      "grad_norm": 0.5544880487158579,
+      "learning_rate": 2.648341034428603e-07,
+      "loss": 0.3513,
+      "step": 3494
+    },
+    {
+      "epoch": 2.7191390041493775,
+      "grad_norm": 0.49098380666720476,
+      "learning_rate": 2.633819121517434e-07,
+      "loss": 0.3311,
+      "step": 3495
+    },
+    {
+      "epoch": 2.719917012448133,
+      "grad_norm": 0.4727675912451536,
+      "learning_rate": 2.619336055785132e-07,
+      "loss": 0.3427,
+      "step": 3496
+    },
+    {
+      "epoch": 2.720695020746888,
+      "grad_norm": 0.4920719109041581,
+      "learning_rate": 2.6048918491099026e-07,
+      "loss": 0.2961,
+      "step": 3497
+    },
+    {
+      "epoch": 2.7214730290456433,
+      "grad_norm": 0.47526376918775703,
+      "learning_rate": 2.590486513338114e-07,
+      "loss": 0.3104,
+      "step": 3498
+    },
+    {
+      "epoch": 2.722251037344398,
+      "grad_norm": 0.4956372590315787,
+      "learning_rate": 2.5761200602842485e-07,
+      "loss": 0.3709,
+      "step": 3499
+    },
+    {
+      "epoch": 2.7230290456431536,
+      "grad_norm": 0.4998636023395081,
+      "learning_rate": 2.561792501730881e-07,
+      "loss": 0.3625,
+      "step": 3500
+    },
+    {
+      "epoch": 2.7238070539419086,
+      "grad_norm": 0.5037176160231542,
+      "learning_rate": 2.5475038494287064e-07,
+      "loss": 0.3904,
+      "step": 3501
+    },
+    {
+      "epoch": 2.724585062240664,
+      "grad_norm": 0.52060308198054,
+      "learning_rate": 2.5332541150965007e-07,
+      "loss": 0.3523,
+      "step": 3502
+    },
+    {
+      "epoch": 2.725363070539419,
+      "grad_norm": 0.5323763089833673,
+      "learning_rate": 2.519043310421132e-07,
+      "loss": 0.362,
+      "step": 3503
+    },
+    {
+      "epoch": 2.7261410788381744,
+      "grad_norm": 0.5551094935268409,
+      "learning_rate": 2.504871447057533e-07,
+      "loss": 0.3442,
+      "step": 3504
+    },
+    {
+      "epoch": 2.7269190871369293,
+      "grad_norm": 0.6423866441835647,
+      "learning_rate": 2.4907385366286897e-07,
+      "loss": 0.3413,
+      "step": 3505
+    },
+    {
+      "epoch": 2.7276970954356847,
+      "grad_norm": 0.48074652673245544,
+      "learning_rate": 2.4766445907256575e-07,
+      "loss": 0.3051,
+      "step": 3506
+    },
+    {
+      "epoch": 2.7284751037344397,
+      "grad_norm": 0.4879563152190972,
+      "learning_rate": 2.4625896209075294e-07,
+      "loss": 0.2913,
+      "step": 3507
+    },
+    {
+      "epoch": 2.729253112033195,
+      "grad_norm": 0.5133084480570629,
+      "learning_rate": 2.448573638701429e-07,
+      "loss": 0.3969,
+      "step": 3508
+    },
+    {
+      "epoch": 2.73003112033195,
+      "grad_norm": 0.543368803608402,
+      "learning_rate": 2.4345966556025055e-07,
+      "loss": 0.3814,
+      "step": 3509
+    },
+    {
+      "epoch": 2.7308091286307055,
+      "grad_norm": 0.47283839788194904,
+      "learning_rate": 2.4206586830739343e-07,
+      "loss": 0.3407,
+      "step": 3510
+    },
+    {
+      "epoch": 2.7315871369294604,
+      "grad_norm": 0.47697415049410297,
+      "learning_rate": 2.406759732546865e-07,
+      "loss": 0.314,
+      "step": 3511
+    },
+    {
+      "epoch": 2.732365145228216,
+      "grad_norm": 0.5702362689212277,
+      "learning_rate": 2.3928998154204905e-07,
+      "loss": 0.3924,
+      "step": 3512
+    },
+    {
+      "epoch": 2.733143153526971,
+      "grad_norm": 0.47570718264804607,
+      "learning_rate": 2.3790789430619354e-07,
+      "loss": 0.2993,
+      "step": 3513
+    },
+    {
+      "epoch": 2.733921161825726,
+      "grad_norm": 0.5274090316788099,
+      "learning_rate": 2.3652971268063595e-07,
+      "loss": 0.3457,
+      "step": 3514
+    },
+    {
+      "epoch": 2.7346991701244816,
+      "grad_norm": 0.6254970380724734,
+      "learning_rate": 2.351554377956855e-07,
+      "loss": 0.4244,
+      "step": 3515
+    },
+    {
+      "epoch": 2.7354771784232366,
+      "grad_norm": 0.45344053283513136,
+      "learning_rate": 2.3378507077844614e-07,
+      "loss": 0.2744,
+      "step": 3516
+    },
+    {
+      "epoch": 2.7362551867219915,
+      "grad_norm": 0.5117193065656884,
+      "learning_rate": 2.324186127528222e-07,
+      "loss": 0.3342,
+      "step": 3517
+    },
+    {
+      "epoch": 2.737033195020747,
+      "grad_norm": 0.49471757577458864,
+      "learning_rate": 2.310560648395055e-07,
+      "loss": 0.3079,
+      "step": 3518
+    },
+    {
+      "epoch": 2.7378112033195023,
+      "grad_norm": 0.46568636240183536,
+      "learning_rate": 2.2969742815598605e-07,
+      "loss": 0.3284,
+      "step": 3519
+    },
+    {
+      "epoch": 2.7385892116182573,
+      "grad_norm": 0.4836626167751069,
+      "learning_rate": 2.283427038165431e-07,
+      "loss": 0.2869,
+      "step": 3520
+    },
+    {
+      "epoch": 2.7393672199170123,
+      "grad_norm": 0.5469894458271374,
+      "learning_rate": 2.2699189293224788e-07,
+      "loss": 0.3436,
+      "step": 3521
+    },
+    {
+      "epoch": 2.7401452282157677,
+      "grad_norm": 0.6167372495237607,
+      "learning_rate": 2.2564499661096472e-07,
+      "loss": 0.3977,
+      "step": 3522
+    },
+    {
+      "epoch": 2.740923236514523,
+      "grad_norm": 0.5564866684181,
+      "learning_rate": 2.243020159573428e-07,
+      "loss": 0.3127,
+      "step": 3523
+    },
+    {
+      "epoch": 2.741701244813278,
+      "grad_norm": 0.47821900386238053,
+      "learning_rate": 2.229629520728227e-07,
+      "loss": 0.3144,
+      "step": 3524
+    },
+    {
+      "epoch": 2.742479253112033,
+      "grad_norm": 0.46736573200981807,
+      "learning_rate": 2.216278060556326e-07,
+      "loss": 0.3577,
+      "step": 3525
+    },
+    {
+      "epoch": 2.7432572614107884,
+      "grad_norm": 0.5045431452772361,
+      "learning_rate": 2.2029657900078717e-07,
+      "loss": 0.3188,
+      "step": 3526
+    },
+    {
+      "epoch": 2.744035269709544,
+      "grad_norm": 0.5175980347261141,
+      "learning_rate": 2.1896927200008633e-07,
+      "loss": 0.329,
+      "step": 3527
+    },
+    {
+      "epoch": 2.7448132780082988,
+      "grad_norm": 0.47418610878181744,
+      "learning_rate": 2.1764588614211545e-07,
+      "loss": 0.3357,
+      "step": 3528
+    },
+    {
+      "epoch": 2.7455912863070537,
+      "grad_norm": 0.47237669426785844,
+      "learning_rate": 2.1632642251224412e-07,
+      "loss": 0.3417,
+      "step": 3529
+    },
+    {
+      "epoch": 2.746369294605809,
+      "grad_norm": 0.4820793838276029,
+      "learning_rate": 2.15010882192625e-07,
+      "loss": 0.3269,
+      "step": 3530
+    },
+    {
+      "epoch": 2.7471473029045645,
+      "grad_norm": 0.4824432187708331,
+      "learning_rate": 2.1369926626219338e-07,
+      "loss": 0.3788,
+      "step": 3531
+    },
+    {
+      "epoch": 2.7479253112033195,
+      "grad_norm": 0.4967545440490051,
+      "learning_rate": 2.1239157579666546e-07,
+      "loss": 0.34,
+      "step": 3532
+    },
+    {
+      "epoch": 2.7487033195020745,
+      "grad_norm": 0.46019975526066415,
+      "learning_rate": 2.110878118685389e-07,
+      "loss": 0.2918,
+      "step": 3533
+    },
+    {
+      "epoch": 2.74948132780083,
+      "grad_norm": 0.6636397780670802,
+      "learning_rate": 2.097879755470883e-07,
+      "loss": 0.4186,
+      "step": 3534
+    },
+    {
+      "epoch": 2.7502593360995853,
+      "grad_norm": 0.6007038003837566,
+      "learning_rate": 2.0849206789837094e-07,
+      "loss": 0.378,
+      "step": 3535
+    },
+    {
+      "epoch": 2.7510373443983402,
+      "grad_norm": 0.5193676575684704,
+      "learning_rate": 2.0720008998522056e-07,
+      "loss": 0.3654,
+      "step": 3536
+    },
+    {
+      "epoch": 2.751815352697095,
+      "grad_norm": 0.6485943545290985,
+      "learning_rate": 2.0591204286724508e-07,
+      "loss": 0.3896,
+      "step": 3537
+    },
+    {
+      "epoch": 2.7525933609958506,
+      "grad_norm": 0.47873712937673857,
+      "learning_rate": 2.0462792760083395e-07,
+      "loss": 0.3034,
+      "step": 3538
+    },
+    {
+      "epoch": 2.753371369294606,
+      "grad_norm": 0.5140651376399792,
+      "learning_rate": 2.0334774523914692e-07,
+      "loss": 0.3365,
+      "step": 3539
+    },
+    {
+      "epoch": 2.754149377593361,
+      "grad_norm": 0.5676648618470439,
+      "learning_rate": 2.020714968321219e-07,
+      "loss": 0.3741,
+      "step": 3540
+    },
+    {
+      "epoch": 2.754927385892116,
+      "grad_norm": 0.47175825308305686,
+      "learning_rate": 2.007991834264672e-07,
+      "loss": 0.3209,
+      "step": 3541
+    },
+    {
+      "epoch": 2.7557053941908713,
+      "grad_norm": 0.4523820899747808,
+      "learning_rate": 1.9953080606566588e-07,
+      "loss": 0.3164,
+      "step": 3542
+    },
+    {
+      "epoch": 2.7564834024896268,
+      "grad_norm": 0.45445228701225204,
+      "learning_rate": 1.9826636578997362e-07,
+      "loss": 0.3118,
+      "step": 3543
+    },
+    {
+      "epoch": 2.7572614107883817,
+      "grad_norm": 0.4321929030843774,
+      "learning_rate": 1.9700586363641473e-07,
+      "loss": 0.303,
+      "step": 3544
+    },
+    {
+      "epoch": 2.7580394190871367,
+      "grad_norm": 0.4823841848456945,
+      "learning_rate": 1.9574930063878504e-07,
+      "loss": 0.3657,
+      "step": 3545
+    },
+    {
+      "epoch": 2.758817427385892,
+      "grad_norm": 0.5083895074407834,
+      "learning_rate": 1.944966778276497e-07,
+      "loss": 0.3344,
+      "step": 3546
+    },
+    {
+      "epoch": 2.7595954356846475,
+      "grad_norm": 0.5419145226977079,
+      "learning_rate": 1.932479962303424e-07,
+      "loss": 0.3709,
+      "step": 3547
+    },
+    {
+      "epoch": 2.7603734439834025,
+      "grad_norm": 0.46053473740734163,
+      "learning_rate": 1.92003256870964e-07,
+      "loss": 0.305,
+      "step": 3548
+    },
+    {
+      "epoch": 2.7611514522821574,
+      "grad_norm": 0.5201704261310988,
+      "learning_rate": 1.907624607703823e-07,
+      "loss": 0.3536,
+      "step": 3549
+    },
+    {
+      "epoch": 2.761929460580913,
+      "grad_norm": 0.5390410179840013,
+      "learning_rate": 1.8952560894623107e-07,
+      "loss": 0.3363,
+      "step": 3550
+    },
+    {
+      "epoch": 2.7627074688796682,
+      "grad_norm": 0.5243677107510115,
+      "learning_rate": 1.8829270241290943e-07,
+      "loss": 0.3755,
+      "step": 3551
+    },
+    {
+      "epoch": 2.763485477178423,
+      "grad_norm": 0.5206519434705965,
+      "learning_rate": 1.8706374218158018e-07,
+      "loss": 0.3282,
+      "step": 3552
+    },
+    {
+      "epoch": 2.764263485477178,
+      "grad_norm": 0.5771468955640193,
+      "learning_rate": 1.8583872926017098e-07,
+      "loss": 0.3609,
+      "step": 3553
+    },
+    {
+      "epoch": 2.7650414937759336,
+      "grad_norm": 0.5127435187037712,
+      "learning_rate": 1.846176646533704e-07,
+      "loss": 0.3471,
+      "step": 3554
+    },
+    {
+      "epoch": 2.765819502074689,
+      "grad_norm": 0.571364196327506,
+      "learning_rate": 1.8340054936262952e-07,
+      "loss": 0.3124,
+      "step": 3555
+    },
+    {
+      "epoch": 2.766597510373444,
+      "grad_norm": 0.5529001161006795,
+      "learning_rate": 1.82187384386161e-07,
+      "loss": 0.3587,
+      "step": 3556
+    },
+    {
+      "epoch": 2.7673755186721993,
+      "grad_norm": 0.5491208406363741,
+      "learning_rate": 1.8097817071893674e-07,
+      "loss": 0.3025,
+      "step": 3557
+    },
+    {
+      "epoch": 2.7681535269709543,
+      "grad_norm": 0.5543394914628731,
+      "learning_rate": 1.7977290935268842e-07,
+      "loss": 0.3743,
+      "step": 3558
+    },
+    {
+      "epoch": 2.7689315352697097,
+      "grad_norm": 0.5307422967341597,
+      "learning_rate": 1.7857160127590645e-07,
+      "loss": 0.3659,
+      "step": 3559
+    },
+    {
+      "epoch": 2.7697095435684647,
+      "grad_norm": 0.4986593461713161,
+      "learning_rate": 1.773742474738377e-07,
+      "loss": 0.3246,
+      "step": 3560
+    },
+    {
+      "epoch": 2.77048755186722,
+      "grad_norm": 0.5045468258175748,
+      "learning_rate": 1.7618084892848775e-07,
+      "loss": 0.3748,
+      "step": 3561
+    },
+    {
+      "epoch": 2.771265560165975,
+      "grad_norm": 0.46164085875794564,
+      "learning_rate": 1.7499140661861868e-07,
+      "loss": 0.3013,
+      "step": 3562
+    },
+    {
+      "epoch": 2.7720435684647304,
+      "grad_norm": 0.48744352354444787,
+      "learning_rate": 1.7380592151974407e-07,
+      "loss": 0.3446,
+      "step": 3563
+    },
+    {
+      "epoch": 2.7728215767634854,
+      "grad_norm": 0.43575913233443514,
+      "learning_rate": 1.7262439460413727e-07,
+      "loss": 0.3187,
+      "step": 3564
+    },
+    {
+      "epoch": 2.773599585062241,
+      "grad_norm": 0.5074835230732381,
+      "learning_rate": 1.7144682684082093e-07,
+      "loss": 0.3328,
+      "step": 3565
+    },
+    {
+      "epoch": 2.7743775933609958,
+      "grad_norm": 0.46066579493546095,
+      "learning_rate": 1.7027321919557361e-07,
+      "loss": 0.3065,
+      "step": 3566
+    },
+    {
+      "epoch": 2.775155601659751,
+      "grad_norm": 0.5547128005215446,
+      "learning_rate": 1.691035726309237e-07,
+      "loss": 0.3739,
+      "step": 3567
+    },
+    {
+      "epoch": 2.775933609958506,
+      "grad_norm": 0.5627015567981491,
+      "learning_rate": 1.6793788810615276e-07,
+      "loss": 0.3803,
+      "step": 3568
+    },
+    {
+      "epoch": 2.7767116182572615,
+      "grad_norm": 0.5479397907064614,
+      "learning_rate": 1.6677616657729266e-07,
+      "loss": 0.3512,
+      "step": 3569
+    },
+    {
+      "epoch": 2.7774896265560165,
+      "grad_norm": 0.599187560862996,
+      "learning_rate": 1.65618408997123e-07,
+      "loss": 0.3551,
+      "step": 3570
+    },
+    {
+      "epoch": 2.778267634854772,
+      "grad_norm": 0.5333332371427961,
+      "learning_rate": 1.644646163151753e-07,
+      "loss": 0.3213,
+      "step": 3571
+    },
+    {
+      "epoch": 2.779045643153527,
+      "grad_norm": 0.5684967206666619,
+      "learning_rate": 1.6331478947772706e-07,
+      "loss": 0.3386,
+      "step": 3572
+    },
+    {
+      "epoch": 2.7798236514522823,
+      "grad_norm": 0.47763004225879624,
+      "learning_rate": 1.6216892942780504e-07,
+      "loss": 0.333,
+      "step": 3573
+    },
+    {
+      "epoch": 2.7806016597510372,
+      "grad_norm": 0.5456157508678409,
+      "learning_rate": 1.610270371051803e-07,
+      "loss": 0.3732,
+      "step": 3574
+    },
+    {
+      "epoch": 2.7813796680497926,
+      "grad_norm": 0.4954148453740823,
+      "learning_rate": 1.5988911344637204e-07,
+      "loss": 0.3097,
+      "step": 3575
+    },
+    {
+      "epoch": 2.7821576763485476,
+      "grad_norm": 0.45687404456926245,
+      "learning_rate": 1.5875515938464369e-07,
+      "loss": 0.3025,
+      "step": 3576
+    },
+    {
+      "epoch": 2.782935684647303,
+      "grad_norm": 0.5194165894443754,
+      "learning_rate": 1.5762517585000247e-07,
+      "loss": 0.3599,
+      "step": 3577
+    },
+    {
+      "epoch": 2.783713692946058,
+      "grad_norm": 0.5512490146787179,
+      "learning_rate": 1.5649916376920037e-07,
+      "loss": 0.3599,
+      "step": 3578
+    },
+    {
+      "epoch": 2.7844917012448134,
+      "grad_norm": 0.4688680594708169,
+      "learning_rate": 1.5537712406573092e-07,
+      "loss": 0.299,
+      "step": 3579
+    },
+    {
+      "epoch": 2.7852697095435683,
+      "grad_norm": 0.48296248600762853,
+      "learning_rate": 1.5425905765983184e-07,
+      "loss": 0.3416,
+      "step": 3580
+    },
+    {
+      "epoch": 2.7860477178423237,
+      "grad_norm": 0.482441862296252,
+      "learning_rate": 1.53144965468478e-07,
+      "loss": 0.3468,
+      "step": 3581
+    },
+    {
+      "epoch": 2.7868257261410787,
+      "grad_norm": 0.4668824770403465,
+      "learning_rate": 1.5203484840539018e-07,
+      "loss": 0.354,
+      "step": 3582
+    },
+    {
+      "epoch": 2.787603734439834,
+      "grad_norm": 0.5286718966449306,
+      "learning_rate": 1.5092870738102505e-07,
+      "loss": 0.3729,
+      "step": 3583
+    },
+    {
+      "epoch": 2.788381742738589,
+      "grad_norm": 0.5510889977972728,
+      "learning_rate": 1.4982654330257972e-07,
+      "loss": 0.3471,
+      "step": 3584
+    },
+    {
+      "epoch": 2.7891597510373445,
+      "grad_norm": 0.5076785350222012,
+      "learning_rate": 1.4872835707399058e-07,
+      "loss": 0.3361,
+      "step": 3585
+    },
+    {
+      "epoch": 2.7899377593360994,
+      "grad_norm": 0.5661834863716257,
+      "learning_rate": 1.4763414959592825e-07,
+      "loss": 0.3344,
+      "step": 3586
+    },
+    {
+      "epoch": 2.790715767634855,
+      "grad_norm": 0.5427685496335461,
+      "learning_rate": 1.465439217658049e-07,
+      "loss": 0.348,
+      "step": 3587
+    },
+    {
+      "epoch": 2.79149377593361,
+      "grad_norm": 0.49088462191769094,
+      "learning_rate": 1.4545767447776527e-07,
+      "loss": 0.3803,
+      "step": 3588
+    },
+    {
+      "epoch": 2.792271784232365,
+      "grad_norm": 0.5527570878872342,
+      "learning_rate": 1.4437540862269007e-07,
+      "loss": 0.3621,
+      "step": 3589
+    },
+    {
+      "epoch": 2.79304979253112,
+      "grad_norm": 0.49603829192848475,
+      "learning_rate": 1.4329712508819703e-07,
+      "loss": 0.3138,
+      "step": 3590
+    },
+    {
+      "epoch": 2.7938278008298756,
+      "grad_norm": 0.4743201727582203,
+      "learning_rate": 1.4222282475863435e-07,
+      "loss": 0.3098,
+      "step": 3591
+    },
+    {
+      "epoch": 2.7946058091286305,
+      "grad_norm": 0.5101731138144961,
+      "learning_rate": 1.4115250851508556e-07,
+      "loss": 0.3691,
+      "step": 3592
+    },
+    {
+      "epoch": 2.795383817427386,
+      "grad_norm": 0.4384143413380726,
+      "learning_rate": 1.4008617723536622e-07,
+      "loss": 0.3251,
+      "step": 3593
+    },
+    {
+      "epoch": 2.796161825726141,
+      "grad_norm": 0.501430149397751,
+      "learning_rate": 1.390238317940229e-07,
+      "loss": 0.3201,
+      "step": 3594
+    },
+    {
+      "epoch": 2.7969398340248963,
+      "grad_norm": 0.5407614033584263,
+      "learning_rate": 1.3796547306233477e-07,
+      "loss": 0.3846,
+      "step": 3595
+    },
+    {
+      "epoch": 2.7977178423236513,
+      "grad_norm": 0.5009266969447356,
+      "learning_rate": 1.3691110190831025e-07,
+      "loss": 0.294,
+      "step": 3596
+    },
+    {
+      "epoch": 2.7984958506224067,
+      "grad_norm": 0.4716593574445943,
+      "learning_rate": 1.3586071919668707e-07,
+      "loss": 0.3838,
+      "step": 3597
+    },
+    {
+      "epoch": 2.799273858921162,
+      "grad_norm": 0.4418842191867218,
+      "learning_rate": 1.3481432578893228e-07,
+      "loss": 0.2972,
+      "step": 3598
+    },
+    {
+      "epoch": 2.800051867219917,
+      "grad_norm": 0.5570948309470493,
+      "learning_rate": 1.3377192254324156e-07,
+      "loss": 0.3387,
+      "step": 3599
+    },
+    {
+      "epoch": 2.800829875518672,
+      "grad_norm": 0.6046745951670048,
+      "learning_rate": 1.3273351031453774e-07,
+      "loss": 0.3447,
+      "step": 3600
+    },
+    {
+      "epoch": 2.8016078838174274,
+      "grad_norm": 0.5636563314907799,
+      "learning_rate": 1.3169908995447068e-07,
+      "loss": 0.3605,
+      "step": 3601
+    },
+    {
+      "epoch": 2.802385892116183,
+      "grad_norm": 0.5016206431806672,
+      "learning_rate": 1.3066866231141395e-07,
+      "loss": 0.3496,
+      "step": 3602
+    },
+    {
+      "epoch": 2.803163900414938,
+      "grad_norm": 0.4717612145207449,
+      "learning_rate": 1.2964222823047046e-07,
+      "loss": 0.3439,
+      "step": 3603
+    },
+    {
+      "epoch": 2.8039419087136928,
+      "grad_norm": 0.4747315257233071,
+      "learning_rate": 1.286197885534657e-07,
+      "loss": 0.3102,
+      "step": 3604
+    },
+    {
+      "epoch": 2.804719917012448,
+      "grad_norm": 0.5707726214793507,
+      "learning_rate": 1.2760134411894887e-07,
+      "loss": 0.3669,
+      "step": 3605
+    },
+    {
+      "epoch": 2.8054979253112036,
+      "grad_norm": 0.5267604134985627,
+      "learning_rate": 1.2658689576219408e-07,
+      "loss": 0.3396,
+      "step": 3606
+    },
+    {
+      "epoch": 2.8062759336099585,
+      "grad_norm": 0.46662831408686345,
+      "learning_rate": 1.2557644431519524e-07,
+      "loss": 0.3168,
+      "step": 3607
+    },
+    {
+      "epoch": 2.8070539419087135,
+      "grad_norm": 0.5960878607918403,
+      "learning_rate": 1.2456999060667163e-07,
+      "loss": 0.3815,
+      "step": 3608
+    },
+    {
+      "epoch": 2.807831950207469,
+      "grad_norm": 0.5501687430028869,
+      "learning_rate": 1.235675354620608e-07,
+      "loss": 0.386,
+      "step": 3609
+    },
+    {
+      "epoch": 2.8086099585062243,
+      "grad_norm": 0.44700483618217596,
+      "learning_rate": 1.2256907970352229e-07,
+      "loss": 0.3004,
+      "step": 3610
+    },
+    {
+      "epoch": 2.8093879668049793,
+      "grad_norm": 0.5949571826775938,
+      "learning_rate": 1.2157462414993715e-07,
+      "loss": 0.3623,
+      "step": 3611
+    },
+    {
+      "epoch": 2.8101659751037342,
+      "grad_norm": 0.5044399030269813,
+      "learning_rate": 1.2058416961690134e-07,
+      "loss": 0.4023,
+      "step": 3612
+    },
+    {
+      "epoch": 2.8109439834024896,
+      "grad_norm": 0.489017200425495,
+      "learning_rate": 1.1959771691673506e-07,
+      "loss": 0.3143,
+      "step": 3613
+    },
+    {
+      "epoch": 2.811721991701245,
+      "grad_norm": 0.5397707112469494,
+      "learning_rate": 1.1861526685847168e-07,
+      "loss": 0.3553,
+      "step": 3614
+    },
+    {
+      "epoch": 2.8125,
+      "grad_norm": 0.47388908643766803,
+      "learning_rate": 1.1763682024786338e-07,
+      "loss": 0.3087,
+      "step": 3615
+    },
+    {
+      "epoch": 2.813278008298755,
+      "grad_norm": 0.5427883520644431,
+      "learning_rate": 1.1666237788737989e-07,
+      "loss": 0.3504,
+      "step": 3616
+    },
+    {
+      "epoch": 2.8140560165975104,
+      "grad_norm": 0.4943684468251194,
+      "learning_rate": 1.1569194057620636e-07,
+      "loss": 0.3232,
+      "step": 3617
+    },
+    {
+      "epoch": 2.8148340248962658,
+      "grad_norm": 0.5504235005804442,
+      "learning_rate": 1.1472550911024172e-07,
+      "loss": 0.3572,
+      "step": 3618
+    },
+    {
+      "epoch": 2.8156120331950207,
+      "grad_norm": 0.5315221055053853,
+      "learning_rate": 1.1376308428210193e-07,
+      "loss": 0.3537,
+      "step": 3619
+    },
+    {
+      "epoch": 2.8163900414937757,
+      "grad_norm": 0.5345173953834147,
+      "learning_rate": 1.1280466688111503e-07,
+      "loss": 0.3885,
+      "step": 3620
+    },
+    {
+      "epoch": 2.817168049792531,
+      "grad_norm": 0.4644597868410316,
+      "learning_rate": 1.1185025769332392e-07,
+      "loss": 0.3032,
+      "step": 3621
+    },
+    {
+      "epoch": 2.8179460580912865,
+      "grad_norm": 0.49706729814867967,
+      "learning_rate": 1.1089985750148247e-07,
+      "loss": 0.3615,
+      "step": 3622
+    },
+    {
+      "epoch": 2.8187240663900415,
+      "grad_norm": 0.48209793856371796,
+      "learning_rate": 1.0995346708505827e-07,
+      "loss": 0.3451,
+      "step": 3623
+    },
+    {
+      "epoch": 2.8195020746887964,
+      "grad_norm": 0.48325913818050453,
+      "learning_rate": 1.090110872202288e-07,
+      "loss": 0.3211,
+      "step": 3624
+    },
+    {
+      "epoch": 2.820280082987552,
+      "grad_norm": 0.48919138309250204,
+      "learning_rate": 1.0807271867988411e-07,
+      "loss": 0.3021,
+      "step": 3625
+    },
+    {
+      "epoch": 2.8210580912863072,
+      "grad_norm": 0.5810260405568397,
+      "learning_rate": 1.0713836223362195e-07,
+      "loss": 0.3831,
+      "step": 3626
+    },
+    {
+      "epoch": 2.821836099585062,
+      "grad_norm": 0.5389908698702055,
+      "learning_rate": 1.062080186477521e-07,
+      "loss": 0.349,
+      "step": 3627
+    },
+    {
+      "epoch": 2.822614107883817,
+      "grad_norm": 0.484678719118101,
+      "learning_rate": 1.0528168868529088e-07,
+      "loss": 0.3055,
+      "step": 3628
+    },
+    {
+      "epoch": 2.8233921161825726,
+      "grad_norm": 0.5312649200202941,
+      "learning_rate": 1.0435937310596556e-07,
+      "loss": 0.3428,
+      "step": 3629
+    },
+    {
+      "epoch": 2.824170124481328,
+      "grad_norm": 0.48248038342819405,
+      "learning_rate": 1.0344107266620773e-07,
+      "loss": 0.3156,
+      "step": 3630
+    },
+    {
+      "epoch": 2.824948132780083,
+      "grad_norm": 0.5183630384956427,
+      "learning_rate": 1.0252678811915884e-07,
+      "loss": 0.3437,
+      "step": 3631
+    },
+    {
+      "epoch": 2.825726141078838,
+      "grad_norm": 0.47289527225286304,
+      "learning_rate": 1.0161652021466628e-07,
+      "loss": 0.3363,
+      "step": 3632
+    },
+    {
+      "epoch": 2.8265041493775933,
+      "grad_norm": 0.551979519484257,
+      "learning_rate": 1.0071026969928011e-07,
+      "loss": 0.3648,
+      "step": 3633
+    },
+    {
+      "epoch": 2.8272821576763487,
+      "grad_norm": 0.48184283580870574,
+      "learning_rate": 9.980803731626076e-08,
+      "loss": 0.3236,
+      "step": 3634
+    },
+    {
+      "epoch": 2.8280601659751037,
+      "grad_norm": 0.5641548545881121,
+      "learning_rate": 9.890982380556858e-08,
+      "loss": 0.337,
+      "step": 3635
+    },
+    {
+      "epoch": 2.828838174273859,
+      "grad_norm": 0.5214243526275569,
+      "learning_rate": 9.801562990386981e-08,
+      "loss": 0.3304,
+      "step": 3636
+    },
+    {
+      "epoch": 2.829616182572614,
+      "grad_norm": 0.5368956039333335,
+      "learning_rate": 9.712545634453508e-08,
+      "loss": 0.3841,
+      "step": 3637
+    },
+    {
+      "epoch": 2.8303941908713695,
+      "grad_norm": 0.43145602382132325,
+      "learning_rate": 9.62393038576348e-08,
+      "loss": 0.2802,
+      "step": 3638
+    },
+    {
+      "epoch": 2.8311721991701244,
+      "grad_norm": 0.4872674977928451,
+      "learning_rate": 9.535717316994541e-08,
+      "loss": 0.3354,
+      "step": 3639
+    },
+    {
+      "epoch": 2.83195020746888,
+      "grad_norm": 0.4898737788451149,
+      "learning_rate": 9.447906500494153e-08,
+      "loss": 0.3224,
+      "step": 3640
+    },
+    {
+      "epoch": 2.832728215767635,
+      "grad_norm": 0.5429913865272162,
+      "learning_rate": 9.360498008279983e-08,
+      "loss": 0.3407,
+      "step": 3641
+    },
+    {
+      "epoch": 2.83350622406639,
+      "grad_norm": 0.5214335282777287,
+      "learning_rate": 9.2734919120398e-08,
+      "loss": 0.3532,
+      "step": 3642
+    },
+    {
+      "epoch": 2.834284232365145,
+      "grad_norm": 0.5212212109241927,
+      "learning_rate": 9.186888283131246e-08,
+      "loss": 0.3675,
+      "step": 3643
+    },
+    {
+      "epoch": 2.8350622406639006,
+      "grad_norm": 0.5054654059470541,
+      "learning_rate": 9.100687192581948e-08,
+      "loss": 0.3173,
+      "step": 3644
+    },
+    {
+      "epoch": 2.8358402489626555,
+      "grad_norm": 0.5719444609733476,
+      "learning_rate": 9.014888711089409e-08,
+      "loss": 0.3184,
+      "step": 3645
+    },
+    {
+      "epoch": 2.836618257261411,
+      "grad_norm": 0.522976213344346,
+      "learning_rate": 8.929492909020898e-08,
+      "loss": 0.3416,
+      "step": 3646
+    },
+    {
+      "epoch": 2.837396265560166,
+      "grad_norm": 0.4861871163586463,
+      "learning_rate": 8.844499856413391e-08,
+      "loss": 0.3444,
+      "step": 3647
+    },
+    {
+      "epoch": 2.8381742738589213,
+      "grad_norm": 0.4975007547044733,
+      "learning_rate": 8.759909622973628e-08,
+      "loss": 0.3139,
+      "step": 3648
+    },
+    {
+      "epoch": 2.8389522821576763,
+      "grad_norm": 0.5462634959389019,
+      "learning_rate": 8.675722278077947e-08,
+      "loss": 0.3626,
+      "step": 3649
+    },
+    {
+      "epoch": 2.8397302904564317,
+      "grad_norm": 0.4917339034203323,
+      "learning_rate": 8.591937890772229e-08,
+      "loss": 0.3223,
+      "step": 3650
+    },
+    {
+      "epoch": 2.8405082987551866,
+      "grad_norm": 0.5489488473983776,
+      "learning_rate": 8.508556529771895e-08,
+      "loss": 0.3471,
+      "step": 3651
+    },
+    {
+      "epoch": 2.841286307053942,
+      "grad_norm": 0.5526504198757111,
+      "learning_rate": 8.425578263461798e-08,
+      "loss": 0.3781,
+      "step": 3652
+    },
+    {
+      "epoch": 2.842064315352697,
+      "grad_norm": 0.49401155217780296,
+      "learning_rate": 8.343003159896334e-08,
+      "loss": 0.3143,
+      "step": 3653
+    },
+    {
+      "epoch": 2.8428423236514524,
+      "grad_norm": 0.6391762926110897,
+      "learning_rate": 8.260831286798999e-08,
+      "loss": 0.3539,
+      "step": 3654
+    },
+    {
+      "epoch": 2.8436203319502074,
+      "grad_norm": 0.53335379872217,
+      "learning_rate": 8.179062711562825e-08,
+      "loss": 0.3477,
+      "step": 3655
+    },
+    {
+      "epoch": 2.8443983402489628,
+      "grad_norm": 0.5240172705241548,
+      "learning_rate": 8.097697501249835e-08,
+      "loss": 0.3451,
+      "step": 3656
+    },
+    {
+      "epoch": 2.8451763485477177,
+      "grad_norm": 0.5110324104324577,
+      "learning_rate": 8.01673572259154e-08,
+      "loss": 0.3016,
+      "step": 3657
+    },
+    {
+      "epoch": 2.845954356846473,
+      "grad_norm": 0.4778195884137782,
+      "learning_rate": 7.936177441988324e-08,
+      "loss": 0.3089,
+      "step": 3658
+    },
+    {
+      "epoch": 2.846732365145228,
+      "grad_norm": 0.5557999701326752,
+      "learning_rate": 7.856022725509726e-08,
+      "loss": 0.3492,
+      "step": 3659
+    },
+    {
+      "epoch": 2.8475103734439835,
+      "grad_norm": 0.46063529305060563,
+      "learning_rate": 7.776271638894328e-08,
+      "loss": 0.3564,
+      "step": 3660
+    },
+    {
+      "epoch": 2.8482883817427385,
+      "grad_norm": 0.45875488552873217,
+      "learning_rate": 7.696924247549642e-08,
+      "loss": 0.3493,
+      "step": 3661
+    },
+    {
+      "epoch": 2.849066390041494,
+      "grad_norm": 0.496828927961454,
+      "learning_rate": 7.617980616552167e-08,
+      "loss": 0.3244,
+      "step": 3662
+    },
+    {
+      "epoch": 2.849844398340249,
+      "grad_norm": 0.551151358215076,
+      "learning_rate": 7.539440810647059e-08,
+      "loss": 0.4025,
+      "step": 3663
+    },
+    {
+      "epoch": 2.8506224066390042,
+      "grad_norm": 0.5746696601432817,
+      "learning_rate": 7.461304894248511e-08,
+      "loss": 0.2878,
+      "step": 3664
+    },
+    {
+      "epoch": 2.851400414937759,
+      "grad_norm": 0.5916571765984496,
+      "learning_rate": 7.383572931439431e-08,
+      "loss": 0.3751,
+      "step": 3665
+    },
+    {
+      "epoch": 2.8521784232365146,
+      "grad_norm": 0.5122422764822142,
+      "learning_rate": 7.306244985971267e-08,
+      "loss": 0.3463,
+      "step": 3666
+    },
+    {
+      "epoch": 2.8529564315352696,
+      "grad_norm": 0.5529551609028788,
+      "learning_rate": 7.229321121264232e-08,
+      "loss": 0.3038,
+      "step": 3667
+    },
+    {
+      "epoch": 2.853734439834025,
+      "grad_norm": 0.460094016573869,
+      "learning_rate": 7.152801400407194e-08,
+      "loss": 0.3414,
+      "step": 3668
+    },
+    {
+      "epoch": 2.85451244813278,
+      "grad_norm": 0.4765904025994527,
+      "learning_rate": 7.076685886157398e-08,
+      "loss": 0.2951,
+      "step": 3669
+    },
+    {
+      "epoch": 2.8552904564315353,
+      "grad_norm": 0.48204162161043174,
+      "learning_rate": 7.000974640940683e-08,
+      "loss": 0.4188,
+      "step": 3670
+    },
+    {
+      "epoch": 2.8560684647302903,
+      "grad_norm": 0.5369390806257512,
+      "learning_rate": 6.925667726851381e-08,
+      "loss": 0.3557,
+      "step": 3671
+    },
+    {
+      "epoch": 2.8568464730290457,
+      "grad_norm": 0.5046967197023763,
+      "learning_rate": 6.850765205652088e-08,
+      "loss": 0.3307,
+      "step": 3672
+    },
+    {
+      "epoch": 2.8576244813278007,
+      "grad_norm": 0.5177567492224863,
+      "learning_rate": 6.776267138773829e-08,
+      "loss": 0.332,
+      "step": 3673
+    },
+    {
+      "epoch": 2.858402489626556,
+      "grad_norm": 0.522733854056261,
+      "learning_rate": 6.702173587315842e-08,
+      "loss": 0.3512,
+      "step": 3674
+    },
+    {
+      "epoch": 2.859180497925311,
+      "grad_norm": 0.4652173646383546,
+      "learning_rate": 6.628484612045737e-08,
+      "loss": 0.3396,
+      "step": 3675
+    },
+    {
+      "epoch": 2.8599585062240664,
+      "grad_norm": 0.6117456268694628,
+      "learning_rate": 6.555200273399177e-08,
+      "loss": 0.3473,
+      "step": 3676
+    },
+    {
+      "epoch": 2.860736514522822,
+      "grad_norm": 0.5621151821932401,
+      "learning_rate": 6.48232063147991e-08,
+      "loss": 0.3127,
+      "step": 3677
+    },
+    {
+      "epoch": 2.861514522821577,
+      "grad_norm": 0.5748928232333419,
+      "learning_rate": 6.409845746060017e-08,
+      "loss": 0.3565,
+      "step": 3678
+    },
+    {
+      "epoch": 2.8622925311203318,
+      "grad_norm": 0.567338357477908,
+      "learning_rate": 6.337775676579505e-08,
+      "loss": 0.3087,
+      "step": 3679
+    },
+    {
+      "epoch": 2.863070539419087,
+      "grad_norm": 0.5709456213385671,
+      "learning_rate": 6.266110482146149e-08,
+      "loss": 0.3559,
+      "step": 3680
+    },
+    {
+      "epoch": 2.8638485477178426,
+      "grad_norm": 0.44722452223584047,
+      "learning_rate": 6.194850221536097e-08,
+      "loss": 0.3639,
+      "step": 3681
+    },
+    {
+      "epoch": 2.8646265560165975,
+      "grad_norm": 0.5099099335427437,
+      "learning_rate": 6.123994953193047e-08,
+      "loss": 0.3256,
+      "step": 3682
+    },
+    {
+      "epoch": 2.8654045643153525,
+      "grad_norm": 0.4386314710040238,
+      "learning_rate": 6.053544735228623e-08,
+      "loss": 0.3134,
+      "step": 3683
+    },
+    {
+      "epoch": 2.866182572614108,
+      "grad_norm": 0.5708712161920274,
+      "learning_rate": 5.983499625422273e-08,
+      "loss": 0.3678,
+      "step": 3684
+    },
+    {
+      "epoch": 2.8669605809128633,
+      "grad_norm": 0.537948558803821,
+      "learning_rate": 5.913859681221268e-08,
+      "loss": 0.3234,
+      "step": 3685
+    },
+    {
+      "epoch": 2.8677385892116183,
+      "grad_norm": 0.534239890616134,
+      "learning_rate": 5.8446249597405855e-08,
+      "loss": 0.3244,
+      "step": 3686
+    },
+    {
+      "epoch": 2.8685165975103732,
+      "grad_norm": 0.5379724128578584,
+      "learning_rate": 5.775795517762639e-08,
+      "loss": 0.3555,
+      "step": 3687
+    },
+    {
+      "epoch": 2.8692946058091287,
+      "grad_norm": 0.5644051986301399,
+      "learning_rate": 5.707371411737717e-08,
+      "loss": 0.375,
+      "step": 3688
+    },
+    {
+      "epoch": 2.870072614107884,
+      "grad_norm": 0.4933297468924029,
+      "learning_rate": 5.6393526977835954e-08,
+      "loss": 0.338,
+      "step": 3689
+    },
+    {
+      "epoch": 2.870850622406639,
+      "grad_norm": 0.5091187542715657,
+      "learning_rate": 5.57173943168543e-08,
+      "loss": 0.3399,
+      "step": 3690
+    },
+    {
+      "epoch": 2.871628630705394,
+      "grad_norm": 0.5302497316062066,
+      "learning_rate": 5.5045316688960296e-08,
+      "loss": 0.3544,
+      "step": 3691
+    },
+    {
+      "epoch": 2.8724066390041494,
+      "grad_norm": 0.5143654305242016,
+      "learning_rate": 5.437729464535579e-08,
+      "loss": 0.3587,
+      "step": 3692
+    },
+    {
+      "epoch": 2.873184647302905,
+      "grad_norm": 0.5387414642977265,
+      "learning_rate": 5.371332873391644e-08,
+      "loss": 0.382,
+      "step": 3693
+    },
+    {
+      "epoch": 2.8739626556016598,
+      "grad_norm": 0.48824248461646647,
+      "learning_rate": 5.305341949919052e-08,
+      "loss": 0.3146,
+      "step": 3694
+    },
+    {
+      "epoch": 2.8747406639004147,
+      "grad_norm": 0.52314230309471,
+      "learning_rate": 5.23975674824001e-08,
+      "loss": 0.3294,
+      "step": 3695
+    },
+    {
+      "epoch": 2.87551867219917,
+      "grad_norm": 0.49069905220992,
+      "learning_rate": 5.174577322143992e-08,
+      "loss": 0.3687,
+      "step": 3696
+    },
+    {
+      "epoch": 2.8762966804979255,
+      "grad_norm": 0.5163372107343186,
+      "learning_rate": 5.109803725087625e-08,
+      "loss": 0.328,
+      "step": 3697
+    },
+    {
+      "epoch": 2.8770746887966805,
+      "grad_norm": 0.474811774398204,
+      "learning_rate": 5.0454360101946376e-08,
+      "loss": 0.3486,
+      "step": 3698
+    },
+    {
+      "epoch": 2.8778526970954355,
+      "grad_norm": 0.5480665396391425,
+      "learning_rate": 4.9814742302560225e-08,
+      "loss": 0.3736,
+      "step": 3699
+    },
+    {
+      "epoch": 2.878630705394191,
+      "grad_norm": 0.47209956643347756,
+      "learning_rate": 4.9179184377298186e-08,
+      "loss": 0.3343,
+      "step": 3700
+    },
+    {
+      "epoch": 2.8794087136929463,
+      "grad_norm": 0.4719388835844545,
+      "learning_rate": 4.854768684740885e-08,
+      "loss": 0.3086,
+      "step": 3701
+    },
+    {
+      "epoch": 2.8801867219917012,
+      "grad_norm": 0.4710154276187187,
+      "learning_rate": 4.792025023081404e-08,
+      "loss": 0.3103,
+      "step": 3702
+    },
+    {
+      "epoch": 2.880964730290456,
+      "grad_norm": 0.5531214257001363,
+      "learning_rate": 4.7296875042102117e-08,
+      "loss": 0.3992,
+      "step": 3703
+    },
+    {
+      "epoch": 2.8817427385892116,
+      "grad_norm": 0.5724950345442729,
+      "learning_rate": 4.6677561792532464e-08,
+      "loss": 0.3462,
+      "step": 3704
+    },
+    {
+      "epoch": 2.882520746887967,
+      "grad_norm": 0.6151569667120081,
+      "learning_rate": 4.606231099003211e-08,
+      "loss": 0.3591,
+      "step": 3705
+    },
+    {
+      "epoch": 2.883298755186722,
+      "grad_norm": 0.5297001258387803,
+      "learning_rate": 4.5451123139195754e-08,
+      "loss": 0.319,
+      "step": 3706
+    },
+    {
+      "epoch": 2.884076763485477,
+      "grad_norm": 0.47879434074393423,
+      "learning_rate": 4.484399874128742e-08,
+      "loss": 0.3282,
+      "step": 3707
+    },
+    {
+      "epoch": 2.8848547717842323,
+      "grad_norm": 0.49719072864223574,
+      "learning_rate": 4.42409382942377e-08,
+      "loss": 0.3292,
+      "step": 3708
+    },
+    {
+      "epoch": 2.8856327800829877,
+      "grad_norm": 0.5714299309542418,
+      "learning_rate": 4.364194229264374e-08,
+      "loss": 0.4044,
+      "step": 3709
+    },
+    {
+      "epoch": 2.8864107883817427,
+      "grad_norm": 0.4355661072889359,
+      "learning_rate": 4.3047011227769775e-08,
+      "loss": 0.2917,
+      "step": 3710
+    },
+    {
+      "epoch": 2.8871887966804977,
+      "grad_norm": 0.5326785122057734,
+      "learning_rate": 4.2456145587546074e-08,
+      "loss": 0.3287,
+      "step": 3711
+    },
+    {
+      "epoch": 2.887966804979253,
+      "grad_norm": 0.5087981350733108,
+      "learning_rate": 4.1869345856568874e-08,
+      "loss": 0.3201,
+      "step": 3712
+    },
+    {
+      "epoch": 2.8887448132780085,
+      "grad_norm": 0.43208961709144134,
+      "learning_rate": 4.128661251609933e-08,
+      "loss": 0.3408,
+      "step": 3713
+    },
+    {
+      "epoch": 2.8895228215767634,
+      "grad_norm": 0.4262815122061262,
+      "learning_rate": 4.070794604406403e-08,
+      "loss": 0.328,
+      "step": 3714
+    },
+    {
+      "epoch": 2.8903008298755184,
+      "grad_norm": 0.5247861663382949,
+      "learning_rate": 4.013334691505333e-08,
+      "loss": 0.3582,
+      "step": 3715
+    },
+    {
+      "epoch": 2.891078838174274,
+      "grad_norm": 0.5340158798974021,
+      "learning_rate": 3.9562815600323046e-08,
+      "loss": 0.3462,
+      "step": 3716
+    },
+    {
+      "epoch": 2.891856846473029,
+      "grad_norm": 0.5443596954097727,
+      "learning_rate": 3.899635256779222e-08,
+      "loss": 0.3174,
+      "step": 3717
+    },
+    {
+      "epoch": 2.892634854771784,
+      "grad_norm": 0.5669367651572692,
+      "learning_rate": 3.843395828204255e-08,
+      "loss": 0.3555,
+      "step": 3718
+    },
+    {
+      "epoch": 2.8934128630705396,
+      "grad_norm": 0.5456593195096175,
+      "learning_rate": 3.787563320431897e-08,
+      "loss": 0.3775,
+      "step": 3719
+    },
+    {
+      "epoch": 2.8941908713692945,
+      "grad_norm": 0.5427397398354736,
+      "learning_rate": 3.7321377792530756e-08,
+      "loss": 0.35,
+      "step": 3720
+    },
+    {
+      "epoch": 2.89496887966805,
+      "grad_norm": 0.5536832830814692,
+      "learning_rate": 3.677119250124761e-08,
+      "loss": 0.4009,
+      "step": 3721
+    },
+    {
+      "epoch": 2.895746887966805,
+      "grad_norm": 0.5111549106063823,
+      "learning_rate": 3.6225077781701365e-08,
+      "loss": 0.3363,
+      "step": 3722
+    },
+    {
+      "epoch": 2.8965248962655603,
+      "grad_norm": 0.4956230925029381,
+      "learning_rate": 3.568303408178708e-08,
+      "loss": 0.3413,
+      "step": 3723
+    },
+    {
+      "epoch": 2.8973029045643153,
+      "grad_norm": 0.4612603924333075,
+      "learning_rate": 3.514506184605748e-08,
+      "loss": 0.3231,
+      "step": 3724
+    },
+    {
+      "epoch": 2.8980809128630707,
+      "grad_norm": 0.5144499059004731,
+      "learning_rate": 3.4611161515730184e-08,
+      "loss": 0.3389,
+      "step": 3725
+    },
+    {
+      "epoch": 2.8988589211618256,
+      "grad_norm": 0.549700887937716,
+      "learning_rate": 3.4081333528681037e-08,
+      "loss": 0.3495,
+      "step": 3726
+    },
+    {
+      "epoch": 2.899636929460581,
+      "grad_norm": 0.5302035445161656,
+      "learning_rate": 3.355557831944578e-08,
+      "loss": 0.3392,
+      "step": 3727
+    },
+    {
+      "epoch": 2.900414937759336,
+      "grad_norm": 0.5575072727403761,
+      "learning_rate": 3.30338963192206e-08,
+      "loss": 0.322,
+      "step": 3728
+    },
+    {
+      "epoch": 2.9011929460580914,
+      "grad_norm": 0.5132638133395562,
+      "learning_rate": 3.251628795586048e-08,
+      "loss": 0.375,
+      "step": 3729
+    },
+    {
+      "epoch": 2.9019709543568464,
+      "grad_norm": 0.5102460419554073,
+      "learning_rate": 3.200275365388028e-08,
+      "loss": 0.4123,
+      "step": 3730
+    },
+    {
+      "epoch": 2.902748962655602,
+      "grad_norm": 0.4717080964877508,
+      "learning_rate": 3.14932938344531e-08,
+      "loss": 0.3286,
+      "step": 3731
+    },
+    {
+      "epoch": 2.9035269709543567,
+      "grad_norm": 0.4912838272139408,
+      "learning_rate": 3.0987908915409705e-08,
+      "loss": 0.3006,
+      "step": 3732
+    },
+    {
+      "epoch": 2.904304979253112,
+      "grad_norm": 0.4406557433689068,
+      "learning_rate": 3.048659931124076e-08,
+      "loss": 0.3151,
+      "step": 3733
+    },
+    {
+      "epoch": 2.905082987551867,
+      "grad_norm": 0.5467470722822243,
+      "learning_rate": 2.9989365433091836e-08,
+      "loss": 0.3376,
+      "step": 3734
+    },
+    {
+      "epoch": 2.9058609958506225,
+      "grad_norm": 0.5045357331342721,
+      "learning_rate": 2.949620768876782e-08,
+      "loss": 0.3433,
+      "step": 3735
+    },
+    {
+      "epoch": 2.9066390041493775,
+      "grad_norm": 0.49660617084236525,
+      "learning_rate": 2.9007126482730185e-08,
+      "loss": 0.3731,
+      "step": 3736
+    },
+    {
+      "epoch": 2.907417012448133,
+      "grad_norm": 0.4813162054947047,
+      "learning_rate": 2.8522122216096403e-08,
+      "loss": 0.2855,
+      "step": 3737
+    },
+    {
+      "epoch": 2.908195020746888,
+      "grad_norm": 0.49387927900052125,
+      "learning_rate": 2.804119528664162e-08,
+      "loss": 0.3453,
+      "step": 3738
+    },
+    {
+      "epoch": 2.9089730290456433,
+      "grad_norm": 0.4806273784534852,
+      "learning_rate": 2.756434608879477e-08,
+      "loss": 0.3394,
+      "step": 3739
+    },
+    {
+      "epoch": 2.909751037344398,
+      "grad_norm": 0.562954022371209,
+      "learning_rate": 2.7091575013643013e-08,
+      "loss": 0.358,
+      "step": 3740
+    },
+    {
+      "epoch": 2.9105290456431536,
+      "grad_norm": 0.5329471148416466,
+      "learning_rate": 2.6622882448926746e-08,
+      "loss": 0.3429,
+      "step": 3741
+    },
+    {
+      "epoch": 2.9113070539419086,
+      "grad_norm": 0.5258153113086734,
+      "learning_rate": 2.6158268779042927e-08,
+      "loss": 0.341,
+      "step": 3742
+    },
+    {
+      "epoch": 2.912085062240664,
+      "grad_norm": 0.5656268246466076,
+      "learning_rate": 2.5697734385042306e-08,
+      "loss": 0.3614,
+      "step": 3743
+    },
+    {
+      "epoch": 2.912863070539419,
+      "grad_norm": 0.5501532406566578,
+      "learning_rate": 2.5241279644629968e-08,
+      "loss": 0.3956,
+      "step": 3744
+    },
+    {
+      "epoch": 2.9136410788381744,
+      "grad_norm": 0.4862532482046553,
+      "learning_rate": 2.4788904932165903e-08,
+      "loss": 0.3146,
+      "step": 3745
+    },
+    {
+      "epoch": 2.9144190871369293,
+      "grad_norm": 0.5448334980591117,
+      "learning_rate": 2.4340610618662774e-08,
+      "loss": 0.3687,
+      "step": 3746
+    },
+    {
+      "epoch": 2.9151970954356847,
+      "grad_norm": 0.4767442660860527,
+      "learning_rate": 2.3896397071787593e-08,
+      "loss": 0.3676,
+      "step": 3747
+    },
+    {
+      "epoch": 2.9159751037344397,
+      "grad_norm": 0.49887408065712857,
+      "learning_rate": 2.3456264655860595e-08,
+      "loss": 0.3162,
+      "step": 3748
+    },
+    {
+      "epoch": 2.916753112033195,
+      "grad_norm": 0.5453648214248696,
+      "learning_rate": 2.3020213731854147e-08,
+      "loss": 0.3513,
+      "step": 3749
+    },
+    {
+      "epoch": 2.91753112033195,
+      "grad_norm": 0.5050915068409529,
+      "learning_rate": 2.2588244657393843e-08,
+      "loss": 0.3407,
+      "step": 3750
+    },
+    {
+      "epoch": 2.9183091286307055,
+      "grad_norm": 0.49078935764697773,
+      "learning_rate": 2.2160357786757404e-08,
+      "loss": 0.353,
+      "step": 3751
+    },
+    {
+      "epoch": 2.9190871369294604,
+      "grad_norm": 0.4851958810204349,
+      "learning_rate": 2.173655347087411e-08,
+      "loss": 0.3016,
+      "step": 3752
+    },
+    {
+      "epoch": 2.919865145228216,
+      "grad_norm": 0.5351370618501007,
+      "learning_rate": 2.131683205732593e-08,
+      "loss": 0.3396,
+      "step": 3753
+    },
+    {
+      "epoch": 2.920643153526971,
+      "grad_norm": 0.5270378619694414,
+      "learning_rate": 2.0901193890345284e-08,
+      "loss": 0.3396,
+      "step": 3754
+    },
+    {
+      "epoch": 2.921421161825726,
+      "grad_norm": 0.5185363647547258,
+      "learning_rate": 2.04896393108156e-08,
+      "loss": 0.348,
+      "step": 3755
+    },
+    {
+      "epoch": 2.9221991701244816,
+      "grad_norm": 0.48344642595651005,
+      "learning_rate": 2.0082168656272993e-08,
+      "loss": 0.3099,
+      "step": 3756
+    },
+    {
+      "epoch": 2.9229771784232366,
+      "grad_norm": 0.5660125097775914,
+      "learning_rate": 1.9678782260901808e-08,
+      "loss": 0.3742,
+      "step": 3757
+    },
+    {
+      "epoch": 2.9237551867219915,
+      "grad_norm": 0.47744383129815837,
+      "learning_rate": 1.9279480455537956e-08,
+      "loss": 0.3112,
+      "step": 3758
+    },
+    {
+      "epoch": 2.924533195020747,
+      "grad_norm": 0.5007426907477518,
+      "learning_rate": 1.888426356766726e-08,
+      "loss": 0.3805,
+      "step": 3759
+    },
+    {
+      "epoch": 2.9253112033195023,
+      "grad_norm": 0.475234611686458,
+      "learning_rate": 1.8493131921424877e-08,
+      "loss": 0.3273,
+      "step": 3760
+    },
+    {
+      "epoch": 2.9260892116182573,
+      "grad_norm": 0.5686131875888422,
+      "learning_rate": 1.8106085837596432e-08,
+      "loss": 0.3911,
+      "step": 3761
+    },
+    {
+      "epoch": 2.9268672199170123,
+      "grad_norm": 0.45717084574405914,
+      "learning_rate": 1.772312563361578e-08,
+      "loss": 0.2819,
+      "step": 3762
+    },
+    {
+      "epoch": 2.9276452282157677,
+      "grad_norm": 0.5660962757592848,
+      "learning_rate": 1.7344251623566677e-08,
+      "loss": 0.3892,
+      "step": 3763
+    },
+    {
+      "epoch": 2.928423236514523,
+      "grad_norm": 0.46923323491667407,
+      "learning_rate": 1.6969464118180568e-08,
+      "loss": 0.3275,
+      "step": 3764
+    },
+    {
+      "epoch": 2.929201244813278,
+      "grad_norm": 0.5752868683891098,
+      "learning_rate": 1.6598763424837682e-08,
+      "loss": 0.3477,
+      "step": 3765
+    },
+    {
+      "epoch": 2.929979253112033,
+      "grad_norm": 0.5421236395690737,
+      "learning_rate": 1.623214984756705e-08,
+      "loss": 0.3448,
+      "step": 3766
+    },
+    {
+      "epoch": 2.9307572614107884,
+      "grad_norm": 0.5241969451071878,
+      "learning_rate": 1.5869623687045366e-08,
+      "loss": 0.3261,
+      "step": 3767
+    },
+    {
+      "epoch": 2.931535269709544,
+      "grad_norm": 0.5254931007387905,
+      "learning_rate": 1.5511185240597027e-08,
+      "loss": 0.3601,
+      "step": 3768
+    },
+    {
+      "epoch": 2.9323132780082988,
+      "grad_norm": 0.45232645054594195,
+      "learning_rate": 1.5156834802193542e-08,
+      "loss": 0.3019,
+      "step": 3769
+    },
+    {
+      "epoch": 2.9330912863070537,
+      "grad_norm": 0.5699909429374762,
+      "learning_rate": 1.480657266245411e-08,
+      "loss": 0.4086,
+      "step": 3770
+    },
+    {
+      "epoch": 2.933869294605809,
+      "grad_norm": 0.48002358597252404,
+      "learning_rate": 1.4460399108644496e-08,
+      "loss": 0.3329,
+      "step": 3771
+    },
+    {
+      "epoch": 2.9346473029045645,
+      "grad_norm": 0.45563959259256576,
+      "learning_rate": 1.4118314424678702e-08,
+      "loss": 0.3547,
+      "step": 3772
+    },
+    {
+      "epoch": 2.9354253112033195,
+      "grad_norm": 0.45715047005145565,
+      "learning_rate": 1.378031889111453e-08,
+      "loss": 0.343,
+      "step": 3773
+    },
+    {
+      "epoch": 2.9362033195020745,
+      "grad_norm": 0.571863826105338,
+      "learning_rate": 1.3446412785158013e-08,
+      "loss": 0.3221,
+      "step": 3774
+    },
+    {
+      "epoch": 2.93698132780083,
+      "grad_norm": 0.5494398248951325,
+      "learning_rate": 1.3116596380661207e-08,
+      "loss": 0.3185,
+      "step": 3775
+    },
+    {
+      "epoch": 2.9377593360995853,
+      "grad_norm": 0.505474674482504,
+      "learning_rate": 1.2790869948121621e-08,
+      "loss": 0.3467,
+      "step": 3776
+    },
+    {
+      "epoch": 2.9385373443983402,
+      "grad_norm": 0.5557835248004093,
+      "learning_rate": 1.2469233754683341e-08,
+      "loss": 0.345,
+      "step": 3777
+    },
+    {
+      "epoch": 2.939315352697095,
+      "grad_norm": 0.4823757347055994,
+      "learning_rate": 1.2151688064133138e-08,
+      "loss": 0.3416,
+      "step": 3778
+    },
+    {
+      "epoch": 2.9400933609958506,
+      "grad_norm": 0.4802695085823668,
+      "learning_rate": 1.183823313690602e-08,
+      "loss": 0.3216,
+      "step": 3779
+    },
+    {
+      "epoch": 2.940871369294606,
+      "grad_norm": 0.5448722109243355,
+      "learning_rate": 1.1528869230080231e-08,
+      "loss": 0.3228,
+      "step": 3780
+    },
+    {
+      "epoch": 2.941649377593361,
+      "grad_norm": 0.5310617495312986,
+      "learning_rate": 1.1223596597380038e-08,
+      "loss": 0.322,
+      "step": 3781
+    },
+    {
+      "epoch": 2.942427385892116,
+      "grad_norm": 0.5365938752852897,
+      "learning_rate": 1.0922415489172944e-08,
+      "loss": 0.3105,
+      "step": 3782
+    },
+    {
+      "epoch": 2.9432053941908713,
+      "grad_norm": 0.46118141986883177,
+      "learning_rate": 1.062532615247136e-08,
+      "loss": 0.3599,
+      "step": 3783
+    },
+    {
+      "epoch": 2.9439834024896268,
+      "grad_norm": 0.39948275911944414,
+      "learning_rate": 1.0332328830932048e-08,
+      "loss": 0.2885,
+      "step": 3784
+    },
+    {
+      "epoch": 2.9447614107883817,
+      "grad_norm": 0.4403448430397249,
+      "learning_rate": 1.0043423764855565e-08,
+      "loss": 0.3508,
+      "step": 3785
+    },
+    {
+      "epoch": 2.9455394190871367,
+      "grad_norm": 0.4788204350212106,
+      "learning_rate": 9.758611191186263e-09,
+      "loss": 0.3501,
+      "step": 3786
+    },
+    {
+      "epoch": 2.946317427385892,
+      "grad_norm": 0.4752325260265481,
+      "learning_rate": 9.477891343511736e-09,
+      "loss": 0.3388,
+      "step": 3787
+    },
+    {
+      "epoch": 2.9470954356846475,
+      "grad_norm": 0.42682223852952905,
+      "learning_rate": 9.201264452063374e-09,
+      "loss": 0.318,
+      "step": 3788
+    },
+    {
+      "epoch": 2.9478734439834025,
+      "grad_norm": 0.5196430770765703,
+      "learning_rate": 8.928730743715807e-09,
+      "loss": 0.2982,
+      "step": 3789
+    },
+    {
+      "epoch": 2.9486514522821574,
+      "grad_norm": 0.5955727759554366,
+      "learning_rate": 8.660290441986352e-09,
+      "loss": 0.373,
+      "step": 3790
+    },
+    {
+      "epoch": 2.949429460580913,
+      "grad_norm": 0.5455725405351639,
+      "learning_rate": 8.395943767035564e-09,
+      "loss": 0.3087,
+      "step": 3791
+    },
+    {
+      "epoch": 2.9502074688796682,
+      "grad_norm": 0.5761952810713062,
+      "learning_rate": 8.135690935665019e-09,
+      "loss": 0.362,
+      "step": 3792
+    },
+    {
+      "epoch": 2.950985477178423,
+      "grad_norm": 0.5106155461297214,
+      "learning_rate": 7.879532161321757e-09,
+      "loss": 0.3901,
+      "step": 3793
+    },
+    {
+      "epoch": 2.951763485477178,
+      "grad_norm": 0.4940348034066626,
+      "learning_rate": 7.627467654091613e-09,
+      "loss": 0.3484,
+      "step": 3794
+    },
+    {
+      "epoch": 2.9525414937759336,
+      "grad_norm": 0.5042995272641554,
+      "learning_rate": 7.379497620705333e-09,
+      "loss": 0.3419,
+      "step": 3795
+    },
+    {
+      "epoch": 2.953319502074689,
+      "grad_norm": 0.5039095641338385,
+      "learning_rate": 7.1356222645346806e-09,
+      "loss": 0.3261,
+      "step": 3796
+    },
+    {
+      "epoch": 2.954097510373444,
+      "grad_norm": 0.5643728283467632,
+      "learning_rate": 6.895841785591884e-09,
+      "loss": 0.3345,
+      "step": 3797
+    },
+    {
+      "epoch": 2.9548755186721993,
+      "grad_norm": 0.47372770400172853,
+      "learning_rate": 6.660156380532967e-09,
+      "loss": 0.3243,
+      "step": 3798
+    },
+    {
+      "epoch": 2.9556535269709543,
+      "grad_norm": 0.47743183129646155,
+      "learning_rate": 6.428566242654422e-09,
+      "loss": 0.3626,
+      "step": 3799
+    },
+    {
+      "epoch": 2.9564315352697097,
+      "grad_norm": 0.4441065154705959,
+      "learning_rate": 6.201071561893201e-09,
+      "loss": 0.3194,
+      "step": 3800
+    },
+    {
+      "epoch": 2.9572095435684647,
+      "grad_norm": 0.5059995153038832,
+      "learning_rate": 5.977672524828393e-09,
+      "loss": 0.3818,
+      "step": 3801
+    },
+    {
+      "epoch": 2.95798755186722,
+      "grad_norm": 0.4527343368909635,
+      "learning_rate": 5.758369314680101e-09,
+      "loss": 0.2863,
+      "step": 3802
+    },
+    {
+      "epoch": 2.958765560165975,
+      "grad_norm": 0.5610629510160414,
+      "learning_rate": 5.543162111308342e-09,
+      "loss": 0.4065,
+      "step": 3803
+    },
+    {
+      "epoch": 2.9595435684647304,
+      "grad_norm": 0.5204996327069903,
+      "learning_rate": 5.3320510912152624e-09,
+      "loss": 0.3431,
+      "step": 3804
+    },
+    {
+      "epoch": 2.9603215767634854,
+      "grad_norm": 0.49719373524501254,
+      "learning_rate": 5.125036427542363e-09,
+      "loss": 0.326,
+      "step": 3805
+    },
+    {
+      "epoch": 2.961099585062241,
+      "grad_norm": 0.672785582325523,
+      "learning_rate": 4.922118290071609e-09,
+      "loss": 0.3358,
+      "step": 3806
+    },
+    {
+      "epoch": 2.9618775933609958,
+      "grad_norm": 0.5014468797417876,
+      "learning_rate": 4.723296845225434e-09,
+      "loss": 0.3529,
+      "step": 3807
+    },
+    {
+      "epoch": 2.962655601659751,
+      "grad_norm": 0.5409112476262806,
+      "learning_rate": 4.5285722560667326e-09,
+      "loss": 0.3662,
+      "step": 3808
+    },
+    {
+      "epoch": 2.963433609958506,
+      "grad_norm": 0.5051143802722214,
+      "learning_rate": 4.337944682298312e-09,
+      "loss": 0.3259,
+      "step": 3809
+    },
+    {
+      "epoch": 2.9642116182572615,
+      "grad_norm": 0.48826847243510835,
+      "learning_rate": 4.151414280262334e-09,
+      "loss": 0.3071,
+      "step": 3810
+    },
+    {
+      "epoch": 2.9649896265560165,
+      "grad_norm": 0.5591920521745861,
+      "learning_rate": 3.968981202940314e-09,
+      "loss": 0.3994,
+      "step": 3811
+    },
+    {
+      "epoch": 2.965767634854772,
+      "grad_norm": 0.5652109464897953,
+      "learning_rate": 3.79064559995479e-09,
+      "loss": 0.3566,
+      "step": 3812
+    },
+    {
+      "epoch": 2.966545643153527,
+      "grad_norm": 0.4711574843116491,
+      "learning_rate": 3.616407617567097e-09,
+      "loss": 0.2969,
+      "step": 3813
+    },
+    {
+      "epoch": 2.9673236514522823,
+      "grad_norm": 0.5220862900550356,
+      "learning_rate": 3.446267398676817e-09,
+      "loss": 0.3243,
+      "step": 3814
+    },
+    {
+      "epoch": 2.9681016597510372,
+      "grad_norm": 0.5585447913887929,
+      "learning_rate": 3.280225082824551e-09,
+      "loss": 0.3453,
+      "step": 3815
+    },
+    {
+      "epoch": 2.9688796680497926,
+      "grad_norm": 0.45646166340329813,
+      "learning_rate": 3.1182808061891444e-09,
+      "loss": 0.3169,
+      "step": 3816
+    },
+    {
+      "epoch": 2.9696576763485476,
+      "grad_norm": 0.5401797481651897,
+      "learning_rate": 2.960434701588799e-09,
+      "loss": 0.4202,
+      "step": 3817
+    },
+    {
+      "epoch": 2.970435684647303,
+      "grad_norm": 0.45318908723896545,
+      "learning_rate": 2.806686898479405e-09,
+      "loss": 0.3179,
+      "step": 3818
+    },
+    {
+      "epoch": 2.971213692946058,
+      "grad_norm": 0.44586723445275683,
+      "learning_rate": 2.657037522957873e-09,
+      "loss": 0.3075,
+      "step": 3819
+    },
+    {
+      "epoch": 2.9719917012448134,
+      "grad_norm": 0.49190988677034014,
+      "learning_rate": 2.511486697757137e-09,
+      "loss": 0.3395,
+      "step": 3820
+    },
+    {
+      "epoch": 2.9727697095435683,
+      "grad_norm": 0.5138291386970586,
+      "learning_rate": 2.3700345422511537e-09,
+      "loss": 0.3407,
+      "step": 3821
+    },
+    {
+      "epoch": 2.9735477178423237,
+      "grad_norm": 0.5163232913627499,
+      "learning_rate": 2.232681172451012e-09,
+      "loss": 0.3399,
+      "step": 3822
+    },
+    {
+      "epoch": 2.9743257261410787,
+      "grad_norm": 0.5128049743783412,
+      "learning_rate": 2.0994267010060467e-09,
+      "loss": 0.3437,
+      "step": 3823
+    },
+    {
+      "epoch": 2.975103734439834,
+      "grad_norm": 0.4829226055122599,
+      "learning_rate": 1.9702712372049458e-09,
+      "loss": 0.3207,
+      "step": 3824
+    },
+    {
+      "epoch": 2.975881742738589,
+      "grad_norm": 0.5214496292736825,
+      "learning_rate": 1.8452148869740893e-09,
+      "loss": 0.31,
+      "step": 3825
+    },
+    {
+      "epoch": 2.9766597510373445,
+      "grad_norm": 0.5557905470988078,
+      "learning_rate": 1.7242577528781002e-09,
+      "loss": 0.3576,
+      "step": 3826
+    },
+    {
+      "epoch": 2.9774377593360994,
+      "grad_norm": 0.5429870324759501,
+      "learning_rate": 1.607399934118181e-09,
+      "loss": 0.3283,
+      "step": 3827
+    },
+    {
+      "epoch": 2.978215767634855,
+      "grad_norm": 0.5194423620296792,
+      "learning_rate": 1.4946415265365554e-09,
+      "loss": 0.3524,
+      "step": 3828
+    },
+    {
+      "epoch": 2.97899377593361,
+      "grad_norm": 0.4496298633297231,
+      "learning_rate": 1.38598262261036e-09,
+      "loss": 0.2958,
+      "step": 3829
+    },
+    {
+      "epoch": 2.979771784232365,
+      "grad_norm": 0.5475991173357746,
+      "learning_rate": 1.281423311456087e-09,
+      "loss": 0.3469,
+      "step": 3830
+    },
+    {
+      "epoch": 2.98054979253112,
+      "grad_norm": 0.48612827390463315,
+      "learning_rate": 1.1809636788273626e-09,
+      "loss": 0.3517,
+      "step": 3831
+    },
+    {
+      "epoch": 2.9813278008298756,
+      "grad_norm": 0.4648214102396428,
+      "learning_rate": 1.0846038071166132e-09,
+      "loss": 0.2867,
+      "step": 3832
+    },
+    {
+      "epoch": 2.9821058091286305,
+      "grad_norm": 0.5289201231171929,
+      "learning_rate": 9.92343775351734e-10,
+      "loss": 0.3852,
+      "step": 3833
+    },
+    {
+      "epoch": 2.982883817427386,
+      "grad_norm": 0.4035240121391826,
+      "learning_rate": 9.041836592005304e-10,
+      "loss": 0.2755,
+      "step": 3834
+    },
+    {
+      "epoch": 2.983661825726141,
+      "grad_norm": 0.45915363519292185,
+      "learning_rate": 8.20123530966832e-10,
+      "loss": 0.3639,
+      "step": 3835
+    },
+    {
+      "epoch": 2.9844398340248963,
+      "grad_norm": 0.5496404012583406,
+      "learning_rate": 7.40163459591603e-10,
+      "loss": 0.3499,
+      "step": 3836
+    },
+    {
+      "epoch": 2.9852178423236513,
+      "grad_norm": 0.5249830551315079,
+      "learning_rate": 6.64303510654607e-10,
+      "loss": 0.3452,
+      "step": 3837
+    },
+    {
+      "epoch": 2.9859958506224067,
+      "grad_norm": 0.4833273004060676,
+      "learning_rate": 5.925437463716322e-10,
+      "loss": 0.3087,
+      "step": 3838
+    },
+    {
+      "epoch": 2.986773858921162,
+      "grad_norm": 0.5672538085567336,
+      "learning_rate": 5.248842255967113e-10,
+      "loss": 0.3705,
+      "step": 3839
+    },
+    {
+      "epoch": 2.987551867219917,
+      "grad_norm": 0.5472942600062126,
+      "learning_rate": 4.613250038193462e-10,
+      "loss": 0.3126,
+      "step": 3840
+    },
+    {
+      "epoch": 2.988329875518672,
+      "grad_norm": 0.5358744743029673,
+      "learning_rate": 4.0186613316839375e-10,
+      "loss": 0.3977,
+      "step": 3841
+    },
+    {
+      "epoch": 2.9891078838174274,
+      "grad_norm": 0.45393063569085,
+      "learning_rate": 3.465076624087349e-10,
+      "loss": 0.3183,
+      "step": 3842
+    },
+    {
+      "epoch": 2.989885892116183,
+      "grad_norm": 0.5001517523168303,
+      "learning_rate": 2.952496369418301e-10,
+      "loss": 0.351,
+      "step": 3843
+    },
+    {
+      "epoch": 2.990663900414938,
+      "grad_norm": 0.5505795562311313,
+      "learning_rate": 2.4809209880682915e-10,
+      "loss": 0.4127,
+      "step": 3844
+    },
+    {
+      "epoch": 2.9914419087136928,
+      "grad_norm": 0.5529171836031039,
+      "learning_rate": 2.050350866805717e-10,
+      "loss": 0.3036,
+      "step": 3845
+    },
+    {
+      "epoch": 2.992219917012448,
+      "grad_norm": 0.5583868181614494,
+      "learning_rate": 1.660786358748112e-10,
+      "loss": 0.3637,
+      "step": 3846
+    },
+    {
+      "epoch": 2.9929979253112036,
+      "grad_norm": 0.48804266902554416,
+      "learning_rate": 1.312227783406561e-10,
+      "loss": 0.3632,
+      "step": 3847
+    },
+    {
+      "epoch": 2.9937759336099585,
+      "grad_norm": 0.5270668388749897,
+      "learning_rate": 1.0046754266468395e-10,
+      "loss": 0.336,
+      "step": 3848
+    },
+    {
+      "epoch": 2.9945539419087135,
+      "grad_norm": 0.5480274857069526,
+      "learning_rate": 7.381295407060673e-11,
+      "loss": 0.363,
+      "step": 3849
+    },
+    {
+      "epoch": 2.995331950207469,
+      "grad_norm": 0.5115775432046697,
+      "learning_rate": 5.1259034418715825e-11,
+      "loss": 0.2884,
+      "step": 3850
+    },
+    {
+      "epoch": 2.9961099585062243,
+      "grad_norm": 0.5639104302579364,
+      "learning_rate": 3.2805802207547255e-11,
+      "loss": 0.4094,
+      "step": 3851
+    },
+    {
+      "epoch": 2.9968879668049793,
+      "grad_norm": 0.5318881957671844,
+      "learning_rate": 1.8453272569995962e-11,
+      "loss": 0.3707,
+      "step": 3852
+    },
+    {
+      "epoch": 2.9976659751037342,
+      "grad_norm": 0.4998559157815779,
+      "learning_rate": 8.201457278311786e-12,
+      "loss": 0.314,
+      "step": 3853
+    },
+    {
+      "epoch": 2.9984439834024896,
+      "grad_norm": 0.5037032983196463,
+      "learning_rate": 2.050364740213695e-12,
+      "loss": 0.3233,
+      "step": 3854
+    },
+    {
+      "epoch": 2.999221991701245,
+      "grad_norm": 0.4478490156477908,
+      "learning_rate": 0.0,
+      "loss": 0.3228,
+      "step": 3855
+    },
+    {
+      "epoch": 2.999221991701245,
+      "step": 3855,
+      "total_flos": 2098305569914880.0,
+      "train_loss": 0.4127941138317613,
+      "train_runtime": 46401.1105,
+      "train_samples_per_second": 7.977,
+      "train_steps_per_second": 0.083
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3855,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2098305569914880.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}